One Foggy Window on a Beautiful World


1. Introduction

Writing an essay on Plato feels much like writing reflections on War and Peace in Russian high school. On the one hand, you know that millions of other students have had to do it before you. You are also fairly certain that what your teacher cares about is not your take on Natasha or Napoleon but whether you have read the thing, or at least significant chunks thereof. At the same time, you still secretly hope to say something clever and memorable, something that has not been said much better a thousand times before.

That hope is, of course, futile – perhaps even more so in the case of Plato than in the case of Tolstoy. After all, War and Peace has only been around for a fraction of the time that people have been commenting on Plato’s dialogues (cue Whitehead’s footnotes quote). Even more important in my case, while I was able to read Tolstoy in the original, I have no knowledge of Greek, classical or otherwise.

This is why, before outlining the structure of the essay, I feel that it is necessary to say a few words about the nature of the task I can reasonably set myself.

The question this essay will try to answer is why, according to Plato’s dialogues, aiming at knowledge unifies the soul. My acquaintance with Plato’s texts, let alone the relevant scholarship, is too cursory to justify a stance on the unitary/revisionist debate, or indeed any other difference of scholarly opinion regarding the unity of views expressed in the dialogues. For this reason, I will specify the key question as follows: if I were to hold the view that aiming at knowledge unifies the soul, what support could I find for it in Plato?

In other words, this essay does not claim to offer an opinion on what Plato might or might not plausibly have meant by having Socrates or his interlocutors say certain things. Instead, my goal is simply to engage with Plato’s texts in a way that can stimulate my own philosophical reflection.

Having said that, I do make one important assumption regarding Plato’s authorial intentions: whatever his characters say, I will strive to take it more literally than my 21st-century sensibilities might otherwise lead me to. In other words, I will not treat as mere metaphor or illuminating allegory certain ideas that are hard to square with my background beliefs concerning ontology, physics, language, knowledge, or personhood. By this, I do not wish to imply that Plato himself literally believed everything that his characters say. Rather, this is an attempt to do justice to the intellectual context in which the dialogues were written, as described by e.g. Lloyd (1999). Thus, for the purposes of this essay, anything that, in Classical Greece, would have been a legitimate hypothesis about the way things are ought to be taken at face value.

With those caveats in place, the structure of the essay is fairly straightforward. I will begin by discussing definitions of knowledge, soul and unity of the soul, inspired by the dialogues. I will then attempt to show how, given those definitions and other textual support from Plato, it is possible to argue that aiming at knowledge unifies the soul.

2. What is knowledge?

In defining knowledge, I will first consider what knowledge is not. The Theaetetus is arguably the most famous discussion of what knowledge is not in the history of Western thought, so it makes sense to take my cue there.

Along with Theaetetus, a gifted young man on whom Socrates practises his intellectual midwifery, we are gradually shown that knowledge is “neither perception nor true judgement, nor an account added to true judgement” (Theaetetus 210a-b). In other words, seeing or hearing something is not the same as knowing it. Believing something that happens to be true is not the same as knowing it. Even believing something that happens to be true plus being able to justify your belief, either by reference to the elements of whatever you believe (207a) or by reference to its differentness from everything else (210a), is not the same as knowing. Here is a reason for the last claim: in order to provide an account in terms of elements or differences, you first need to have knowledge of the elements and differences involved, which means that you will also need to provide accounts for those elements and differences by referring to their own elements or differentness, and very soon you will end up with an infinite regress of accounts on your hands.

This does not necessarily mean that, once you do know the nature of a thing, you are not expected to be able to give an account of this nature. For instance, Republic 534b-c suggests that a person does not “know the good itself or any other good” unless she can offer an account of the Form of the good to herself or to others by “distinguishing” it from everything else. A (fragmentary) true belief about the good (i.e. “some image of it”) will not do because it comes “through opinion, not knowledge”. To the best of my understanding, an ability to provide an exhaustive account of the Form of a thing is therefore a necessary consequence rather than a necessary condition of true knowledge.

