II. INTRODUCTION TO THE MISPERCEPTION ARGUMENT

The misperception argument begins with Socrates' claim that ‘in contrast to the pleasure of the rational man, the pleasure of others is neither true at all nor pure, but in a way shadow-painted’ (583b3–5). Socrates does not here identify these others more specifically than by contrasting them with the rational man. In the true filling argument, however, he refers to them as ‘irrational men’ (586c3). Accordingly, I will refer to them either as ‘irrational men’ or in the generic singular as ‘the irrational man’.

In the true filling argument Socrates contrasts philosophers with men whose appetitive and spirited souls govern their lives. Thus, it is natural to assume that the philosophers of the true filling argument are identical to the rational man of the misperception argument and that the appetitive and spirited men of the true filling argument are identical to the irrational men of the misperception argument. Strictly speaking, this must be true. However, when Socrates speaks of the irrational man in the misperception argument, he principally has somatic pleasure-seekers in mind, that is, men whose appetites govern their lives.

Observe that Socrates speaks of irrational men's pleasure as not at all true, impure and in a way shadow-painted: ‘the pleasure of others is … not at all true’. Contrast this with Socrates' claim, later in the argument, that irrational men do not have healthy opinions about pleasure: ‘Should you wonder, then, if those inexperienced in truth do not have healthy opinions regarding many other things as well’ (584e7–8). The belief of the irrational man may be representationally and specifically semantically false – so we would characterize it – but it is questionable whether the untruth of the irrational man's pleasure itself is to be explained in this way. Relevant to this question is the question of how being impure, untrueFootnote ² and in a way shadow-painted relate to one another. We will answer these questions in the following sections.

III. THE IRRATIONAL MAN'S MISPERCEPTION OF CALM AS PLEASANT

At 583c3–6 Socrates introduces a distinction between pain, pleasure and a condition of neither having pleasure nor being pained:

1. (1) Pain is the opposite of pleasure (583c3–4).

2. (2) There is a condition of neither having pleasure nor being pained (583c5–6).

Socrates then clarifies the nature of the condition of neither having pleasure nor being pained and its relation to pleasure and pain:

1. (3) < the condition of neither having pleasure nor being pained, > Footnote ³ which is in the middle between both of these, <namely, pleasure and pain,> is a certain calm of the soul in relation to them <again, pleasure and pain> (583c7–9).

It is unclear what Socrates means by claiming that the condition of neither having pleasure nor being pained is ‘in the middle between both of these’, namely, pain and pleasure. He might mean that during a process in which there is a transition from pain to pleasure or vice versa, the condition of neither having pleasure nor being pained intervenes. Alternatively, he might mean that on some dimension, the condition of neither having pleasure nor being pained relates to the poles of pleasure and pain as a midpoint. In that case, we would like to know what dimension Socrates has in mind. In the absence of further evidence, I propose temporarily to leave the meaning of this aspect of (3) undetermined.

In (3) Socrates also refers to the condition of neither having pleasure nor being pained as a certain ἡσυχία ‘in relation to’ pain and pleasure. I take this to entail that pleasure and pain are both conditions of absence of calm. Hereafter I will refer to the condition of neither having pleasure nor being pained as ‘calm’. Precisely how we should understand absence of calm is questionable. For example, perhaps Socrates understands pleasure and pain to be conditions of agitation. I return to the problem below.

Socrates also indicates that calm is psychic: it is a condition ‘of the soul’. This is significant for interpreting Socrates' conception of pain and pleasure. It would be odd for Socrates to conceive of calm as psychic, but not to conceive of pain and pleasure as involving the soul. Accordingly, we should assume that Socrates conceives of all three conditions as involving the soul. This in turns raises the question of how we should understand the condition of calm as involving the soul. Is Socrates referring to what the experience of calm is like? Or is Socrates referring to an objective condition of the psyche: the psyche's being at rest? At this point in the argument no evidence indicates which interpretation to prefer or whether Socrates would distinguish between the two. I will return to this question below.

The first step of the misperception argument thus begins with these basic distinctions between pain, pleasure and calm. At once, however, Socrates draws attention to events that appear to undermine these distinctions:

1. (4) Sick men say that being healthy is most pleasant (583c10–d1).

2. (5) In general, men who are in the grip of great pain say that nothing is more pleasant than the cessation of being in pain (583d3–5).

In both (4) and (5), men who are in pain claim that the states of health and cessation of pain are pleasant. It is clear from the ensuing argument that Socrates himself views the states of health and cessation of pain as states of calm rather than pleasure. Thus, Socrates understands pained men as claiming that calm is most pleasant. He continues:

1. (6) And in many other circumstances, men who are in pain praise not being in pain and the calm following pain as most pleasant, rather than <praising> having pleasure <as most pleasant> (583d6–9).

The claim of pained men that cessation of pain is even more pleasant than having pleasure is ostensibly self-contradictory. But I take their claim to mean that cessation of pain is more pleasant than familiar sorts of pleasure, for example, pleasures associated with eating, drinking and sex. Pained men's praise of cessation of pain challenges Socrates' basic distinctions, and the ostensible conflict between these positions invites resolution. Glaucon offers the following:

1. (7) Perhaps on this occasion, <namely, when calm follows pain,> calm becomes pleasant and beloved (583d10–11).

Glaucon's response assumes that the claims of pained men are true, that is, that calm following pain is indeed pleasant. This implies that Socrates' basic distinctions are false.

At 583e1–8, Socrates responds with an argument whose purpose is to show that Glaucon's explanation in (7) is untenable. I will refer to Socrates' argument as ‘the puzzling argument’. The puzzling argument is a reductio. It begins by assuming the following premise drawn from the content of the preceding premise-set:

1. (8) Whenever a man ceases being pained, the calm following the pain will be pleasant.

Given this, by parity of reasoning Socrates suggests that:

1. (9) Whenever a man ceases having pleasure, the calm following the pleasure will be painful (583e1–3).

From (8) and (9), Socrates infers:

1. (10) That which is between both pleasure and pain will at some point be both pleasure and pain (583e4–6).

Yet he insists:

1. (11) It is impossible for that which is neither <pleasure nor pain> to become both <pleasure and pain> (583e4–8).

Thus, Glaucon's explanation at (7) is untenable. More precisely, Socrates maintains that (8) and (9) are false.

Socrates' puzzling argument is puzzling in at least two respects.Footnote ⁴ First, it is puzzling how Socrates can derive (10) from the conjunction of (8) and (9).Footnote ⁵ (8) and (9) claim that under certain conditions calm is ‘pleasant’ or ‘painful’, but in (10) Socrates infers that under certain conditions calm is ‘pleasure’ or ‘pain’. Thus, Socrates appears to conflateFootnote ⁶ predication and identity.Footnote ⁷ Second, it is puzzling what justifies Socrates' assertion of (11), namely, that it is impossible for what is neither pleasure nor pain to become both pleasure and pain. Is Socrates committed to either one of the following principles?

1. (P1) It is impossible for not-F to become F.

2. (P2) It is impossible for what is neither F nor the opposite of F to become both F and the opposite of F.

Let's consider (P1) and (P2) in turn. (P1) is reasonable in so far as ‘F’ is taken to refer to a universal or type, conceived realistically: universals or types so conceived do not change. Thus, (11) does follow from (P1). However, it then becomes difficult to see what function (9) and (10) serve in the argument. Given (P1), (8) suffices to undermine Glaucon's explanation. In other words, Socrates could merely argue that it is impossible for what is not pleasure, namely calm, to become pleasure. Consequently, it seems that if Socrates' assertion of (11) depends upon either (P1) or (P2), it depends upon (P2). On the other hand, it is hard to see what substantive work (P2) does beyond (P1). (P1) is all that is needed for (11). This is simply because (P2) is a conjunction of two interpretations of (P1), one on which ‘F’ stands for one of a pair of opposites (for example, pleasure) and the other on which ‘F’ stands for the other member of the pair (for example, pain).

