6.1 The Logic of Discrimination

We have seen (in Section 4.7) that Aristotle’s notion of perceptual discrimination (κρίνειν) conveniently encapsulates the difference between how (i.e. with what result) perceivers and non-perceptive entities, respectively, are affected by perceptual objects – that is, exactly the difference for which the perceptive soul is supposed to account. But we have also seen that different views are available regarding how exactly discrimination is intended to fit within the assimilation model of perception as it is worked out primarily in An. 2.5. Furthermore, we have seen that these differences go hand in glove with differences in understanding how the perceptive soul fits in – that is, in what way it is supposed to be the sought-after first principle of perception. It is now time to dig deeper into the notion of perceptual discrimination (Sections 6.1, 6.2, and 6.3) and the account of it offered in An. 2.11 (Sections 6.4 and 6.5). The proposed interpretation of the latter will then lead us directly (in Sections 7.2 and 7.3) to a more concrete elaboration of the third potential approach to the role of the perceptive soul outlined in Section 5.5.

Let us begin from the question of meaning ascribed in the De Anima to κρίνειν. Theodor Ebert showed decisively how misleading it is to translate the verb as ‘to judge’. By ‘judging’ we primarily understand ‘passing judgment on’ something, or ‘contending’ – ‘asserting’ something in a way that can turn out to be either right or wrong. However, κρίνειν never means that – at least not in the De Anima.Footnote ¹ Even in what Ebert (following Bonitz)Footnote ² calls the ‘judicial sense’ (sensus iudicialis) of κρίνειν, it is better rendered as, for instance, ‘deciding’ a case.Footnote ³ Furthermore, according to Ebert, this is just a special use of κρίνειν; its ‘primary sense’ has even less to do with ‘judging’: what it means is rather ‘to discriminate’ or ‘to discern’.Footnote ⁴

In addition to these general observations, Ebert also made two claims that relate more specifically to Aristotle’s account of perception, suggesting that: (a) meanings dependent on the sensus iudicialis in Aristotle’s ‘psychological writings’ are exceptional and insignificant for the question of ‘what is done in perceiving’;Footnote ⁵ (b) in its primary ‘discriminative’ meaning, κρίνειν is a ‘three-place predicate … “S discriminates x from y” … symmetrical as to the second and the third place’.Footnote ⁶ Scholars have largely accepted both of these claims,Footnote ⁷ but each of them can be called into question.

The second claim implies that ‘what is done in perceiving’ is primarily discriminating two qualities from each other. One criticism of this implication runs as follows: it would seem that this meaning of discrimination cannot be ‘cognitively basic’, because it does not account for how we cognize each of the two qualities in the first place.Footnote ⁸ One could come to Ebert’s aid by biting the bullet and insisting that there is simply no more primary sense of cognizing x than discriminating it from some y and vice versa. But that idea is objectionable.

Let us consider the kind of scenario on which the phenomenon of ‘change blindness’ has been explored.Footnote ⁹ Assume that someone replaces the colour y from which, on Ebert’s view, I am discriminating the colour x with another colour z. One can argue that, irrespective of whether I notice the change, on Aristotle’s view, it will make no difference for the content and the meaning of my discrimination of x. If this is correct, there is something wrong with accepting y rather than z into the formula of discrimination (before the change). It is wrong, I contend, not because x was not discriminated from y, but rather because it was equally well discriminated from z, even though z was not actively perceived. One upshot is that, if discrimination is to be analysed as a three-place formula, then this formula cannot be ‘symmetrical as to the second and the third place’. This observation accords with the etymology of κρίνειν. The original meaning (IE *krei-) seems to be sifting or winnowing on a sieve.Footnote ¹⁰ This suggests that y need not be noticed, and that there need be no determinate y at all. If κρίνειν originally means something like ‘to select the valuable part (the grain) from an indiscriminate mass’,Footnote ¹¹ then it is natural to expect that y involves all colours, while none of them is necessarily being perceived.

There are reasons to think that Ebert would reject this thought experiment. He insists that there is a ‘specific type of propositions involved in the propositional attitudes resulting from perception of special sensibles’.Footnote ¹² The information contained in the basic acts of perceptual discrimination is non-copulative in the sense of involving exclusively items of the same ‘ontological group’, namely the modal-specific qualities themselves. Ebert argues that this information is entirely free from our everyday ‘belief that [for instance] colours somehow “stick” to their objects’.Footnote ¹³ In other words, he emphatically endorses the idea that, according to Aristotle, we primarily perceive the modal-specific qualities alone, and that these are only attached to their putative bearers somehow additionally – namely, by our ‘beliefs’. This allows Ebert to insist that a change in one part of the perceptual field necessarily involves changes in other parts. If, in contrast, we accept the idea (defended in Section 4.2) that, according to Aristotle, we also perceive, at least equally primarily, the bearers of perceptual qualities, then it becomes natural to believe that the change from y to z will have no consequences for our discrimination of x.

We shall later consider further reasons for rejecting Ebert’s formula. However, before inquiring deeper into the logic of perceptual discrimination, we must first discuss the other claim flagged above – that is, Ebert’s marginalization of the judicial sense of κρίνειν, which has not hitherto been questioned by scholars.

6.2 The Judicial Sense of Perceptual κρίνειν

According to Ebert, there is one single passage in the De Anima in which the meaning of perceptual κρίνειν is determined by the judicial context. This is the passage from An. 2.6, already quoted in Section 4.2, where Aristotle contrasts for the first time the infallibility of modal-specific perception with the fallibility of other kinds of perceiving:

Recent translations tend to give κρίνειν in this passage the discriminative meaning, partly as a result of Ebert’s influence, it seems. However, Ebert himself took this passage to be the exception to the rule, arguing that κρίνειν here cannot be rendered in terms of ‘discriminating/discerning’.Footnote ¹⁴ The idea seems to be that one can very well say that ‘S judges about/of x’,Footnote ¹⁵ but it is not obvious what ‘S discriminates/discerns about/concerning x’ would mean,Footnote ¹⁶ and the preposition περί cannot be simply ignored.Footnote ¹⁷ Ebert, in any case, analysed this passage convincingly as employing κρίνειν in a well-established meaning derived from the judicial context, such that the best translation would be ‘each sense decides about these’.