If knowledge is not true belief with an appropriate account – and the discussion in the Theaetetus strongly indicates that it is not – what could it be? More knowledgeable people might be in a position to reconstruct an alternative reductive account of knowledge from Plato’s texts. To me, however, the opposition of belief and knowledge, a recurrent feature of the dialogues, suggests that reduction might be the wrong way forward. Perhaps knowledge is sui generis, a phenomenon that cannot be given a satisfactory definition of the form “A is B”, except in the most trivial sense, whereby we can describe knowledge as a state or a process – or, since we are talking about Plato, a Form.

According to Amber Carpenter, there are three instances where Plato’s characters talk of a form of knowledge, but none of the relevant passages is an explicit attempt to define knowledge (Carpenter, 2015, pp. 180-181). Thus, as I read Carpenter, there seems to be little evidence in the dialogues to suggest that knowledge is to be understood as a Form alongside the Forms of beauty, justice, the good, and so on. This makes sense if we bear in mind that knowledge can have Forms as its object: there is knowledge of beauty, of justice, and of the good. In fact, the Republic has knowledge come in many sorts: “a particular sort of knowledge is of a particular sort of thing” (438c), such as building houses. It seems then that knowledge should not be treated as a Form, since all the other Forms are of things of which they are Forms, not of something else. Perhaps knowledge is best understood as “an activity of soul” (Carpenter, 2015, p. 181, footnote, emphasis in the original), or else a certain kind of relation between the knower and the object known.

Even if we cannot reduce this activity or relation to something else, we can still meaningfully talk of its properties or ways of bringing it about. I have already noted one property of knowledge: there are different kinds of it. As is so often the case in Plato, these kinds are not born equal; some are superior to others. At the top of the epistemic hierarchy is the knowledge of being as such and of “what is really and forever in every way eternally self-same” (Philebus 58a), which would include the Forms of the beautiful, the good, and other things (Cratylus 439c-d, Phaedo 78d). Any knowledge that is not of “things that are forever in the same state” is “second-rate and inferior” (Philebus 59c).

In fact, so inferior is it that at times Plato appears to doubt that ever-changing sensibles can be known at all (e.g. Republic 529b-c). This is perhaps only to be expected if we recall that sensibles are encountered through the body and its senses. In the Phaedo, Plato has Socrates mince no words about the body’s epistemic potential: any investigation undertaken through the senses makes you “confused and dizzy”, as if you were “drunk” (79c), and instead of gaining wisdom you end up wallowing “in every kind of ignorance” (82e).

Given such disparaging remarks about the human epistemic condition, it is permissible to wonder – at least rhetorically, as Carpenter does (2015, p. 182) – whether humans can hope to know anything at all. If knowledge is not compatible with sense-perception, it is not clear why we should waste so many papyrus rolls discussing it in this life, stranded as we are in this vale of sensibles and particulars. If my goal as a committed Platonist is perfect wisdom, would I not be better off taking a cup of hemlock and escaping the prison of the flesh to “observe things in themselves with the soul by itself” (Phaedo 66e)?

Fortunately, Carpenter (2015) offers a conceptual remedy against the temptation of immediate suicide for epistemic reasons. She argues that “having a robust – even unattainable – conception of knowledge” does not entail unrelenting skepticism about human cognitive abilities (pp. 183, 203). If we adopt a paradigmatic epistemology, as Carpenter believes Plato does in the Philebus, we need not deny the status of knowledge to what we can learn about the ever-changing world of sensibles. Instead, we can talk of a spectrum of knowledge, where different sorts of knowledge instantiate certain paradigmatic properties, such as exact measures, to a greater or lesser degree (pp. 188-189). Even if every single sort of knowledge available to a (living) human being falls short of a perfect instantiation, the set of criteria provided by this unattainable ideal can come in handy both in philosophy and in the most mundane of matters, such as building houses.

This is good news for the core question of this essay. As the following sections will try to show, the soul as the dialogues describe it is made up of both permanent and transient elements. If striving for knowledge is to unify the soul, the concept of knowledge had better apply to the ever-changing realm of particulars.

3. What is the soul?

I did not end the previous section with an explicit definition. This is because, as far as I understand Plato, an adequate conception of knowledge, whether sui generis or not, has to involve the soul. This provides all the more reason to consider the second key term in the main question of this essay.

