1 Introduction
In De anima 2.6, Aristotle distinguishes between two kinds of sensible items (aisthētá): per se (kath’ hautá) and incidental (katà symbebēkós).1 The per se sensibles are so called because they affect the senses in and of themselves. They include what Aristotle refers to as proper (ídia) and common (koiná) sensible items. Proper sensible items are those by which each individual sense is defined: sight by colour, hearing by sound, smell by odour, taste by flavour, and touch by the tangible features such as temperature, solidity, and humidity. Common sensible items are those that necessarily accompany the proper sensible items and can be perceived by more than one individual sense. These items include shape, size, number, movement, and rest.2 For instance, when we see a colour such as whiteness, we necessarily see it as being of some size and shape, that is, we see a white figure or patch. In a typical case, the colour belongs to a particular object in some place and time. When we touch that object by hand, we can feel its shape and size. In contrast, incidental sensible items are so called because they coincide with the per se sensible items and yet do not affect the sense that perceives the per se sensible items in question. For instance, ‘the son of Diares’ is an incidental sensible item with respect to the sense of sight because he coincides with the white figure and yet he does not affect sight. In fact, the son of Diares is an incidental sensible item with respect to each of the five senses because he affects none of them. We shall see that not all incidental sensible items are of this type; for instance, bitterness is incidental with respect to sight but not with respect to taste.
The foregoing description of Aristotle’s distinction between the two kinds of sensible items is not controversial. What is controversial, however, is the question of how Aristotle understands perception (aísthēsis) of an incidental item. I shall refer to this perception as incidental perception, contrasting it with per se perception, which concerns proper and common sensible items. An influential interpretation suggests that incidental perception is “genuine” perception even when it concerns items such as the son of Diares that cannot affect the senses. The suggestion that incidental perception is genuine perception implies two claims: first, incidental perception is an activity of the perceptual capacity of the soul alone; second, the contents of incidental perception extend beyond those of per se perception.3 I shall refer to this interpretation as the “perceptual interpretation.”
The reason why many scholars are committed to the perceptual interpretation is the assumption that incidental perception plays a substantial role in Aristotle’s cognitive theory.4 They think that incidental perception allows us to sense individual substances that fall outside the scope of the individual senses and of the intellect.5 However, I argue that this assumption is controversial because it is not clear that Aristotle confines the scope of individual senses to sensible features in opposition to sensible substances, for instance, to a white colour in opposition to a white thing or white surface.6 But this is not the main reason why I am suspicious of the perceptual interpretation. Even if Aristotle confined the scope of individual senses to sensible features, there is another and more serious problem with the perceptual interpretation. The problem is that the perceptual interpretation cannot reasonably cover all those cases that Aristotle refers to as incidental, some of which he discusses outside the De anima, the text from which the perceptual interpretation draws most of its support. Even if the perceptualist could explain the incidental perception of the son of Diares in non-conceptual terms, for example, by reference to a memory of his sensible features, she faces a difficulty in explaining incidental perceptions of universals, which, according to Aristotle, are objects of knowledge and thus intelligible items.7 In Metaphysics 13.10, he claims: “Sight sees the universal colour incidentally because the particular colour that it sees is a colour.”8 How could one have a genuine perception of the whiteness as a universal (tò kathólou)? In Analytica posteriora 1.31, Aristotle resolutely denies that this is possible because according to him the cognition of a universal requires cognition of a cause.9 The perceptual capacity alone cannot inform us of the causes of things even if we might perceive, in some sense of the term, facts such as the fact that the moon is being eclipsed.10 If there is an incidental perception of the universal whiteness in some substantive sense of “incidental,” the sense of sight ought to be assisted by the intellectual capacity.11
The foregoing considerations suggest that the perceptual interpretation is problematic on textual and philosophical grounds. In what follows, I shall show that there is an alternative interpretation that avoids the problems outlined above. The alternative is subtler than what could be called the “intellectual interpretation”: the view that incidental perception is a thought or inference based on a per se perception.12 Even if this view can explain the incidental perception of the son of Diares and of the universal whiteness, for example, it cannot explain incidental perceptions of the proper sensible items of another sense, such as the incidental seeing of something bitter, which does not require an intellectual capacity. I argue instead that if we attempt to develop a substantive account of incidental perception based on Aristotle’s minimal remarks on the matter, we have to incorporate insights from both the perceptual and the intellectual interpretation: some cases of incidental perception can be explained by reference to the perceptual capacity, whereas others require reference to the intellectual capacity.13 In Aristotle’s implied view, then, there is no fixed set of incidental sensible items nor is there one single kind of incidental perception. Basically, Aristotle uses the term “incidental” (katà symbebēkós) in a negative sense to indicate that an individual sense does not perceive the incidental sensible item even if that item coincides with the per se sensible item that the sense perceives; instead the incidental item may but need not be perceived or cognised per se by another sensory or cognitive capacity, including the capacity for phantasía and the capacity for thinking. That means that Aristotle uses the term “incidental” to indicate the limits of a single sense. Aristotle also uses the term in a relative sense because he takes a sensible item to be either per se or incidental depending on the capacity to which it is attributed. Bitterness is a per se sensible item to the taste, but incidental to the sight. The same holds for intelligible items such as universals, which are (per se) objects of the intellect, but incidental objects to a sense. I argue that this is the way in which Aristotle introduces incidental perception in De anima 2.6. I shall call this understanding of incidental perception “deflationary.” However, when Aristotle, in De anima 3.1 and 3.3, proceeds to discuss incidental perceptions that have complex contents such as “The white thing is the son of Diares,” or “The yellow thing is bile,” he refers to these perceptions in positive and absolute terms. In other words, he assumes that these perceptions have contents that go beyond the per se sensible items, and that these perceptions can be explained only by reference to more than one cognitive capacity. I shall call this understanding “inflationary.” It is only in the inflationary sense that a substantive account of incidental perception can be given.
2 Incidental Sensible Items and Incidental Perception
Aristotle gives an account of incidental sensible items in De anima 2.6. There he suggests the following:
Something is said to be a sensible item incidentally if, for example, the white should be the son of Diares. One perceives him incidentally, because [1] he coincides with the white, [2] which14 one perceives. For this reason, [3] one is not affected by the sensible item insofar as it is the son of Diares.15
I argue that this account is best interpreted in the deflationary manner outlined above. Perceiving the son of Diares incidentally only implies seeing something white with which the son of Diares coincides. That means that the only thing that is perceived here is the white that one sees by sight. The son of Diares is not seen because he does not affect the sense of sight, and one can see only things which affect sight. But he need not be perceived by another sensory or cognitive capacity either, because no such requirement is set here for incidental perception. To perceive the son of Diares incidentally, then, the percipient need not have any appearance (phantasía) or memory (mnḗmē) of the son of Diares, nor any intellection (nóēsis) of the identity of the white figure seen. Likewise, I suppose, one perceives incidentally any other item that coincides with the white, and which does not affect the sense of sight, for instance, the citizen of Athens, or a democrat. If incidental items include relative features, such as being the son of someone, or standing one kilometre from the Parthenon, they must be innumerable.16 Perceiving an infinite number of items incidentally is not a problem, however, because these items do not constitute sensory contents for the sense in question; the only contents that it has derive from the per se sensible items that are finite in each case. In the deflationary sense, then, incidental perception can be true or false only in a derivative and purely stipulative sense: an incidental perception is true only if the corresponding per se perception is true, and it is false only if the per se perception is false. However, it does not make sense to say that an incidental perception might be false while the related per se perception is true. In the inflationary sense, by contrast, incidental perception can be false even when the corresponding per se perception is true, as we shall see below.
I argue that the reason why Aristotle introduces incidental sensible items and incidental perception is that he wishes to show the limits of an individual sense and in this way determine the kinds of cognitive functions that are possible on account of that sense without the contribution of another sense or some other cognitive capacity.17 His point is that even if each object has several features, both sensible and intelligible, an individual sense can single out only those that fall into its scope, that is, those items that affect the sense, both proper and common. It follows that the sense in question ignores others, namely incidental items, which affect some other sensory or cognitive capacity. To say that one perceives an F incidentally (by a given sense), then, means that one does not perceive that F per se (by that sense); instead, one perceives some per se sensible item with which the F coincides. To make this point about the limits of an individual sense, Aristotle need not refer to another capacity by which one can perceive or cognise the item in question per se. When a further capacity is involved, it is still true that one does not perceive the F per se on account of the original sense. Before I proceed to discuss this interpretation in more detail, I should like to make three terminological notes.
First, Aristotle’s reference to the white (tò leukón) and thereby to the sensible item (aisthētón) is ambiguous between the white colour and the white thing (or surface), that is, that which is white. Logically speaking, the claim “The white should be the son of Diares” makes sense only if “the white” is understood as signifying the white thing, but it is possible to understand “the white” as a sensible item in either of the two ways, that is, as the white colour or the white thing.18 Again, the latter phrase does not imply that the white and the thing are distinct visible items. Rather, we see the white thing only insofar as it is white, and yet we see that item in some place and time.19 I shall return to this matter shortly. Second, there are several ways in which an F can be incidental (katà symbebēkós) to a G.20 In the present case, the son of Diares coincides (symbébēke) with the white either because both are attributes that belong to the same thing, that is, a man, or because the son of Diares is an attribute of the white thing, or because the white thing is an attribute of the son of Diares. In any case, the son of Diares can be said to be a sensible item incidentally only because the white with which it coincides is a sensible item per se.21 Third, the qualification “insofar as” (hêi) specifies the respect in which the sensible item in question is supposed to be causally efficacious or inefficacious in relation to the sense. As a result, the percipient is supposed to perceive or fail to perceive the sensible item in that respect.