Ebert is drawing here on a group of passages from Aristotle’s Politics, listed already by Bonitz, in which Aristotle uses the expression κρίνειν περί with the genitive in the sense of having the authority to decide some issue. More specifically, such an authority is standardly ascribed to the judge in matters that are not unambiguously determined (‘decided’) by the law.Footnote ¹⁸ The question of authority (who is κύριος, who is entitled to ἄρχειν, who is ὁ κρίνων or κρινῶν) often comes to the fore in these contextsFootnote ¹⁹ – so much so that on some occasions the correct translation of κρίνειν seems to be just ‘to have (judicial) authority’, as in a curious fragment from the Protrepticus where Aristotle says that the rational part of the soul ‘by nature governs and has (judicial) authority over us’ (κατὰ φύσιν ἄρχει καὶ κρίνει περὶ ἡμῶν).Footnote ²⁰ Unlike in nature, political matters often feature more options about how the authority to decide various issues can be distributed.Footnote ²¹ However, Aristotle does not, of course, take the question of distribution to be arbitrary; there is a certain normativity involved, which implies that the judge, for instance, is correctly recognized as the authority to decide certain kinds of issues. So, the question of who is ὁ κρίνων or κρινῶν in a given matter is often to be understood as asking about the ultimate authority to which one ought to defer should any uncertainty arise.Footnote ²²

Ebert argues that, in An. 2.6, Aristotle ‘makes each sense, as it were, sit as a judge, an arbiter upon its corresponding special sensibles’.Footnote ²³ This is an important observation that has not, I think, been sufficiently appreciated. It can, however, be further shown that (1) An. 2.6 is not the single exception: there is at least one other important passage in the De Anima, overlooked by Ebert, which draws clearly on the judicial sense. Moreover, (2) the clear-cut dividing line drawn by Ebert between ‘the primary sense’ of κρίνειν on the one hand and meanings derived from its ‘judicial sense’ on the other is misleading. Rather, the judicial overtones come more or less into the foreground throughout the De Anima, without ever entirely disappearing; recognizing this fact is important if we are to understand correctly what is at stake when Aristotle characterizes perception as a case of κρίνειν.Footnote ²⁴

The quoted passage from An. 2.6 deserves detailed attention because it is the passage in which Aristotle introduces the notion of κρίνειν into his account of perception (only two other passages mention κρίνειν in the rest of Book 2). However, this term does not appear suddenly or unexpectedly. Aristotle has already employed the notion of κρίνειν when addressing LKL in An. 1.5,Footnote ²⁵ and this is hardly accidental because Theophrastus uses κρίνειν fairly often for what his predecessors thought (or should have thought) is achieved in perceiving.Footnote ²⁶ In An. 1.5, Aristotle raises an objection against the assumption that, if the perceiver is to know (γνωρίζειν) – that is, to κρίνειν – two opposites, such as black (dark) and white (pale), she must contain them both. In the Empedoclean picture, for instance, it means that she must contain passages of both water and fire. This assumption, Aristotle argues, is idle, for one of the opposites would be sufficient:

The idea seems to be that the straight edge,Footnote ²⁷ so to speak, ‘makes’ (or serves to make) authoritative, and, indeed, objective assessments whether something is straight or bent – much like a judge is the ultimate authority in deciding whether something is just or unjust. This agrees with the original meaning of κρίνειν as sieving: the sieve has an undisputable, perfectly objective authority with regard to isolating grain. Mutatis mutandis, Aristotle expects a successful theory of perception to explain how the senses can be the ultimate and infallible authority with regard to modal-specific qualities.Footnote ²⁸ Moreover, he seems to be implying that at least some proponents of the LKL principle were already attempting to offer this kind of explanation, albeit not successfully.

So, what kind of authority exactly is being ascribed to the senses here? Theophrastus’ De Sensibus articulates a contrast between two different understandings of perceptual κρίνειν as a part of the overarching contrast between LKL and the idea that contrary is known by contrary (CKC).Footnote ²⁹ The Empedoclean version of LKL, as construed by Theophrastus, provides a perfectly objective and impartial account of perceptual κρίνειν: the passages of water are such that only effluences of water fit into them and the passages of fire are such that only effluences of fire fit into them, so that there is effectively no room for error or anything ‘subjective’ with regard to identifying the black (dark) and the white (light) colour of external objects. The Anaxagorean version of CKC, in contrast, makes perceptivity directly dependent on the actual state of the subject: we perceive and discriminate, for instance, the drinkable out in the world by the briny in us (and so only as long as there is something briny and thirsty in us). Anaxagoras, to be sure, still ascribes a remarkable authority to the senses in determining the objective features of things (the account is not relativist or sceptical); only the range of accessible features becomes dependent on the current state of the perceiver’s body, because we perceive only what we lack.Footnote ³⁰

This background provides a further reason for Aristotle’s focus on LKL in the De Anima:Footnote ³¹ in contrast to CKC, it can be presented as attempting to develop an account of perceptual κρίνειν that is entirely objective and impartial, a task that Aristotle is interested in taking over. As presented by Theophrastus, the Anaxagorean CKC would not be helpful in this respect, because it does not appreciate the impartiality of the rudimentary perceptual κρίνειν. That seems to be why Aristotle makes a different contrast central to the discussion he undertakes of his predecessors’ views on knowing – namely, the contrast with the Anaxagorean account of νοῦς as entirely impassive.Footnote ³² It is that position, and not CKC, which provides a genuine alternative to LKL with regard to accounting for the objective and impartial cognition achieved in perceiving. This alternative is not absent in Theophrastus’ presentation – but only marginal attention is paid to it. Theophrastus alludes to an alternative interpretation of Anaxagoras on κρίνειν when discussing Clidemus in chapter 38. Here, the idea seems to be that none of the senses really κρίνει anything; they only send the affections further to something else, namely νοῦς, which alone is capable of κρίνειν.Footnote ³³ Again, the account seems to be modelled on a judicial situation in which certain questions are deferred to a judge for decision. The presumption is, in any case, that the cognition achieved by Anaxagorean νοῦς is at least as objective and impartial as the one provided by LKL. That seems to be why Aristotle contrasts in the De Anima precisely this kind of account with LKL: both accounts can be presented as agreeing on what is achieved in perceiving – namely, a perfectly objective and impartial cognition – but disagreeing on how it is achieved.Footnote ³⁴

If this observation is correct, then it sheds additional light on the two intuitions that we identified (in Sections 1.4 and 2.7) behind the two horns of the key puzzle in An. 1.2. They can now be spelled out as two intuitions about the source and the nature of the judicial authority to be ascribed to perceivers. The Unbiasedness/Externality Requirement can be seen as drawing on the idea that, for something to have judicial authority over a certain question, it is necessary for that thing to be absolutely unbiased and not to be involved in it in any way whatsoever. It seems to be the same kind of intuition as the one occasionally articulated by Aristotle that no one rightly κρίνει things concerning himself. This is, for instance, why medical doctors are often unable to treat themselves properly.Footnote ³⁵ The Acquaintance/Contact Requirement, on the other hand, would be drawing on a very different intuition about the source of a judicial authority – namely, that in some areas at least, it is not possible to decide the question without being intimately acquainted with the matter, for it is only by having, or being able to have, personal experience with what is at stake that one is able to arrive at the requisite understanding.Footnote ³⁶ One model for this intuition about the source of judicial authority would be the idea of σύμβολα (which Aristotle associates with Empedocles).Footnote ³⁷ To decide on some matter, I already need, so to speak, to have one of the σύμβολα in my possession (say, a past impression of a scar) because that is the only way to identify the decisive σύμβολον (a scar) in the matter under consideration. Similarly, the fire in the eye is like one half of a die (tessera hospitalis) that perfectly fits into the other half, represented by all the fire found in the external world, so that perceptual discrimination is like a case of a guest-friend (ξένος) recognizing his guest-friend (ξένος).