The first thing to note about the soul is that it exists. I hesitate to use the words “substance” or “dualism” in an essay on Plato, but the soul is clearly an entity that is separate from the body (e.g. Phaedo 79b) – to the extent that it might have existed before the body (e.g. Phaedo 76d-e, Meno 86a), and it may well outlive the body (Phaedo 84b, among numerous other passages). Moreover, whatever the soul is made of, soul stuff is very different from body stuff: it is invisible “to human eyes” (Phaedo 79b).

The soul has its own unique qualities. Most obviously, it can possess, or occupy, bodies (Phaedo 105d, Philebus 64b). In fact, the soul is the thing that, when “present in a body, makes it living” (Phaedo 105c). Other qualities of the soul include “justice, courage, intelligence, memory, munificence”, moderation and an unspecified number of others (Meno 88a).

Crucially, the soul – at least when unencumbered by the body – seems to enjoy privileged access to “the truth about reality” (Meno 86b). It is the soul alone that can know being itself; the sole alone can grasp the nature of “beautiful and ugly, good and bad” (Theaetetus 186a). This is because, being invisible, the soul shares the realm of the intelligible and invisible with the Forms. Thus, in order to have the purest sort of knowledge about reality, the soul needs to “withdraw from the senses” and “gather itself together by itself” (Phaedo 83a), shutting out everything sensible and visible. As we saw in the previous section, this might be a tall order for a soul dwelling in a body, i.e. for a human being, but “the power of dialectic” (Philebus 57e) can educate the soul about the deceitful nature of the senses and in this way gently nudge it towards freedom even before it sheds its mortal shell (Phaedo 83a). In fact, without such nudging, the soul cannot escape (here I am really tempted to say “Samsara”) the realm of the sensible even when the body is gone. “Riveted” to the flesh by the nail of unreflected pleasure and pain, the soul will pine for the sensible world. As a result, it will “fall back into another body” to live and suffer all over again (Phaedo 83d), and so it will go on – presumably until the wretched soul is lucky enough to be reborn as a philosophically inclined slave-owning male with plenty of leisure time and like-minded friends.

As an upshot of all this, it seems clear that a robust incorporeal soul is an integral part of a Plato-based conception of knowledge. Whether we describe knowledge as an activity, a state or a relation, it has to be something that links such a soul and the object of knowledge. Ideally, this link comes about without any interference from the body, but even in inferior sorts of knowledge, allowed on a paradigmatic epistemology discussed earlier, there will be an element of what the soul “considers alone and through itself”, such as the notions of “like and unlike” or “same and different” (Theaetetus 185e-186a).

However, there is a slight problem with the notion of the soul presented so far (and largely derived from the Phaedo): it seems to make the key question of this essay either unnecessary or at best poorly worded. How can aiming at knowledge unify the soul if the soul, indestructible as it apparently is, does not seem to need much unification in the first place? Perhaps, as Phaedo 83 suggests, it makes more sense to speak of the soul cleansing itself or extricating itself from the realm of the sensible by aiming at knowledge. While the soul is chained to the body and constantly yanked at by fears and desires, it might seem to it that it lacks unity, but once that illusion has been shattered by the might of dialectic, the liberated soul will realise that it has been unified all along.

Put differently, soul-unification only becomes a genuine problem if the soul consists of distinct parts that are not necessarily in sync with each other. Is there evidence in the dialogues that the soul has parts which can lack unity? That is the question of the next section.

4. What does it mean for the soul to be unified?

It is hard to think of a clearer metaphor for internal conflict than that of a civil war in the soul. Luckily for the topic of this essay, that is precisely the analogy we find at Republic 439-441. Not only are we told here that the soul has three distinct parts; we also learn that they can be at loggerheads with each other.

Since the universe of the dialogues is thoroughly hierarchical, it stands to reason (no pun intended) to begin an inventory of the soul’s components with the rational part, which is superior to the other two. It is the rational part that can calculate (Republic 439d) and is therefore capable of measure, which is “a property of the first rank”, according to Philebus 66a. In fact, at least three out of the top five ranks of properties of knowledge listed at Philebus 66a-c appear to be directly associated with the rational part of the soul. Only ranks 4 and 5, which are said to include, respectively, “the soul’s own properties” and “the soul’s own pure pleasures”, seem to allow a role for the parts of the soul that are not strictly concerned with dialectic, mathematics, or the never-changing Forms of things.