The account given in the above quotation concerns incidental sensible items such as the son of Diares that do not affect any individual sense. The first issue that we need to address is whether the account given can be extended to the other kinds of incidental items, whether sensible or intelligible, that Aristotle acknowledges elsewhere. In De anima 3.1, he refers to the proper sensible items of another sense. They differ from the son of Diares in that they affect some sense per se. In Metaphysics 13.10, as seen above, Aristotle claims that sight sees a universal colour incidentally. The universal colour must be an intelligible item when the intellect apprehends it. It might be asked whether it is an incidental sensible item when it is attributed to the sight. Even if Aristotle does not refer to the universal colour as an incidental sensible item, he would be consistent in doing so based on his claim about the incidental seeing and on his assumption that an item may be called an incidental sensible even if it does not affect any sense. That is the case with the son of Diares. That the ‘son of Diares’ is a particular substance or an attribute of a substance need not explain why it is classified as being an incidental sensible. For the purposes of my argument, however, it does not matter whether the universals are referred to as being incidental intelligibles or incidental sensibles in relation to the perceptual capacity. Either way, Aristotle’s point is that we do not perceive the universals per se. There are good reasons, then, to extend the foregoing account to the proper sensible items of another sense and to the universals, which, for convenience, I shall call incidental sensibles. Taking all these kinds of incidental sensible items into account, then, the three conditions that Aristotle suggests for the perception of an incidental sensible item can be generalised as follows:
-
[1’] The incidental sensible item coincides with a proper sensible.
-
[2’] There is a per se perception of that proper sensible item.
-
[3’] The individual sense that perceives the proper sensible item in question is not affected by the incidental sensible item insofar as it is incidental.22
Hence, I do not assume that incidental perception should be explained by reference to the common sense, for example. That the generalised account applies to the proper sensible items of another individual sense can be shown by way of an example. The sense of sight can incidentally perceive something bitter if the bitter coincides with the white, sight sees the white, and sight is not affected by the bitter item insofar as it is bitter. The generalised account also applies to universals because, for instance, the sense of sight can perceive incidentally colour (as a universal) if the universal colour coincides with, that is, is instantiated in some particular colour such as the white, sight sees the white, and sight is not affected by the universal colour insofar as it is universal. That is why Aristotle has good reason to distinguish between perceiving a particular (i.e., an F) and having a perception of the universal F when he discusses concept acquisition in Analytica posteriora 2.19.23 He says: “Although we perceive particulars, perception is of universals.”24 I am suggesting, then, that Aristotle takes the perception of the universal to be an incidental perception in the deflationary sense of the term. In this context, he makes it clear that unless memory and empeiría are involved, perception is not sufficient for apprehending the universal, be it a universal feature or a universal quantified proposition.25 In effect, then, Aristotle claims that we perceive a particular, but we do not perceive a universal. Here, again, he puts effort into showing the limits of perception. By analogy, it could be said that we see a white thing, but we do not see the son of Diares. The seeing is of the son of Diares, but that only means that we see the white thing that happens to be that person.
In none of the foregoing cases does the incidental sensible item fall within the range of the individual sense in question. This claim can be clarified as follows: if sight sees the white per se, and the white coincides with the son of Diares, for example, it does not follow that sight sees the son of Diares per se, i.e., as the son of Diares rather than as a white figure. It only follows that sight incidentally sees the son of Diares.26 Indeed, the son of Diares is not perceived as the son of Diares unless a more sophisticated cognitive capacity is involved. For instance, the dog of the son of Diares presumably recognises his master (i.e., the son of Diares) based on his per se sensible features but she does not perceive him as the son of Diares because she lacks the required intellectual capacity. In the absence of such a capacity, then, incidental perception is merely a way of speaking: it is a report on what coincides with the per se sensible items. This is what I have referred to as the deflationary case.
The foregoing interpretation has two major advantages. First, it succeeds in taking into account all the incidental sensible items that Aristotle acknowledges in his treatises. Second, in accordance with the text, it does not place any further requirement for perceiving an incidental sensible item. In particular, it does not require that incidental perception necessarily involve more than one individual sense. Even if the individual senses, according to Aristotle, are integrated to constitute the perceptual capacity as a whole, there is no need to adduce this assumption to explain incidental perception. By contrast, those who favour the perceptual interpretation typically make that assumption. In their view, sight perceives incidentally neither the son of Diares nor a bitter object unless it is assisted by memory, which explains why the white thing seen is recognised as being the son of Diares and why the yellow object seen is associated with the bitter object that has been tasted previously. The perceptualists, then, are convinced that a further capacity beyond an individual sense is required for incidental perception. However, this conviction is justified only if incidental perception is understood in the inflationary sense. To review some further reasons for the perceptual interpretation, and to justify the suggested alternative, we need to study the quoted passage in more detail.
It is not clear how the son of Diares should be understood in this passage. The interpretation of the term depends on one’s understanding of the white (tò leukón). The perceptual interpretation takes being the son of Diares as being an individual substance and being the white as an attribute of this substance.27 This interpretation is required if one assumes that Aristotle introduces incidental perception to explain how we can perceive an individual substance rather than a mere attribute of this substance. This assumption is a key motivation for the perceptual interpretation. According to this interpretation, seeing the white per se does not imply seeing the white thing per se, because we see the white thing only insofar as it is white. What we see, then, is just an attribute. Hence, the white thing is interpreted as being logically analogous to the son of Diares as follows: if sight sees the white per se, and the white coincides with the white thing, it does not follow that sight sees the white thing per se. It only follows that sight sees the white thing incidentally. Thus understood, we would perceive individual substances only incidentally, though based on perceptions of proper sensible features that are attributes of those substances.
I argue in contrast that it is not necessary to interpret the white as an attribute of the son of Diares. An alternative interpretation, as suggested above, is that Aristotle’s understanding of the white implies the white thing. According to this interpretation, being the son of Diares is an attribute of that which is white. That Aristotle applies the relative term “son” rather than the mere proper name “Diares” suggests that he has in mind the attribute rather than the substance.28 However, each of the foregoing two interpretations is possible insofar as Aristotle’s technical terminology is concerned.29 I prefer the alternative interpretation because I do not see why Aristotle should limit the scope of the per se perceptions to the attributes of individual substances. In other words, I do not see why Aristotle should assert that sight, for example, sees the white colour per se, and yet deny that it sees the white thing per se. Let me dwell on this matter for a while because it helps to contrast incidental perception with per se perception.
Even though Aristotle claims in De anima 2.6 that colour is the per se sensible item for sight,30 he qualifies that claim in the subsequent chapter by saying:
The visible is colour, and that [i.e., the colour] is on the surface of what is visible per se – per se not by definition, but because it contains within itself the cause of its being visible.31
Here Aristotle extends the scope of the per se visible beyond colour to the thing or surface that has within itself the cause of its being visible. He does not assume that the thing or surface is a distinct per se sensible item in addition to the colour. Rather, he merely gives a fuller account of what constitutes a per se sensible item. Aristotle needs such an account because he wishes to distinguish between two cases of seeing, namely seeing a coloured thing in light and seeing a fluorescent thing in darkness.32 In the first case, the thing or its surface is per se visible because it has a colour that is the cause of its being visible when there is light. In the second case, the thing or its surface is per se visible because it contains fiery or fluorescent matter that is the cause of its being visible when there is no light. That means that Aristotle does not see any problem in calling not only colour and fire, but also coloured and fiery things (that is, surfaces) per se sensible items.33 They are per se sensible in the same sense of the word, because in the present context the reference to a coloured or fiery thing is just a more complete way of referring to colour or fire.34 Thus, Aristotle does not commit a category mistake here. Moreover, he makes no suggestion that coloured and fiery things are analogous to incidental sensible items such as the son of Diares. Hence, he does not define the sense of sight by reference to the per se sensible item in a way that excludes the coloured or fiery thing. That is why he refers to sensible items as particulars (tà kath’ hékasta) in the first place.35 In doing so, he assumes that the individual senses perceive particular things (individual substances) per se. Therefore, it is incorrect to judge that, since Aristotle limits the scope of the individual senses to disparate sensible features such as colours and sounds, he faces a challenge of explaining how we can perceive objects that have those features.36 For the same reason, it is incorrect to judge that he introduces incidental perception to explain how we can perceive such objects. The motivation for the perceptual interpretation is therefore contestable.
