It has long been observed that the key texts in which Aristotle explains how we reach knowledge of first principles can also be understood as accounts of concept development.Footnote 1 So Posterior Analytics (An. post.) 2.19 and Metaphysics (Metaph.) A.1 lay out a progression of cognitive states leading to a highest state of knowledge: first perception, then memory, then experience, then technē or epistēmē or nous. Each stage of development marks a step towards grasping the universal principles of reality. If we take a concept to be a mental content that figures in the representation of the world in some way or other,Footnote 2 we might say that as our cognition advances our concepts become representative of increasingly universal and fundamental features of reality.
What primarily concerns me in this paper is the manner of the progression, that is how one cognitive state of the soul follows from another, or put differently, how one stage of conceptual competence leads to the next. There is a philosophical issue here which Aristotle is sensitive to, of how the mind manages to extract the appropriate information from perception, if it doesn’t already have access to the relevant concepts. In the modern literature on concept formation, this is often referred to as the circularity problem. Innatism is one response to this problem. However, Aristotle’s position is, at least in comparison to Plato’s theory of recollection, empiricist in outlook. It is clear that he takes the concepts with which we grasp reality to have their origin in perception. It is more difficult to say, as we shall see, how the soul is able to organise and classify perceptual information to develop the concepts grasped in knowledge.
I shall begin and end with An. post. 2.19. Aristotle has asked where our knowledge of the first principles of demonstrative knowledge comes from. Rehearsing Meno’s paradox he excludes the two options that we are born with this knowledge and that we develop it not having some prior knowledge. His answer instead is that, in perception, we have a kind of prior knowledge (gnōsis) from which knowledge of principles develops.Footnote 3 As he explains:
T1 Necessarily, therefore, we have some capacity (dunamis), but do not have one of a type which will be more valuable than these [sc. principles] in respect of precision. And this evidently belongs to all animals; for they have a connate discriminatory capacity, which is called perception. And when perception is present in them, in some animals retention of the percept comes about, but in others it does not come about. Now for those in which it does not come about, there is no knowledge outside perceiving (either none at all, or none with regard to that of which there is no retention); but for some perceivers,Footnote 4 it is possible further to hold (echein) the percept in their souls. And when many such things come about, a difference comes about, so that some come to have an account from the retention of such things, and others do not. So from perception there comes memory, as we call it, and from memory (when it occurs often in connection with the same thing), experience; for memories that are many in number form a single experience. And from experience, or from the whole universal that has come to rest in the soul (the one apart from the many, whatever is one and the same in all those things), there comes a starting point (archē) of craft and of understanding – if it deals with coming into being, craft, if it deals with what is, understanding (epistēmē). Thus the states neither belong in us in a determinate form, nor come about from other states that are more cognitive; but they come about from perception – as in a battle when a rout occurs, if one man makes a stand another does and then another, until a starting point is reached. And the soul is such as to be capable of undergoing this. What we have just said but not said clearly, let us say again: when one of the undifferentiated things makes a stand, there is a primitive universal in the mind (for though one perceives the particular, perception is of the universal – e.g. of man but not of Callias the man); again a stand is made in these, until what has no parts and is universal stands – e.g. such and such an animal stands, until animal does, and in this a stand is made in the same way. Thus it is clear that it is necessary for us to become familiar with the primitives by induction; for perception too instils the universal in this way.’
It is important here that perception be a natural capacity and not also an acquired state. If perception itself was acquired, one could ask how, and then one would have to point to some other prior knowledge, given the principle that learning proceeds from previously existing knowledge. To say that perception is a natural capacity is to avoid this regress. The other states that develop after perception, meanwhile, are referred to as hexeis, acquired states. This includes, first of all, memory, which is a retention (monē) of a percept (aisthēma), and then experience (empeiria) which arises from many memories of the same thing. Experience is here associated with the emergence of reason, logos, particularly it seems to be grasping a difference (diaphora) that characterises the various things remembered.
Aristotle describes the development of these states (perception, memory, experience, knowledge) also as a development of their conceptual contents. The contents develop along two axes, one of greater unity, another of greater universality. So a plurality of memories lead to experience, a plurality of experiences to knowledge, but this is also a process by which the whole universal is represented in the soul. The expression ‘the one over the many’ seems to offer a unifying description: as cognition progresses the universal concept unifies the many things represented by the cognition at the previous level. The comparison of the soldiers in a rout might also be taken to indicate both unification and universalization, all the soldiers stop and assume a common stance, arranged in a singular formation.
Cognitive development alongside concept acquisition: which explains which? It would be in the spirit of Aristotle’s general preference for prioritizing the object of a cognitive state to say that we have a series of cognitive states that are characterised by their distinct contents.Footnote 5 However, the question here is not simply what each of the cognitive states is, which would be determined by their contents, but how they develop from each other. The venue of the soldiers stopping is our soul and we need to explain how the battle proceeds on this ground. How do we advance then from the content of a lower cognitive state to that of a higher? This is where the circularity problem emerges.
It may seem that Aristotle gives this question short shrift: ‘And the soul is such as to be capable of undergoing this.’ (100a13–14) But it is not clear whether this claim is meant to forestall any further questions about how the soul is able to universalise and unify experiences under higher concepts: the explanatory buck stops here, one might say, or whether he is rather identifying the subject of the changes as the soul. If the latter, Aristotle may intend to attribute to the soul a process that he thinks can and should be explained in its own right, the process which the image of the soldiers is supposed to illustrate. In that case, we are supposed by a proper interpretation of the image to be able to understand the process of concept formation. When Aristotle concludes that it is the soul that is able to do this, he is then not forestalling an explanation of the process but rather attributing the process, as it is explained, to the soul. I shall explore the second option as the more helpful.