Given this, our second puzzle turns out to consist of two puzzles that are closely related. One is what justifies Socrates' assertion of (11). The other is why Socrates argues against Glaucon's explanation in (7) using the conjunction of (8) and (9), as opposed to using merely (8).

Let us return to the problem of the conflation of predication and identity. What if we introduce indefinite articles into (10), thus rendering (10) as:

1. (10r) That which is between both pleasure and pain will at some point be both a pleasure and a pain.

The inference of (10r) from (8) and (9) is now more reasonable: in certain contexts a given condition instantiates pleasure and thus is pleasant; in certain other contexts that condition instantiates pain and thus is painful. Given this, we should also emend (11) by adding indefinite articles:

1. (11r) It is impossible for that which is neither pleasure nor pain to become both a pleasure and a pain.

This solves the first puzzle.Footnote ⁸

Let us turn to the second and third puzzles. I grant that (11r) may depend upon some more general principle, but before attempting to determine what that principle is, let us attend to the subject of (10r). Socrates describes the subject precisely as follows: ‘What we just now said was between both <pleasure and pain >, calm’ (583e4–5). That is, Socrates uses the word ‘calm’, but he also reminds Glaucon that calm is between pleasure and pain. Recall Socrates' emphatic statement of this in (3): ‘< the condition of neither having pleasure nor being pained,> which is in the middle between both of these, <again, pleasure and pain >’. I suggest that Socrates' reminder of this point in (10r) is important. Clearly, there are many things that are neither F nor the opposite of F that can instantiate F and the opposite of F in different circumstances. For example, the act-type of giving a person money, which is not justice or injustice, can be just in certain circumstances and unjust in other circumstances. However, giving a person money is not – I presume – ‘in the middle between’ justice and injustice in the sense in which Socrates intends this phrase here.Footnote ⁹ Yet in (3) Socrates emphatically states that calm is in the middle between pleasure and pain.

As we noted above, it is unclear precisely what Socrates means when he says that calm is in the middle between pleasure and pain. Assume, following the second of the two possible interpretations suggested above, that calm is in the middle between pleasure and pain in the sense that all three essentially stand in certain relations in a single dimension, and specifically that calm is situated between the poles of pleasure and pain in this dimension. In that case, what is in the middle cannot be at either pole because the relations in which the entities stand in this dimension are fixed. In other words, calm is necessarily between pleasure and pain in this dimension.Footnote ¹⁰

My talk of pleasure, pain and calm situated and essentially related in a dimension is informed by Socrates' own talk in the ensuing broader misperception argument. At 584d1–585a7 Socrates uses the analogy of spatial positions to characterize the distinction and relations among pleasure, pain and calm. On a vertical line segment, the top point is analogous to pleasure, the midpoint is analogous to calm and the bottom point is analogous to pain. Socrates does not clarify the nature of the dimension in which pleasure, pain and calm are situated. For convenience, we may offer Socrates the following description: pleasure, pain and calm are situated in the dimension of affectivity.

Granting this, let us return to the suggestion that the puzzling argument has no logical need for (9) and (10) or (10r). On that view, the puzzling argument needs only the following weaker version of (11r):

1. (11w) It is impossible for what is not pleasure to become a pleasure.

However, this is not entirely true. Just as it is possible for the act-type of giving a person money to instantiate justice or injustice, it is possible for some entity E that is not pleasure to become a pleasure, so long as E is not essentially situated within the dimension of affectivity. Given this, Socrates could indeed dispense with (9) and (10) or (10r), but only so long as he emended (11) or (11r) to something stronger than (11w) such as:

1. (11s) It is impossible for what is not pleasure, but is essentially situated at a fixed point in the dimension of affectivity, to become pleasure.

But given this, it should now be easier to appreciate why Socrates does in fact construct the puzzling argument using (9) and (10) or (10r) as well as (11) or (11r): the clause ‘but is essentially situated at a fixed point in the dimension of affectivity’ is doing a lot of work. My suggestion is that (9)–(11) or (9)–(11r) are doing similar, if less explicitly sophisticated, work. Precisely, the use of (9)–(11) or (9)–(11r) serves to emphasize Socrates' point that calm is essentially situated in the same dimension as pain and pleasure and that the relations between these entities within this dimension are fixed.

This, then, resolves the closely interrelated second and third puzzles of the puzzling argument. Given this, let us note the following consequence of the preceding resolution. Two paths are potentially available to advance the inquiry. One path, open to Glaucon, would be to defend the pained man by contesting Socrates' basic distinctions. In particular, Glaucon could demand justification of Socrates' initial assertions that pleasure, pain and calm are essentially situated in a set of fixed relations in a single dimension. The other path, which Socrates has some obligation to pursue, is to explain why pained men tend to make the false claims they do. As the broader argument in fact ensues, Glaucon concedes the conclusion of the puzzling argument and thus the soundness of Socrates' basic distinctions, and Socrates proceeds to explain the confusion of the pained, irrational man.

IV. SOCRATES' EXPLANATION OF THE IRRATIONAL MAN'S MISPERCEPTION

Socrates' explanation of the misperception of the irrational man begins with the following three claims:

1. (12) That which becomes pleasant in the soul and that which becomes painful <in the soul> are both a sort of motion (583e9–11).

2. (13) Yet that which is neither pleasant nor painful is calm and in the middle of both of these (584a1–3).

3. (14) Therefore, it is incorrect to think, <as the irrational man does,> that not being in pain is pleasant and that not having pleasure is painful (584a4–6).

(12)–(14) have both backward- and forward-looking functions. (12)–(14) look backward to the preceding stretch of argumentation, (1)–(11), and now in conjointly intrinsic and objective terms specify how calm differs from pleasure and pain: pleasure and pain are kinetic conditions, whereas calm, as the name indicates, is an absence of motion.Footnote ^11,Footnote ¹² (14) thus corroborates the conclusion of the puzzling argument that the irrational man misperceives and that calm does not become pleasant or painful.

It remains to explain the source of the irrational man's misperception. In contributing to this explanation, (12)–(14) look forward. Central to Socrates' explanation is a distinction between the appearance of the affective condition, which is a subjective component, and the objective component, which, as (12)–(13) state, is a kinetic or a-kinetic condition. Socrates continues:

1. (15) <It is not that calm> is <pleasant or painful >, but that calm appears (φαίνεται) pleasant beside that which is painful and that calm appears painful beside that which is pleasant (584a7–8).

In other words, Socrates explains the irrational man's misperception as based on a misleading appearance. Socrates explicitly characterizes the misleading appearance as illusory: ‘with respect to the truth of pleasure, there is nothing healthy in these phantasms, but a certain magic (γοητεία)’ (584a9–11). Let us say, then, that the irrational man is subject to a hedonic illusion and thus experiences a quasi-pleasure.

In the true filling argument Socrates refers back to the irrational man's quasi-pleasures as ‘shadow-painted and tainted by juxtaposition’ (586b8–c1). The invention of the technique of shadow-painting is attributed to Apollodorus of Athens in the fifth century. Apollodorus seems to have transmitted it to Zeuxis, among other students, who apparently composed a treatise on painting technique.Footnote ¹³ Shadow-painting involves the juxtaposition of darker and lighter shades on a two-dimensional surface to create the illusion of depth.Footnote ¹⁴ For example, Pliny the Elder relates the story that Zeuxis painted a bunch of grapes so realistic a bird pecked at it (HN 35.36). Analogously, in Republic 9 Socrates suggests that the juxtaposition of pain and absence of pain engenders a hedonic illusion.

How should we understand the juxtaposition of pain and absence of pain that engenders the hedonic illusion? When at (4)–(6) Socrates reports the views of the irrational, he speaks of what sick people say when they are sick, of what men in the grip of great pain say, and of the praise that men in pain heap on not being in pain. In each case, the irrational speak while in pain. Thus, the irrational are describing as pleasant not a state they are currently in, but a future state. Thus, the future state presently appears to them as pleasant. So the irrational are imagining or envisioning the future state, and this image is of something pleasant.