Nevertheless, we need to further clarify the nature of incidental objects other than those that are proper sensible items of another sense. It is evident that universals are intelligible items. Is it correct to say that being the son of Diares is also an intelligible item? I think so, and I argue that this is Aristotle’s view. That is because we cannot apprehend this object unless we have some conceptual understanding of what it is to be a son of someone. Non-rational animals lack such understanding. It follows that they can perceive the son of Diares incidentally only in the deflationary sense, which does not require the intellectual capacity. In general, it is reasonable to suggest that incidental objects other than those that are proper sensible items of another sense are, properly speaking, intelligible items. That is because Aristotle divides the existing things into two main kinds: the sensible and the intelligible.37 Furthermore, he takes the division exclusively and exhaustively. Therefore, if an item is not sensible, it is intelligible, and conversely. However, the sensible and the intelligible coincide if that which is sensible is numerically the same thing as that which is intelligible.38
There is further evidence for this line of interpretation. In Physics 5.1, Aristotle refers to colour as an incidental intelligible item.39 Colour is presumably an incidental intelligible item because it is not the kind of entity that can affect the intellectual capacity, not at least insofar as it is some particular colour of a particular substance. Can we nevertheless have a genuine thought (nóēsis) about such a colour, as the perceptual interpretation would require by analogy? I do not think Aristotle thinks so. That is because, as seen above, he takes colour to be a proper sensible item to sight. It would, then, be superfluous to assume that there is another capacity by which we apprehend the very same item. However, it does not follow that we could not have genuine thoughts about colour in general, namely the universal colour, which we apprehend by giving an adequate account.40 Note that Aristotle would not refer to such an item as an incidental intelligible item because he takes universals to be proper (and hence per se) intelligible items. That is true even though he does not use the terms “proper” (ídion) and “per se” (kath’ hautó) in referring to universals as intelligible items.41 By analogy with incidental perception, then, Aristotle could suggest that we can think of a particular colour incidentally if
[i] that colour coincides with the universal colour (which it does, in a way),
[ii] we think per se of the universal colour, and
[iii] the particular colour does not affect the intellectual capacity.
The foregoing considerations suggest that, unless a further sensory or cognitive capacity is involved, incidental perception is just a manner of speaking in much the same way as incidental thought. The qualification ‘incidental’ suggests that the objects referred to are proper objects of another capacity. There is no implication that the other capacity ought to be active with respect to the object in question when we have an incidental perception or incidental thought. Indeed, the requirements that Aristotle puts on incidental perception and, by analogy, incidental thought imply that, if one has a per se perception or a per se thought, one also has, in the deflationary sense, an incidental perception or an incidental thought about those features of the object that coincide with the features perceived or thought per se. For instance, if we think of the universal colour per se, we think of a particular colour incidentally. However, that does not imply that we see that colour, or have a phantasm of the colour simultaneously. We may but need not do so.42 The same applies to incidental perceptions. If we see a white thing per se and thereby perceive the son of Diares incidentally, we do not need to have an appropriate act of the intellect regarding him.
3 Sensible Items as Causes
We have seen above that Aristotle does not assign any causal power to the incidental objects of perception and thought with respect to the sensory or intellectual capacity in question. That is because he defines incidental items by reference to their lack of causal power with respect to the sensory or intellectual capacity in question. This point is put forward as conditions [3’] and [iii] above. In the following, I shall clarify these conditions by relating Aristotle’s distinction between per se and incidental sensible items to the distinction he makes between per se and incidental efficient causes. My argument is that according to Aristotle incidental sensible items are merely incidental efficient causes, which means that they do not play a role in producing perceptions.43 To see that he is committed to these claims, we need to study Physics 2 and Metaphysics 6.2, where he discusses incidental causes in general.44
In Physics 2.3, Aristotle introduces a distinction between per se and incidental causes. As an example of a per se efficient cause, he refers to the doctor and the expert as causes of health, and to the sculptor as the cause of a statue. The doctor and sculptor are per se efficient causes of health and a statue, respectively, because they produce these things in the capacity of being a doctor (or an expert) and of being a sculptor. As an example of an incidental efficient cause, by contrast, Aristotle refers to Polyclitus, who coincides with the sculptor. This is what he says:
Furthermore, there is the incidental cause and its [different] kinds [i.e., the efficient and the final]; thus, the cause of a statue is in one way Polyclitus and in another a sculptor, because being Polyclitus coincides with the sculptor.45
Here, as in the case of perception, Aristotle uses the term “incidental” in a negative sense to indicate that which does not constitute an efficient cause. He is thus again pointing out the limits of the subject matter. I shall suggest below that Polyclitus being an incidental cause of a statue is analogous to the son of Diares being an incidental cause of the perception of a white thing. Before I go into the analogy, we need to study the case of Polyclitus in more detail.
In the foregoing citation, Aristotle places two requirements on someone named Polyclitus being an incidental cause of a statue: (1) he must coincide with a sculptor, and (2) the sculptor is the per se efficient cause of the statue. That means that Polyclitus does not produce the statue insofar as he is named Polyclitus, but insofar as he is a sculptor. In other words, his being named Polyclitus does not contribute to the production of this particular statue even if his being named Polyclitus co-occurs with his being a sculptor. That is because his skills in sculpting that he applies in the present case suffice to explain why the statue takes a certain form rather than another form. Regardless of whether he is an inexperienced sculptor or an extraordinarily skilled one who has succeeded in establishing a brand bearing his name, he informs the statue in the capacity of a sculptor rather than as a person named Polyclitus, with his skill rather than with his name. Indeed, he could change his name without losing his capacity to produce a statue. That is why Polyclitus, insofar as he is named Polyclitus, does not exert per se efficient causal power in producing the statue. He is merely an incidental efficient cause of that statue.
The foregoing consideration suggests that incidental efficient causes are not basic explanatory causes. Their causal efficacy is based on per se efficient causes, which are universal in the sense that they apply to all cases that are similar in some relevant respect. In this way, then, incidental efficient causes are reducible to per se efficient causes: if they have any causal power whatsoever, they exert that power by coinciding with per se efficient causes. However, the foregoing consideration suggests more than that. If incidental efficient causes make no difference to the product, it is questionable whether they can be regarded as explanatory causes at all. Rather, they would be efficient causes only by name and entirely eliminable from scientific discourse. In fact, there is some evidence suggesting that Aristotle takes this line of argument.
In Physics 2.5, Aristotle extends his discussion of incidental causes to final causation. There he identifies chance (týchē) as an incidental cause, observing that: “Strictly speaking, it is the cause of nothing.”46 Aristotle’s example is a man who is engaged in collecting subscriptions for a feast. Suppose the man goes to the agora for another purpose, say to buy vegetables, but by chance meets there a person who wishes to subscribe, and collects the money from her.47 Since the man goes to the agora to buy vegetables, and going to the agora and collecting the money coincide, it can be said that his wish to buy vegetables is an incidental final cause of collecting the money, whereas his wish to collect subscriptions is the per se final cause if it is operative in the present case (which Aristotle does not confirm, however). Aristotle adds that there are innumerable other incidental causes for going to the agora and collecting the money, such as a wish to see somebody, a wish to follow or avoid someone, or a wish to see a spectacle.48 He identifies a common pattern in cases of efficient and final causation because he compares the present case to a flute-player as an incidental efficient cause of a house.49 In Aristotle’s view, then, there are innumerable incidental causes not only for the man going to the agora, but also for the house being built: in addition to the flute-player, we may refer to a citizen of Athens, a taxpayer, a fan of Aristophanes, a bipedal featherless animal, an ensouled body, etc. All this suggests that Aristotle is not willing to assign any explanatory role to incidental causes. They do not have any per se explanatory force in a scientific account which holds “always or for the most part.”50
Against this background, it is easy to see why incidental sensible items cannot be explanatory causes. The only exception to this rule are the proper sensible items of another sense, but even they affect only that other sense, not the one to which the incidental perception is attributed. Therefore, it is reasonable to infer that all the explanatory work that is required for explaining sense perception is done by the proper and common sensible items (together with sound senses as well as favourable external and internal conditions for perceiving). From what has been said thus far, however, it does not follow that the proper and common sensible items could not explain incidental perceptions if there are such perceptions. That is because one might suggest that incidental perceptions are incidental effects of the proper and common sensible items affecting the senses. Aristotle rejects this suggestion, however. He does not think that the proper and common sensible items could produce incidental perceptions as side-effects. That is because he believes that in general incidental effects do not have per se efficient causes. Here is the evidence from Metaphysics 6.2:
For one who makes a house does not produce all the things which coincide with the house that is coming to be, for they are indefinite. There is nothing to prevent the house that is produced from being pleasing to some, harmful to others, beneficial to others, and different, so to speak, from everything that is. The art of house-building does not produce any of those things.51
Aristotle thus claims that the builder does not cause those things that co-occur with the house, such as the pleasure that the owners take from the house or the pain that the neighbours feel if the house blocks their sea-view. The builder’s capacity is thus restricted to producing the house, and does not extend beyond that. Understandably, there are some other per se causes that explain why people take pleasure or pain from the house, including the house and the capacities for perception and desire.