Aristotle’s answer in outline is the following: at each level of cognition the soul discerns what he calls a ‘difference’ amongst its contents. When we perceive, the soul grasps differences among colored objects, pale ones, for example; when we experience we gather memories of various pale objects into one, for example white people who were ill; when we have a craft (technē) we select from these various cases of pallor those that were cases of, say, food-poisoning. All our cognitive capacities are discerning (kritika) in the sense that they are able to pick out different kinds of difference. Each step after perception identifies a difference within the contents of the previous level. The higher-level capacity only becomes able to operate given the activity of the lower-level capacity, beginning with perception. One might think of the process as a kind of bootstrapping:Footnote 6 given the concepts provided by a lower-level form of cognition an ability to recognise new, higher-level concepts emerges. So, for example, the capacity of experience develops out of memories as the ability to discern similarity relations between objects remembered. While one might say that we have already been endowed with a general ability to discern, or a joint perceptual-intellectual ability, this develops from perception to specific competencies, memory, experience, craft or science, only through being provided with the contents from lower-level cognition. The specific higher abilities unfold with the provision of perceptual contents from ‘the bottom-up’.Footnote 7 At least this is the picture for which I shall argue.Footnote 8
To set out this picture in more detail I turn next to our other main text Metaph. A.1. There are three cognitive steps to consider: from perception to memory, from memory to experience, and from experience to craft or knowledge. I shall argue that the development of knowledge, given a perceptual base, can be taken as the development of the rational capacity through a series of acquired states each of which provides us with a distinctive capacity which in turn allows us to develop the following state.
1 Cognitive Development Step 1. From Perception to Memory
T2 Animals come to be possessing perception by nature. In some of them memory does not come about from perception, while for others it does. And that is why they are wiser or better at learning than the ones that are not capable of remembering. Wise without learning are those that are not able to hear sounds (for example the bee and if there is any other such kind of animal), while those learn which in addition to memory also have this sense. The other animals live by means of imagining and remembering things, and have little by way of experience. Human kind, however, live by craft and reasonings. Human beings acquire experience from memory, for many memories of the same thing constitute the capacity (dunamis) of a single experience.
The passage sets out a hierarchy of cognitive states. In each case the lower cognition is a precondition for the higher. You only have memory if you have perception, you only have experience if you have memory, you only have craft (technē) or science (epistēmē) if you have experience.
Before trying to explain the cognitive progression plotted in T2, it is important to make some observations about the terms Aristotle uses here and in T1 to describe what sorts of thing are ordered in the progression. T2 talks about capacities (dunameis) such as the senses and the power of experience. T1 (100a9–10), meanwhile, refers to hexeis, ‘states’, although perception is also referred to as a dunamis, while Aristotle’s general claim ‘that the soul is capable of being so affected’ may suggest that the different kinds of cognition manifest the soul’s capacity or capacities. Tradition has also taken nous to be the capacity whereby we are able to grasp first principles.Footnote 10
We might initially think that states and capacities are mutually exclusive. This impression could arise from several texts. So Metaphysics Δ.15 says that capacities are relatives and one may infer that because relatives are a distinct category from states, states being a kind of qualification, nothing can be both a capacity and a state. This, however, is to misunderstand Aristotle’s categories: it is quite common for one item to fall under several categories. Knowledge, for example, occurs both as a relative in Categories (Cat.) 7 and as disposition in Cat. 8.Footnote 11 Another passage, Nicomachean Ethics (Eth. Nic.) 2.5 1105b20, divides up the things that arise in the soul as either capacities, affections or states, and argues that virtue is a state and not one of the other two. One might wrongly infer from this that nothing can be both a capacity and a state. One distinguishing mark that Aristotle uses here is that states are acquired, and this fits the fact that we become good through habituation. Capacities, in contrast, can be said to be natural and unalterable, like a stone’s capacity to fall down. However, it is clear that this distinction does not exclude that some things can be both states and capacities. So health is both a standard example of a state (see Metaph. Δ.20), but it can also occur as a natural capacity. Some people are born healthy (Cat. 8, 9a15–21), that is having the capacity to resist corrupting influences, others acquire health through healthy activities. When we talk of health as a state, we can also take it be the efficient cause of health (Physics 2.3, cf. Eudemian Ethics 1.8 1218b20–22), so being healthy in this way is also the capacity to participate in healthy activities like walking.
Consider also craft (technē), one of the terms in the cognitive progression of Metaph. Α.1 and An. post. 2.19. Crafts are states. This is true generally of them as a form of knowledge, for Cat. 8 says that ‘Comprised among what we call “states” (hexeis) are virtues and all kinds of knowledge’ (8b29–30). Metaph. Θ.2 is quite clear about technai being also capacities (1046b1–4). However, it is not hard to combine the two descriptions of technē. Eth. Nic. 6.4 (1140a10) after all defines the technē as poiētikē hexis, that is a state which is capable of producing. If you ask what technē is a hexis of, the answer, again following from Eth. Nic. 6, is that it is a hexis of the rational part, to logistikon or to logon echon. Technē, one recalls, is one of the virtues of the intellect, alongside the other kinds of knowledge. Describing technē as a virtue, again, goes well with thinking of it as a hexis, since a hexis in one of the three senses listed in Metaphysics Δ.20 is that by which you are well or ill-disposed either in oneself or in relation to something. Technē would indeed be that by which you are well disposed in relation to something, namely the characteristic product of the technē.