The two intuitions are thus very different, but what they have in common is their concern with the kind of authority that can decide a matter under consideration in a perfectly just manner – that is, to assess it objectively as it truly is. Aristotle shares this concern (unless we say he is projecting it onto his predecessors) and seems to think that there is something right about both intuitions, although each of them was muddied by errors committed by earlier thinkers. The demand of unbiasedness, on the one hand, must not be absolutized in the Anaxagorean manner, because that would cripple the alleged judge from any contact with the thing on which she is supposed to decide. The ‘symbolic’ idea of κρίνειν, on the other hand, reifies the requisite likeness in such a way that it leads to unacceptable results, as exposed in An. 1.5. Aristotle already suggests an amendment in the quoted passage (411a2–7): the judicial authority must be rather like that of the straight edge which allows us to identify reliably not only straight but also bent things. This amendment, as we shall see, is then further developed in the final section of An. 2.11. Aristotle’s aim here is, arguably, to provide an account of perceptual κρίνειν that would preserve what is true about each disputing side, while preventing their respective shortcomings.

More could be said about the ‘judicial’ overtones of perceptual κρίνειν. They become even more prevalent against the background of Aristotle’s engagement, particularly in Metaph. Γ.5–6 (and Κ.6), with the ‘narrow’ Protagoreanism, as developed and discussed in Plato’s Theaetetus, which is largely about the judicial authority of the senses.Footnote ³⁸ The Protagoreans, on the one hand, insist that every perceiver contains a perfect criterion (κριτήριον) and measure (μέτρον), which makes her an ultimate infallible authority (κριτής, ὁ κρίνων) in the matter of what is perceived, albeit only insofar as it is perceived by her and so appearing to her.Footnote ³⁹ Socrates, on the other hand, draws into question the very authority of the senses to decide about truth.Footnote ⁴⁰ Aristotle emphatically asserts that the senses, in a healthy condition, are a perfect criterion (κριτήριον) and measure (μέτρον) for deciding (κρίνειν) about truth, the assessments of which cannot be further questioned in any meaningful way (contra Socrates), while insisting that what the senses access is an objective, and not merely a relative, truth (contra Protagoras).Footnote ⁴¹

The senses, in a healthy condition, are an objective and impartial authority, albeit only in the narrowly limited domain of modal-specific qualities. Indeed, Aristotle insists that the senses are the ultimate authority (κυρίον) in this domain in a sense that Socrates wishes to reserve for expert knowledge. Although our senses may lead us astray in causing us to believe, for instance, that the wine that is so sweet tonight will still be sweet tomorrow, there is no error possible about the perceived sweetness of sweet things itself:

Aristotle’s account of perceptual κρίνειν and its first principle in the De Anima can, then, be understood as effectively addressing an obvious objection that one may raise against his optimistic epistemology of perception as outlined in Metaph. Γ.5–6 and Κ.6. The grievance is that such an objective and impartial κρίνειν is, on the level of perception at least, simply impossible, and that no such authority can meaningfully be ascribed to the senses.Footnote ⁴³ In the De Anima, Aristotle blunts at least one edge of this criticism by identifying the first explanatory principles on the basis of which this kind of perceptual κρίνειν can be scientifically explained.

If this is, indeed, the background against which the notion of κρίνειν in An. 1.5 should be understood, then it confirms the centrality of the judicial sense of the verb. However, even without this wider context, the passage quoted from An. 1.5 is sufficient to suggest that it is more difficult than Ebert claimed to distinguish between the meanings of κρίνειν that are and those that are not derived from the sensus iudicialis. In the first sentence, κρίνειν can very well be translated as ‘to discriminate’, but in the second sentence the word κριτής (in whatever way it is translated) clearly evokes the judicial context. Thus, the passage strongly suggests that there is a continuity between the two meanings that Ebert wanted to separate.Footnote ⁴⁴ This continuity agrees well with the original meaning of κρίνειν as sieving, which is a (non-propositional) kind of discrimination involving a perfectly objective authority.

Aristotle’s judicial use of κρίνειν in An. 2.6 is, thus, far from being an exception that could be put aside when we attempt to understand his account of perceptual κρίνειν and of ‘what is done in perceiving’. The notion of κρίνειν introduced here rather encapsulates the task that Aristotle wishes to accomplish in what follows, including, crucially, the account of the ‘discriminative mean’ in An. 2.11.

6.3 Discrimination of the Imperceptible

As noted above, Aristotle’s use of the κρίνειν vocabulary in the De Anima is surprisingly sparse given how important a concept it is to his theory (as is broadly agreed by interpreters). Within An. 2.7–3.2, there are no more than two occurrences of it, besides the passage on the discriminative mean in An. 2.11 and the final section of An. 3.2 (on which, see Sections 6.4 and 6.6, respectively). These two occurrences are worth briefly commenting on here because they have a direct bearing on the question of the logic of perceptual discrimination, as well as on how the verbs κρίνειν and αἰσθάνεσθαι relate to each other. More specifically, these passages suggest that κρίνειν is a broader notion.

In both passages (An. 2.10, 422a20–31 and 3.2, 425b20–2), Aristotle employs the verb κρίνειν to articulate a thought that he elsewhere expresses without mentioning κρίνειν. The idea is that, for instance, sight is of both the visible and the invisible, hearing is of both the audible and the inaudible, and so on. Aristotle makes this claim four times within An. 2.7–3.2 and he never phrases it in terms of, say, sight seeing the invisible.Footnote ⁴⁵ However, in both passages in which he employs κρίνειν, he affirms that sight discriminates (κρίνει) – or that we discriminate (κρίνομεν) by sight – ‘the invisible’, namely darkness. This nuance should not be overlooked: the notion of discriminating under individual sense modalities seems to be broader than the notion of perceiving under them – as understood throughout An. 2.5–12 where, for instance, ‘perceiving by sight’ is equivalent to ‘seeing’. It is true that at An. 3.2, 425b20–2, Aristotle employs the notion of ‘perceiving by sight’ in a broader sense than ‘seeing’ that appears to be co-extensive with the notion of ‘discriminating by sight’.Footnote ⁴⁶ However, this seems to be an innovation vis-à-vis An. 2.5–12. Throughout these early chapters, ‘perceiving by sight’ is treated as equivalent to ‘seeing’, and so it would be wrong to say that sight perceives darkness, although sight is of darkness and discriminates it. One of the four passages making this claim (An. 2.11, 424a10–15) immediately follows Aristotle’s account of perceptual discrimination at An. 2.11, 423b31–424a10, and so the described mismatch between perception and discrimination is not unlikely to have a bearing on it.