By analogy, I argue, Aristotle takes the proper and common sensible items’ capacities to be restricted to inducing perceptions of proper and common items. Furthermore, he takes those capacities to induce the per se perceptions in question “always or for the most part.” That is not the case with incidental sensible items. Being the son of Diares makes a difference only to those animals that have the capacity for both seeing and intellection. When a human being associates certain per se sensible items with incidental items such as the son of Diares, there is some further per se cause to explain why she makes such an association. For instance, the percipient may have a desire to find the son of Diares at the agora, and when she gets a glimpse of someone who looks like him, she may use the capacity for reasoning to judge that the white thing is indeed the son of Diares. However, this judgement is not a sense perception, but rather a belief that is based on a per se perception. I have referred to such cases as “inflationary” in opposition to “deflationary” cases, which do not require a further cognitive capacity. In the remaining sections of the present chapter, I shall examine whether either reading, the inflationary or the deflationary, can make sense of the key claims that Aristotle makes about incidental perception. In particular, I shall examine two claims: first, that the individual senses perceive each other’s proper objects incidentally, and second, that one can be mistaken about incidental sensible items.
4 The Senses Perceiving Each Other’s Proper Objects Incidentally
I will begin with the first claim, which Aristotle makes in De anima 3.1. There are two questions that we need to address: Why does he make that claim and what does it imply? I argue that the deflationary reading that I have given of De anima 2.6 is applicable to De anima 3.1, too. If this is correct, Aristotle wishes to emphasise the limits of each sense: each sense perceives its proper sensible items only, not the proper items of the other senses, even if those items coincide. Thus understood, saying that the senses perceive each other’s proper sensible items incidentally means that each sense perceives its proper sensible items only even if those items coincide with the proper items of the other senses. This implies that an individual sense can incidentally perceive another sense’s proper object without the contribution of that sense. My argument, then, is that the two senses need not cooperate for there to be an incidental perception.52 However, they do cooperate when we perceive these items per se at the same time, or discriminate (krínein) between them, and yet Aristotle does not refer to such perceptions and discriminations as incidental when he discusses them in detail in De anima 3.2 and 3.7 as well as in De sensu et sensibilibus 7.53 Here is the key evidence from De anima 3.1:
The senses perceive one another’s proper objects incidentally, not insofar as they are the senses they are, but insofar as they are one, whenever perception occurs of the same thing at the same time, for example of bile that it is bitter and yellow (for it most surely does not belong to another sense to say that the two are one).54
The perceptualists tend to read this passage as saying that the senses perceive one another’s proper objects incidentally only if they cooperate. That is taken to be what Aristotle suggests by saying: “insofar as they are one, whenever perception occurs of the same thing at the same time.” Incidental perception, then, is to be attributed to a capacity that is comprised of the individual senses. Some refer to it as the common sense, others as the perceptual capacity in its entirety.55 I argue that the text does not compel this reading, and that there is an alternative that better captures Aristotle’s intentions in the context.
The alternative that I introduced above is more modest in that it does not require cooperation of the senses. The proposal, then, is that the senses perceive one another’s proper objects incidentally only if there is a single object such as bile that has features belonging to the scope of each sense in question. Since taste perceives that which is bitter, and sight perceives that which is yellow, there is a single object for the two, namely bile, which is bitter and yellow, and yet the two senses discern a different feature of that object. Aristotle says that the senses are one because they have a common object. By saying that this happens “whenever perception occurs of the same thing at the same time,” he means either “whenever each perception occurs of the same thing at the same time,” or “whenever one perception, namely discrimination of bitter and yellow, occurs of the same thing at the same time.” However, as mentioned above, Aristotle does not refer to discrimination as being an incidental perception. That is why I prefer the former reading. Either way, the two perceptions are independent of each other, and the incidental perception refers to the deflationary type: perceiving another sense’s proper feature only requires that a sense perceives its own proper sensible feature that as a matter of fact coincides with the other sense’s proper feature. However, no necessary link such as necessary covariation between the two sensible features is required. There is thus no requirement for cooperation of the senses insofar as incidental perception is concerned.
The foregoing considerations show why the alternative interpretation is a reasonable interpretation of the present passage. This interpretation requires further support, however, because it might be objected that I have misconstrued the structure and contents of incidental perception. In fact, I have gone so far as to imply that in the deflationary case there is no structure and content to incidental perceptions because they are not perceptions at all. Indeed, I have suggested, incidental perception is just a way of saying what an individual sense does not perceive. The objection, then, is simply that Aristotle surely intends that incidental perception be more than that.
Thomas Johansen, for instance, suggests that in the passage quoted above, Aristotle takes the content of the incidental perception (in his terms, accidental perception) to be either “the bile is both bitter and yellow” or “the bitter is (the) yellow.”56 Either way, the incidental perception has contents that go beyond the proper and common sensible items. Johansen claims: “What makes the perception accidental seems to be that [sic] the introduction of a relation between the special perceptibles, sameness or difference, which is not a proper perceptible.”57 It follows that incidental perception has complex contents that are comparable, though not identical, to linguistic predications of the type “S is P.” It also follows, as Johansen sincerely admits, that we do not acquire such contents solely by individual senses. He says: “Insofar, then, as we understand vision as taking on the form of the sense-object it is hard to see how the accidental perceptibles become part of the content of vision.”58 He continues by claiming that we do not entertain these contents by the individual senses either:
insofar as accidental perceptibles are remembered, as implied by saying that they are acquired by learning, it seems plausible that we should not look to the special senses as the locus of their representation but rather to the common sense which is responsible for memory and other imagination-based activities.59
I am not convinced by this line of interpretation. Johansen is too hasty to suggest that we need to introduce the notion of relation to explain the incidental perception in question. It is worth noting that in De anima 3.2 and 3.7 as well as in De sensu 7, where Aristotle explains how we can perceive several proper sensible items at a time and how we can discriminate between them, he does not introduce such a notion.60 Indeed, there is no indication that he would take the contents of such perceptions and discriminations to be analysable in terms of relations between proper sensible items. That is true even if he assumes that a perception of several sensible features implies there being a unity of those features. That means that we either perceptually associate one feature with another one, or differentiate them. Neither of these basic acts requires that we perceive those features as standing in relation to each other either by a sameness or a difference relation. Of course, we can form perceptual beliefs about such relations, but that is not a matter of sense perception, or so Aristotle thinks.
Johansen is not alone in promoting this line of interpretation. There are many who argue that incidental perception requires the contribution of phantasía or memory.61 They assume that incidental sensible objects other than those that are proper (and common) cannot affect the individual senses, and yet these items can be “perceived” incidentally. Therefore, it is argued, incidental perception ought to derive its content from phantasía and memory. The assumption, then, is that we can retain a phantasm of the son of Cleon from an earlier meeting with him, and that is why we can recognise (that is, perceive incidentally) the white thing that we see by sight as being the son of Cleon. The same applies to seeing the bitter incidentally in those cases in which the sense of taste is not involved: we see a yellow item per se by sight, and recognise it as being bitter based on a phantasm that we have retained from an earlier tasting of the flavour of the object.
I think this interpretation is reasonable if incidental perception is understood in the inflationary sense. However, the passage quoted above does not require this understanding. I have two further objections that are more general in nature. First, the perceptual interpretation is in one way too broad in scope. It attempts to explain incidental perception of objects such as the son of Diares by reference to phantasía or memory even if the texts cited in support do not mention the two. Since these texts can be given an alternative interpretation that does not appeal to phantasía or memory, as I have tried to show, the perceptual interpretation seems to be textually unmotivated. Second, the interpretation is in another way too narrow in scope; there are some incidental items that we cannot perceive in the suggested way. In Metaphysics 13.10, Aristotle says: “Sight sees the universal colour incidentally because the particular colour that it sees is a colour.”62 There is no way in which the perceptual interpretation can deal with this passage other than saying that incidental perception is also informed by the capacity for thinking. In fact, that would be a reasonable explanation for the incidental perception of the son of Diares, too, if the perception in question is understood in the inflationary sense. However, that is not a persuasive move because it seems to contradict the key assumption underlying the perceptual interpretation, namely that incidental perception is an activity of the perceptual capacity of the soul alone. Furthermore, if incidental perception were “penetrated” (as the contemporary idiom has it) by intellection, the distinction Aristotle draws between perceiving and thinking would be seriously compromised, if not entirely obliterated.63 Again, since there is an alternative interpretation that can explain incidental perception of universals, there is no need to accept the perceptual interpretation in the first place.
In addition to the foregoing objections, I should like to note that, given Aristotle’s understanding of memory as being based on sense perception rather than thinking, we do not retain a phantasm of the son of Cleon insofar as he is the son of Cleon. Indeed, since memory, according to Aristotle, concerns intelligible items only incidentally, it is not based on a conceptual understanding of the object.64 Therefore, we retain the phantasm of the son of Cleon only insofar as he is, for example, a white, moving figure. This is a deflationary reading of the memory of the son of Cleon. However, memory can be combined with a further cognitive function, in which case we are considering it in inflationary terms. For instance, we can take sensual pleasure from remembering the son of Cleon if the per se perceptions that gave rise to the phantasm come with sensual pleasure, or if the phantasm is associated with some other pleasant phantasms regarding him. That is how Aristotle can account for those sensible items that are perceived as desirable: food, for example, can be analysed as a certain unity of per se sensible items that tends to elicit pleasure of a certain type when the animal is hungry.65 Accordingly, perceiving food can be analysed in terms of a per se perception that comes with feeling pleasure. This kind of perception can be called incidental, but it is incidental in the suggested inflationary sense of the word. By analogy, incidental phantasía can be analysed as a per se phantasía with, for example, pleasure or conceptual understanding.