As Physics 7.3 (246a10–17) shows, the notion of ‘hexis’ is particularly suited to expressing the perfection (or imperfection) of a natural disposition.Footnote 12 We may reasonably think that the person who acquires health through healthy activities is also perfecting an underlying capacity to be healthy, perhaps the same as the one who is born healthy maintains effortlessly. If there was no underlying capacity the healthy person could perfect through his exercises, then no amount of running around would make him healthy. In the same way, it seems right to take a cognitive state, such as technē, to be the perfection of our underlying rational ability. Technē is after all an intellectual virtue. ‘State’ and ‘capacity’ are then not always mutually exclusive, but may in some cases, like health or technē, be descriptions of the same entity.
If this view of technē is correct, can we see the other states mentioned in T1 and T2 as developed capacities also? If so, how do the various states-cum-capacities relate to each other in our cognitive development? Let us begin with memory. Aristotle expresses the development of memory in terms of memory arising in some animals, not in others from perception. Does perception here mean the act of perceiving or the capacity to perceive? In favour of the first, it seems that an individual instance of memory arises from an individual act of perception. Aristotle’s theory of memory in De Memoria (Mem.), as we shall see, takes a memory to be a stored perceptual image, so this seems fine. However, reading memory as an instance of perception, even if taken as any instance, is at odds with the general scope of the claim that memory, singular, arises in certain animals and not in others. If one takes memory instead to refer to the capacity to remember, one might think of the capacity as arising, egginetai, from the possession of perception. This can in turn be understood in either of two ways. Either i) the capacity is taken to arise after (ek) perception, or ii) the capacity develops out of perception. This latter option is more in line, as we shall see, with the account of memory in Mem.
To recapitulate: T2 allows us (though hardly requires us) to read the claim that memory arises out of perception as a claim about the capacity of memory and as a claim about the functional dependence of this capacity on that of perception. However, there are now two ways to see this functional dependence. One is to say that you already have the capacity to perceive and you have the capacity to remember, you just have to activate the one before you can activate the other. Another way is to say that you don’t yet have the capacity to remember before you have activated the capacity to perceive. In this case the capacity to remember would develop out of the activity of perception.
To decide the matter, we need more information about memory. Luckily Mem. is on hand. Chapter 1 of Mem. starts by characterizing memory in terms of its proper object, what can be remembered. Aristotle argues that what one remembers belongs to the past. He thinks that remembering involves a sensory image, a phantasma, and that one does not remember objects of thought as such (450a24–5). Since memory is also of the past, it belongs to the same capacity as the one that perceives time, that is, the common sense. In remembering, the perceptual image is used as an image by which one perceives the past experience rather than as an object in its own right. Aristotle sums up his position as follows:
T3 It has been said what memory and remembering are, namely that it is a state (hexis) of an image (phantasma), taken as a likeness (eikōn) of that which the image is; and [it has been said] to which part within us it belongs, that it belongs to the primary faculty of sense-perception, and that by which we perceive time.
What does it mean to say that memory is ‘a state (hexis) of an image (phantasma)’? Earlier (450a13) Aristotle referred to the presentation as an affection (pathos) of the sensory capacity. Memory might be said to be the state of the sensory capacity when it holds (echein)Footnote 13 or retains this affection,Footnote 14 hence a hexis; remembering is the using of the image as a likeness of the past.
However, remembering, in the sense of having committed something to one’s memory, is also presented as a capacity (dunamis), the capacity to call forth the phantasma in this way:
T4 Remembering is the presence within of the power (dunamis) which excites the changes and this in such a way that the man moves of himself and because of changes that he possesses, as has been said
The state of having committed something to memory (memnēsthai) involves then the capacity to move oneself to recall that thing. Or as Richard Sorabji puts it, ‘the state is the ability to stir up this image’.Footnote 15 Having remembered is not as such the ability to recollect (anamimneskesthai) since recollecting is a particular way of bringing about remembering, by a rationally controlled method such as a mnemonic. Yet we can say that recollection presupposes the capacity to recall.Footnote 16 The upshot is that you can acquire a capacity by developing a state. By committing something to memory you also develop the capacity to recall it. This is in itself a fairly obvious thought: when you acquire a memory of something you also acquire the ability to remember that thing.Footnote 17 The state, the having of the memory, may in that sense be said to constitute the capacity.Footnote 18
Now, the account of memory suggests a certain pattern of acquiring a capacity: affections of one capacity bring about a state which in turn constitutes a further capacity. This pattern I take to be iterable in a way that explains, given a certain underlying capacity, how other derived capacities can develop. I want to suggest that we can see this pattern not just in the development of memory from perception, but also in that of experience from memory and craft from experience:
Underlying capacity1: affection1 (pathos)Footnote 19: state1 (hexis) : capacity2 (dunamis): affection2 :state2: capacity3….
One question that this pattern immediately raises is what the underlying capacity is that Aristotle could have in mind in T1 and T2. Clearly perception is a starting point and we have seen that memory is a state of the perceptual capacity. In An. post. 2.19, T1, Aristotle is explicit that the capacity of perception is the starting point for cognitive development. And there is an important reason for this. Perception, as De an. 2.5 tells us, is ready for activity by the time we are born. Perception is our starting point because no prior act of learning is needed for us to be able to perceive. No other cognitive mechanism is required to bring us up to the point where we can start perceiving. We are born in a state of second-level potentiality with respect to perceiving. All we need to fully actualise perceiving is exposure to something perceptible.Footnote 20 Perception contrasts with intellectual thinking where our state at birth is one of mere first-level potentiality. We must learn before we can exercise knowledge, that is, before we are in a state of second-level potentiality with regard to knowledge corresponding to our perceptual condition at birth. This is also why our perceptual capacity is in a better position to initiate cognitive progress than our intellectual capacity. Perceptual information can help trigger learning, that is, the development of the intellectual capacity to the level of second potentiality, exactly because the perceptual capacity is ‘born ready’.
