1. The Project and the Plan

Plato knows that knowledge is not justified true belief. He wrote a book to prove it – the Theaetetus. So, what is it? Well, it is both one thing and many, a coherent clan (genos). Of course, the strategy of the Theaetetus is not the only one possible. If we cannot define knowledge, epistêmê, by breaking it down into its elements, we may be able to give an account of it by saying what it does, its job in the economy of human and other life. Epistêmê is the most inclusive noun Plato has for knowledge.¹ In this essay, knowledge and its cognates serve as catchall phrases for the whole clan, as well as translations for epistêmê and its cognates. (I translate genos “clan” to avoid associations with genus, or kind.) As we will see, phronêsis, roughly: wisdom, and nous, roughly: intellect are crucial members of the clan in his eyes. Nous is either truth or most like it and most true (65 D 2-3 cf. 64 B 2-3 – Henceforth, numbers with no further reference are those of the Stephanus pages of the Philebus, more closely specified by paragraphs and lines as necessary).

Plato’s Socrates aims to show in the dialogue that human life must consist of knowledge and pleasure if it is to be good, and part of this demonstration is an account of knowledge whereby knowledge can be more or less pure, more or less blended with extraneous elements such as sensation and practice. I say “blended,” rather than “mixed,” since a blend in English tends to be good, unlike a mix. All meixeis in the Philebus, as products of a maker, are good.² The cosmos of the Philebus is organised for the good and part of that cosmos, both changing and unchanging, is knowledge. Obvious examples here are technai, roughly: arts, which serve ends, most obviously in producing artefacts. So, we ask what is the good of knowledge? Now, knowledge may be many things – pure, exact, and clear are attributes both Plato and others have seen in it. But one thing it always is, is true. Knowledge is true.³ In some sense, truth is, as in the Theaetetus, part of the definition of knowledge. But the Theaetetus fails to produce a list of the components of knowledge. Instead of making truth an element in a definition, we are now proposing that truth is the end of knowing, what it is good for. And if knowledge is in its essence true, then knowledge cannot fail to achieve this end. If something is knowledge, then it achieves its end of being true. Formal and final cause coincide.

Our task here is to show how this is achieved in the Philebus, that is, to approach knowledge as a complex, which is, in its parts and as a whole, end-directed. My claim is that Socrates thinks that knowledge, as such, is directed at truth (alêtheia, to alêthes). Knowledge has to be true. If we were discussing language here, which we are not, we would say that “know” is a factive verb: any subordinate clause depending on it must express a truth. This fact about the verb, and the fact that knowledge must be true are of course connected; but, again, this connection is not our topic here. Things are complicated by the fact that alêtheia and alêthês in the Philebus also apply to things, being genuine or real. In knowledge, one may well think that these senses coincide: knowledge is genuine, only when it is true. It is not my purpose here to discuss further the senses of alêtheia. It is enough for this occasion to suggest a neglected strategy for the determination of knowledge by dialectic.

Since learning, teaching and investigating in the Philebus involve taking a clan (genos) and dividing it up exhaustively, knowledge must be treated this way by the anatomy in the Philebus. Thus, knowing has precisely that attribute which its objects have (cf. Charmides 168C 10-11); so, knowledge of knowledge is possible. Thus, a very rapid line of thought. In the Philebus, knoweldge is indeed complex; in fact this does not solve the problem, since dialectic must know what knowledge is. One part of knowledge must tell us what the whole of knowledge is, including itself. Telling us how to achieve knowledge is the function of the heavenly tradition. It tells us what knowing and getting to know is.

This paper falls basically into three parts, after this brief introduction, which outlines some of my assumptions and preoccupations in presenting this paper in this volume. The first two parts belong together: as answers to three questions, which Socrates poses, they present the gift of the gods, a procedure by which we are to teach, learn and conduct investigations, and the four genera which complete the gift of the gods by adding a maker to blend boundaries with the boundless, to produce blends. These blends are the unities cum pluralities we are to investigate using the procedure the gods have given us. The third part applies this procedure to knowledge and its clan, by following the swift anatomy of knowledge given towards the end of the dialogue.

The expression Finding my way home applies to two things: the soul becoming reunited with its real nature, and me finding my way home each day. The first applies to all soul, the second only to humans. The ambiguity in the phrase “finding my way home” aligns roughly with these two senses: the process of finding, coming to find, in contrast to the completed activity, having found. Even when the soul searches for herself, she is already there, what she really is, if distorted. If the soul was not already anchored in her essence, she would never find herself. Obviously, finding my way home, as one does every day, requires knowledge: right opinion is part of the clan of knowledge. As Protarchus, Socrates’ dialogue partner in the Philebus, says, we need false measures, inaccurate geometry, if we are ever to find our way home (62 B 5-9). These too are parts of the knowledge humans need to live well. I find my way home each day by attaching concepts to things I sense, and thus recognizing them – friends, Herms, policemen, lamps, steps, doors – or statues in the woods, which I may mistake for a human (38 C 5-D 10). Sensation itself is not informative; it adds nothing to knowledge, it merely allows us to attach knowledge to the here and now, things at times and in locations.