Before considering that passage, however, let me add two preliminary observations concerning the logic of perceptual κρίνειν. First, the two passages insisting that sight discriminates darkness speak against the idea of κρίνειν as a three-place predicate: when I am ‘discriminating’ darkness, I am obviously not discriminating it from any actively perceived colour or light that would, thus, be symmetrically discriminated from darkness. Second, in the first of these passages (An. 2.10, 422a20–31) the meaning of κρίνειν seems, once more, to be directly connected to the judicial context. Again, the question is about distributing authorities, and the claim that ‘sight κρίνει this [i.e. darkness] as well’ could plausibly be paraphrased as ‘sight is also the authority to decide whether it is dark’. Indeed, given that this is the first occurrence of κρίνειν after An. 2.6, Aristotle’s claim is naturally read in direct connection with it: sight is not only the authority to decide about colours, but also the authority to decide about invisibility (darkness), and similarly for the other sense modalities. If this is correct, then the passage confirms that the sensus iudicialis plays a more central role in Aristotle’s argument than is usually assumed.Footnote ⁴⁷

6.4 Discrimination within the Assimilation Model

Let us now finally turn to the part of An. 2.11 in which Aristotle introduces the notion of a ‘discriminative mean’. Most of An. 2.11 has been dedicated to defending the thesis that flesh is not the organ, but merely the medium of touch. The concept of the ‘discriminative mean’ is then introduced in the context of tactile blind spots.Footnote ⁴⁸ Aristotle argues for the existence of these blind spots (at 423b26–424a4) drawing on the classification of tangible qualities as distinguishing features of bodies qua bodies (i.e. qualities that all bodies necessarily have to some degree),Footnote ⁴⁹ combined with the assimilation model of perception as developed in An. 2.5. The fact that Aristotle proceeds in this way is convenient for our purposes because the passage explicitly embeds the account of perceptual discrimination within the assimilation model:

The first thing to notice about this passage is that, although the account of the discriminative mean in (ii) is developed specifically with respect to touch, it seems to be intended to apply generally to all sense modalities, as suggested by the generalization in (iii) and confirmed by Aristotle’s later uses of μεσότης.Footnote ⁵⁰ Just as the notion of mediation was introduced as applying primarily to sight and hearing and then extended, somewhat surprisingly, to all other senses, so the notion of the discriminative mean is introduced as applying primarily to touch before being extended to the remaining sense modalities.Footnote ⁵¹

Another salient feature of our passage is the exceptional clarity with which Aristotle distinguishes the material and the formal aspects of perceivers – that is, the bodily perceptive organs on the one hand, and the perceptive capacity of the soul, on the other. In (i) he explicitly distinguishes between the organ (αἰσθητήριον) of touch and the sense of touch (ἁφή), which is what makes the organ capable of touching (ἁπτικόν). The assimilation model from An. 2.5 is then clearly applied to the perceptive organ rather than the perceptive capacity, which aligns perfectly with An. 1.3–4.Footnote ⁵² This suggests that Aristotle is willing to tell us something important about the respective roles of the body and the soul.Footnote ⁵³

Furthermore, there is no reason to assume that he is any less careful in this respect in (ii) and (iii). Yet, most translators and interpreters disregard the difference between μεσότης τις and τὸ μέσον in (ii): most often they are translated by the same word,Footnote ⁵⁴ and even when not, the difference usually plays no role in interpretation.Footnote ⁵⁵ Because it is αἴσθησις (apparently in the sense of the perceptive capacity, as a generalization of ἁφή) that is identified with μεσότης, however, (i) makes it natural to expect that the intended example of τὸ μέσον would be the organ as determined by the μεσότης, just as the organ of touch is ἁπτικόν owing to the ἁφή operative in it.Footnote ⁵⁶ This is not to say that μεσότης could not be treated in a more relaxed manner as a case of τὸ μέσον; rather, the point is that the present context strongly suggests that Aristotle expects the reader to be sensitive to the difference between that which is in a mean state and the mean itself.Footnote ⁵⁷

The final phrase of (i) clearly recalls the assimilation model from An. 2.5, and it has been interpreted in very different ways, depending on the literalist, spiritualist, or analogical understanding of the role of the perceptive organs from which interpreters have begun. However, the present passage brings into focus again how each of these approaches falls flat.Footnote ⁵⁸ It is notoriously difficulty for spiritualism to account for the phenomenon of blind spots introduced in (ii) as a consequence of the assimilation model: if only a ‘spiritual’ acting is involved in perception, the claim that ‘that is why we do not perceive what is equally warm and cold’ would be a clear non sequitur.Footnote ⁵⁹ If, on the other hand, the assimilation model was just about acquiring literal likeness, then the change that is supposed to underlie perception would effectively make it impossible. On becoming literally like the perceptual object, the organ would become unable to be affected by it and thus, as Aristotle clearly implies here, unable to perceive it.Footnote ⁶⁰ The change supposed to bring perception about would, in fact, only produce a perceptual blind spot.Footnote ⁶¹ Finally, the analogical reading does not fare any better on this point (as observed in Section 4.6). If the way in which a perceptual object F acts on the perceiver consists in making it G (i.e. making it embody the same ratio in a different pair of contraries) and if the result of this acting is that the organ comes to be G, then, again, the change that is supposed to underlie perception makes it effectively impossible, for it implies that the organ cannot be further affected by the perceptual object F in the required way, and so that the organ cannot perceive F.

The interpretation of Aristotle’s first general account developed in Chapters 1–3 offers a viable alternative (spelled out in Section 4.6), and Aristotle’s phrasing in the present passage provides additional support for it. The perceptual object F is making (ποιεῖ) FFootnote ⁶² the organ, which is F in capacity (δυνάμει ὄν). Aristotle says nothing here about the result of this making. However, combining this claim with the assertion that perceiving is a kind of being affected has an important implication that could have hardly escaped Aristotle’s notice: throughout the time of perceiving F, the respective organ must remain F in capacity; indeed, to continue being affected, and so perceiving, the organ must remain unlike the object (i.e. not F in fulfilment).