Now it is clear, I think, that Aristotle’s concern in Metaph. Α.1, T2, is not with perception or memory as states of the perceptual capacity as such, but rather in the way perception contributes to our intellectual development. That is, Aristotle’s perspective on perception and memory is how they affect our intellectual capacity. So consider how Aristotle distinguishes in T2 between those animals that might be wise without learning and those who become so through learning. For the latter, hearing is required. Since being wiser as a result of having memory clearly involves being wise as a result of gathering perceptual information, we have to allow for these animals in some sense to have learnt. But it would not be learning in the sense associated with receiving instructions from others. This raises the question of which animals Aristotle can be thinking of as ‘better at learning’ if not simply humans. For why privilege learning by hearing if learning can mean more broadly acquiring and retaining perceptual information? Sight, as Aristotle just told us, does the best job here! The answer to this question is given in De Sensu 1 (437a3–17). Sight as such does indeed give us the most information about differences, but if we allow for accidental perception, hearing wins. The explanation is that speech by being heard is the cause of learning. The differences of sound that hearing as such informs us of are not instructive, but only when taken as symbols, and that’s the job of the intellect. Here then it is accidental perception that teaches us. However, this point is limited to beings in possession of phronēsis.Footnote 21
The key point for our purposes is that already in the move from perception to memory, Aristotle has, by concentrating on learning through hearing, focused on the impact that perception has on creatures with some sort of phronēsis or intellect. And this shows that Aristotle throughout is concerned with the growth of our underlying intellectual capacity. While memory itself is a perceptually based capacity, the focus in Metaph. Α.1, is on the retention of accidental perceptibles which form the basis of learning.
2 Step 2. From Memory to Experience
Aristotle thinks that memory differs from experience.Footnote 22 The claim that other, I take it ‘non-human’, animals live only a little (mikron) by experience might be taken as a meiosis: animals do not live by experience at all or only to a negligible extent.Footnote 23 In any case, Aristotle ignores non-human animals in the rest of his account of experience here.
Aristotle presents the steps from memory to experience and from experience to craft as remarkably similar. As we continue reading Metaph. Α.1, this becomes clear from the two sentences in bold:
T5 Human beings acquire experience from memory, for many memories of the same thing produce the power (dunamis) of a single experience. Indeed, experience seems to be similar almost to knowledge and craft (technē), and knowledge and craft come about through experience in humans. For experience produces craft, as Polus rightly says, and inexperience chance. Craft comes about whenever from many thoughts belonging to experience arises a single universal judgement about similar things. For having the judgement that this thing here helped Callias when he was suffering from this here disease and Socrates too and that it in this way in each case helped many is a matter of experience. But it is proper to craft to judge that this helped all such people differentiated as a single kind suffering from this disease here, for example to the phlegmatic or bilious when they are burning with fever.
There are clearly several ways one could understand the unification of memories in experience. Unifying in itself does not imply greater generality: one might imagine the same perception repeated, say of the postman arriving in the morning at nine. In that case one might remember individual instances of this event without generalising that the postman always or generally arrives at nine in the morning.
One way to think about this scenario is in terms of the wax tablet model of memory that Aristotle inherits from the Theaetetus: the same perception is fitted into a past memory trace, coinciding in a manner where you may no longer be sure which day you saw the postman arrive at nine. In that case there would be a sort of unification between the several memories. Perhaps the repetition creates a memory track, each perception like a footstep imprinting this path on the mind; indeed, this may not be too far from how repetition actually reinforces memory. There is clearly a certain shape that gets imprinted here, which deselects individual variation. Perceiving the postman on many individual occasions will tend to leave an imprint of him which does not reflect individual features of particular mornings – whether he was smiling one morning, pale on another occasion. Or to return to the footprint analogy, the repeated imprint of the same foot will after a while blur the differences between individual imprints but there will still only be one particular imprint in the soil. This reinforcement of repeated information is not, then, a process of universalising, but rather one of homogenising. Universalising is extracting a general characteristic shared by all such experiences, while homogenising would be erasing the differences between individual experiences to create a sort of uniform individual.
Homogenised imprints could form a sufficient basis for projection. One could imagine a future scenario sufficiently similar to a past scenario simply on the model of past perception. It was no doubt in some such way that the memorists thought agency could be explained without recourse to universal concepts.Footnote 24 And it may be how Aristotle thinks animals are able to live by imagination and memory without experience. There is a sufficient fit or similarity between the dog’s impression of the postman today and its memories for it to recognise the man, even if some details of today’s impressions differ from previous ones. There is no need for the dog to see this man as an instance of some general property also instantiated by some previous perception.
Humans, in contrast, live by experience, which arises from memory. How so? Aristotle, as we have just seen, says that ‘Many memories of the same thing produce (apoteloun) the capacity (dunamis) of a single experience’. The claim calls for elucidation in several ways. First, what does ‘produce the dunamis of’ mean? One could conceivably take dunamis here as ‘padding’: in that case ‘the capacity of experience’ would just mean ‘experience’. So in An. post. 2.19 (100a4–5) he says simply that ‘many memories are one experience in number’. Yet dunamis seems appropriate in the context of explaining experience in terms of what it can do, namely collecting similar observations under one judgment (hupolēpsis). So in T5 it is a function of experience to suppose that the medicine helped the various individuals in the various cases. The judgment (hupolēpsis) referred to here in the singular seems to exemplify the exercise of the dunamis referred to. The difference between the flat statement of the Posterior Analytics, T1, ‘many memories are one experience in number’ and T5’s statement is that T5 is referring to the unifying of memories as the particular function of an experience.