Less obviously, and more Platonically, the soul, necessary for all knowing, including nous, intellect, can appear more or less as she ought to be, more or less as she is when unadulterated. The soul’s real essence lies in her love of wisdom (Republic, Book X, 611 E1), in contrast, to the way she now is maimed by long contact with the body, in the image of the sea god Glaucus (611 CD):

All we have seen of (the soul) is something like Glaucus who lives in the ocean, if people were to see him. They would no longer find it easy to make out the shape he started with, because some of the original parts of the body have been broken off, others have been worn away and completely eroded by the waves, while things like shells, seaweed and stones have grown onto him. As a result he no longer resembles his original nature. He looks like some wild beast. It’s the same with us, looking at the soul when it is afflicted with all these evils (sc. caused by associating with the body).⁴

Finding my way back to where I have come from, the soul may rid itself of the extraneous accretions clogging up my cognition, acquired by the soul being affected by the body. The true nature of the soul lies in its relation to what does not change (cf. Phaedo 79 D the soul is akin to what is always, cf. also Timaeus 90 A, and Republic 519 AB). The art which realizes this true nature in Plato is, of course, dialectic, not only the highest of the high, but also the exercise of the love of wisdom, “philosophy,” in other words of the love of truth which will emerge as the proper ruler in the soul, that is, what we do everything for (Philebus 58 D 4-8). “Love” does not imply a desire of something distant, but hands on, day to day, thorough organization of a life, a drive that requires exercise for its existence, not a mere longing. In this drive, dialectic has various functions to fulfil. Things – desires, sensations, actions, habits, get blended with knowledge, when knowledge occupies a place in incorporated souls. But, besides finding my way back to an unadulterated soul, I do find my way home, as a body-soul composite, a living body.

The editors have chosen the title Les architectures du savoir for this collection of papers. This phrase suggests a concrete structure, built for a purpose or end. And, indeed, correct opinion, knowledge, art, nous, memory and wisdom are said in the Philebus to have a certain structure, albeit an abstract one, a general one, so possessing instances. And, of course, knowing, remembering reckoning and practicing arts are things we do, purposeful activities: if we did nothing, we would have nothing to talk about. Not only do I think that the model in the Philebus shows how Plato thinks we should deal with definitions of knowledge, it is also, and even more characteristically for ancient philosophy, clear why knowledge is the decisive constituent of the good life. In Phileban terms, this is because knowledge, in its varieties, but finally in the dialogue itself, dialectic, imposes measure on pleasure. Dialectic tells us the answer to the question posed in the dialogue, insofar as we assent to the answers to Socrates’ questions given by Protarchus. Knowledge is not the determinant of the good life in dominant forms of ethics in modern times, and so may count as a (“pagan”) Greek contribution to thinking about the good life. It is, to use an architectural image, the keystone, which, although the highest part of the arch, is fundamental.

The subject of the Philebus is the human good. Knowledge may make humans good, as Verity Harte⁵ has argued recently, but that is not the point at issue in this paper. Knowledge may still have its own good or end above and beyond human existence, as it were, simply speaking and not in relation to us. Plato’s Socrates attaches great importance to a life without body, and so without pleasure or pain. For knowledge, or, at least, its pure forms, nous, intellect and phronêsis, wisdom, are there even without body, and so even apart from human life. So, their good is something apart from human life. The suggestion is that knowledge must be true, and this is its end, and this is what can guarantee its stability under certain circumstances: its end is internal to it. The relation between knowledge and truth is stable; you might say, necessary or essential. The further members of this clan knowledge arise from blending truth with other attributes, e.g. sensation and practice, thus making the knowledge impure and impinging on its exactness and clarity. For if forms of knowledge are pure, and unmixed with extraneous elements such as practice or sensation, then nothing affects their exactness, or their clarity. Each thing, as known, is just what it is, and none other. Without the admixture of anything extraneous, the presence of each and every thing to nous is complete. In this way, only what is knowledge pure and simple, and so only true, has the stability of unchanging things, being above and beyond change. Plato’s dialectic has these things, above all, as its objects.

Dialectic may claim to be able to talk about everything, but this should not mean that it has the license to call into question the deliverances of the (other?) arts and sciences. Epistemology has to have an object, if it itself lays claims to truth. So, in the Apology 22 DE, Socrates is happy to admit that the craftsmen do know what they are doing, and talking about, if they confine themselves to their area of expertise. When they stray from this, they are like the poets and the politicians, neither of which groups in fact proved to have knowledge at all, under Plato’s Socrates’s cross-examination. It makes sense to think about what it means to know for those who know; but that requires that knowledge is justifiably ascribed in the first place, in this case, to the builders of the Parthenon, to Phidias, shipbuilders, the makers of sandals & co. Art, technê, in Plato and Aristotle is a knowledge concept; that much is true about its fundamental meaning for them. It is instructive to remember Aristotle’s account in the Nicomachean Ethics: technê is a productive disposition guided by a true account (VI 4 1140a 10). Sir John Lyons,⁶ in his Cambridge PhD, indispensable for any investigation of knowledge terms in Plato, speaks of an essential connection between epistasthai and technê: knowing, in Plato’s Greek, is just what you do to an art, when you have learnt it. Furthermore, specialized knowledge tends to be a technê for Plato. This is relevant to a core concern of the Philebus in distinguishing the parts of knowledge. While pleasures are opposed to one another (13-14), there is no knowledge opposed to other knowledge. Knowledge, all knowledge, is coherent. This of course points to the importance of truth for knowledge. If truth is necessary to knowledge, and truths cannot contradict one another, then nor can knowledge. This of course also requires that truth, and knowledge not be involved in change. This makes true opinion (“correct opinion”) about changing things, and thus itself susceptible to change, not knowledge, however good it may be in guiding one from Athens to Larissa (Meno 97 AD). But by guiding us just as well, if not just as reliably as knowledge, even correct opinion is in the same group as knowledge. There is no bad knowledge, no knowledge is excluded from the good life.