The present passage is not the only additional support for the proposed interpretation of the assimilation model. The same idea can, arguably, be found in a convoluted sentence from An. 2.10 describing how the tongue, as the alleged organ of taste, is affected by flavours:Footnote ⁶³

The syntax of this passage is uncertain and there are at least three ways to construe the sentence. The position of δέ calls for the insertion of a comma after σωζόμενον and suggests that this participle should be understood as a part of the μέν clause. This leaves the meaning of σωζόμενον underdetermined. (1) On one reading, it tends to be translated along the lines of ‘without being harmed’ and the sentence is taken to be saying at best something very vague.Footnote ⁶⁴ But Aristotle is more likely to have something more precise in mind, namely, (2) that the organ is moistened in such a way that it is preserved as being capable of being moistened: that is, upon being moistened, it does not lose its capacity for being moistened.Footnote ⁶⁵ The δέ clause (‘not being moist’) would then seem to be not simply describing the state before being moistened, but rather the state that needs to be preserved also when the tongue is being moistened. (3) Finally, a third construal that would elucidate this idea even more clearly would place a comma before σωζόμενον and take it already to be a part of the δέ clause. (This does not seem quite right stylistically, but it can hardly be excluded as a possibility in a manuscript of Aristotle’s work.) The sentence would then be saying that the organ of taste is moistened as something (a) that can be moistened (μέν), but (b) that, upon being moistened, is preserved as not moist (δέ).

What follows at 422b5–10 seems to support (2) or (3) against (1). Aristotle explains here that the tongue cannot properly taste anything if it is itself too moist: what happens then is that we do not really taste, but rather experience an after-taste on our tongue. This suggests that, although tasting involves the tongue being moistened, it must at the same time be preserved as itself not being (too) moist, because otherwise the moistening would eventually lead to the tongue becoming incapable of being moistened, and thus incapable of tasting, for it would acquire an (excessive) moisture (and, with it, apparently a flavour) of its own. Under (2) or (3), the passage fits perfectly with the interpretation of preservative πάσχειν proposed in Chapters 1–3,Footnote ⁶⁶ as well as the reading of (i) in the quoted passage from An. 2.11 proposed above.

Returning to the 2.11 passage, I have been arguing that Aristotle describes here what happens in perceiving in terms of the perceptual object making the perceptive organ like itself, with the organ, however, having its unlikeness preserved as a condition of being further affected, and so of continued perceiving. If that is true, then (i) spells out only a part of the assimilation model as developed in An. 2.5: it does not say anything explicit about the likeness acquired by the perceiver for the time of perceiving. And there is a good reason for this omission: it is, I shall argue, because Aristotle is going to tell us something new and important about the likeness in (ii). Put very roughly: the perceptive organ must not be assimilated to the perceptual object in the sense of acquiring a material quality of its own, because only then can it receive a quality of the external perceptual object. My suggestion in the following section will be that the notion of the ‘discriminative mean’ in (ii) is intended precisely to explain how the latter kind of likeness (further spelled out in An. 2.12 in terms of forms without the matter) comes to be present in the perceptive organ without compromising its neutrality.Footnote ⁶⁷

6.5 The Discriminative Mean

In (ii), Aristotle argues that the perceptive capacity (αἴσθησις), such as touch,Footnote ⁶⁸ is a certain μεσότης of the respective contrariety; this is why something or someone κρίνει the perceptual objects: καὶ διὰ τοῦτο κρίνει τὰ αἰσθητά. The grammatical subject is not expressed, but it clearly refers to the same kind of thing as ‘τὸ μέσον’ in what follows: τὸ γὰρ μέσον κριτικόν. Usually, interpreters assume that the grammatical subject of κρίνει is the μεσότης from the preceding sentence; this leads them, understandably, to treat ‘μεσότης’ as a case of ‘τὸ μέσον’. However, this is not the only possible reading.Footnote ⁶⁹ The sentence with κρίνει can also be taken as an absolute construction, ‘that is why one discriminates’, playing the same role as the first person plural at the beginning of (ii) ‘that is why we do not perceive’.Footnote ⁷⁰ Alternatively, the grammatical subject may be τὸ ἁπτικόν from 423b30, to which Aristotle has already referred back by the demonstrative pronoun ἐκεῖνο at 424a2.

Both of these readings suggest that the intended instance of τὸ μέσον capable of κρίνειν on account of μεσότης is not this μεσότης itself. Rather, it is the organ in which this μεσότης – that is, the perceptive capacity (αἴσθησις) in question – is primarily present, as outlined in the first sentence of (i).Footnote ⁷¹ This reading is supported by the beginning of (iii) where Aristotle talks of ‘that which is going to perceive white and black’ (τὸ μέλλον αἰσθήσεσθαι λευκοῦ καὶ μέλανος). He says that this must be neither black nor white in fulfilment but both in capacity, and he clearly compares this with what he has already said in (i) about the organ of touch (it is neither hot nor cold in fulfilment). Accordingly, τὸ μέλλον αἰσθήσεσθαι must be the perceptive organ. Now, there would be a strange shift if Aristotle were not also referring to the organ in the preceding sentence concerning τὸ μέσον. Rather, throughout the whole passage, he seems to be carefully distinguishing between that which can perceive, can touch, can κρίνειν, on the one hand, and that on account of which it can do so, on the other. The passage thus seems to be perfectly in line with the ‘Rylean passage’ in treating discrimination as something performed by an animal on account of its soul.Footnote ⁷² And this is no accident for, I shall argue, Aristotle is telling us something very important here about how the soul makes the animal capable of discriminating and perceiving.

Aristotle claims that the perceptive capacity is ‘a sort of mean’. This claim has been interpreted in various ways – of which I shall mention only two.Footnote ⁷³ According to Klaus Corcilius, Aristotle characterizes the perceptive soul here as a neutral value of each range, which is ‘a metaphysical given’ for AristotleFootnote ⁷⁴ and is localized in a non-metaphorical way ‘somewhere in the heart’.Footnote ⁷⁵ As such, the soul can serve literally as the end point of perceptual motions transmitted, according to An. 1.4, ‘up until’ the soul.Footnote ⁷⁶ Upon their arrival at this spatial point in the heart, ‘a sort of contact between body and soul’ is established and thus also a literal ‘contiguity of the soul with the incoming … motions’.Footnote ⁷⁷ Such a juxtaposition generates a contrast between the neutral value (i.e. the soul) and the value of the incoming motion. Because the former value is neutral, the contrast just is the manifestation of the incoming value: it is the perceptual object in activity – the form without the matter of the external object acting on the senses. It is in this context that Corcilius suggests an alternative to Ebert’s three-place model of perceptual discrimination: the perceptive soul qua mean discriminates the quality of the perceptual object from itself.Footnote ⁷⁸ Much can be said in favour of this account. It accomplishes what has rarely been achieved by providing a concrete model of the perceptive soul’s causality that is neither circular nor compromises the soul’s impassivity. I shall, accordingly, discuss this view further below.