Contrast this with Glaucon's explanation at (7), which assumes that calm becomes pleasant. In this case, Glaucon speaks not in terms of someone envisioning the future, but of what the future state will actually be like. Similarly, at (9) in the puzzling argument, Socrates says that when one who is having pleasure ceases, the calm following the pleasure will be painful. Again, he speaks not in terms of someone envisioning the future, but of what the future state will actually be like. Similarly in (10), Socrates says that the calm will be both pleasant and painful. Now of course Socrates rejects Glaucon's explanation at (7) and thinks that (8)–(10) are false. But Socrates' rejection of (7)–(10), which entails rejection of the view that calm actually becomes pleasant or painful, is compatible with the view that calm, when it occurs, appears pleasant or painful. This possibility is corroborated by Socrates' analogy between affective conditions and spatial positions at 584d1–585a7. Socrates says:

In this case, the mover is not predicting where he will stand as he moves from the bottom. Rather, he believes he is at the top when he is standing at the midpoint.

Analogously, as we have seen, Socrates claims:

1. (16) ‘Whenever <people inexperienced in pleasure> are conveyed from pain to the middle state, they strongly believe they are in a state of filling and pleasure’ (585a2–3).

Observe that the hedonically inexperienced are characterized here as being ‘conveyed’ to the middle state, not merely predicting what the middle state will be like.

Further consideration of Socrates' view of how the juxtaposition of pain and absence of pain produces a hedonic appearance can be derived from Socrates' contrast between pure pleasures and quasi-pleasures. At 584b1–c3 Socrates introduces what he calls ‘pure’ pleasures:

Socrates then contrasts pure pleasure with two kinds of quasi-pleasure. First, at 584c4–7, he says:

1. (17) Yet by far the most numerous and greatest of the so-called pleasures that extend through the body to the soul are of this kind, releases from pains.

The phrase διὰ τοῦ σώματος ἐπὶ τὴν ψυχήν indicates that Socrates is here referring to quasi-pleasures that are bodily. Pure olfactory pleasures are bodily. So Socrates first contrasts pure bodily pleasures with bodily quasi-pleasures. Socrates' use of ‘greatest’ in (17) suggests a quality of hedonic experience. Compare Socrates' use of μέγεθος at 584b7: Socrates explains that olfactory pleasures, which are not preceded by pain, ‘suddenly become especially great’. In (17) the quality of the hedonic experience appears to be coterminous with the release from pain, not a prediction or envisioning of how the release will be.

Socrates then compares bodily with psychic quasi-pleasures:

1. (18) Anticipatory pleasures and pains that arise in expectation of and prior to those <states> to come are also the same <as the quasi-pleasures of (17)> (584c9–11).

How are anticipatory pleasures or rather anticipatory quasi-pleasures the same as bodily quasi-pleasures? I suggest that Socrates conceives of these quasi-pleasures as arising from an occurrent state of pain. One imagines or envisions a future state in which there is a cessation of the pain; the future state appears pleasant; and one takes pleasure in this image or vision.Footnote ¹⁵

In sum, Socrates variously conceives of the juxtaposition of pain and absence of pain as productive of a hedonic appearance. In (4)–(6) those in pain envision the cessation of pain as pleasant. In (18) the envisioned cessation of pain as pleasant is enjoyed in anticipation. In (8)–(10), (16) and (17) calm preceded by pain produces a hedonic appearance. Thus, when in (15) Socrates claims that ‘calm appears pleasant beside that which is painful’, he can be charitably interpreted to mean both that calm actually following pain appears pleasant and that cessation of pain envisioned as following pain when one is in pain appears pleasant. The former exemplifies an experiential illusion; the latter exemplifies a misleading vision or, so to speak, mis-imagination.

There is more to be said about Socrates' explanation of the misperception of the irrational man, in particular, about Socrates' conception of the hedonic appearance. But before we discuss this, let us turn to the contents of the true filling argument.

V. THE TRUE FILLING ARGUMENT

In the misperception argument Socrates hints at his view that pleasure requires πλήρωσις. Specifically in (16), in the analogy of spatial positions, he says that the inexperienced strongly believe they are in a state of ‘filling and pleasure’. In the true filling argument, Socrates makes this point more explicitly. More precisely, he suggests that pleasure requires a filling with what is ‘naturally appropriate’.Footnote ¹⁶ Moreover, Socrates maintains that there are three species of naturally appropriate filling related to the appetitive, spirited and rational parts of the soul respectively. Socrates argues that the fillings related to the appetitive and spirited parts of the soul are less true than rational filling. Since a truer filling is constitutive of a truer pleasure, Socrates concludes that rational pleasure, which is equivalent to philosophical pleasure, is the truest pleasure.

Socrates begins the true filling argument by clarifying two species of inanition and correlative filling:

1. (19) Hunger, thirst and the like are inanitions of the body's state (585a8–b2).

2. (20) Ignorance and lack of intelligence are inanitions of the state of the soul (585b3–5).

3. (21) He who partakes of nourishment, such as food, drink and relish, is filled (585b6–8, with 585b13).

4. (22) He who possesses true judgement, knowledge and understanding is filled (585b6–8, with 585b14–c1).

Socrates ultimately – premise (30) below – clarifies that being filled with what is naturally appropriate is pleasant. Accordingly, in view of (19) and (21), we will assume that nourishment is the naturally appropriate content for the body. Regarding the naturally appropriate content of the soul, although Socrates initially speaks of true belief, knowledge and understanding (585b14–c1), he otherwise focusses on knowledge. So, for the sake of simplicity, I will assume that knowledge is the naturally appropriate content of the soul.

Socrates now suggests that some fillings are truer than others:

1. (23) Truer filling fills with ‘what is more’ (585b9–11).

The concept of what is more in (23) is clarified by the contents of subsequent premises, in particular (24). Socrates suggests that what is more is more immutable. Thus, (23) suggests that a true filling is a filling whose contents are relatively immutable. In fact, Socrates subsequently claims, in premise (29) below, that truth of filling is a function of containers as well as contents. In that case, we may take (23) as an abbreviated and preliminary formulation. (23), like (29), is a crucial and faulty premise of the true filling argument. It requires further discussion, which I offer below.

Socrates now claims:

1. (24) That which is connected to that which is always the same and immortal and to truth and is itself of such a kind and comes to be in such a kind has more being than that which is connected to that which is never the same and mortal and is itself of such a kind and comes to be in such a kind (585c1–6).

(24) is obviously complex. One fundamental question for the interpretation of (24) is whether reference is being made to four or rather six entities. Consider the following two interpretations, in which I have added symbols to clarify the various referents:

1. (24₄) That (x1) which is connected to that (x2) which is always the same and immortal and to truth and is itself (x1) of such a kind and comes to be in such a kind (x2) has more being than that (y1) which is connected to that (y2) which is never the same and mortal and is itself (y1) of such a kind and comes to be in such a kind (y2).

2. (24₆) That (x1) which is connected to that (x2) which is always the same and immortal and to truth and is itself (x1) of such a kind and comes to be in such a kind (x3) has more being than that (y1) which is connected to that (y2) which is never the same and mortal and is itself (y1) of such a kind and comes to be in such a kind (y3).

I suggest that (24) refers to six entities; thus, (24₆) is the correct interpretation. One reason for preferring (24₆) to (24₄) is that the distinct phrases ‘being connected to’ and ‘becoming in’ suggest that what x1 or y1 is connected to and becomes in are two different things. Further considerations below will confirm that (24₆) is the correct interpretation.

Assuming that (24₆) is the correct interpretation of (24), the core claim that (24) makes is:

1. (24a) x1 has more being than y1.

But (24) also makes the following claims:

1. (24b) x1 is connected to x2.

2. (24c) x2 is always the same and immortal, and is truth.

3. (24d) x1 is always the same and immortal, and is truth.

4. (24e) x3 is always the same and immortal and truth.Footnote ¹⁷

5. (24f) x1 comes to be in x3.

And similarly:

1. (24g) y1 is connected to y2.

2. (24h) y2 is never the same and is mortal.