The lowest rung in the tripartite structure of the soul belongs to the appetitive part. It is explicitly described as an “irrational” “companion of certain indulgences and pleasures” (Republic 439d) and is thus placed in direct opposition to the rational part. As the name suggests, instead of doing mathematics or contemplating the Form of the good, the appetitive part “lusts, hungers, thirsts, and gets excited by other appetites” (439d). Presumably, it is the appetitive part which, enamoured of fleeting earthly sensations, acts as an ally of the body and drags the entire soul down, making it “full of body” (Phaedo 83d) and chaining it to the sensible world, while the rational part yearns for the perfect realm of the intelligible. This, in a nutshell, is the reason for the “civil war” (Republic 440b, 440e) that can sometimes rage in the soul.

The good news for the rational part is that in this conflict the third component of the soul, its spirited part, is said to side with reason (440e), unless it has been “corrupted by a bad upbringing” (441a). The spirited part is the one “by which we get angry” (439e). When properly aligned with the rational part, it is triggered by injustice and, crucially, by any attempt of the appetitive part to overthrow the legitimate rule of the rational part.

Civil war is a bad thing, clearly to be avoided. In other words, Republic 439-441 seems to imply that unity of the soul is indeed a pressing concern. For a soul to function properly, its three parts have to act as a well-disciplined unit, with the spirited and appetitive parts fully at the beck and call of the rational part. This need for unity is echoed in the Philebus, where a key goal of the investigation is to sketch “the design of … an incorporeal order that rules harmoniously over a body possessed by a soul” (64b).

Where there is a need for unity, there is a potential role for knowledge in achieving it. However, before discussing that role, I would like to argue that the need for unity of the soul may be even more acute than suggested by the passages discussed above. As I understand Republic 439-441, its primary concern is with the state of the soul at a given time. Needless to say, synchronic unity of the soul is important, but as a fascinating passage in the Symposium suggests, we might also need something to make sure that the soul stays unified over time.

At Symposium 207-208, in a rare case of Plato having a woman speak, Diotima of Mantinea tells Socrates that a person does not consist of the same things throughout her lifetime even if we always use the same name to refer to the person in question. All bodily parts of a living human being, such as her “hair and flesh and bones and blood” (207e), are constantly renewed. However, it is not just the body that changes; the soul does not stay the same either. Diotima lists a number of features of the soul that can come and go, namely “manners, customs, opinions, desires, pleasures, pains or fears” (207e). Crucially, it transpires that even our knowledge does not stay the same (208a). Not only does the soul acquire new knowledge; it is said to forget knowledge it already has. The whole reason why we need to study is to replace the knowledge we lose with fresh memories, “so that it seems to be the same”. “Seems” is a key word here: only the divine actually stays “the same in every way”, while all things mortal – apparently including some parts of the soul – only imitate sameness by leaving behind something that is similar to them (208a-b).

Most of the soul’s impermanent features listed by Diotima can be readily identified with the two non-rational parts of the soul, as discussed in the Republic. To be sure, opinions might be in the domain of the rational part, but it is possible to argue that the rational part only develops opinions in the first place because it is affected by its irrational counterparts and by the sensible world in general. However, the remark that the soul changes with regard to knowledge does complicate matters if, as suggested above, we want to understand the soul’s unity in terms of the rational part ruling over fully compliant non-rational parts. How can such an arrangement ensure genuine unity if the rational part itself does not seem to stay self-same throughout a person’s lifetime? If the rational part has a particular piece of knowledge today but loses it tomorrow, the legitimacy of its rule appears to be compromised.

Given these concerns, the final section of the essay will look at how aiming at knowledge can unity the soul both at a time and over time.

5. How can knowledge unify the soul?

In his article on the soul’s unity in the Republic, Eric Brown (2012) makes a distinction between unearned and earned unity. As Brown reads the Republic, the soul must enjoy a degree of unearned unity simply in virtue of staying alive because the basic activities necessary for the survival of a human being require at least some coordination among the different parts of the soul. In other words, as long as the tripartite soul wants to pursue any common goal at all – be it to eat some olives or to procure a donkey, or to lie down with a handsome lad on the cusp of adulthood – it has no choice but to act as a unit.