2. The Gift of the Gods

2.1. Socrates’ Questions

In order to investigate pleasure, and its kin, and then knowledge and its kin, Socrates first asks some questions, which are then answered, as I think, firstly by the gift of the gods, and then by the doctrine of the four clans (“genera”, “genê”). These are hotly debated stretches of text, and what follows is a summary account of a reading that aims to integrate what there is, and its investigation into the account of knowledge in the Philebus.

Socrates first dismisses questions here about the unity of perishable things, individuals like Protarchus (14 C 11-D 3, D 4-E 4). His topic is not changing things, but things above and beyond change, and, thus, also their relation to changing things. Moving on to the topic in hand, Socrates then asks (15 AB) three questions, about units that do not come to be and pass away. He gives as examples the unit human, the unit ox, the unit fine thing, the unit good thing. These are units about which “much zeal in making divisions turns into controversy” (15 A 6-7). The way these units are presented suggests dialectic, in the sense of there being one questioner, and an answerer. So, there are three questions that Socrates thinks this approach gives rise to:

• Should we assume (hupolambanein) these units (henades) exist by being in truth (alêthôs) self-standing things (monades)? 15 B 1-2
• Then: how are these, in being always just one unit, liable neither to coming to be nor passing away, each a unit even if they are not involved in coming to be, each one a single stability (bebaiotês)? 15 B 2-4
• And finally, in investigating the things not subject to change, we need to relate them to those things that are subject to change: thus, we ask, are the units one and the same, while present both in themselves and in all the many things? 15 B 4-9

These are the questions that the gift of the gods (16 C 5-8), and the doctrine of the four clans (“genera”, “genê”) will answer. Plato uses a hypothesis, one of the fundamental tools he developed for circumventing apparently unanswerable questions, notably in the Meno and the Phaedo. In the first dialogue, we assume that virtue is knowledge. From this, it follows that it can be taught, thus answering the question of the dialogue, viz. how we come by virtue. In the second, we assume that there are ideas, because of which individuals are what they are: the beautiful is beautiful because of the beautiful itself. This hypothesis is the safest one possible, according to Plato’s Socrates, thus offering a way out of a situation in which recourse to statements has to be made in the absence of the things themselves. Because we select our hypothesis, on the basis of its advantages for explanation and so knowledge, we are not simply bound to follow the deliverances of the senses, or to make arbitrary assumptions. We choose the hypothesis in the light of the good it brings to the explanation.

2.2. The Answers

In the Philebus, the gift of the gods gives the answer to the first question, I think as a hypothesis. For the fancy dress of tradition may well simply clothe a hypothesis, as suggested by the first question just mentioned. The method is based on what there is, a boundary (peras), which is “naturally grown together” (sumphuton), naturally fused with something unbounded (apeiron) (16 C 10). Something such as human or ox are parts of the unit animal. Animal is thus both a plurality and a unit. The analysis of animal will tell us how many there are. That is, how many parts there are of the unit animal, not the number of changing individuals. Thus, the first question is whether these units are indeed, “in truth”, independent, stand-alone things (monades). The second question concerns the certainty of their unity, quite apart from any question of their coming to be or passing away. Quite how does animal encompass human and ox, and not fall apart? This same model is then extended, in the doctrine of the four genera, to the relation between the units, which stand alone (monades), and things coming to be, relations produced by nous in such a way that ends are fulfilled among changing things. As to the third question, the point is that a bounded thing bounds the parts that make it up, or the more and less, the unbounded which needs bounding so that it can undergo coming to be, and reach being (cf. 26 D 8-9). Ordering coming to be does no harm to being, in fact by providing boundaries to coming to be, it makes it possible.

Finite beings depend on the gods, and the light of our knowledge comes from them. And tradition has passed it down to us; this combination of tradition and divinity is likely to make hackles rise. A divine tradition would appear to leaves the basis of knowledge hanging in mid-air. But let us make this hypothesis, that is to say, accept the tradition as handed on by Plato’s Socrates. Our subject here is knowledge, and things in its clan, so one point to bear in mind is whether knowledge itself, “investigation, learning and teaching” (16 E 3), is amenable to the treatment which the tradition prescribes, supposing, for a moment, that we understand what has been handed down.

Let us sketch Socrates’ account of the tradition, which he asserts has passed on the gift of the gods, handed down from a time when humans lived more closely with the gods. (16 C 9-E 2). Things which are said (by us, of course) to be always, are constituted from unity and plurality. The reason for the constitution of all things out of unity and plurality lies in “boundaries” and the “boundless”. This would appear to be an explanation for what we have already been told: there are units, which each breaks down into a plurality. Because there are these things, the tradition says, our procedure should have a determinate form. In other words, any investigation follows an assumption about the way things are, about what is the case. What this means more fully will emerge in due course. Briefly, here and now, we can see that it allows for the greatest variety combined with unity. We analyze a unit into all its parts. In investigation, we want not to miss anything: the greatest variety must be taken into account, but merely hitting on something, and not fitting it into a greater unity will tell us nothing. So, we can see already that the gift from the gods requires systematic investigation of us. We need to break things into their parts, but also see where they belong in a whole. We go up (towards generality), as well as down (towards particularity) as far as we can in each case. Not only that, we need to know, of any whole, how many parts there are (16 D 6). Counting the members of any clan requires identifying these members (17 B 8, 17 D 1, 19 B 3)⁷ Counting, not measuring is what the Gods’ gift primarily requires. We start from any idea, that is any unit to be analyzed, and we are assured we will find one if we look (16 D 1-2). This has, even from a low, “empirical” level, plausibility: the world is chunky, presents itself to us, and presents us to ourselves, in determinate parts. The process of analysis will show us if we have at the outset chosen the idea wisely: systematic investigation requires breaking units down into their constituent parts, and if we have begun from a good unit, a good idea, the analysis will take us far. And when this process is done, and the parts enumerated, well, the rest is the unbounded ocean of material things. As such, they cannot be investigated, understood or taught. As such, they can be sensed and recognized using what we know.