The foregoing considerations suggest that incidental perception can be understood not only in the deflationary way, in which the reference is to an individual sense, but also in the inflationary way, in which two capacities cooperate. In the following section, I proceed to discuss the latter understanding in more detail, focusing on Aristotle’s claim that one can be mistaken about incidental sensible items. My argument will be that this claim is best understood in the inflationary sense. Furthermore, I argue that the mistakes that Aristotle explicitly refers to are best understood in intellectual terms. However, if he acknowledges mistakes about the proper sensible items of another sense, he is consistent in explaining them in perceptual terms.
5 Mistakes about Incidental Sensible Items
In several places, Aristotle claims that we can be mistaken about incidental sensible items. Here are the key pieces of evidence:
Each sense discerns these [proper sensible items] and is not deceived that there is colour or that there is sound – as opposed to what or where the coloured or sounding thing is.66
This is also why one is deceived when, should something be yellow, one thinks it is bile.67
The perception of proper sensible objects is true, or is subject to error to the least degree. Second comes perception regarding the attributes, and already here it is possible to be mistaken. For one is not mistaken that there is white; but only as to whether this or that other thing is what is white. Third, there is perception of the common objects which follow upon the attributes to which the proper objects belong (I mean, for instance, motion and magnitude, which are attributes of the sensible objects). It is about these that perceptual error is most likely to arise.68
As usual, Aristotle is very brief in giving these accounts. The verb that he uses for misperceiving is “to be deceived” (apatâsthai) in the first two passages, and “to be in error” (pseúdesthai) in the third. To all appearances, these verbs cover mistakes about all three kinds of sensible items, that is, proper, common, and incidental, if we assume that in the third passage Aristotle corrects himself in allowing for our being mistaken even about the proper sensible items. Indeed, he seems to imply in the first passage that we do not misperceive the proper sensible items, but in the third he makes a cautious qualification by saying that these perceptions are “subject to error to the least degree.” However, he need not assume that what it is to be deceived is the same in all three cases. That is because Aristotle accounts for the content of each type of perception in somewhat different terms.
The content of the perception of proper sensible items is given in existential terms “There is an F,” which can be understood as implying the mere presence of the object F, and the perception in question can be analysed in objective terms as “S perceives an F.” Being deceived about F, then, might be understood as perceptually affirming the reality of an F when there is no F present to the senses; instead, there is a G that appears to be F. By “perceptually affirming,” I am not referring to an act of assent that is distinct from the act of perceiving an F. On the contrary, I take it to be essential to all perceiving that each sense by default affirms the reality of what it perceives unless there is some more authoritative capacity to contradict their report.69 It is not clear whether the perception of common sensible items can be given an existential analysis along the same lines. An alternative analysis would be associative “The F with a G,” where F is a proper sensible item and G a common sensible item. This should be contrasted with the predicative analysis “The F is G,” which I take to require conceptual understanding, and to be a function of the intellectual capacity. I write F here with a definite article to indicate that the perception of a common sensible item is supposed to be based on a perception of a proper sensible item. Thus understood, misperceiving a common sensible item would be understood as mis-associating the proper sensible item with a common sensible item, which is to be contrasted with intellectually mistaking the proper item as being the common item. For the present purposes, however, we need not decide how Aristotle might understand being deceived in the cases of proper and common sensible items. Rather, we are interested in how he explains misperceiving the incidental sensible items. The above three passages suggest that he accounts for the contents of incidental perception in predicative terms, for example: “The yellow is bile,” or “The yellow is in a glass.” However, I shall suggest below that incidental perceptions of the proper sensible items of another sense are best analysed in associative terms. That is how those who prefer the perceptual interpretation tend to understand all incidental perceptions, even if some perceptualists account for them in predicative terms, failing to make a clear distinction between the predicative and the associative.70
The foregoing considerations about per se perceptions suggest the following relational understanding: there is a yellow item which S perceives incidentally to be bile; formally, there is an F which S perceives incidentally to be G. The yellow item is present to S through a per se perception. Based on this perception, S perceives incidentally that the item in question is bile. I argue that when this is not understood in the deflationary sense, it is a perceptual belief, because apprehending bile requires some conceptual understanding of the matter in question.71 It would be more precise to say that S takes the item seen as being bile. In other words, there is a yellow item about which S believes that it is bile.
It is easy to see why Aristotle thinks there are false incidental perceptions in addition to those that concern the proper sensible items of another sense: he thinks that one can mistake the yellow item for bile even if it is not bile. For instance, when there is yellow soda in a glass and one sees the yellow, she can mistake it for bile, or for being in a bottle. In Aristotle’s view, then, misperception of an incidental item involves mistaking a proper (or common) sensible item for something other than what it is. I argue that this is a misbelief about a proper (or common) sensible item rather than a misperception that could be accounted for in terms of a per se perception combined with phantasía or memory. Indeed, if incidental perception has predicative structure with conceptual contents, it is most reasonable to interpret it as a belief rather than as a perception which lacks conceptual contents, however complex it is otherwise. Therefore, I think, those who favour the perceptual interpretation owe us an explanation for how “being bile” or “being in a bottle” can constitute contents for a perception. The only explanation that I know from literature is this: even if the capacity for perceiving bile or a bottle requires that one has acquired the concept of bile and the concept of bottle by learning and thus by using the intellectual capacity, one may nevertheless perceive bile and a bottle through a single perceptual act without the intellectual capacity being involved in this perception.72 Perception, then, may have in part intellectual contents even if it is not combined with an intellectual act. The problem with this explanation is, however, that it separates intellectual contents from intellectual acts. There is no evidence that Aristotle would take this line of argument. On the contrary, there is much evidence that he takes the two to be inseparable. Thus, the perceptualists seem to apply the term “perception” too broadly when they suggest that all incidental perceptions are genuine perceptions.
I am thus suggesting that in the cases considered above, we should understand misperception of an incidental item as a misbelief about a proper or common sensible item. Unless we take it as a misbelief, we face a difficulty in explaining why Aristotle, in the second passage, explains the error in question by using a term that refers to thinking rather than perceiving. He says that the error arises “when, should something be yellow, one thinks it is bile.” It is natural to understand the expression “one thinks” (oíetai) as indicating a misbelief about the yellow thing, which, in the present case, is sweet rather than bitter.73 The passage in question immediately follows Aristotle’s discussion of the senses’ incidental perceptions of one another’s proper sensible items, which we have studied in the previous section of the present chapter. In that context, the misbelief arises because we judge the yellow thing to be bile based on seeing the yellow only. That judgement is based on seeing the yellow only because the alleged incidental perception of sweet does not contribute anything to seeing the yellow. Thus, I argue, Aristotle uses this example of a misbelief to show that there is no content to the incidental perception in question.74 If we had tasted the object in question and learned that it is sweet, we would have realised that it cannot be bile because bile is bitter rather than sweet. It is worth noting that there is nothing in the context to suggest that Aristotle would explain the mistake in question solely by reference to the senses of taste and sight, either individually or jointly. That is because neither of the two apprehends bile insofar as it is bile. They cannot apprehend bile because bile qua bile does not affect these senses. Hence, if there is a mistake about the nature of the yellow perceived, it ought to be explained by reference to a higher cognitive capacity, namely the capacity for thinking, as Aristotle correctly implies by using the expression “one thinks.” I assume that an analogous analysis can be given of a false incidental perception of the son of Diares, as follows: there is a white thing about which one believes that it is the son of Diares.
There is further evidence for the suggested understanding of “one thinks.” In De insomniis 1, Aristotle claims: “Mis-seeing and mishearing implies that one sees or hears something real, but not what one thinks it is.”75 The context suggests that this claim is not to be taken as a general account of mis-seeing and mishearing. Aristotle only has in mind cases in which we perceive some feature of an object correctly, but make a mistake regarding another feature.76 That is why this is analogous to the case discussed above. In the present context, Aristotle refers to the sun appearing to be a foot across, “even to those who are healthy and have knowledge.”77 I understand him to be suggesting that in this case, we are seeing correctly a real object, the sun or its yellow colour, but are deceived about the size. There are two things involved in being deceived here. First, the sense of sight is deceptive because it does not discern accurately the size of distant objects. People who are healthy and in their right minds are aware of this limitation of sight and do not believe that the sun is a foot across even though it looks to be so. Second, in some cases, when people are seriously ill, for instance, they may be oblivious to that fact and be misled into thinking uncritically that the sun is indeed a foot across. That mistake is not a sensory misperception but rather a false belief about what one really sees or hears. Something similar happens to those who are asleep, or so Aristotle suggests. Since belief (dóxa), or rather the capacity for forming beliefs, is held in check during sleep, it uncritically follows dream phantasms.78
I am thus suggesting that the incidental perceptions in question are beliefs rather than perceptions. It might be asked why Aristotle refers to them as perceptions in the first place if they are not genuine perceptions. There is not much that one can say based on direct evidence. Nonetheless, I argue that Aristotle would have good reason for doing so. That is because he would be consistent in separating beliefs that are based on actual per se perceptions from beliefs that are not based on such perceptions even if they might be about per se sensible items. In perceptual beliefs such as “The white thing is the son of Cleon,” the subject term refers to an item that is currently perceived by sight. In non-perceptual beliefs, by contrast, the subject term does not single out an object of actual sense perception. In De anima 3.6, for instance, Aristotle gives as an example the thought “Cleon is white.”79 Per se perception is not involved in this thought even if the thought is about a particular substance that can be perceived insofar as it is white. In this context, Aristotle claims that it is the intellect that combines the two items to make an assertion that is either true or false.80 He does not make a corresponding claim about the perceptual belief “The white thing is the son of Cleon,” perhaps because even if that belief is the work of the thinking capacity, it involves the sense of sight being active. Hence, it is a belief that is based on a per se perception.