3. (24i) y1 is never the same and is mortal.

4. (24j) y3 is never the same and is mortal.

5. (24k) y1 comes to be in y3.

I suggest that the core claim, (24a), depends upon the following ontological principles:

1. (O1) That which is is true.

2. (O2) That which is is immutable.

(O1)'s identification of truth and being is one among several conceptions of what I will refer to as ‘ontological truth’ that occur in the true pleasure argument. I will say more about this and the other conceptions of ontological truth below.

Note also that (24h)–(24j) are not precisely opposites of (24c)–(24e). Socrates does not say that y1, y2 and y3 are untruth. Since untruth entails non-being, that would mean that y1, y2 and y3 were non-existent. Instead, y1, y2 and y3 have an intermediate ontological status.

Now, (24) itself is a general claim. In the following premises, (25)–(28), Socrates specifically contrasts the containers and contents involved in nutritional filling with those involved in epistemic filling. In doing so he is in effect applying (24) to these specific entities. To facilitate understanding of premises (25)–(28) and their function in the argument, it may be helpful to bear in mind the following application of (24):

Knowledge, which is connected to beings, which are always the same and immortal and are truth, and which is itself of such a kind and comes to be in the soul, which is of this kind, has more being than nutriment, which is connected to becomings, which are never the same and mortal, and which is itself of such a kind and comes to be in the body, which is of this kind.

Consider now premises (25)–(28):

1. (25) The being (οὐσία) of what is always the same participates in being (οὐσία) as much as the being of knowledge (585c7–9).

Socrates does not explicitly mention Forms here, but I assume he has these entities in mind. Given this, Socrates' point is the following: that which is immutable participates in being to a given extent; knowledge is of what is immutable; therefore, knowledge participates in being to that extent.Footnote ¹⁸

1. (26) The being of what is always the same participates in truth (ἀληθείας) as much as the being of knowledge (585c10–11).

(26) follows from the ontological principle (O1). Admittedly, the fact that Socrates claims (26) as well as (25) might be taken to suggest that he does not identify being and truth, but rather that he takes truth to be a property of being. However, I can see no further reason to prefer this thesis. I will assume that the function of (26) is to clarify that since knowledge participates in being as much as its objects do, knowledge participates in truth as much as its objects do.Footnote ¹⁹

1. (27) The kinds of thing related to somatic treatment participate less in truth and being than the kinds of thing related to psychic treatment (585d1–4).

I assume that ‘the kinds of thing related to somatic treatment’ in (27) refers to becomings and nutriment, the correlates of being and knowledge in (25) and (26).

1. (28) The body participates less in truth and being than the soul (585d5–6).

(28) thus completes the specification of the three components – x1, x2, and x3, and y1, y2, and y3 – involved in (24) interpreted as (24₆).

On the basis of the application of (24) to nutritional and epistemic filling entailed by premises (25)–(28), Socrates now expands his point regarding more and less true fillings at (23):

1. (29) That which is filled with things that are more and which itself is more is more filled than that which is filled with things that are less and which itself is less (585d7–10).Footnote ²⁰

Whereas in (23) Socrates correlates truth of filling with truth of contents, here in (29) Socrates correlates truth of filling with both contents and containers. The phrase ‘that … which itself is more’ refers to the container. From (29) it can now be inferred that epistemic filling is truer filling than nutritional filling.

Observe that Socrates' argument does not require (29). A premise akin to (23) would suffice to show that epistemic filling is truer than nutritional filling, for instance:

1. (29₂₃) That which is filled with things that are more is more filled than that which is filled with things that are less.

I presume that Socrates expresses (29) simply because (29₂₃) does not adequately capture his view. Evidently, his view is that the truth of a filling is not solely a function of the contents, but conjointly of the container and the contents.

Having clarified the conception of true filling, Socrates now relates pleasure to filling by specifying that pleasure involves a filling with what is naturally appropriate:

1. (30) Being filled with what is naturally appropriate is pleasant (585d11).

From (30) Socrates infers:

1. (31) Therefore, being filled with things that really are makes one more really and truly enjoy true pleasure (585d12–e1).

From (31) it follows that rational filling is the truest pleasure. Thus, philosophical pleasure, which is rational pleasure, is the truest pleasure.

VI. SEMANTIC REMARKS ON ἈΛΗΘΕΙΑ AND ITS COGNATES

The preceding certainly cannot stand as a satisfactory interpretation of the true filling argument. A number of premises cry out for explication. Above all, the central concept of truth requires elucidation. Indeed, clarification is needed here both to understand the true filling argument and to make sense of Socrates' conception of pleasure.

The word ἀλήθεια and its cognates παναληθές, ἀληθές, ἀληθεστέρον, ἀληθέστατον, ἀληθῶς, ἀληθινῶς and ἀληθεστέρως – all of which I will refer to as ‘alethic terms’ – occur 23 times in the true pleasure argument, in the following order:

1. (a) At the beginning of the misperception argument, Socrates characterizes the pleasure of the irrational man as ‘not at all true’ (οὐδὲ παναληθής) (583b3).

2. (b) Socrates says that in the phantasms of the irrational men's quasi-pleasures, there is nothing healthy relative to ‘the truth’ (ἀλήθειαν) of pleasure (584a10).

3. (c) In response to Socrates' example of olfactory pleasures as pure in so far as they are not preceded by pains, Glaucon replies that Socrates has spoken ‘most true things’ or ‘most truly’ (ἀληθέστατα) (584b9).

4. (d) In his analogy of spatial positions, Socrates claims that one who moves from the bottom to the midpoint believes he is at the top because he has not seen what is ‘truly top’ (ἀληθῶς ἄνω) (584d9).

5. (e) In the same analogy, Socrates suggests that if the mover were brought back down to the bottom, he would believe that he was at the bottom and would, in this case, believe ‘true things’ (ἀληθῆ) (584e2).

6. (f) In the same analogy, Socrates says that the mover has true and false beliefs because he is inexperienced in ‘that which is truly top’ (τοῦ ἀληθινῶς ἄνω ὄντος) (584e5).

7. (g) In the same analogy, Socrates says that when irrational men experience pain, analogously to the mover being brought to the bottom, they believe ‘true things’ (ἀληθῆ) (585a1).

8. (h) In the true filling argument, Socrates says that a ‘truer’ (ἀληθεστέρα) filling is a filling with that which is more (585b9).

9. (i) Socrates says that the kind of filling that the soul undergoes is a filling with ‘true belief’ (δόξης ἀληθοῦς), knowledge and understanding (585b14).

10. (j) Socrates claims that that which is connected to what is always the same and immortal and ‘truth’ (ἀληθείας) is more (585c2).

11. (k) Socrates claims that the being of that which is always the same partakes equally of being and knowledge and ‘truth’ (ἀληθείας) (585c10).

12. (l) Socrates claims that the therapy pertaining to the body partakes less ‘of truth’ (ἀληθείας) and being than the therapy pertaining to the soul (585d3).

13. (m) Socrates claims that being more really filled makes one more really and (1) ‘truly’ (ἀληθεστέρως) enjoy (2) ‘true pleasure’ (ἡδονῇ ἀληθεῖ) (585e1).

14. (n) Socrates claims that when one is less (1) ‘truly’ (ἀληθῶς) and stably filled, one has a share of less trustworthy and less (2) ‘true’ (ἀληθοῦς) pleasure (585e4).

15. (o) In the true filling argument, Socrates refers back to the analogy of spatial positions and claims that those preoccupied with non-philosophical fillings, never ascend to what is ‘truly the top’ (τὸ ἀληθῶς ἄνω) (586a4).

16. (p) Socrates claims that pleasures mixed with pains are imitations of ‘true pleasure’ (τῆς ἀληθοῦς ἡδονῆς) and shadow-painted (586b8).

17. (q) In reference to Stesichorus' ode, Socrates speaks of the Greeks and Trojans fighting over the copy of Helen, through their ignorance of ‘that which is true’ or ‘truth’ (τοῦ ἀληθοῦς) (586c5).