On the side of things known, “what is now” (23 B 4), we have bounds and the boundless, aspects naturally fused together (16 C 10). These two factors come together to form a whole, an idea, comprising the parts it falls into, as a whole. An example is given of the gift of the gods in action, devoted to sounds, which breaks down into two arts, (technai), mousikê, music and grammatikê, the art of letters (18 D 2, 26 A 3-4, cf. 53 A 3, C 5), the latter taken here to be the art dealing with all of the sounds Greek includes, and so too how to represent them. This art forms a bounded whole, for example, all the phonemes (of Greek), which are the parts that the sounds treated by the art of letters fall into. Clearly, the whole and its parts belong together naturally. And the whole bounds the parts, sets them off from everything else: the subject of mousikê is a unit. In turn, the parts, each bounded, determinate and set off from their fellows, have features which set them off from one another. We will not pursue these examples in detail,⁸ but they are significant for our project here in that they make clear that arts, and so all forms of knowledge, are analyzable in terms of the gift of the gods. But they should also be subsumable under greater wholes, such as those we will encounter as the analysis of knowledge and its kin.

2.3. The Four Genera

Socrates says that for their discussion, he and Protarchus need “weapons” apart from those already used (23 B 9-10). Thus, the four genera, the “four clans” must include boundary and the boundless as two clans in the same sense as before, if Socrates is not to be guilty of equivocation (23 B 8-9, C 9-10, cf. 16 C 9-10). What is to be divided up into these four clans is “what is now” (23 C 4), thus both things subject to change, and those necessary for change and so not subject to it. Boundary (peras) and the boundless (apeiron) are what the technai are about, the things that are said to be. Besides boundary, and the boundless, the new kinds introduced here are the blend of boundless and boundary, and what blends them, to make up the four clans, “genera” (23 C 12-D 2, D 7-8). This is clearly always a question of types, not tokens, that is, we are speaking generally, even if it is about things that come to be and pass away.

For present purposes, we will not present this doctrine in detail, beyond what is needed for the understanding of knowledge & co., our subject, and the subject of the detailed, if rapid, anatomy towards the end of the dialogue. Some readers feel that the four genera are introduced only to be dropped, and that the dialogue cannot be regarded as an example of its own method. Yet it is clear that the kinds of knowledge are blends, of sensation, practice, with knowledge. These include arts, technai, correct opinion, true calculations, memory, and then again forms of knowledge, epistêmai, alongside nous and phronêsis. These all fulfil ends, most strikingly in the case of the technai, and so are put together for this purpose by the maker, nous, who is the cause of their being, of their achieving some good.

3. The Anatomy of Knowing in the Philebus

3.1. Boundaries, the Boundless and Pure Knowledge

We now have the instruments, or weapons, as Socrates calls them, to reconstruct the anatomy of knowledge. Prominent among these are boundaries and the boundless, the clans a maker puts together to form blends. How should we take the close association between knowledge and boundaries? Boundaries and the boundless when fused in their end-directed blends are the things we can know. The blends knowledge enters into are with sensation, practice, for example, – these are things which may be more or less – intensively, extensively – and are susceptible to other predicates of more and less, that is, those predicates which are susceptible of being in greater or lesser ratios. The more and the less is the characteristic of the boundless (24 B 4-8). In the case of knowledge these attributes are clarity, stability, purity, truth. These are predicates with counterparts in many accounts of knowledge; but it is not my purpose here to pursue these connections. Furthermore, these predicates are connected to one another. A thorough account would lay out just how these connections work. Here a few hints must suffice. Like pleasure a small part of knowledge can be pure and true. Truth (61 E 4) and purity (55 C 7, 57 B 1) are the criteria for ranking both pleasure and knowledge. Truth and purity are intimately connected in the Philebus. The question about the truth of white, i.e. what is truly white, is answered in terms of its purity: what is truly white, is purely white (52 D 6-53 C 2). Otherwise, one might say, it is not really white, it is grey. Generally, truth and knowledge go together (59 B 8). But in dialectic, the discipline concerned with being, what is truly and forever the same we have “by far truest of all kinds of knowledge” (58 A). It has the highest degree of truth, along with clarity and exactness (58 C 4).

Socrates talks about impure knowledge (55 D 7-8). Along with clarity, purity in knowledge correlates with, at least, exactness (57 B 9-C3). Some forms of knowledge are indeed called exact (57 E 3). Purity is such a predicate, since knowledge can be more or less pure, that is, with or without extraneous parts added. The purity of knowledge will be connected with its exactness. And knowledge can be more or less stable (bebaion). Since knowledge depends on what is, i.e. what is changeless, the less non-knowledge parts it contains the more stable it will be; so too for clarity, if clarity is what allows the unclouded presence of the thing known to the knower.

Here is a swift sketch of some plausible relations between these attributes of knowledge:

Truth belongs to something when it is genuinely what it is.

Purity belongs to something when it is only just that thing that it is.

• Impurities would impugn Truth, since then the thing is not genuinely what it is: you can point to parts that are not it. So Purity follows from Truth.

Stability belongs to something when it is unchangingly what it is.