The foregoing considerations do not imply that Aristotle always refers to perceptual beliefs as incidental perceptions. In De anima 3.7, he discusses a case in which a thinker is moved “outside of perception, whenever one is presented with phantasms.”81 “For instance,” he says, “one who perceives a beacon, because it is fire, and by common perception sees it moving, recognises that it is an enemy’s.”82 Here the verb “recognises” (gnōrízei) signifies a belief rather than perception. This case is “outside of perception,” first, because even if recognition is based on one’s seeing the fire moving, it is not an activity of sight. Instead, it is a belief that the moving fire belongs to an enemy. Second, recognition is “outside of perception” because it involves some phantasm about an enemy, for instance, an unpleasant feeling about an earlier encounter with the enemy.
Thus far, I have not considered mistakes about the proper sensible items of another sense. It is difficult to determine Aristotle’s view on the matter because he does not discuss these cases. His lack of interest in the matter can be partly explained by the fact that he thinks we misperceive the proper sensible items “to the least degree.” However, that does not imply that he thinks the per se perceptions are infallible in all circumstances. For example, an ill person may taste a sweet object as bitter if her tongue is filled with bitter fluid.83 A further difficulty is that the alleged mistakes about the proper sensible items of another sense seem to overlap false discriminations of proper sensible items, and yet, as I have noted, Aristotle does not refer to discriminations as incidental perceptions. That is why we cannot be sure if he wishes to make a distinction between the two. However, we may set aside this difficulty, which is basically a terminological one, and examine how a mistake about the proper sensible item of another sense could be analysed. There are basically two kinds of cases. First, one might see a yellow object correctly, but taste it incorrectly as being bitter. Here the mistake is based on a failure of taste. Second, one might see correctly a yellow object (which is sweet) and taste another object correctly as being bitter, but make a mistake in associating the yellow object with the bitter object. In this case, the mistake can be attributed to neither sight nor taste individually, but rather to the cooperation of the two senses. In either case, the mistake can be explained in perceptual terms. This is, then, where I follow the perceptual interpretation. However, I do not assume that the associative mistake in question implies a predicative mistake.
6 Conclusion: The Cognitive Role of Incidental Perception
I will conclude my discussion by making some observations on the cognitive role of incidental perception. If the proposed interpretation is correct, Aristotle has two different ways of understanding incidental perception: the deflationary and the inflationary. I suggested that it is only in the inflationary sense that incidental perception can play some cognitive role. As an example of such a role, I made some remarks on how Aristotle understands the perception of the universal as an incidental perception when he accounts for concept acquisition in Analytica posteriora 2.19. I pointed out that according to Aristotle we do not grasp universals on account of the senses alone, but rather on account of several cognitive capacities that help us to proceed step by step in our inquiry towards understanding the universals.
Aristotle rarely draws a distinction between per se perception and incidental perception outside his psychological treatises. Even when he does so, it is difficult to determine exactly what cognitive role, if any, he assigns to incidental perception. In some cases, if my interpretation is correct, he does not assign such a role, because he wishes to show merely that a given sense does not perceive the incidental item in question per se. That is the case with the son of Diares in De anima 2.6. By contrast, many interpreters have assumed that Aristotle discusses this particular case because he wishes to show that incidental perception is required for cognition of a particular object as opposed to its attribute. In the present chapter, however, I hope to have shown that incidental perception, as it is conceived of in the De anima, need not play that role.
Nonetheless, the fact that Aristotle uses the term aísthēsis in an unqualified way in his logical and ethical treatises has led some scholars to reasonably assume that some occurrences of the term can be interpreted in the incidental sense. It has been suggested, for instance, that in contrasting practical reason (phrónēsis) as a kind of perception with the perception of proper sensible items,84 Aristotle has in mind incidental perception.85 This interpretation is not entirely out of place, if we understand the perception in question as a distinctive kind of cognition that is based on a per se perception. Practical reason can thus be incidental perception in the inflationary sense. I should like to emphasise, then, that even if practical reason concerns particular items that coincide with per se sensible items, it does not follow that according to Aristotle practical reason is an activity of the perceptual capacity of the soul.
However, when Aristotle claims, for example, that we perceive that the moon is eclipsed,86 he does not seem to use the verb “perceive” (aisthánesthai) in either of the two incidental senses, the deflationary and the inflationary, that we know from the De anima. The predicative content “the moon is eclipsed” is analogous to the content “Cleon is white,” which, as shown in the previous section, does not require an actual per se perception. Rather, it constitutes content for a belief. That is all that Aristotle needs in the context of Analytica posteriora 1.31, where he contrasts “perceptions” of facts with knowledge of their causes. We can replace the verb “perceive” with the verb “believe” or “observe” without changing the meaning of the claims that Aristotle makes. Of course, nothing prevents us from referring to such beliefs or observations as incidental perceptions, but the evidence from the De anima suggests that Aristotle did not acknowledge such a broad understanding of incidental perception.
Acknowledgements
This research has been funded by the Riksbankens Jubileumsfond. I would like to express my gratitude to Pavel Gregoric, Simo Knuuttila, Péter Lautner, Seyed Mousavian, Filip Radovic, David Sanson, and the members of the Representation and Reality programme who made very generous comments and suggestions on this chapter in April 2019. In finishing the chapter, I greatly benefited from the comments that I received from the organisers and participants of the Tartu Workshop in Ancient Philosophy in October 2019 and the Helsinki History of Philosophy Seminar in November 2019, as well as from an anonymous referee. I am also very grateful to the editor Juhana Toivanen and the language editor Jordan Lavender who helped me to improve my writing and thus further clarify my arguments.
Aristotle, De anima, ed. W. D. Ross (Oxford: Clarendon Press, 1961), 418a8–25. Instead of “incidental,” some interpreters use the term “accidental”; these terms have the same meaning in this context.
De An. 2.6, 418a17–18. In de An. 3.1, 425a16, there is, according to most manuscripts, a more complete list, which includes “unity” (
This interpretation originates from Stanford Cashdollar, who claims that: “incidental perception is a case of aisthēsis alone” (Stanford Cashdollar, “Aristotle’s Account of Incidental Perception,” Phronesis 18 (1973): 156). Several others have followed Cashdollar, though in somewhat different ways, including Deborah Modrak, Aristotle: The Power of Perception (Chicago: The University of Chicago Press, 1987), 69; Paolo Crivelli, Aristotle on Truth (Cambridge: Cambridge University Press, 2004), 108; Pavel Gregoric and Filip Grgic, “Aristotle’s Notion of Experience,” Archiv für Geschichte der Philosophie 88 (2006): 11; Gregoric, Aristotle on the Common Sense, 199–201; Thomas Kjeller Johansen, The Powers of Aristotle’s Soul (Oxford: Oxford University Press, 2012), 182–84; Ron Polansky and John Fritz, “Aristotle on Accidental Perception,” in Aristotle – Contemporary Perspectives on His Thought, ed. D. Sfendoni-Mentzou (Berlin: De Gruyter, 2018), 125–50; Victor Caston, “Aristotle on Perceptual Content” (unpublished manuscript).
For instance, Cashdollar claims: “[…] incidental perception is of no less importance than perception of special and common objects in Aristotle’s psychology and thought in general.” (Cashdollar, “Incidental Perception,” 156.) More recently, an even stronger claim has been put forward by Polansky and Fritz: “Accidental perception is significantly more complicated than the simpler types of perception that serve as its basis, yet it seems ultimately to be the most important sort of perceiving that animals capable of it engage in.” (Polansky and Fritz, “Accidental Perception,” 149.)
See, e.g., Gregoric and Grgic: “It seems that experience of a thing necessarily requires perception of at least one incidental sensible, namely that thing. John’s experience of the table requires that he is perceptually aware of something in addition to a brown colour and a square shape, that is of an object which happens to be brown and square, and which English speakers would call ‘table’. The table is not perceptible in itself, but incidentally, because some features that are perceptible in themselves happen to belong to it, i.e., because the table is brown and square. […] At this point we must insist that incidental perception is indeed perception, rather than some other type of cognition, inference or whatever else has been suggested. We would agree that the ability to perceive incidental sensibles requires development, and that this development may need co-operation among various cognitive capacities, notably representation and memory. However, once the ability is sufficiently
developed, incidental sensibles are indeed perceived.” (Gregoric and Grgic, “Aristotle’s Notion of Experience,” 11.)