In a recent paper, Roberto Grasso offers a different interpretation of μεσότης. He claims that it refers to a state of the perceptive organ resulting from two opposite agencies cancelling each other out: (1) the agency of the perceptual object assimilating the organ to itself and (2) the counterbalancing agency of the organ. (2) neutralizes (1), and so retains the organ in its neutral state. The μεσότης is just this state ‘that results from opposite extremes’ – a ‘zero vector sum’.Footnote ⁷⁹ I shall argue below that Aristotle does indeed have a homeostatic mechanism in mind in our passage; however, I do not share Grasso’s view that the passage is simply about physiology.Footnote ⁸⁰ As noted above, Aristotle seems to be promising to tell us something about the role of the soul. And, indeed, it is difficult to read the sentence about μεσότης as not concerning the perceptive soul. It is αἴσθησις that Aristotle identifies with ‘a sort of mean’, and although the word αἴσθησις can also refer to the activity of perceiving and is occasionally used to denote the ensouled organ, the present context strongly suggests that Aristotle has the perceptive capacity in mind (see 423b31).Footnote ⁸¹ It is difficult to see how the claim that αἴσθησις is a mean could be read as a claim about the preserved physiological state.Footnote ⁸²

However, if Aristotle does have the perceptive capacity in mind, how should we interpret its status ‘as a sort of mean’ and its role in perceptual discrimination? One problem with the first mentioned interpretation is that, as on most other interpretations, it disregards the difference between ‘μεσότης’ and ‘τὸ μέσον’. If I am right about Aristotle’s intention behind this difference, then it is not the perceptive capacity itself that is said to discriminate. Rather, Aristotle is saying that the organ (or the living being endowed with this organ) discriminates on account of the perceptive capacity operating in it ‘as a sort of mean’ and so determining the organ as ‘being in a mean state’. If this is the case, then there is some hope that we need not take the perceptive capacity as ‘a sort of mean’ to be a static constant whose mode of presence in the perceptive organ cannot be further analysed. Instead, a story can be told regarding how the μεσότης determines the organ as τὸ μέσον and thus makes it capable of discriminating. One reason for preferring this interpretation, I submit, is that it need not assume that the soul is non-metaphorically located in a spatial point ‘somewhere in the heart’.Footnote ⁸³

Such a non-metaphorical localization of the soul, which underlies the idea of a contact between the body and the soul and a contiguity between the soul and the incoming motion, seems to be at odds with the spirit of An. 1.3–4. Aristotle’s very first argument against the soul itself undergoing a change was that it would imply that the soul is in a place.Footnote ⁸⁴ That was considered to be patently absurd, and Aristotle never suggests that the idea of the soul itself having a position (θέσις), like a point, is more acceptable than that of it having a place (τόπος). Moreover, Aristotle’s definition of touch in Phys. 5.3 implies that only extended entities, not points, can be in touch with other beings.Footnote ⁸⁵ And, although his commitment to there being a touch between the mover and the moved leads him to introduce the notion of a non-reciprocal touching in GC 1.6 as a way of accounting for the case of unmoved movers,Footnote ⁸⁶ it is not at all obvious that this should be understood in terms of the unmoved mover having a determinate position.Footnote ⁸⁷ Accordingly, ascribing such a position to the soul would still be objectionable as a way of assimilating it to spatial entities (although not bodies) and would thus go against the thrust of An. 1.3–4.Footnote ⁸⁸

In my understanding of the passage, Aristotle is proposing an entirely different model of the soul’s involvement in perceptual discrimination. The soul figures in it as ‘a sort of mean’ not in the sense of a static constant, literally contiguous with the incoming motions, but rather in the sense of dynamically determining the perceptive organ as being in a mean state and so being discriminative, as it ‘comes to be the other extreme’ with respect to each object. This account can fill precisely the gap in the assimilation model left open in (i): it contains the key to understanding the sense in which the perceiver is assimilated to the perceptual object acting on her, while remaining neutral, and thereby capable of being further affected.

The persisting neutrality of the organ is, I suggest, captured exactly by its description as τὸ μέσον.Footnote ⁸⁹ The most obvious case, which Aristotle seems to have primarily in mind here, is that of perceiving the heat and the cold. The cold air acts on my body by literally cooling it down. If I were a plant or a kettle of water, this acting would result simply in the lowering of my temperature, not in perception. The reason for this would be exactly my lack of the requisite mean, as Aristotle says in An. 2.12: the reason why plants do not perceive, although they are cooled and warmed, is that ‘they do not have a mean (μεσότης)’; they ‘are affected with the matter’, rather than receiving the forms of perceptual objects.Footnote ⁹⁰ Having the requisite mean, I contend, implies precisely that the air acting on my body does not result in the lowering of my temperature, but rather in something else. Presumably, my temperature remains exactly the same (with smaller or larger local oscillations) owing to a homeostatic reaction that counterbalances the agency of the external object acting on me.Footnote ⁹¹ This homeostatic mechanism is what guarantees that I can continue being affected by the same object and thus can continue perceiving it.Footnote ⁹²

But there is more to be said. It is not just that my organ remains in its neutral state: when doing so, ‘it comes to be the other extreme’ with respect to the perceptual object acting on it, and so it discriminates this object. Aristotle seems to be capturing here the sense in which the perceiver is assimilated to the perceptual object acting on her, while remaining neutral with respect to it. The sense of ‘coming to be’ that Aristotle has in mind can hardly be a mere Cambridge change. He cannot mean just that τὸ ἁπτικόν ‘comes to be the other extreme’ with respect to anything we might wish to compare it with (just as something tepid ‘becomes’ warm in comparison with ice, and cold in comparison with boiling water). This would not explain why τὸ ἁπτικόν is κριτικόν, which is exactly what Aristotle is trying to explain. Rather, it seems, when affected by something cold, τὸ ἁπτικόν ‘comes to be the other extreme’, and exactly the other extreme, in the sense that, to remain a μέσον, it must undergo a certain amount of internal heating that corresponds exactly to the amount of cooling caused by the external object.

The key point here is that the counterbalancing reaction is – in a way that is yet to be explored – governed by the perceptive capacity that serves, in this precise sense, as the mean of the organ that defines the normative state in which the organ is always retained (or to which it returns as quickly as possible). This homeostatic mechanism provides a model for the involvement of the soul in perceiving that is not circular, does not compromise the soul’s impassivity, and need not ascribe a literal spatial position to the soul. Rather, the perceptive soul can be understood as being a kind of unmoved mover, which ‘meets’ the incoming motions not by becoming literally contiguous with them, but by its ‘acting’ on the organ, with the result that this ‘acting’ is directly determined by the agency of the perceptual object. This, I submit, is how the relevant quality comes to be present in the perceiver exactly as a measured quality of the external object rather than as a quality of the perceiver herself.Footnote ⁹³

An important point to notice about this model is that the soul cannot be absolutely impassive (like Anaxagoras’ or Aristotle’s divine νοῦς). Crucially, it is coincidentally affected by perceptual objects insofar as its organ is affected. Moreover, its activity is occasioned by the agency of the perceptual object and, indeed, is fully determined by it to the extent that the counterbalancing reaction does nothing more and nothing less than perfectly neutralize the agency of the perceptual object. In this sense, the activity can be described as receptive: it consists in nothing other than ‘receiving’ (in an inverted form comparable to a wax imprint) the quality of an external perceptual object. However, it is crucial that the reception does not take place in the perceptive soul itself. The soul does not ‘absorb’ the quality of the object into itself, not even in the etiolated sense of undergoing a kind of ‘spiritual’ alteration. Rather, the quality is received in the soul’s very acting on the organ, and so the soul is not affected in any non-coincidental way that would compromise its impassivity.Footnote ⁹⁴ We shall return to this point in Sections 7.4 and 7.5 when discussing the passages in An. that may seem to compromise the perceptive soul’s impassivity.