• If changing things have to be complex, then Purity implies Stability.

Clarity belongs to something when it is what it is without obstructing elements.

• Purity implies Clarity: if nothing else is in the thing, then there are a fortiori no obstructing elements.

Exactness belongs to something when it is bounded.

• A lack of bounds would impugn Truth and Purity, and hence Clarity and Exactness. Unclear boundaries make the decision as to Truth and Purity impossible, so they impugn these attributes. Exactness and truth lie in the arts’ use of measure and number (57 D 3).

Epistêmê is introduced by Protarchus at 19 D 4, and then taken on readily by Socrates in his anatomy, which is said by him to be of nous and epistêmê (55 C 4-5). Just as with pleasure, some parts of knowledge are purer than others (57 A 11-B 2). The purest part, we see, is dialectic (58 AB). Plato’s Socrates here relies on the judgement of the sensible, “those who have even a small grasp of nous” (58 A 4-5) to confirm that dialectic is the “truest gnôsis,” the truest form of cognition, dealing with what is, simply speaking. Here, nothing occludes truth.

A boundary must bound something. Boundaries cannot occur alone. In other words, in terms of the gift of the gods, nothing is simply unit, it also has plurality within it, not merely in the sense of having instances, also all generalities have lesser generalities within them. So, all reality, “beings,” are composite, blends. And as we have seen, this is an end-directed cosmos: boundaries and the boundless blend insofar as they attain ends, as conceived by the maker. The cause of blending is sophia or nous, human or divine. Even goodness, as we will see, is not simple. Blending does not compromise the boundary: on the contrary, the legs, beak, wings, and tail of the bird, say the swan, constitute that bird in the first place. So too with knowledge: music would not work without the pains needed to gain strength (rhômê) in your playing. This is so, even if practice (so many hours, so intense), in itself, is blind, not knowledge at all. We remember that only unlike things can be blended. So, no blending of the same part of knowledge with the same part of knowledge.

Knowledge has parts, since it can blend with other things, thus making arts, for example. The blending produces parts. Nonetheless, the question is not only one of parts and whole, but also purity and impurity, as we have already indicated. There is no one genos, in the sense of a class, to which all the parts of knowledge belong. Comparison with the Phileban analysis of pleasure is instructive. Here too there is not one kind into which all pleasures fall, as in a genus encompassing species.⁹ There is one central case, the refilling of emptiness in the body, to which all pleasures must relate in a variety of ways. No body, no pleasure. But there are some pleasures which belong only to the (incorporated) soul. The structure of pleasure has similarities with Aristotle’s pros hen structure (cf. Metaphysics IV 2, G.E.L. Owen’s “focal meaning”). Aristotle’s example is health: many things can be healthy – a walk, a ruddy complexion, an apple. But each in its own way relates to a healthy body. Some healthy things produce it, some are signs of it, and others preserve it. In the case of knowledge, the members of the clan to which the others relate are the pure parts of nous and phronêsis.¹⁰

So, the arts are not parts of nous. The branches of knowledge and the arts arise through impurity. The god-given procedure requires us to break things down into their constituent units. But it gives us no guidance as to the differentiation, exactly how we should divide one part from another. And the answer in the case in hand is that the parts of knowledge are blended with non-knowledge in such a way as to attain ends. The process of blending, at the hands of a maker, (to dêmiourgoun 23 D 7-8, 26 E 1-27 B 1), is the contribution of the four genera. As with the example of white used to illustrate the search for pure, rather than extreme, pleasure, we look for something with only knowledge in it. What is knowing pure and simple? Socrates says, let us use the most honorable names: nous and phronêsis (50 E 5-58 D 1).

The way Socrates begins dividing up knowledge to start with is puzzling: “So, I reckon one bit is productive of knowledge concerned with things to be learnt, the other of knowledge concerned with education and upbringing.” (55 D 1-3). Verity Harte¹¹ sees in this a distinction between what produces things and knowledge that makes us good, by training. The dialogue is about human goodness, so it is desirable that we are told what makes us good; clearly, pleasure does not, so knowledge must. Fair enough. But the criteria of discrimination between these kinds are difficult to grasp:

1. One produces epistêmê relating to doctrines (mathêmata).
2. The other produces epistêmê relating to training (paideia) and upbringing (trophê).

What is the nature of this distinction? Not one between things known. Are these divisions not external to the sciences, i.e. relative? They appear to allow a distinction based on what the knowledge can be used for, in other words, based on the good. This conforms to the way we have suggested that the four clans work: wholes, blends, are produced from boundaries and the unbounded, to attain ends. But we must be carfeul here, for this is not to say that utility is the only end; we will return to utility shortly. Furthermore, an art may well both produce knowledge and be useful in upbringing. So this distinction simply points to the uses knowledge may be put to. It does indeed look rather as though it is a distinction between things you can do with cognition: either you can use it to contribute to the sciences, or else you can train people, bring them up as academics, or at the least people who know their letters. Or you simply bring them up, but well, speaking simply. This is of course not necessarily a distinction between kinds of knowledge, but what you do with knowledge. This is then relevant in the dialogue, not as a contribution to saying what knowledge is, and so what the good of it is, but how knowledge contributes specifically to human life. Furthermore, the contribution to making humans better at doing what they do is not spelled out.¹²

The analysis of epistêmê and nous follows that of pleasure. When taking pleasure apart, into its parts, Socrates is sensibly trying to isolate the pure parts of pleasure (50 E 5-52 D 1). The hedonist Philebus does not dream that the important thing is pure pleasure, not extreme (“the greatest”) pleasure. For extreme pleasure in fact includes pain, something the hard work of an investigation brings to light. The importance of purity as opposed to extremity is made clear using a comparison: pure white, even if a small part of white is simply white, and nothing else (52 D 6-53 B 6). So, the arts are to be analyzed, beginning with the manual ones, analyzing what is purer and what is less pure in them. Now on the workmanship side there are impurities. The first move is to distinguish those parts of them that are such as to guide or command (hegemenokai, 55 D 5-8): these are their “mathematical” parts, counting, measuring and weighing.