That is not to say that colour and surface are distinct per se sensible items in the way in which colour and shape are distinct, one being a proper sensible item and the other common, or in the way in which colour and place are distinct, one being a proper sensible item and the other incidental. The point is rather that seeing a colour implies seeing a coloured surface. Thus understood, the latter seeing has no more content than the former one; it is just a different, more complete way of saying the same thing.
De An. 2.5, 417b22–23.
Aristotle, Metaphysics, ed. W. D. Ross (Oxford: Clarendon Press, 1924), 13.10, 1087a19–20.
Aristotle, Analytica priora et posteriora, ed. W. D. Ross (Oxford: Clarendon Press, 1964). For arguments that Aristotle’s considerations in APo. 1.31 are compatible with the further claim that “Although we perceive particulars, perception is of universals” (APo. 2.19, 100a16–b1), see, e.g., David Bronstein, Aristotle on Knowledge and Learning (Oxford: Oxford University Press, 2016), 245; Marc Gasser-Wingate, “Aristotle on Perception of Universals,” British Journal for the History of Philosophy 27:3 (2019): 446–67. I shall give my interpretation in section two below.
APo. 1.31, 87b40–88a2.
Caston, “Perceptual Content,” denies that the incidental perception of a universal require an intellectual capacity, and the application of a concept. In his interpretation, even non-rational animals are capable of perceiving objects as certain kinds of objects. He says that he agrees with Jonathan Barnes, Aristotle: Posterior Analytics, 2nd ed. (Oxford: Clarendon Press, 1993), who comments on APo. 2.19, 100a17, thus (on p. 266): “[…] the process Aristotle describes produces universals; but it starts from perception and perception is of particulars – how, then, can the gap between particulars and universals be jumped? Aristotle’s answer is that perception in fact gives us universals from the start (cf. A 31, 87b29). He means that we perceive things as As; and that this, so to speak, lodges the universal, A, in our minds from the start – although we shall not, of course, have an explicit or articulated understanding of A until we have advanced to Stage (D) [i.e. the final stage of inquiry]. (It should be noted that this account is intended to hold for all perceivers: it is not peculiar to human perception, nor does it involve the intellect in any way. Even a fly sees an F.)” (Emphasis original.) However, Barnes gives no reason to assume that the fly could see something as F (i.e., as belonging to a kind, as a universal) by simply seeing an F, i.e., an instance of F, not the F itself. The gap, then, remains to be jumped.
The interpretation that incidental perception is a thought or inference rather than perception is proposed in various ways by, e.g., John I. Beare, Greek Theories of Elementary Cognition (Oxford: Clarendon Press, 1906), 286; Charles Kahn, “Sensation and Consciousness in Aristotle’s Psychology,” Archiv für Geschichte der Philosophie 48 (1966): 48; Irwing Block, “Aristotle and the Physical Object,” Philosophy and Phenomenological Research 21 (1960): 94; Andreas Graeser, “Aristotle’s Framework of Sensibilia,” in Aristotle on Mind and the Senses, ed. G. E. R. Lloyd and G. E. L. Owen (Cambridge: Cambridge University Press, 1978), 90; and more recently by Stephen Herzberg, Wahrnehmung und Wissen bei Aristoteles: Zur epistemologischen Funktion der Wahrnehmung (Berlin: De Gruyter, 2011), 137–55. I assume that it is also implied in the Greek commentary tradition even if it is not properly discussed there. See, e.g., Alexander of Aphrodisias, De anima, ed. I. Bruns (Berlin: Reimer, 1887), 41.8–10; id., Quaestiones et solutiones, ed. I. Bruns (Berlin: Reimer, 1892), 3.8, 93.23–94.9; Themistius, In libros Aristotelis De anima paraphrasis, ed. R. Heinze (Berlin: Reimer, 1899), 58.5–16; Simplicius(?), In libros Aristotelis De anima commentaria, ed. M. Hayduck (Berlin: Reimer, 1882), 127.25–128.10.
This is the interpretation given by, e.g., Thomas Aquinas, Sentencia libri De anima, ed. R.-A. Gauthier (Rome: Commissio Leonina / Paris: Vrin, 1984), 2.13, 120b161–122b222. My interpretation differs from his in that I do not assume that one can perceive an intelligible item incidentally only when the perceptual capacity is assisted by the intellectual capacity. This interpretation follows from the distinction that I make between two ways of understanding an incidental perception.
The reference of “which” (
De An. 2.6, 418a20–24. All translations from Aristotle’s texts are mine unless otherwise indicated.
Cashdollar, “Incidental Perception,” 158, opposes this interpretation by claiming that it makes incidental perception “trivial.” However, it is not at all trivial if we acknowledge that Aristotle introduces incidental perception to indicate the limits of the individual senses, as I shall suggest below.
Hence, I do not think that the reason why Aristotle introduces incidental sensible items and incidental perception is that he wishes to explain perception of particular substances. If this were his intention, the explanation he provided would remain inadequate because saying that we perceive the son of Diares incidentally by seeing the white thing with which he coincides does not really explain anything. All that we learn by this account is that we see the white thing. That is why those who favour the perceptual interpretation adduce further considerations to supplement Aristotle’s account. See, e.g., Modrak, who claims: “The sensory basis for the perception of an incidental object does not fully determine the content of the perception. The physical characteristics of the son of Diares suffice to bring about a perception (in a healthy percipient under normal conditions) of a white thing having a certain shape and magnitude. However, only a percipient acquainted with the son of Diares would be able to perceive this white shape as the son of Diares. Similarly, only a person who has experienced flavors can perceive a white thing as sweet through sight.” (Modrak, The Power of Perception, 69.)
In APr. 1.27, 43a26–27, Aristotle says that “[…] for example, Cleon and Callias, and the particular and sensible items” are not truly predicated of anything else. In APo. 1.22, 83a4–9, however, he qualifies this position by saying that “The white is wood” is either not a case of predication at all, or is predication in a qualified sense, namely incidentally. He clarifies the example in this way: “When I say that the white is wood, then I say that that which coincides with the white is wood, and not that the white is the underlying subject for the wood” (APo. 1.22, 83a4–7). That means that the logical subject of the incidental predication “The white is wood” is “the white thing,” i.e., “that which is white,” rather than “the white colour” understood as the underlying subject (
In APo. 1.31, 87b29–30, Aristotle makes this point by saying that “one necessarily perceives a this (
Metaph. 5.7, 1017a19–22.
For a similar use of the term “incidentally” (
Judging from the explanatory “For this reason” (
For a discussion on Aristotle’s view on concept acquisition, see Börje Bydén, “Aristotle’s Light Analogy in the Greek Tradition,” in Forms of Representation in the Aristotelian Tradition, Volume Three: Concept Formation, ed. C. Thomsen Thörnqvist and J. Toivanen (Leiden: Brill, 2022), 34–77.
APo. 2.19, 100a16–b1.
For a comparable interpretation without reference to incidental perception, see Bronstein, Knowledge and Learning, 245: “Perception is of the universal, not because the universal is perceivable, but because the universal is instantiated in particulars each of which is perceivable, and in virtue of this the universal is encoded in the representations we receive when we perceive particulars. Perceiving particulars is necessary but not sufficient for reaching universals. We need the perceptual faculty as a whole (including imagination and memory) and experience, which provide the basis for our advance.”
In Sophistici Elenchi 24, Aristotle notes that in general, “[…] it is not necessary that what is true of the accident is true of the thing” (Sophistici Elenchi, in Topica et Sophistici elenchi, ed. W. D. Ross (Oxford: Clarendon Press, 1958), 179a36–37). He gives an example about Coriscus and a man who is approaching (Sophistici Elenchi 24, 179b1–3). Suppose that Coriscus is the man who is approaching. Now it does not follow that if one knows Coriscus, one also knows the man who is approaching, i.e., that Coriscus is the man who is approaching. Even if Aristotle does not use the notion of incidental knowledge in this context (nor elsewhere), he would be consistent in suggesting that one knows the man who is approaching only incidentally, namely by knowing Coriscus.
Even if Graeser, “Aristotle’s Framework,” 90, eventually dismisses the perceptual interpretation, he calls Aristotle’s framework of perceptual language “an inverse ontology,” meaning that “genuine substances are treated as attributes and non-substances are treated as genuine subjects” (ibid., 74).
See also Joseph Owens, “Aristotle on Common Sensibles and Incidental Perception,” Phoenix 36 (1982): 227–28n17.
For instance, Hicks, Aristotle: De anima, 363, correctly notes that a
De An. 2.6, 418a13.
De An. 2.7, 418a29–31.