Notice, next, how Aristotle’s language ascribes agency to the memories: they produce (apoteloun) the capacity of a single experience. This need not mean that the memories have sole causal responsibility for the capacity of experience, but it is implied that they contribute to the completion of the capacity.Footnote 25 Presenting the memories as active in the creation of the capacity of experience is what we would expect if the pattern we found in the relationship between perception and memory generalises: it is the activity of the lower-level cognition that helps generate the higher-level cognitive capacity. This certainly is the most straightforward, literal reading of ‘the memories produce the capacity of experience.’
Of course, producing the capacity of experience does not happen without an underlying potentiality in the human soul, and this may be a reason for not ascribing sole causal responsibility to the memories. An underlying potentiality is assumed, on the current reading the first-level potentiality of the intellect.
Secondly, the experience seems to be singular because the memories are ‘of the same thing’ (tou autou pragmatos). For this to be true, for the sameness in the object of the memories to be productive of the capacity of the single experience, it seems that this sameness must obtain prior to the unity of the experience. So we would not want to say that experience brings a singular concept to the various memories and that thereby they achieve a sameness. For then the causation would go in the wrong direction. However, just as we could say that various similar perceptions reinforce a single memory, as for example in the postman case, we can also say that various memories of the same thing by their similarity coincide, and it is this coincidence that provokes the mind to think of them as one. That is what the rational soul is like: it responds to causally similar inputs by distinguishing them as one kind of thing. While the soul’s underlying rational capacity to respond appropriately is presupposed, it is because of the ordering of the input, because of the memories’ objective similarity, that the specific capacity of a single experience arises in the soul.
Experience seems then to confirm our pattern:
Underlying capacity1: affection1 (pathos)Footnote 26: state1 (hexis) : capacity2 (dunamis): affection2 :state2: capacity3….
Recall the case of memory. Memory, we saw, is the state of having retained perceptual imprints. This state of the perceptual faculty could also be seen as a dunamis to recall. Similarly, we can say now that many memories affect the rational soul to produce a state of experience, where this state is referred to as a dunamis, specifically the capacity to recognise and gather together in one judgment observations of similar phenomena.
3 Step 3. From Experience to technē
‘Craft comes about whenever from many thoughts belonging to experience arises a single universal judgement about similar things.’ I noted earlier the similarity between this and the claim that ‘many memories of the same thing produce the capacity of a single experience’. One might seek to assimilate the two further by identifying the many thoughts of experience with the many memories that produce the capacity of one experience. The theory would then be that the single judgment of experience unifies the many memories. But this cannot be what Aristotle intends here. For the contents that belong to experience are not memories, but many thoughts (ennoēmata). Ennoēma refers literally to what one has in one’s nous or noēsis, that is, the object of one’s thought. The term deserves attention: this is its first attested use and Metaph. 981a6 is its only occurrence in Aristotle.Footnote 27 In Plato we have the cognate ennoēsis, ‘reflecting’, which would be the process of entertaining an ennoēma. Later the Stoics, stressing the relation of ennoēma to nous, use the term to mean a rational appearance.Footnote 28 Given that memories belong to the perceptual soul for Aristotle, and not, as we saw, the intellect, it is unlikely that Aristotle would use this term to refer to the memories. The term rather suggests some rational presentation of what we remember. This impression is confirmed by Aristotle’s example: ‘For having the judgement that this thing here helped Callias when he was suffering from this disease here and Socrates too and that it in this way in each case helped many is a matter of experience.’ (981a8–9) The content is propositional, as appropriate to rational judgements (hupolēpseis), and the propositions refer to ‘this disease here’, that is, they identify a general concept across the various memories. We could in contrast imagine somebody who remembered Socrates taking an aspirin and his getting better but not reflecting that the aspirin helped him, nor again having thought of the case of Callias’ being similarly helped. The phrasing of these contents, then, reflects some rational processing as befitting an ennoēma, and not a mere memory. So in bringing together the memories, experience has also conceptualised their similarity in certain terms such as this disease, this medicine, and helped in this way, etc.
In An. post. 2.19, T1, Aristotle talks of animals in which ‘an account’ (logos) came about from the retention of perceptual impressions. And he says that in experience ‘the whole universal has come to rest in the soul’. This makes it explicit that experience in contrast to mere memory marks a stage of rational development where the perceptual impressions have been gathered under a universal.Footnote 29 We can see how the medical example fits this description: the many memories of various sick people have been collected in one judgment. This would be a kind of account (logos) based on their similarity with respect to a common concept such as ‘this disease here’. Not an account in the sense of a causal account or a definition of the disease, that would be the remit of craft or science, but an account staking out the universal phenomenon nonetheless.
One might object that this account builds so much rationality into experience that it challenges the role of craft. It is after all craft that gives us a single hupolēpsis about the universal, ‘differentiated as a single kind’. In the medical example, experience is described as referring to individuals on particular occasions, but there is also a reference to certain constants: this disease here and this remedy here. That is to say, both disease and remedy are identified by the judgment of experience as the same across the different cases.