So, three technai guide the other technai – arithmêtikê, metrikê and statikê, the arts of counting, measuring and weighing (55 E 1-3): what makes them all technai remains unsaid. The art of counting is the general art, including the other two: counting is necessary for both measuring e.g. with a ruler, as well as for weighing with scales. At 56 D 4-E 3 arithmêtikê emerges even more clearly as the important one – unequal units are given to the many in contrast to the equal one’s used by philosophers, a distinction that is then made for other quantitative arts, as though these fall under arithmêtikê” (56 E 7-57 A 4).

The numerable parts of the arts guide the other parts, tell them what to do. Then we say, take away measurements and counting, and weighing and what is left is pretty miserable (55 E 1-3). The rest is described as follows: “After that, guessing would be left, and taking care of sensations using experience and a kind of massage, using the powers of hit and miss, which many call arts, and which acquire their strength from laborious pains.” (55 E 5-56 A 1)

Here, the last thesis is not Plato’s Socrates’ view but that of the many who call these things technai, where Socrates does not, reserving that term for arts that really are knowledge, and so have their strength by being true (cf. 56 A 1). Other questions rear their heads. Is Plato right to be so rude about practice and the “intuition” that grows with it? As both its friends and its detractors say, there is no accounting for it! Think of the tongue tied musician, who may be able to train, but cannot teach. He is silent when challenged to give an account. Plato is confident that the knowledge can be sheared off the practice. What are the shears that do this? I would think, accounts delivering generalities. In music, part is a matter of measurement, part a matter of practice; art here is the actual activity of the art. The same applies to the art of navigating, farming, building, medicine and military strategy (56 B 1-2). Each of these is divided in two, practice and the numerical arts.¹³

This division is performed on music, one example of an art dealing with sounds that we have already met. Here, music, on the one hand, does badly; for it is not clear, and pretty unstable (56 A 7). This analysis shows that we are not involved in a classifcation of kinds of knowledge, in that parts of the one knowledge may reappear later, indeed may reappear unblended later: a kind of arithmetic, “the art of numbering,” as one may translate it clumsily and literally, is part of architecture. Of course, the same knowledge may not be blended with the same knowledge – that is just more of the same, so no posibility of blending; but different parts of knowledge do indeed blend beautifully. The use of measuring tools, and hence of number, is crucial to carpentry and many other arts: measurement, using measures (the square, the ruler), provides exactness (56 B 4-C 2). Music is then contrasted with architecture (56 C 5-6), to its disadvantage in that the latter uses tools to measure, ensure the exactness of, its work.

It is clear that this division of the arts is not one purely in terms of their objects, nor in terms of their ends, but also in terms of their relations to arithemetic.¹⁴ What are we to make of this? I suggest that this is because knowledge of counting and measuring are pure, in the sense that they are not mixed with sensation, practice and guesswork for their strength. There are in fact attributes of the philosopher’s arithmetic which are helpful here. Knowledge is ranked as to whether it is exact, stable, clear, true (56 A, 57 BD, 58 C). So, what can knowledge be mixed with? Sensations since these do not tell us anything, nor can practice (meletê) and experience. Only a logos (when spoken) can tell us something. Pointing to things says nothing.¹⁵

One of the debates about the interpetation of the Philebus is whether individuals can be investigated by the divine method and whether they then are constituted by the four genera. Now individuals, usually so called, are what feature in the Philebus as beings subject to change, comprised in the natural cosmos. The position of the anatomy of knowledge towards the “investigation into nature” is clear (58 E 4-59 A 10): it does not concern beings that are always. It is thus precluded from the most “exact truth,” so neither nous nor epistêmê about this will have “the greatest truth” (to alêthestaton 59 B 4-5). This indicates that changing things, which is what is meant mostly by individuals, will be treated, but not of course as such. They will be treated generally. Thus also, Aristotle thinks that individuals such as Socrates are substances. But of course, he does not discuss them as individuals, either in his dialectic or his first philosophy, although of course he gives a much discssued answer to the question, what makes all of them, generally, individuals.

Thus nous and epistemê are the last analysis determined by their subject; depending on its nature, ands whether it is susceptible to being true. The contrast with natural investigations (“if someone were to investigate nature” 58 E 4-59 A 5, A 11-B 2) lies in the qualities of knowledge, namely stability, (bebaion), purity (katharon), and truth (alethes, cf. 59 C 2-6). The art dealing with things that are always in the same state, and which behave “most unblendedly” (59 C 4) has these qualities to the greatest degree. The talk of blending is not a chance addition here. Knowledge pure and simple deals with things that do not change. It can be blended, but when blended, it loses purity, and so clarity, stability, exactness and truth.