18. (r) At the end of the true filling argument, Socrates says that if they are led by reason, the appetitive and spirited parts of the soul can grasp the (1) ‘truest’ pleasures (ἀληθεστάτας), in so far, that is, as it is possible for their pleasures to grasp (2) ‘true’ (ἀληθεῖς) pleasures, and that they will do so since they follow (3) ‘truth’ (ἀληθείᾳ) (586d8–e1).

19. (s) Likewise, Socrates says that if the baser parts of the soul follow the wisdom-loving part, they will enjoy the best and ‘truest’ (ἀληθεστάτας) pleasures possible for them (587a1).

Prima facie, these instances fall into two groups. The first group – (c), (e), (g), (i) and (q) – includes those instances in which truth is attributed to semantic entities such as beliefs and statements.Footnote ²¹ The second group includes the remainder. The first group appears to employ the cognates of ἀλήθεια representationally. At least – and this is an important qualification – this is how we who are committed to the correspondence theory of truth understand phrases and clauses such as ‘true belief’, ‘what you claim is true’ and ‘you have spoken truly’. In other words, the content of the belief or utterance represents a fact or state of affairs; and the content corresponds to the world. I will refer to this as a ‘representational’ truth conception. The remaining cases – namely, those in (a), (b), (d), (f), (h), (j), (k), (l), (m), (n), (o), (p), (r) and (s) – I will refer to as employing an ‘ontological’ conception of truth.

Among the cases of ontological ἀλήθεια and its cognates, a substantive distinction can be drawn between what I will call ‘independent’ and ‘dependent’ uses. In the independent use – which occurs in (j), (k), (l) and at (r3) – the noun ἀλήθεια is employed, and in its employment the noun does not complement any other noun. In (k), for example, the being of what is always the same is said to partake of ‘truth’ (ἀλήθειας).Footnote ²² In its dependent uses, ἀλήθεια and more commonly its adjectival and adverbial cognates are employed, and these terms modify or complement a nominal, adjectival or verbal expression. In (g), for example, Socrates speaks of ‘truer filling’ (πλήρωσις ἀληθεστέρα) and in (m2) of ‘true pleasure’ (ἡδονῇ ἀληθεῖ).

I will first comment on the independent uses. In (j), (k) and (l) Socrates explicitly relates ἀλήθεια to what is always the same and immortal and to being.Footnote ²³ In my discussion of the true filling argument, specifically in (O1), I suggested that Socrates identifies truth with being (οὐσία). Some support for this suggestion derives from a line in the true filling argument. After Socrates has gained Glaucon's assent to the claim that the being of what is always the same and immortal partakes of truth, he implies that that which partakes of truth partakes of being (585c12).Footnote ²⁴ Strictly speaking, this only implies that truth entails being. But consider the following passages from elsewhere in the Republic:

I suggest that in these passages the conjunction τε καί is epexegetical. In support of this suggestion, consider also the discussion of artificial imitations in Book 10. In one passage these imitations are said to be at a ‘third remove’ from ‘being’ (599a1; cf. 599d2, 602c2). In another passage, one such artefact is said to be an obscure thing in relation to ‘truth’ (ἀλήθειαν, 597a10–11).Footnote ²⁵

In short, I maintain that Socrates uses ἀλήθεια in the independent ontological sense and οὐσία or τὸ ὄν interchangeably. Granted this, what is the independent ontological sense of ἀλήθεια? One possibility is that Socrates means ‘the totality of Forms’. That is, he uses ἀλήθεια as a mass noun, whereby each Form is a portion of ἀλήθεια. Alternatively, Socrates may be using ἀλήθεια to designate a property of Forms, precisely, that in virtue of which they are the kind of entity they are. Blake Hestir distinguishes two aspects of ontological ἀλήθεια that correspond to these two independent uses of ontological ἀλήθεια:

I suggest that in (j), (k), (l) and (r3), that is, in all of its instances in the true filling argument, Socrates deploys the independent use of ontological ἀλήθεια in accordance with Hestir's second aspect. That is, Socrates deploys the independent use of ontological ἀλήθεια to designate the being in virtue of which the Forms are the kind of entity they are. Consider (k) in particular: ‘the being of that which is always the same’ partakes of ‘being’ and so of ‘truth’. The Forms are that which is always the same, and Socrates specifically refers here to their ‘being’. He says that their being partakes of being and so of truth. At least the second instance of οὐσία must therefore refer to Hestir's second aspect of ontological ἀλήθεια. Moreover, I incline to agree with Hestir that stability – both existentially (that is, eternality) and with respect to alteration (that is, immutability) – is the fundamental concept that informs Socrates' claim. That is, Plato here treats οὐσία and thus ontological ἀλήθεια as stability. Recall (O2): that which is is immutable.

I turn now to the dependent use of ontological ἀλήθεια and more commonly its adjectival and adverbial cognates. In this case, the alethic term functions as a logical attribute of some noun. So far as I know, Gregory Vlastos was the first to discuss this use in Plato.Footnote ²⁷ I find Vlastos's account mistaken in crucial ways. However, because it is a seminal contribution, I will discuss it before turning to the more recent and, I think, more tenable position of Jan Szaif.

Vlastos never uses the phrase ‘ontological truth’; none the less, the concept pervades his discussion. His paper begins with the claim that Greek, unlike English, can form a noun, οὐσία, and an adverb, ὄντως, from the verb εἶναι, whereas English requires the use of a different verbal root, ‘real’ as in ‘reality’ and ‘really’ (p. 1). Vlastos asserts that ‘“true” is a fairly common meaning of “real”’ in English as ἀληθές is in Greek; that ‘one can speak of “a true friend”, “true gold”, “true courage”, and so forth’; and that ‘in all cases “real” can be substituted for “true” with little change of sense’ (p. 3).

Vlastos claims that when we speak of a ‘true friend’, we mean that the person so described has ‘those very properties in virtue of which they can be truly so described: e.g. of Jones as sympathetic, responsive, considerate, loyal, and whatever else we expect of a man whom we would call “a friend” …’ (ibid.).Footnote ²⁸ Observe that Vlastos's formulation explains the sense of ‘true’ as ‘real’ using the concept of a true sentence or description. Again, Vlastos writes: ‘Jones is a “real” friend because … he has those attributes in virtue of which sentences applying these predicates to him are true …’ (ibid.). One wonders whether the ontological sense of ‘true’ as ‘real’ must be formulated using the representational sense of ‘true’. Instead, why not claim that a true friend is something that has all the properties of a friend? It may be objected that ‘true friend’ is not synonymous with ‘friend’, but that the reformulation makes ‘true’ redundant. Although Vlastos does not consider the reformulation I am offering, he is aware of the redundancy problem: ‘There would be no point in saying that Jones is a “real” friend, as distinct from just saying that he is a friend, unless we were tacitly contrasting him with people who talk and act like friends, and for a time pass for friends, but then turn out to be fakes …’ (ibid.). Granted this, we can still formulate the meaning of ‘true friend’ independently of the representational truth conception: ‘a true F’ means ‘something that has all the properties of an F, in contrast to something that does not’. In other words, part of the meaning of the dependent ontological use of ‘true’ is contrastive. Granted this, Vlastos should have avoided formulating his account in terms of representational truth.

Vlastos also claims that when we speak of a true friend, the following counterfactual is implied: ‘< such claims> would be found to be true if put to the test’ (ibid.). Again, we should reformulate this point independently of the reference to representational truth: someone's being a true friend entails that if that person's friendship were tested, the person would pass the test. Thus, a true friend is reliably a friend. Vlastos thinks that it is this ‘implicit reference to reliable truth … that saves “real” from redundancy when used in this sense’ (ibid.). As we have seen, this is mistaken, for in this context ‘true’ has a contrastive meaning. Moreover, it is necessary to distinguish between what the word ‘true’ here means and entailments of F that are independent of the meaning of ‘F’. For example, ‘heptangular’ means ‘having seven angles’; and although having seven angles entails having seven sides, ‘heptangular’ does not mean ‘having seven sides’. Again, it is necessary to distinguish what the word ‘true’ here means from the relations in which true Fs stand, such as that given by the counterfactual Vlastos mentions. Again, a true friend may be one who would prove true if put to the test and thus prove reliable, but ‘true F’ does not mean ‘reliably F’.