However, there is clearly a difference between the way the experienced person picks out a disease or a remedy from the way the doctor does so. The doctor understands the patients in terms of their proper disease and understands the remedy in terms of its causal properties to cure this disease. Only the doctor can account for the cause, the why and not just the fact, the that. But the experienced person has a sufficiently developed concept to identify the disease and the remedy.Footnote 30 The point seems clear from the passage: it is only the craftsman who grasps the proper cause. The terms ‘phlegmatic’, etc. are significantly chosen: against a Hippocratic background of humour medicine they are the causes of various diseases. Picking out somebody as phlegmatic is to indicate the imbalance in humours that causes the disease, and to indicate what kind of cure is appropriate.Footnote 31 The experienced person has an account that mentions a cause, for instance that this medicine helped on this occasion, but does not show it as the cause.Footnote 32 It is for this reason that experience represents a lower state of cognition than craft. Aristotle takes having a logos of the cause as the benchmark of knowledge as opposed to lower states of intellect, belief, judgment, experience.
Regarding technē as an emergent capacity we need to say relatively little given the earlier remarks. It is quite clear that technē is a hexis and a dunamis. What is worth underlining is how the state may also be seen arising from experiences. Aristotle wrote: ‘Craft comes about whenever from many thoughts belonging to experience arises a single universal judgement about similar things.’ We noted the parallel with experience arising from memories. The formulation is consistent with seeing technē as the intellect’s response to lower-level cognitive affections. In brief, technē is the second-level potentiality of the rational soul that develops in its encounter with the many thoughts of experience, when it grasps their cause.
4 The Underlying Capacities
In my interpretation, there are capacities that develop as states in response to the affections of lower-level capacities, but there are also underlying capacities that explain how the soul responds to input by developing these capacities. The underlying capacities are just two in number: perception and intellect. The perceptual capacity is particularly involved in the development of memory, the intellectual in experience, craft and science. While these are two basically different capacities, as we see in their distribution amongst humans and animals,Footnote 33 Aristotle does offer a functional description that holds of both: each is able to discriminate (kritikē), that is able to discern differences (diaphorai) within a certain domain. From a rather abstract level, one could see then the entire cognitive development as the perfection of a single discriminatory ability. and Aristotle does on occasion talk of perception and intellect as if they were one such ability,Footnote 34 though this again would be to ignore the fundamental differences that obtain between the objects of perception and rational thought.
Another way of presenting our cognitive development is as a development of states of the underlying capacity. Let’s look particularly at the intellect, that in virtue of which we learn. Hexeis come in degrees according to which they perfect the underlying capacity: you can be more or less skilled at your craft, more or less knowledgeable of grammar, just as you can be more or less healthy, says Aristotle in Cat. 8. This point is relevant to the analysis of degrees of potentiality and actuality in De an. 2.5. Knowledge, we are told, allows for three levels of development, first and second potentiality, and fulfilment (entelecheia), corresponding to a boy who has not yet learnt mathematics, a man who has learnt it but isn’t using his knowledge, and one who has learnt and is currently doing it. All three characters are knowers, the first two in potentiality or capacity (kata dunamin) the last in fulfilment, The transitions between the first and the second potentiality is of a different kind from that between the second potentiality and the fulfilment. The latter is a manifestation of what one already is; the builder building only displays the attributes he already possesses: the mathematician when thinking mathematically shows himself as the mathematician he already is. The first transition happens ‘through a qualitative alteration by means of learning, and ‘after frequent changes from a contrary state’, the other passes by a different process.Footnote 35 The ‘frequent changes from the contrary state’ may mean that one changes from imperfection to a state of perfection. Later Aristotle talks of learning as a kind of transition into the ‘states and nature’ of the thing, by which he must mean that the states are perfective of the thing’s nature. That is, hexis seems to be used, in the manner of Physics 7.3 (417b13–16), to express the perfection of thing’s natural disposition at the level of second potentiality.
When someone has become a trained mathematician, it means that she has perfected her natural, human capacity to know. She now has (echein) the skill and so has the hexis of knowing. This state, prior to actual mathematical thinking, is still kata dunamin, and she still qualifies as able, dunatē. And the reason is of course that she is still only capable of thinking mathematically or architecturally, not actually doing so. It is appropriate then to describe this second potentiality as both a state and a capacity, a capacity at a higher level of achievement than that of the mere first-level capacity all humans have as such. Again, the point is worth underlining that a hexis such as knowledge is an acquired, perfective disposition but also a capacity primed (because perfected) to perform a certain function: actual thinking about mathematics or building or such.
Anthony Kenny has connected what he calls Aristotle’s ‘stratification of powers’ in De an. 2.5 with a view of the soul as composed of more basic potentialities: ‘Something that is a potentiality with respect to a particular exercise may itself be an actuality of a more primitive potentiality.’Footnote 36 On Kenny’s own theory the soul is a second-order capacity, a capacity to acquire capacities. This view is clearly congenial to my interpretation. We can take what I have called underlying capacities to be second-order capacities to develop higher-order capacities. Many memories, as we have seen, constitute the capacity of experience, the ability to bring together in one judgment a range of similar observations. Many thoughts of experience bring about the capacity of technē. But this only happens because we, unlike animals, have intellect, the second-order capacity to acquire experience, science and wisdom. What my interpretation adds to Kenny’s picture is the idea that these second-order capacities develop first-order capacities dynamically and serially: memory depends on perception, experience on memory, and so on.