The question (57 E 6-58 B 5) is then whether anything could possibly be prior to dialectic, that is, in terms of exactness, clarity and truth. Dialectic is taken as a technê – and its objects are: “I think that all who are attached to nous even to a small extent think that the cognition of being (to on) and what is naturally always the same and in truth (ontôs) is the truest.” (58 A 2-5) A competitor is then introduced, rhetoric, but it is introduced only to be chased out again, in that the reasons it has priority for Gorgias do not fit it for a contest with dialectic. Gorgias says that the other arts are willing slaves of rhetoric (58 A 4). He plays a surprisingly important role. One would have thought he had been long left behind by the aged Plato. Protarchos has, supposedly, been taught by him, and so introduces the idea of the utility of art. This suggestion is swept aside by Socrates in a long speech (58 B 9-D 8), using the opportunity to make clear that the criteria he is interested in are not utility or reputation. So Gorgias has two functions – firstly to paint in lurid colours the idea of arts ruling one another, one power enslaving others; in contrast, we have already met the more discreet leadership of the mathematical parts of the arts. Gorgias, we are told, propagates the willing slavery of the other arts under rhetoric. But in terms of the hierarchy of the arts, the crucial thing is that utility is not the criterion used. Being good specifically for humans is irrelevant to the hierarchy, although it is relevant to the main purpose of the dialogue.

Socrates admits that Protarchus has made this mistake because of the lack of precision in Socrates’ own formulation: we are not looking for the science or art that helps us most, but which has as its remit, or end: clarity, exactness and what is truest – even if it is small and helps us little! “My dear Protarchus, I was not looking for which art or which science is distinguished from the rest by being the biggest and the best, and helping us most, but rather which takes care of clarity, exactness and what is truest, even if it is small, and of little use, that is what we now are looking for.” (58 B 8-C 4) Socrates reduces the list of attribues of knowledge that he favours to simply what is truest (58 D 1).

This makes clear that helping us is not a criterion guiding Socrates’ search, despite the overall aim of the dialogue; it is this aim which makes Veristy Harte’s suggestion, already mentioned, that the good of knowledge is making us good, so sensible. However, our interest in our well-being is parochial; and knowledge is not parochial. The good of knowledge is not our good. So we must note that Socrates characteristically has all living things in mind when asking about the good; and that anyone who knows the good, anyone, not just humans, goes for the good (20 D 8-9, cf. 21 B 4-6). Plato may well still think that knowing is constitutive for our goodness, not merely accidentally good in providing us with good effects. The question is then what it is about us that makes that possible, and what it is about knowledge that makes it possible. One thing is that pleasure, without knowledge, is not registered, and so not pursued, and so not good. Another is that knowledge and its clan allow us to work out which pleasures will not disrupt knowledge, and the amounts and kinds of pleasure which are compatible with the drive for truth.

So what is Socrates looking for? “The power of the soul that loves truth and does everything for the sake of this.” The passage reads in translation: “But let us say, by investigating whether there is a power of our soul that both naturally loves truth and does everything for the sake of this, let us say if dialectic possesses the pure part of both nous and phronêsis, in all likelihood, or whether we must look for another knowledge more authoritative than dialectic.” (58 D 4-8)

Here he reaches the central point of Phileban psychology, as I would think: a power of the soul that naturally, that is in its true nature, is possessed of a drive, organising all its activities, towards the truth. We mentioned the love of wisdom at the outset which gives the true image of the soul in the Republic, in contrast to the sea god Glaucus. Here we meet this love or drive again: this is the soul finding its way to its true nature, the way home. Phileban psychology has practical spin to it, despite, as I have mentioned several times, there being no practical – theoretical distinction in knowledge. I think that the practical spin derives simply from the fact that it is our knowledge that is at stake. If this knowledge is to exist, we must do things, be active. Hence the idea of a love of truth is practical in the sense that it has profound practical implications, depending on whether this love is active in me or not. Socrates says that we then do everything for the sake of this, that is, for the sake of acquiring truth. In other words, truth is the organising end of all we strive for. It is the heart of the Phileban soul because it explains why the soul does all it does, inter alia why it has authority over body and hence all else. If you do not want to know what is the case, if you do not “love the truth,” then you are left as a passive hedonist such as Philebus himself. Only the soul can discover, and know what is the case. This explains its authority over the body which knows nothing.

Protarchus agrees that no other art or science can be said to hang on to, cleave to truth more than dialectic (58 E 1-3). The notion of blending allows, as it were, the purity of knowledge to be diluted, such that it may fulfil different purposes. Music needs practice, as does knowing your letters; if they are to reach their ends, these arts cannot be pure, they must be blended. For other things, blending merely does or would get in the way. But as we have mentioned, different members of the clan knowledge may be blended. There may be a blend, for example of different kinds of mathematics.

Most crafts use opinions (58 E 4-59 A 2), and are concentrated on finding out about opinion: Now as we see in the discussion of pleasures, opinions can be true or false (doxa 37 B-38 A). In contrast, there is only talk of truth, indeed: the greatest truth, when dialecic is at issue. This makes clear how important truth with no possible contrast to falsity is for Plato’s view of truth. False, deceptive nous does not make sense. As to truths about ideas, either you understand them or you don’t. You cannot get them wrong. Either you have a grasp of the units animal, the fine the good, or you do not. Perhaps this is the reason that in the Philebus, doxa, opinion, is right, not true (11 B 8, 64 A 5, but cf. 42 AD on false pleasures). Other sciences and arts are susceptible to the wobble of doxa (58 E 4-59 B 11). In contrast, and this is what sets Socrates apart from Gorgias, purity and truth, and unblendedness, so called, concern things that are always in the same state. Socrates is interested in unchanging things, which we might see as the bearers of necessary truths. There are assumptions at work too about the knowledge, tracking, or modelling the attributes of the known, which we cannot go into here.