I dwell on this point because a central claim of Vlastos's paper is that the dependent ontological use of ‘true’ has the ‘sense’ of ‘reliable’. More precisely, Vlastos claims that ‘true’ here has the ‘sense’ of ‘cognitively reliable’ (ibid. 7). In sum, then, Vlastos's claim rests on two errors. The first depends on Vlastos's conception of dependent ontological ‘truth’ in terms of representational truth.Footnote ²⁹ The second is that Vlastos conflates the meaning of ‘true’ in the sense of ‘real’ with the fact that a true F would reliably prove to be an F, and he consequently claims that ‘true’ means ‘cognitively reliable’. The mistakes are evidently interrelated precisely in so far as true and false beliefs and statements about Fs or pseudo-Fs derive from relations to true and untrue Fs. In short, then, we ought to distinguish between the meaning of ‘true’ in its dependent ontological use and the epistemological significance of the fact that there are true and untrue Fs.

We will return to Vlastos's paper briefly below. For the present, let us turn to Jan Szaif's discussion of the logically attributive use of ἀληθές or of equivalent expressions such as ἀληθινόν, (ὡς) ἀληθῶς, and τῇ ἀληθείᾳ.Footnote ³⁰ Szaif claims that this use is equivalent to the logically attributive use of the German ‘wirklich’, ‘echt’ and ‘wahr’, and likewise to the English ‘true’, ‘real’ and ‘genuine’. More precisely, Szaif distinguishes two such logically attributive uses, which he calls ‘statuierende’ and ‘signierende’ (n. 30 [1996], 52, 54). I will render the German adjectives as ‘confirmative’ and ‘paradigmatic’ respectively.

The confirmative use can be seen in following contrasting cases: ‘true gold versus fool's gold’, ‘the true Elvis versus an impersonator’.Footnote ³¹ In such cases, Szaif observes, the attribute ‘true’ does not affect the extension of the nominal expression ‘F’ that it modifies. Thus, a lump of gold is identical to a lump of true gold, whereas a lump of fool's gold is not part of the extension of ‘gold’ (ibid. 52). Contrast this with the paradigmatic use of ‘true’ as in the following contrasting cases: ‘a true Irish setter versus an non-pure-bred Irish setter’ and ‘a true red versus a red that has an admixture of other colours’ (ibid.). A distinctive feature of this paradigmatic logically attributive use is that it permits entities being F in varying degrees. Accordingly, an untrue F may belong to the set of Fs. For example, a colour that is predominantly red but also mixed with some other colour may still be a case of red.Footnote ³² Consequently, the use of ‘true’ here is not truth-conditionally otiose. Rather, it narrows the extension of the type F. In other words, while true and untrue Fs may both be Fs, true Fs are a subset of Fs.

Szaif claims that the confirmative use applies to the distinction between originals and copies or fakes, while the paradigmatic use applies to the distinction between pure and impure cases. Moreover – apropos of Vlastos's view of the cognitive significance of the dependent ontological use of ‘true’ – Szaif notes the following epistemological or cognitive implication of the confirmative use: there is a tendency for the confirmative use to be employed in the context of indicating that untrue Fs are deceptive (cf. ibid. 55). In other words, when one characterizes an F as ‘true’, in the confirmative sense, in contrast to an untrue F, one tends to convey that the untrue F merely appears to be, without actually being, an F. Note that Szaif does not here make Vlastos's mistake of claiming that the epistemological implication of the confirmative use is constitutive of the meaning of ‘true’.

At this point, let me briefly return to Vlastos's discussion. In section two of his paper, Vlastos maintains that Plato conceives of some entities, namely Forms, as truer or more real than others, namely participants, in so far as the latter are always, while the former are never, both F and not-F (n. 27 [1969], 10). Note that by ‘not-F’ here Vlastos intends the contrary, not logical opposite, of F. As Vlastos puts it, the ‘F nature <of participants> is adulterated by contrary characters’ (ibid. 6). For example, a beautiful participant is both beautiful and ugly. As such, Plato conceives of participants as ‘less “pure” Fs’ than true Fs (ibid.).Footnote ³³

Evidently, Vlastos's conception of Plato's dependent ontological use of alethic terms is akin to Szaif's paradigmatic use. Thus, Vlastos's explanation of Plato's ‘degrees of reality’ is akin to Szaif's view that the paradigmatic use permits entities being F in varying degrees. But Vlastos's conception is not equivalent to Szaif's. Again, on Vlastos's view untrue Fs are both F and the contrary of F; but on Szaif's view untrue Fs need not be untrue because they are both F and the contrary of F. Recall the examples of true red versus red mixed with another colour and a true Irish setter versus a non-pure-bred Irish setter. Even granting that the colour red has a contrary – say, the complementary colour green, untrue or impure red needn't be red mixed with green, as opposed to, say, yellow or blue. In the canine case, being an Irish setter simply has no contrary, such that a non-pure-bred Irish setter is a hybrid of an Irish setter and some particular other breed. Accordingly, while Plato does often characterize paradigmatically untrue Fs as both F and the contrary of F, the claim that a certain F is paradigmatically ‘untrue’ does not entail that it is both F and the contrary of F.

What makes an F paradigmatically true? The key term Szaif uses to characterize paradigmatically untrue Fs is ‘eingeschränkt’. For example: ‘die als wirkliches X gekennzeichnete Sache wird vielmehr von solchen Fällen abgehoben, in denen etwas nur in eingeschränkte Weise X ist’ (n. 30 [1996], 53). Accordingly, a paradigmatically true F is an ‘uneingeschränktes’ F (ibid.). The question, then, is what makes a paradigmatically true F completely or unqualifiedly or unrestrictedly an F. Unfortunately, Szaif does not say. We might suggest that a paradigmatically true F fully realizes the properties necessary for being an F. For example, a fully realized red is red that is only constituted by red. Thus, an impure red is not paradigmatically true because it is constituted by colours other than red. However, this does little more than substitute one explanandum for another. What does it mean for something to fully realize a property? In this paper, I must leave the problem unresolved. The examples given in conjunction with our theoretically uninformed intuitions must suffice.

Given the preceding semantic remarks, let us return to the dependent uses of ἀλήθεια and its cognates in the true pleasure argument. There are 13 such uses: in (a), (b), (d), (f), (h), (m1), (m2), (n1), (n2), (o), (p), (r1), (r2) and (s). This number is reducible to three for the following reasons. First, in some cases there are several tokens of the same or at least an equivalent phrase type. For example, ἀληθῶς ἄνω in (d), τοῦ ἀληθινῶς ἄνω ὄντος in (f) and τὸ ἀληθῶς ἄνω in (o) are clearly equivalent. Less obviously, ἡδονῆς ἀλήθειαν in (b) and, for instance, ἡδονῆς ἀληθοῦς in (n) are equivalent.Footnote ³⁴ Further reduction is possible in so far as an alethic adjective or adverb in different degrees modifies the same type of nominal expression. For example, in (s) Socrates speaks of ‘truest pleasures’ (ἡδονὰς ἀληθεστάτας). So we are left with three phrase types where an alethic term modifies: a pleasure term, as in (a), (b), (m1), (m2), (n2), (p), (r1), (r2) and (s); a term for the top position, as in (d), (f) and (o); and a filling term, as in (h) and (n1).

In the filling cases, the alethic term is in the comparative degree: (h) speaks of ‘truer’ (ἀληθεστέρα) filling, (n) of being ‘less truly’ (ἧττον ἀληθῶς) filled. This indicates the paradigmatic use. Likewise, in the pleasure cases, in so far as Socrates uses alethic terms in comparative and superlative degrees and argues that true pleasures correlate with true fillings, the alethic terms must be used paradigmatically. This is so in (m1), (m2), (n2), (r1), (r2) and (s).