The dynamic ordering has two explanations, which one might think of as an efficient and final causal account. The efficient causal explanation concerns the actualisation of the powers of the soul. If we ask ‘Where does knowledge come from?’, one answer is clearly perception, and perception, as we saw, has the distinct advantage that it is primed for activity at birth. Simply by our senses’ being affected by perceptible qualities, as we inevitably are as corporeal beings, we have actual perception, the lowest form of cognition. But our rational capacity is not in the same boat. We do not get to think knowledgably simply by being in the vicinity of the intelligible. Here we need a story of how we get from first to second potentiality. The model gives the explanation: through lower-level affections originating first in perception which create new states and capacities in the rational capacity, the soul is able to unfold higher-order capacities and states. The model thus explains the transference of actuality to the underlying intellectual capacity: perceptual affections are actual and once stored as memories trigger experience, the lowest-level capacity to make judgments about the universal. The model addresses an issue we find elsewhere in Aristotle, for example in his account of the development of the embryo.Footnote 37 The process is a sort of efficient causal bootstrapping whereby the activities of the lower capacities activate new states and capacities whose activities in turn trigger the development of even higher-order states and capacities.
The other way of looking at the cognitive progression is in terms of the development of content. Cognitive growth is also a process of conceptual differentiation. Each new stage grasps a more intelligible content by discerning a difference within the contents of the previous stage. To use another analogy from bodily growth, just as we cannot and could not acquire the form of adult human without having been through an ordered series of physical developments, so the content of higher-order thinking could not have been articulated other than through an ordered series of articulations of lower-order contents. It is a general feature of Aristotle’s teleology that it does not just prescribe a certain end, but that it also understands the processes that lead to that end as ordered in a certain way given that end.Footnote 38 Conceptual development is no different. It is this ordered process of increasing conceptual differentiation that the image of the rout in T1 addresses, as we shall see.
5 Nous and the Rout
I return then finally to T1. Here once again are the lines:
T1 Thus the states neither belong in us in a determinate form, nor come about from other states that are more cognitive; but they come about from perception – as in a battle when a rout occurs, if one man makes a stand another does and then another, until a starting point is reached. And the soul is such as to be capable of undergoing this. What we have just said but not said clearly, let us say again: when one of the undifferentiated things makes a stand, there is a primitive universal in the mind (for though one perceives the particular, perception is of the universal – e.g. of man but not of Callias the man); again a stand is made in these, until what has no parts and is universal stands – e.g. such and such an animal stands, until animal does, and in this a stand is made in the same way. Thus it is clear that it is necessary for us to become familiar with the primitives by induction; for perception too instils the universal in this way.’
This is not the place to develop a detailed interpretation of these notoriously intricate lines,Footnote 39 but just to mark some points of significance for my reading. The first thing to note is that the idea of universals ‘coming to a stand’ naturally goes with the idea of developing states.Footnote 40 A state (hexis) is a settled condition. It does not suffice simply to be affected: the affection needs to be retained. Since perception is not a developed state, the first thing that makes a stand then must be the memory that results from perceiving. Thus memory was the ‘retention (monē)’ and the holding (echein, 100a1) of the percept in the soul. We may think of this as a primitive universal because perception itself selects for a universal feature: ‘for though one perceives the particular, perception is of the universal’. However, the memory is retention of one of ‘the undifferentiated things (adiaphora)’, so I take it that it is not retention of a differentiated kind.Footnote 41
One might at this point reasonably ask, if perception is already of the universal, why we need to go through induction to grasp the universal. Previously, I suggested that experience is the point at which many particular perceptions are collected under one universal. But it seems now that perception can already recognise the percept as falling under a universal, ‘man’. One popular way of explaining Aristotle’s line is to say that while the perception is caused by a particular, its content is of a universal. This is likely to be true, but it does not help with our immediate problem, which is exactly that by having a universal content perception would seem to preempt the role of experience.
In An. post. 1.31 Aristotle explains how perception is not of the universal in the sense in which the universal is that which obtains always and everywhere (87b33). Hence although perception is of a certain universal attribute (cf. hē aisthēsis tou toioude, 87b29), it is not of the universal as such.Footnote 42 So while one may perceive that this or that cat has four legs, we do not perceive universally of cats that they have four legs.
Now if perception is not of the universal as such it stands to reason that perception cannot differentiate the universal from another universal. Perception may perceive man, but it does not begin to differentiate man as a kind from other kinds. Experience in contrast offsets this here disease from other diseases on the basis of a similarity, as we saw in T2. It is in experience that we first represent the kind as a universal, in its general extension, even if experience does not yet grasp, as does knowledge, the differentia that defines the kind.
Compare on the content of perception a passage in Ph. 1.1 also concerned with how we reach the principles of knowledge beginning from perception:
T6 Now what is to us plain and obvious at first is rather confused masses, the elements and principles of which become known to us later by dividing. Thus we must advance from the universals to the particulars; for it is a whole that is best known in sense-perception, and the universal is a kind of whole, comprehending many things within it, like parts. Much the same thing happens in the relation of the name to the formula. A name, e.g. ‘round’, means indeterminately a sort of whole: its definition divides this into the particulars. Similarly, children begin by calling all men ’fathers’, and all women ‘mothers’, but later on distinguishes each of them.
Aristotle’s use here of ‘particular’ to stand for a part of a universal is perhaps surprising.Footnote 43 However, the general point seems clear. While perception is of a universal, ‘father’, for example, it does not mark off this universal from other related universals like ‘man’, as if every man was a father. Perception does not observe the conceptual boundaries between objects. It does not itself have the resources to distinguish different kinds which are both being instantiated in the same object, like ‘father’ and ‘man’.Footnote 44 As in An. post. 2.19, Aristotle makes this point in terms of gradually distinguishing the parts of the concept involved. So, if one comes to understand that a father is a man with children, and a man a certain kind of animal, one has learnt to distinguish the conceptual parts.