3.2. Goodness

The good – what is it? What anything that knows it, pursues and wants to have in its possession (20 D-21 A, cf. 60 A 10-C 4): it is sufficient (hikanon) – i.e. stills desire, puts drives to rest, is complete (teleon): if the good does not satisfy a drive, nothing will. Nothing is missing from it, sc. relative to the desires pursuing it. And “it is choice-worthy for all, whether plant or animal, such that they can live in this way, i.e. having it, throughout their lives.” (22 B 4-6) This account of the good is of course suited to the enquiry of the Philebus: if you know pleasure, you “desire” it, i.e. are driven to it by a life-ordering drive, but this requires knowledge – memory, expectation, awareness – otherwise the attraction of pleasure fades, indeed, its goodness goes. So, such a life must comprise both pleasure and knowledge. The ethical aim, from our perspective, i.e. of beings who have to guide their actions and shape their lives, falls under the general account of the good. There is, as usual in Plato, no special category moral.

3.3. All knoweldge is in our good life

Forms of knowledge are not opposed to one another (13 E 9-14 A 5), in contrast to pleasures (13 A 4-5). Pleasures conflict with one another, kinds of knowledge do not. This presumably means that they do not contradict one another. Not only that, there is no reason to reject any of them, when blending our good life: “Shall we add to the blend the not-stable, not-pure, the false (pseudes) measure, as well as the false circle? Yes, we’d better, if anyone is ever going to find the way home.” (62 B 5-9)

Socrates is referring to the measuring and the geometry used by the arts (56 D 4-6). Indeed, Plato may well be saying that the circle used by an architect is not a real circle at all. Which is quite right: strictly speaking it cannot be the line that is equidistant from a point. So it is deceptive if you are interested in what circles are, their properties as revealed and used in geometry. But not if you are walking home.

3.4. The differentia of the kinds of knowledge

We have now reached a situation in which dialectic has been established as the art able to acquire the pure parts of nous and phronêsis. It would therefore appear to be pure, in the sense that nothing is blended in that would make it unstable, because subject to change, or inexact, because adding elements that are not just what they are, a matter of guesswork, hit-and-miss. Socrates distinguishes two kinds of knoweldge as to truth (es to alêthes 61 D 10-E 4), one is about coming to be, the investigation of nature we have already met, and the other dialectic which we have also discussed. The latter is then truer. In the four genera, the unbounded can be more or less. Is truth unbounded here? A grammatical point that is indicative, if not a knock-down argument: all these terms are used in the comparative and superlative. Along with clarity, certainty and exactness, each part of knowledge has its own measure of these qualities. But do they define each part of knoweldge? Presumably not. To judge by the heavenly tradition, an art needs an idea, one might say, as its object, one might say, as its subject. It looks as though the degrees of these qualities – more or less – depend on their subject: the investigation into nature is wobbly, where dialectic is adamantine, because the one deals with wobbly things, whereas dialectic deals with unchanging things.

But there is a much more fundamental reason for truth being included in the good life, than our getting home. Socrates says: “Anything we do not blend truth (alêtheia) with, could this truly come to be, or truly be, having come to be? (64 B 2-4) Statements like this might encourage one to say that here alêtheia means something other than truth – perhaps genuiness. Without fixed units shaping change, serving as the ends of change, there can be no change. That is change to something that is truly (“really”) what it is. So without beauty itself, no one can grow up to be beautiful. If the ideas are to explain what is the case, then it has to be true that, say, Helen is beautiful.¹⁶ Ideas, true being, e.g. justice itself (cf. 64 A 1) are necessary for anything to come to be just, and then be just at all. So truth, alêtheia, earns itself a place in the human good life, something which can come to acquire and then really, genuinely possess certain predicates. The striving for a good life would be senseless, if no life could genuinely be called just, for example. Truth is then used to pursue goodness to its dwelling, along with beauty and measure. These are the elements that blend to give us goodness.

Nous and its poor relations are true, pure, stable and exact, to varying degrees. These qualities come in degrees. All of this clan of blends, like blends in general, serve purposes – most obviously in the case of technai, but Plato’s willingness to see the technai as epistêmai indicates, as does the general set up of the Maker producing all things, that epistêmai too serve ends, even when most pure, in the case of phronêsis and nous. As Socrates says, their point is not to be useful to us, but to be true, above all. Dialectic is agreed to have the pure parts of nous and phronêsis, which have the love of truth, and the desire to do everything for the sake of truth. And love relates to an end: love moves, and moves the lover towards ends. We have been told (54 A-E) there is a division into ousia, being, and genesis, coming to be, and genesis is for the sake of ousia, for ends: shipbuilding is for the sake of ships. This is my reason for saying that the end of knowledge, in fact, in the pure case, nous and phronêsis, is truth. It looks to us, I think, that knowing, and even more, intellect and wisdom are made up of truth, in part. That is their end, what they strive for, or would do, if they did not necessarily have it in them. Truth is the internal end of all knowledge.

Knowing is not something that happens to us, we are actively involved in this achievement, and it is closely connected to ends and means: knowing can serve as both, and is necessary for us achieving ends at all. And all this achievement of knowing is necessarily connected to the truth. Whatever knowledge is, it is true. It is not accidental that pleasure is classed among means (53 D 3-54 D 7). This means that pleasure is excluded from the good’s portion: it is a thing that pursues, not a thing pursued (54 D 1-2), a coming to be, genesis, not being, ousia. So, it is obvious to place knowledge in the other portion – knowledge is what is pursued. Then we ask: what completes knowledge, makes knowledge knowledge, and the answer is its truth. Then it is genuinely knowledge. And without it, it is not knowledge at all.