De An. 2.7, 419a1–6. For recent studies on the matter, see Katerina Ierodiakonou, “Aristotle and Alexander of Aphrodisias on Colour,” and Pavel Gregoric, “Aristotle’s Transparency: Comments on Ierodiakonou, ‘Aristotle and Alexander of Aphrodisias on Colour’,” in The Parva naturalia in Greek, Arabic and Latin Aristotelianism: Supplementing the Science of the Soul, ed. B. Bydén and F. Radovic (Cham: Springer, 2018), 77–90, and 91–98, respectively.
See also de An. 3.2, 425b18–19, where Aristotle says that sight sees “the colour or that which has colour.”
That is because colour necessarily occurs in some body; see Cat. 2, 1a27–28. I assume that the same applies to fire when the reference is to fiery substances.
De An. 2.5, 417b22.
For a recent formulation of this problem, see Anna Marmodoro, Aristotle on Perceiving Objects (New York: Oxford University Press, 2014), 86.
De An. 3.8, 431b22.
In the final section, I shall say something about how Aristotle would understand perception of facts such as “the moon is eclipsed.”
Aristotle, Physica, ed. W. D. Ross (Oxford: Clarendon Press, 1950), 224b19–20.
For this interpretation of the apprehension of the universal, see, e.g., Bronstein, Knowledge and Learning, 246.
By suggesting that universals are proper intelligible items for the intellectual capacity, I am not suggesting that apart from universals, there are no other per se intelligible items. Aristotle assumes, for instance, that the intellect (noûs) can apprehend individuals such as Cleon when it unites Cleon and being white in asserting that Cleon is white (de An. 3.6, 430b5–6). Aristotle makes no suggestion that this thought is incidental. For discussion, see Mika Perälä, “Aristotle on Singular Thought,” Journal of the History of Philosophy 53:3 (2015): 364–65.
Aristotle notes, though, that the soul, or rather we by the soul, never think without having a phantasm (de An. 3.7, 431a16–17). However, he does not introduce that requirement as a requirement for incidental thought. Furthermore, to be plausible, the requirement should cover cases in which we think of a universal such as the chiliagon for which we do not have a corresponding phantasm. In such a case, Aristotle could suggest that the thought of chiliagon is based upon our having a relevant sort of phantasm, say a phantasm of a triangle. That would be a reasonable suggestion to make because we might form the concept of chiliagon by adding sides to the concept of triangle for which we have a phantasm. That does not imply that to think of chiliagon, we ought to simultaneously entertain the phantasm of a triangle.
In this respect, I agree with most scholars. However, my arguments can be seen as a criticism of Cashdollar, who claims that Aristotle “recognizes the incidental sensible itself as the source of a certain kind of psychic movement which is distinct from but of a class with the other two distinctive movements which result from special and common perception (428b18 ff.)” (Cashdollar, “Incidental Perception,” 159).
For a helpful recent study on incidental causation in Aristotle, see Tyler Huismann, “Aristotle on Accidental Causation,” Journal of the American Philosophical Association 2:4 (2016): 561–75. My interpretation of the passages in question is compatible with his.
Ph. 2.3, 195a32–35.
Ph. 2.5, 197a14.
Ph. 2.5, 196b33–36.
Ph. 2.5, 197a17–21.
Ph. 2.5, 197a15.
Ph. 2.5, 197a19.
Metaph. 6.2, 1026b6–10.
Thus, I oppose the interpretation given by, for instance, Modrak, The Power of Perception, 69, and Marmodoro, Aristotle on Perceiving Objects, 166. See also Ross, De anima, 34, who suggests that a sense can perceive another sense’s proper object incidentally if the two senses have perceived the same object at the same time earlier.
For a full discussion of the interpretation that perceptual discrimination, according to Aristotle, is not an incidental perception, see Mika Perälä, “Aristotle on Perceptual Discrimination,” Phronesis 63:3 (2018): 257–92. For the De sensu et sensibilibus, I am referring to Aristotle, Parva Naturalia, ed. W. D. Ross (Oxford: Clarendon Press, 1955).
De An. 3.1, 425a30–b2, trans. C. Shields, in Aristotle: De anima (Oxford: Clarendon Press, 2016), modified.
For the common sense, see, e.g., Johansen, The Powers, 184–85; for the perceptual capacity, see, e.g., Gregoric, Aristotle on the Common Sense, 199–201. In the present context, the difference between the two is merely terminological.
Johansen, The Powers, 182.
Johansen, The Powers, 182–83, emphasis his.
Johansen, The Powers, 184.
Johansen, The Powers, 184.
For a discussion on Aristotle’s De sensu 7 and especially its medieval reception, see Toivanen’s contribution (chapter five) below.
For recent interpretations, see, e.g., Polansky and Fritz, “Accidental Perception,” 126; Marmodoro, Aristotle on Perceiving Objects, 165–67; Gregoric, Aristotle on the Common Sense, 199–201.
Metaph. 13.10, 1087a19–20.
Some late Platonist commentators of Aristotle assumed that the human capacity for perception is penetrated by reason. See, e.g., Simplicius(?), In de An. 187.29–36.
De memoria et reminiscentia, in Aristotle, Parva naturalia, ed. W. D. Ross (Oxford: Clarendon Press, 1955), 1, 450a24–25; 2, 451a28–29.
Sens. 5, 443b20–26.
De An. 2.6, 418a14–16, trans. Shields.
De An. 3.1, 425b3–4, trans. Shields.
De An. 3.3, 428b18–25, trans. Shields, modified.
In fact, Aristotle says that it is the principle or origin (i.e., the primary perceptual capacity that lies in the heart) that affirms the report from each sense; see De insomniis 3, 461a30–b7 (the text is included in Ross’s edition of the Parva naturalia). However, there is no functional difference between the principle and the sense when only one sense is involved in perceiving the proper sensible items. Furthermore, Aristotle’s understanding of “more authoritative” (
For instance, Cashdollar, “Incidental Perception,” 161–67, confounds the predicative and the associative by referring to what he understands as spontaneous association of perceived items as “perceptive predication.” For a perceptual interpretation that is not based on predicative analysis, see, e.g., Gregoric, Aristotle on the Common Sense, 199–201.
Aristotle has two terms for belief: hypólēpsis and dóxa. By hypólēpsis, he refers to any cognitive state that involves taking something to be the case or taking something to be true or false, including knowledge (epistḗmē), practical reason (phrónēsis), and dóxa (de An. 3.3, 427b24–26). By dóxa, he refers to the kind of cognitive state that does not satisfy the requirements of knowledge, for example universality (see, e.g., APo. 1.33, 88b30–89a3). They include perceptual beliefs. For the present purposes, it is not necessary to decide whether according to Aristotle all cases of taking something to be the case or something to be true or false require predicating something of something; alternatively, there might be cases in which one grasps the truth by simply apprehending a unified item such as an essence. For a recent discussion of belief in Aristotle (and Plato), see Jessica Moss and Whitney Schwab, “The Birth of Belief,” Journal of the History of Philosophy 57:1 (2019): 1–32.
For instance, Johansen, The Powers, 183–84, argues: “Clearly the reason why one can see the white as the son of Diares is that one has at some point learnt who he is; typically by somebody telling you. That somebody is the son of Diares is clearly then something we learn. What persuades Aristotle nonetheless to view such accidental perceptibles as objects of perception is that they are, as we might say, represented in perception. On this reasoning, it is one question whether a certain content has its origin in memory or intellect or some other cognitive activity, it is another question whether the content is perceived or not. We perceive that this white thing is the son of Diares within one perceptual act, not as an act of seeing plus an act of intellection.” (Johansen’s emphasis.) Johansen follows here Cashdollar, “Incidental Perception,” 168. For the same argument, see, e.g., Gregoric and Grgic, “Aristotle’s Notion of Experience,” 11–12.
See also Herzberg, Wahrnehmung und Wissen, 171–72.
By contrast, some interpreters take this example to show that we would not incidentally perceive the yellow thing as being bitter, or being bile for that matter, unless we previously had perceptions of the yellow combined with perceptions of the bitter. That is why we tend to associate the yellow with the bitter (even when there is nothing bitter present to taste), and are prone to incidentally perceive the yellow thing as being bitter, or bile. See, e.g., Ross, De anima, 34; Cashdollar, “Incidental Perception,” 168; Modrak, The Power of Perception, 69; Gregoric, Aristotle on the Common Sense, 200.
Insomn. 1, 458b31–33.
Here, again, I am following Herzberg who suggests: “In allgemeiner Form wird solch ein sinnlicher Irrtum in De Insomniis beschrieben: ‘Denn das falsche Sehen und das falsche Hören geschehen erst dann, wenn man etwas Wirkliches (
Insomn. 1, 458b28.
Insomn. 1, 459a6–8.
De An. 3.6, 430b5.
De An. 3.6, 430b5–6.
De An. 3.7, 431b4, trans. Shields, modified.
De An. 3.7, 431b5–6, trans. Shields.
De An. 2.10, 422b7–10.
Aristotle, Ethica Nicomachea, ed. I. Bywater (Oxford: Clarendon Press, 1894), 6.8, 1142a27–30.
See, e.g., Cashdollar, “Incidental Perception,” 172.
APo. 1.31, 88a1.