There is a difference, then, between the way in which the universal, say man, figures in perception and the way it is represented in experience. Just before the lines quoted from T1, Aristotle said,
And from experience, orFootnote 45 from the whole universal that has come to rest in the soul (the one apart from the many, whatever is one and the same in all those things), there comes a starting point (archē) of craft and of understanding – if it deals with coming into being, craft, if it deals with what is, understanding (epistēmē).
It is experience that marks the point at which, after repeated perceptual encounters, the whole universal as a universal is present in the soul. This might again be illustrated by the medical example in Metaph. A.1: experience gathers the various cases of people suffering from this here disease in one judgment on the basis of similarity. Once we make this point, it is attractive also to take experience as the starting point (archē), in the sense that it is the basis on which we can come to recognise the causal universal. This would be supported by Prior Analytics 1.30 (46a18–28), where Aristotle refers to experience as delivering (paradounai) the starting points or principles (archai) of science.Footnote 46 We recognise the causal universal in response to experience’s sufficiently having identified a kind of universal phenomenon that allows of a causal explanation. Experience gives us the ‘that’ required to proceed to the ‘why’.Footnote 47
Since a universal that works as a cause may also itself have a cause, the process is iterable. As Aristotle says, ‘again a stand is made in these, until what has no parts and is universal stands, e.g. such and such an animal stands, until animal does, and in this a stand is made in the same way’. He presents the progression here as a development of more general terms. This might come as a surprise if one thinks of the process also as one of differentiation. We might expect the process to go from the more general to the specific if we were thinking of differentiation as demarcating a subset within a class. But this would be to misunderstand the notion of a difference (diaphora) in this context: just as noticing that Callias and Socrates et al. are all men may be noticing a difference, where this is something generally true of those individuals, so noticing that the deer and the fish are both animals is to grasp a difference. Aristotle is here thinking of an explanatory hierarchy where the higher terms will also enter into the definition of the lower terms, as animal will enter into the account of such and such an animal, and presumably, living being in the account of animal, and so on until we arrive at terms that have no parts, that is, are not subject to further definitional analysis.Footnote 48 This then is the story of the ordered differentiation of conceptual content that helps explain why reason’s development has to take this path.
6 Nous as a Capacity
As J. Barnes says, An. post. 2.19 asks not one but two questions ‘how do the principles of demonstration become known?’ and ‘what is the state which gets to know them?’ Barnes argues that the chapter’s answers to those two questions are respectively ‘induction from perception’ and ‘nous’. It is important to keep those two questions apart because nous plays no role in answering the first question. The first question is answered by pointing to various capacities, but the second is just a question about a state, Barnes maintains. That nous is a state is clearly correct. However, Barnes also takes it that because nous is a state, it is not also a capacity or faculty, like a quasi-perceptual ability to see the first principles, and for that reason tradition is wrong in thinking of nous as ‘intuition’. Nous, Barnes says, is no more a distinct capacity than is epistēmē: both are just states of whatever underlying faculty is involved.
I think a version of this is correct: the two questions are distinct, and nous does not work as a mechanism of induction. Also, it is correct to say that both nous and epistēmē are states – and virtues – of the intellect as a single underlying capacity. They are not previously given capacities, like the power of perception.Footnote 49 However, denying that nous is this kind of capacity does not mean denying that it is a developed capacity. Some intellectual capacities are acquired as states of reason and are also to be understood as capacities. We saw this in this case of technē, but it is also clear that Aristotle would understand epistēmē both as a state and as capacity. Its definition in Eth. Nic. 6.3 as a hexis apodeiktikē is entirely parallel to that of technē as a hexis poietikē, or phronēsis as a hexis praktikē. And we saw in De an. 2.5 how epistēmē is generally understood as a hexis and a second-potentiality capacity.
Nous is the terminal state of coming to grasp the universal in induction. It is a state and a capacity arising from the activities of lower-level states and capacities. It is the perfection of our inborn intellectual capacity, and so it is not itself an intellectual capacity responsible for propelling induction, the ‘bootstrapping’ process I have attempted to describe in this paper.Footnote 50 While the emphasis is here on nous as a state, it may be helpful also to think of nous as a capacity, specifically the second-potentiality capacity to think about first principles. The attribution of such a capacity need not be trivial. Just as epistēmē as a capacity enables you to give demonstrations on various occasions in relation to various demonstranda, so nous might enable you to recognise this definition or essential concept, as a principle in a range of different explanatory contexts,Footnote 51 as perceptual acuity allows you to recognise this perceptual item in a range of different settings.Footnote 52 But again nous understood as this kind of capacity only occurs at the end of induction, as a result of the final state in the process of grasping the universal. It is not available as a second potentiality already prior to induction. So Barnes’ main point holds against the traditional notion of nous as a capacity.
If the account of this paper is moving along the right lines, then the development of our cognitive powers follows the path of conceptual differentiation in our souls. On this account, Aristotle´s position is neither rationalist nor empiricist in any straightforward sense. It is not simply rationalist since the process of conceptual differentiation does not presuppose a full-fledged intellect or nous from the outset. The intellect has itself to develop gradually into nous through perception, memory and experience, just as also memory and experience have to develop as states and capacities from their starting point in perception. Nor is the position straightforwardly empiricist since it is the intellect itself that at each stage responds to percepts, memories or experiential judgments by recognising ever more general, unified concepts in the deliveries of lower cognition. Our intellect develops naturally through these stages by forming concepts that are ever closer to representing the basic features of reality.
Open access