It is interesting, for the overall reconstruction, to note that the notion of μεσότης, when interpreted in the proposed way, can be directly related to the notion of ἀπάθεια that Aristotle ascribes to ‘that which can perceive’ in An. 3.4, drawing on An. 2.5. Characterizing the perceptive capacity as a mean comes very close to describing it in terms of an impassivity of the organ along the lines of Metaph. Δ.12 and Θ.1 – that is, in terms of a power for resisting a certain (detrimental) kind of being affected, which is a prerequisite, or even a constituent, of a capacity for being affected in an excellent way.Footnote ⁹⁵ It is exactly because the organ resists – in a determinate way owing to the soul – undergoing the change that non-perceptive entities would normally undergo when so affected that it is assimilated to the perceptual object in the requisite way (i.e. that it receives a quality of the external perceptual object not as a quality of its own but exactly as a quality of that object).

This goal is achieved by means of a homeostatic mechanism, through which the agency of perceptual objects is very precisely measured.Footnote ⁹⁶ The quality comes to be present in the ensouled organ as a result of this measurement: as a precisely determined proportion (λόγος) defining the quality in question. We could compare the way of measurement that Aristotle appears to have in mind here with weighing:Footnote ⁹⁷ τὸ ἁπτικόν is like a qualitative balance scale with one pan turned towards the perceptual objects and the other pan turned, so to speak, towards the soul; it resembles the beam of a balance scale continually brought to equilibrium by a mechanism governed by the soul as its unmoved mover. This weighing, in which the ‘weight’ of the perceptual object is continually determined, is the activity of perceiving – as a way of being affected by the perceptual object that is a case of discriminating that object.

The metaphor of weighing in the last paragraph was intentionally embellished as a way of emphasizing one question raised by the proposed account, namely: how exactly should we understand the alleged agency of the soul? It must, of course, not be understood as turning the soul into a quasi-bodily agent – that would be an obvious failure by the criteria set out in An. 1.3–4. Rather, as the nutritive soul, it is an impassive agent, much like an artFootnote ⁹⁸ – with the difference that it resembles not so much the art of carpentry but rather a measuring art.Footnote ⁹⁹ That, of course, does not fully answer the question, which will need to be discussed in its own right in Section 7.3. For now, however, the point is only that there is no prima facie reason for rejecting such an account of the perceptive soul as an efficient cause.Footnote ¹⁰⁰ What needs to be discussed is how exactly the model Aristotle develops for the nutritive soul can be adapted to perception.Footnote ¹⁰¹

Another question that must be addressed (in Section 7.2) is how far the proposed account can be extended from touch to the other sense modalities. Assuming that it works fairly well for the heat and the cold, one might wonder whether there is a way of making it plausible for other perceptual objects, as suggested by (iii) in the quoted passage from An. 2.11.

If we assume, for now, that these questions can be satisfyingly answered, we can see what kind of solution Aristotle provides in An. 2.11 to the key puzzle from An. 1.2 when rephrased in terms of κρίνειν (as proposed in Section 6.2). The judicial authority of perceivers is not based on any ‘symbolic’ relation to their objects – that is, it is not based on a static likeness preceding perception (as in the model of a guest-friend recognizing his guest-friend). Rather, to possess the requisite judicial authority, the senses need to be more unbiased than that: the perceiver must be, and remain throughout the duration of perceiving, neutral with respect to the perceived object and any other perceptual object within the given range.Footnote ¹⁰² However, this neutrality does not preclude the perceiver being in contact with the perceived object, and, indeed, being intimately acquainted with it, by means of a dynamic likeness constantly re-established by a measuring homeostatic reaction to the agency of the perceived object. The perceptive organ measures the perceived object not as a static straight edge (κανών), but rather by constantly receiving, so to speak, the perfect measure of the object from the object itself.

This may explain why Aristotle can say both that the perceiver measures perceptual objects and that she is measured by them.Footnote ¹⁰³ The perceiver measures the objects in the sense that, without the homeostatic reaction governed by the soul there would be no perceptual measuring whatsoever. However, the perceiver is also measured by these objects insofar as she does not contain any fixed measure (like κανών) of them: she is rather receiving the measure from the objects themselves throughout the activity of perceiving them, as they constantly determine, by their agency, the precise counterbalancing reaction governed by the soul.

6.6 The Single Mean and the Core of Perceptual Discrimination

Let us now step back from the details of the proposed reading of An. 2.11, 423b25–424a15. The interpretation of this passage offered in Sections 6.4 and 6.5 fits well with the overall picture of perceptual discrimination in the De Anima that I have been developing. What matters for now is that the passage neither supports the three-place model of perceptual discrimination, nor does it necessarily imply the alternative two-place model proposed by Corcilius in which the soul qua neutral value discriminates the quality of the perceptual object from itself. Instead, what Aristotle has in mind when he speaks of discriminating darkness, for instance, is an act of singling it out as being distinct from all potential objects of vision, that is, colours – plus apparently (intense) light in which no colour can be seen. This will also hold for colours themselves and for the other modal-specific qualities: when a colour is discriminated, it means that it is identified as just being this colour and no other; as on a sieve, it is discerned from the background of all potential colours (and objects of sight in general).

One can worry at this point whether the case of discriminating darkness and the case of discriminating colours can really fall under the same notion of discrimination. After all, we have seen that in An. 2.11 Aristotle embeds the account of perceptual discrimination firmly in the assimilation model and spells it out in terms of coming to be the other extreme. But it is obvious that none of this happens in the case of darkness. So, how could ‘discrimination’ mean the same thing here?