That leaves the instances in (a), (b) and (p). These are difficult cases. In (a) Socrates characterizes the irrational man's quasi-pleasure as οὐδὲ παναληθής. The adjective παναληθής only occurs here in Plato, and prior to Plato only twice.Footnote ³⁵ If we render οὐδὲ παναληθής as ‘not wholly true’, this entails that the irrational man's quasi-pleasure is partly true. But this is hard to square with (b), where Socrates claims that there is ‘nothing’ (οὐδέν) healthy in the phantasms of the irrational man's quasi-pleasure relative to the truth of pleasure, that is, relative to true pleasure. Moreover, the contrast between the irrational man's quasi-pleasure as a phantasm and true pleasure suggests that the former is confirmatively untrue. As such, the irrational man's quasi-pleasure appears to be pleasure, but isn't. Again, in (p) the irrational man's quasi-pleasures are characterized as imitations or copies of true pleasure. Here too there is a contrast between appearance and reality. Accordingly, I believe that the alethic terms in (a), (b) and (p) are used confirmatively.

Finally, the top position cases do not yield a straightforward result. The instances in (d) and (f) occur within the misperception argument and are analogous to the contrast between apparent and real pleasure. Thus, they seem to be used confirmatively. But in (o), the spatial positions analogy is applied to the three types of filling. Those preoccupied with non-philosophical fillings never ascend to what is ‘truly the top’. In this case, we appear to have a paradigmatic use since the midpoint would be more truly the top than the bottom point, just as spirited fillings are truer than appetitive fillings, and even though neither is completely true.Footnote ³⁶

Somewhat tentatively, I conclude that the irrational man's quasi-pleasure is confirmatively untrue. In other words, the irrational man's quasi-pleasure is not a form of pleasure. It is a condition that shares an appearance with (confirmatively true) pleasure, but no more. In contrast, appetitive, spirited and philosophical pleasures are all confirmatively true; that is, they are pleasures. However, among these, only philosophical pleasure is paradigmatically true.

The following section elaborates on these conclusions by clarifying what confirmatively and paradigmatically true pleasures are. Note also that since confirmatively true pleasure simply is pleasure – even though ‘(confirmatively) true pleasure’ does not simply mean ‘pleasure’ – it will often be convenient to speak just of pleasure. To avoid confusion, when I wish to refer to paradigmatically true pleasure instead of confirmatively true pleasure, I will speak of ‘paradigmatic pleasure’.

VII. CONDITIONS OF TRUE PLEASURE

Having clarified Socrates' uses of alethic terms in the true pleasure argument, I will here clarify his conception of confirmatively and paradigmatically true pleasure. I begin with bodily pleasure and pain. In (19) and (21) Socrates refers to somatic depletion and replenishment.Footnote ³⁷ Passage (30) suggests that nutritional replenishment is somatically pleasant. Socrates does not specify, but it is reasonable to infer that nutritional depletion is somatically painful.Footnote ³⁸ In Republic 9, then, Socrates conceives of somatic replenishment as a constituent of bodily pleasure. The point is significant because somatic replenishment is itself a kind of κίνησις. Recall that within the misperception argument, at (3) and then (12), Socrates distinguishes pleasure and pain from calm by claiming that pleasure and pain are both forms of motion. Yet as we also saw, in the misperception argument Socrates identifies these motions as being of the soul. Evidently, then, bodily pleasure and pain involve two motions: one somatic, the other psychic. How should we understand these motions?

Clearly the somatic motion of bodily pleasure is a motion of replenishment. But it is unclear whether the psychic motion of bodily pleasure is a motion of replenishment as well. In considering this question, note that bodily pleasure requires that the somatic motion affect the soul.Footnote ³⁹ This point is explicit in the treatments of pleasure in Timaeus and Philebus. That is, some somatic motions of replenishment are too weak or subtle to affect the soul. Consequently, such motions are not constitutive of pleasures. The point is implicit in Republic 9 in Socrates' treatment of olfactory pleasures. Olfactory pleasures are preceded by nasal depletions; but the depletions are too subtle and gradual to be experienced algesically; thus, they are not constitutive of pains.

This suggests that in the case of bodily pleasure, psychic motion is the objective correlate of the subjective experience of pleasure, that is, of the hedonic appearance. But what of the case of psychic pleasure? In the case of psychic and specifically rational pleasure, a motion of psychic replenishment occurs. But what is the objective correlate of the experience of rational pleasure? Are there two psychic motions, one of replenishment and one that constitutes the objective correlate of experience? Or does one psychic motion play both roles in so far as the soul necessarily experiences its objective state? The text offers no answer to the question.

Conservatively, I will hereafter assume that in the case of bodily pleasure, somatic motion is motion of replenishment that affects the soul and that psychic motion constitutes the objective correlate of subjective hedonic experience. In the case of psychic pleasure, psychic motion is motion of replenishment. But it is unclear whether the soul apprehends this motion just because it is psychic, or whether there is a second psychic motion, not of replenishment, that is constitutive of psychic apprehension.

Note that hereafter I will use ‘psychic motion’ to refer to psychic motion that constitutes the objective correlate of subjective experience, whether this only occurs in the case of bodily pleasure or also in the case of psychic and specifically rational pleasure. I will distinguish the objective psychic motion that constitutes the objective correlate of the subjective experience of pleasure as ‘hedonic psychic motion’. I will continue to use the phrase ‘hedonic appearance’ to refer to the subjective state of hedonic experience. And I will use ‘replenishment’ and ‘depletion’ to refer to motions of replenishment and depletion, be they somatic or psychic.

In the wake of these conclusions, we can comfortably specify the following conditions on pleasure:

1. (C1) Pleasure requires a replenishment.

2. (C2) The replenishment may be psychic or somatic; but if it is somatic, then it must affect the soul.

3. (C3) Thus, pleasure requires a hedonic psychic motion.

4. (C4) Pleasure requires a hedonic appearance, coterminous with the hedonic psychic motion.

Since the irrational man's quasi-pleasure is a calm preceded by pain that produces a hedonic appearance, the irrational man's quasi-pleasure is confirmatively untrue; it is no pleasure at all. It merely shares with pleasure a hedonic appearance; and it is in virtue of this appearance that it is called a quasi-‘pleasure’, as opposed to some other sort of experience.

As we have seen, at 584b1–c8 Socrates appeals to olfactory pleasures to support his claim that pleasure is not a cessation of pain.Footnote ⁴⁰ This suggests that pleasure does not arise from pain. In that case, it is necessary to add the following condition of pleasure:

1. (C5) Pleasure requires that calm immediately precedes the replenishment.

Accordingly, we can also add the following condition:

1. (C6) Pleasure requires that the complex of a replenishing motion, a psychic motion, which may be identical to or produced by the replenishing motion, and an immediately antecedent psychic calm are conjointly productive of the hedonic appearance.

While the textual evidence so far considered compels the addition of (C5) and (C6) to the set of conditions of pleasure, these additional conditions do not sit so well in view of two other considerations. Once again, consider Socrates' contrast of bodily quasi-pleasure with ‘pure’ pleasure: ‘we should … not accept that pure pleasure is the relief from pain’ (584c1–2). I suggest that by ‘pure’, Socrates here means that the hedonic condition is not mixed with pain, the contrary of pleasure. This suggestion is supported by a passage in the true filling argument. At 586b7–8, Socrates refers back to irrational men's quasi-pleasures as ‘mixed with pains’. The irrational man's quasi-pleasure is mixed with pain and thus not pure pleasure. But, reflecting on the paradigmatic use of alethic terms, purity should be a property of paradigmatic pleasure, not merely confirmatively true pleasure. Thus, it is awkward that pure pleasure is contrasted with quasi-pleasure in the sense of confirmatively untrue pleasure.