Arguably, even such basic unity requires some knowledge of the lower sorts, i.e. of the sorts that only partially instantiate the paradigmatic features of knowledge. The appetitive part may desire olives or handsome lads all it wants, but it will not get very far unless the rational part lends it a hand in finding a course of action that leads to a meal or a sexual encounter. Any successful course of action will require notions such as like and unlike or same and different – in other words, it will require some degree of measure, which, as we are told at Philebus 66a, is knowledge par excellence. However, the primitive grasp of intelligible reality sufficient for lowly pursuits pales in comparison with the knowledge it takes to achieve what Brown calls earned unity of the soul.

In contrast to unearned unity, earned unity comes about as a result of striving for virtue and excellence (Brown, 2012). For a soul to be fully unified, the rational part must acquire “the knowledge of what is advantageous for each part and for the whole soul” (Republic 442c). The tighter the rational part’s link to the Forms, i.e. the objects of the highest sort of knowledge, the less vulnerable it is to the influences of the appetitive part and the sensible world, and the firmer its rule over the trinity. As Socrates puts it at Meno 88c, if “virtue is something in the soul and it must be beneficial, it must be knowledge”. When all chips are down, virtue is, quite simply, “a kind of wisdom” (Meno 88d). Therefore, aiming at knowledge and aiming at perfect unity of the soul are essentially identical pursuits. Any activity that contributes to knowledge, such as studying astronomy or engaging in dialectic, necessarily contributes to the soul’s unity.

This would be a nice QED moment were it not for Diotima’s inconvenient reminder that the soul, at least while it occupies a body, can and does lose at least some of the knowledge that it acquires. Sure, the soul may have enough knowledge to qualify as unified today, but what if it forgets some of that knowledge tomorrow? In other words, what is to guarantee the soul’s hard-earned unity over time?

As I see it, there is one reply that immediately presents itself. It is suggested by the very same passage in the Symposium: if the soul is prone to forgetfulness, then we need an insurance policy against loss of knowledge, and that policy is studying (208a). Through constant studying, the soul continually renews its access to intelligible reality and in this way remains unified. This conception of unity over time becomes all the more plausible if, as discussed above, we define knowledge as a Forms-directed activity of the soul, or perhaps as a relation to the Forms that needs constant maintenance by means of philosophy, at least while the soul is tied to a body. Perhaps the soul’s unity over time is like looking at a breathtakingly beautiful view from a tiny room through a large window that keeps clouding up. We try hard to wipe off the fog as fast as we can because we want to see the entire view at once, but no sooner have we cleared a big chunk of the window pane in one place than it starts misting up again somewhere else, and so it goes on and on because, even though we may not realise it, it is our own breath that fills the tiny room and keeps fogging the window, hiding the beautiful view outside.

Now that the essay question has been formally answered and the required length reached, I ought to write a conclusion retracing my steps section by section, pointing out weak links in my arguments and noting areas that would have particularly benefited from a more intimate acquaintance on my part with Plato’s surviving body of work as well as the secondary literature. However, in Plato’s universe unity is virtue, and I will use unity of form and topic as an excuse to write no such conclusion just this once.

The bane of Plato commentators (especially those writing lowly course papers reminiscent of high school compositions on War and Peace) is that, while the master himself got away with churning out wonderfully meandering plays that can start and end in mediis rebus and range from courtroom drama to bawdy homoerotic bromance featuring lots of wine, analysis of the master’s work usually has to take itself very seriously and conform to the conventions of modern Western academic writing. I teach those conventions for a living, and far be it from me to question them. But just this once, I feel that the essay should end abruptly, on a semblance of a conclusion, as the interlocutors promise each other to keep thinking about the matter. Then they say goodbye and continue on their way to watch the festival procession in honour of the goddess, who may or may not exist.



Brown, E. (2012). The unity of the soul in Plato’s Republic. In C. Brittain & R. Barney & T. Brennan (Eds.), Plato and the Divided Self (pp. 53-74). Cambridge: Cambridge University Press.

Carpenter, A. (2015). Ranking knowledge in the Philebus. Phronesis, 60, 188-205.

Lloyd, G. (1999). Magic, reason and experience: Studies in the origins and development of Greek science. Indianapolis, Ind: Hackett Publishing.

Plato, Cooper, J. M., & Hutchinson, D. S. (1997). Complete works. Indianapolis, Ind: Hackett Publishing.


Illustration: Das Gastmahl by Anselm Feuerbach