I think the answer is contained in the passages that have often been taken to support the three-place model of discrimination. The question at An. 3.2, 426b12–14 is what makes us capable of discriminating various qualities – of the same or of different modalities – in relation to each other (πρὸς ἕκαστον), and so to ‘perceive that they differ’. On the most general level, I take Aristotle’s answer to be that it is because the perceptive capacity is a single capacity.Footnote ¹⁰⁴ Later in An. 3.7, the claim is formulated in yet stronger terms: that ‘up until’ which all the perceptual motions, of various modalities, extend is a numerically single mean (μία μεσότης) with more than one being.Footnote ¹⁰⁵ Thus, when Aristotle identified a perceptive capacity such as touch as ‘a sort of mean of the contrariety in perceptual objects’, what he was describing was, effectively, just one of the beings (or aspects) of a more complex entity numerically identical for all sense modalities, and beyond.Footnote ¹⁰⁶

The idea of a numerically single mean seems to be directly connected to Aristotle’s account of perceiving the bearers of perceptual qualities, as discussed in Section 4.2. Furthermore, it can help us understand the discrimination of what is imperceptible. Suppose that I am standing on a mountain in the midst of a cloudy moonless night, and I am, in Aristotle’s terms, perceptually discriminating darkness. What happens, I take it, is this: I am perceiving the air around me by touch and smell, while perhaps also hearing how it moves; that is, I am perceiving this curious three-dimensional thing that is the air surrounding me, and I am perceiving it by the single mean active in me under several modalities at once. In the case of the three aforementioned modalities, the discriminated qualities of the air are all acting on me; and even one of them would be sufficient to make me perceive this thing (i.e. the surrounding air), because even one of them is sufficient to activate the single perceptive mean. As long as this mean is activated, albeit just by a single quality (say, the cold), it produces a holistic discrimination of its bearer. If the air makes no noise, then it cannot be heard, but it can be discriminated (under the auditory modality) as silent; and if there is no light present in it, it cannot be seen (or nothing can be seen through it), but it can be discriminated (under the visual modality) as dark; and so on. Thus, to return to our question: it seems true that all episodes of perceptual discrimination can be fitted into the assimilation model with the organ ‘coming to be the other extreme’ and so receiving a quality not as a quality of its own but as a quality of the external object acting on the perceiver; but it does not follow that the perceiver is assimilated to all the features of the perceptual object that she discriminates. There is sufficient room for discrimination of what is imperceptible, without any need to assume that this is an entirely different kind of discrimination or that it lacks any connection to the assimilation model.Footnote ¹⁰⁷

This consideration also sheds some light on the phenomenon of tactile blind spots and on why, after presenting the account of the discriminative mean, Aristotle returns, in the final lines of An. 2.11 (424a10–15), to the topic of discrimination of what is imperceptible with an emphasis on what is non-tangible. One kind of the non-tangible is ‘that which has an entirely indistinct διαφορά of tangible qualities, such as the air’. Now, air can surely have a very strong tangible διαφορά, for instance, when it is very cold;Footnote ¹⁰⁸ but it can also have no thermal διαφορά at all when it has the same level of heat as the perceiver’s body.Footnote ¹⁰⁹ This applies to other objects as well. When a pot of water has exactly the same temperature as our body, we cannot feel it, for it coincides with our blind spot; but it is nonetheless an object of touch and we can, in line with An. 2.10, discriminate its temperature by touch. When I put my hand into the pot and am affected by its humidity, I use touch to discriminate not only this humidity but also the water’s non-tangible tepidity.Footnote ¹¹⁰

What leads scholars to interpret Aristotle’s account of perceptual discrimination in terms of a three-place model is primarily the final part of An. 3.2, where he discusses the question of how we discriminate various qualities in relation to each other (ἕκαστον τῶν αἰσθητῶν πρὸς ἕκαστον κρίνομεν, 426b13–14), which implies that ‘we perceive that they differ’. What Aristotle analyses here, I take it, is not a special case, but rather something that happens in all (or nearly all) cases of perceiving: we standardly perceive the external object simultaneously under a plurality of its qualitative aspects, which implies that we also, in a way, perceive that these qualities differ from each other (otherwise there would be no perceived plurality at all). In this sense, it is true that in perception we standardly discriminate between actively perceived qualities (of a single modality, as well as of different modalities). However, it does not follow that the basic, and explanatorily primary, kind of discrimination is conceived by Aristotle as a three-place predicate. We have already seen several pieces of evidence telling against this assumption. Moreover, the way in which Aristotle introduces his inquiry into simultaneous perception in An. 3.2 provides no support for such a claim, either. Rather, if anything, it provides an additional piece of evidence against it. Aristotle writes:

Aristotle’s phrasing here is, admittedly, somewhat ambiguous. Elsewhere in the De Anima he speaks, for instance, of black and white both as διαφοραί of colour and as διαφοραί of visible bodies.Footnote ¹¹¹ These are, then, two candidates for the intended ‘underlying perceptual object’ in our passage. Given that the two remaining passages in the De Anima using this latter expression refer unambiguously to the quality rather than the body,Footnote ¹¹² it seems safer to assume that this is also what Aristotle has in mind here. In any case, the term διαφορά does not refer to a difference between two qualities; rather, it refers to what is specific about one quality in contrast to the other qualities of the same kind.Footnote ¹¹³ The only alleged source of support for the three-place model in the quoted passage thus can be taken to be found in Aristotle’s description of sight as ‘[discriminating] white and black’ and of taste as ‘[discriminating] sweet and bitter’. It has been argued that what Aristotle has in mind here are cases of discriminating the actively perceived white colour from the actively perceived black colour, and so on.Footnote ¹¹⁴ However, there is hardly any support for this reading, which would be very difficult to reconcile with the preceding passages on κρίνειν (as we have seen). Rather, even the present passage suggests that Aristotle has something else in mind. First, it would be prima facie awkward if he was claiming that we can discriminate sweetness only from an actively perceived bitterness, implying that no simple perception of sweetness was possible but that our experience is always literally bittersweet.Footnote ¹¹⁵ Furthermore, the μέν … δέ structure of the quoted passage suggests that Aristotle intends to consider the discrimination of multiple qualities in relation to each other (implying a perception of their difference) separately from the basic case of discrimination, which is taken for granted as something that has already been analysed before.Footnote ¹¹⁶

The key meaning of κρίνειν in the basic account, as I have argued, was that of singling out or identifying the quality of the given range possessed by the perceptual object acting on the perceiver – that is, deciding, with the ultimate authority, what the perceived object is like, not for the perceiver, but on its own.Footnote ¹¹⁷ This has nothing to do with distinguishing two perceived qualities from each other. Rather, we should think of contexts like the wind passage from the Theaetetus.Footnote ¹¹⁸ There may arise disagreements about what some perceptual object is like (what colour it has, whether it is cold or warm, and so on), where the question becomes what authority we should call on. Aristotle’s emphatic answer is that the ultimate, objective, and impartial authority with regard to modal-specific aspects of perceptual objects is the perceptive organ in a healthy condition.Footnote ¹¹⁹

In An. 2.11, Aristotle seems to come closest to explaining how this authority is grounded and how it is exercised. In the following chapter, we shall explore the loose ends of this account and ask how it builds up to Aristotle’s second general account of perception in An. 2.12.