The second consideration in relation to which (C5) and (C6) do not sit well pertains to the appetitive pleasures introduced in the true filling argument. These are treated as confirmatively true pleasures, albeit not paradigmatically true. But consider the very common condition of the pleasure of drinking following thirst, which is a pain. In order to remain consistent, Socrates must maintain that appetitive hedonic conditions that are not preceded by pain are confirmatively true pleasures, while appetitive hedonic conditions that are preceded by pain are quasi-pleasures, that is, confirmatively untrue. I suggest that both considerations and their attendant problems are symptomatic of a failure on Plato's part to distinguish clearly between confirmatively and paradigmatically true and untrue hedonic conditions.

Finally, let me turn to the conditions of paradigmatic pleasure. In principle, since confirmatively true pleasure requires the satisfaction of several conditions, more and less paradigmatically true pleasures could be those that more and less fully realize those conditions. But it is not so clear whether any of conditions (C1)–(C6) is conceivable as gradable. For example, what would it mean to have more or less of a hedonic appearance? Perhaps some of these conditions could be construed as gradable in so far as some appearances are more robust or intense than others or in so far as some motions are more or less powerful or swift than others. At any rate, in so far as he distinguishes more and less paradigmatically true pleasures, Socrates focusses only on the following condition:

1. (C7) Paradigmatic pleasure requires paradigmatically true replenishment, which is replenishment whose content and container are stable.

Appetitive and spirited pleasure may satisfy conditions (C1)–(C6), but only rational or philosophical pleasure satisfies (C7) as well.

(C7) claims that paradigmatically true replenishment requires the stability of the container and content. The idea is partially expressed in (23) and (29) of the true filling argument. Once again:

1. (23) Truer filling fills with what is more (where ‘what is more’ is subsequently clarified as ‘more immutable’).

2. (29) That which is filled with things that are more and which itself is more is more filled than that which is filled with things that are less and which itself is less.

But (C7) makes a crucial error. First, why should one think that a paradigmatically true filling requires the stability of the container and content? A container is more or less filled depending on whether the filling content completely fills the deficit. In other words, a true filling is a matter of the relation between the quantity of the deficit of the container and the quantity of the filling content, not the durability of either the container or content.

Granted this, one might charitably suggest that although Socrates speaks of filling or replenishment, he has the concept of restoration in mind. True restoration seems to require durability and thus stability. For example, if one glued two broken bowls with two different adhesives and one bowl came unglued, while the other held, it would be correct to say that the one bowl had been more truly restored than the other. But this charitable move does not save Socrates' argument. In restoration, extent of durability or stability is relative to the kind of entity restored. Compare restoring a torn paper bowl and restoring a broken ceramic bowl. A truly restored paper bowl should be as durable and stable as a normal paper bowl, while a truly restored ceramic bowl should be as durable and stable as a normal ceramic bowl. Thus, the fact that the properly restored ceramic bowl is far more durable or stable than the properly restored paper bowl in no way undermines the fact that the paper bowl is truly restored.

Plato misses this point. Instead, in view of the fact that the soul and its proper content, knowledge, are immortal and immutable respectively, whereas the body and its proper content are mortal and mutable, Plato is misled to think that what makes psychic filling or restoration truer than somatic filling or restoration is its stability or durability. Thus, he falsely claims that that which is filled with what is more is more truly filled. The mistake is fatal to the true filling argument, for Socrates argues that truer pleasure entails truer filling and that philosophical filling is truer than somatic filling.

VIII. THE REPRESENTATIONAL UNTRUTH OF THE IRRATIONAL MAN'S QUASI-PLEASURE

One further sense of ‘truth’ is operative in the true pleasure argument. More precisely, it is operative in Socrates' account of the untruth of the irrational man's quasi-pleasure in the misperception argument. In the course of discussing Socrates' explanation of the irrational man's quasi-pleasure, I spoke of the irrational man as subject to a hedonic illusion. The illusory nature of the irrational man's experience suggests that his quasi-pleasure is representationally untrue, for the content of the hedonic appearance represents the irrational man as having pleasure, whereas in fact he is in a state of calm. In this section, I discuss the representational untruth of the irrational man's quasi-pleasure and attempt to clarify the nature of the hedonic appearance constitutive of his experience.

Within the true filling argument, at 586b7–c5, Socrates refers back to the quasi-pleasures of irrational men as εἴδωλα:

There are 17 additional instances of εἴδωλον in Republic.Footnote ⁴¹ In all cases, an εἴδωλον shares an appearance with something – call it ‘the original’ – but lacks other properties that the original possesses.Footnote ⁴² Thus, the εἴδωλον can be conceived as a defective imitation of the original as well as a misleading or illusory entity. For example, Socrates contrasts εἴδωλα as the products of imitators, for example, the paintings and poems of painters and poets, with the originals that they imitate (599a7, d3, 600e5, 601b9, 605c3). Socrates also speaks of images in water as εἴδωλα of the things of which they are reflections (516a7, 532c2).

Socrates' reference to Stesichorus' treatment of Helen informs our interpretation of the use of εἴδωλον here. In Stesichorus' version of the myth, the gods created an εἴδωλον of Helen and it was this εἴδωλον that went to Troy, rather than Helen herself, who stayed in Egypt.Footnote ⁴³ The Trojans were duped by the εἴδωλον of Helen, and Paris even slept with it. In this context, ‘imitation’ or ‘copy’ is a fitting translation of εἴδωλον. The copy evidently shares Helen's appearance. But qua copy the pseudo-Helen lacks some of Helen's properties. Compare the irrational man's quasi-pleasure. The irrational man's quasi-pleasure shares an appearance with pleasure. Yet it lacks numerous other properties of pleasure. As such, the irrational man's quasi-pleasure is confirmatively untrue.

But Socrates' talk of the εἴδωλον of Helen and of the εἴδωλα of true pleasure suggests another way in which the irrational man's quasi-pleasure is untrue: the irrational man's quasi-pleasure is also representationally untrue. As we have seen and repeatedly stated, the irrational man's quasi-pleasure is illusory. More precisely, the hedonic appearance of the irrational man's quasi-pleasure is illusory. In contrast, true hedonic appearances are not illusory; they correspond to the replenishing motion that is partially productive of them.

As we have seen, Socrates appeals to the visual illusion produced by shadow-painting to explain the hedonic illusion to which the irrational man is subject. The explanation amounts to no more than an analogy between the juxtaposition of darker and lighter shades in the case of visual illusion and psychic calm and immediately antecedent pain in the case of hedonic illusion. That is, Socrates offers no further physiological explanation of why such hedonic illusion occurs. Indeed, I see no way of advancing our understanding of the explanation in this respect. On the other hand, I do see a way of clarifying what may be called ‘the mode of hedonic representation’, that is, the mode of the hedonic appearance. In other words, I would like to clarify how Socrates understands hedonic appearances, be they illusory or veridical. More precisely, I would like to clarify whether Socrates conceives of hedonic appearances in phenomenal or doxastic terms.

My use of the phrase ‘hedonic appearance’ follows Socrates' own language. Recall that at 584a7–8 Socrates says of the irrational man's quasi-pleasure that it is not pleasant, but only ‘appears’ so. At 584a9 he refers to such conditions as ‘phantasms’. Again, at 586c1–2 he says that the juxtaposition of the antecedent pain and occurrent calm produces quasi-pleasures that ‘appear’ intense. In short, the appearance of a hedonic appearance, be it illusory or veridical, is a φαινόμενον. Moreover, a φαινόμενον is a subjective state; it is a φαινόμενον to someone. But it is questionable how Socrates understands such subjective hedonic states.

On the one hand, there are passages where Socrates seems to understand the φαινόμενα in terms of experiential qualities. For instance, as we have just recalled, at 586c1–2 Socrates speaks of the conditions appearing ‘intense’. At 584b7 and c6, Socrates refers to their ‘greatness’ in what I take to be a subjective sense, that is, to their robustness or, again, intensity. Recall also Socrates' use of the word περιωδυνία in (5) in reference to the condition of sick people. Generally speaking, Socrates' discussion encourages us to interpret his account of veridical and quasi-pleasure and pain at least partially in phenomenal terms.

At the same time, it is questionable whether affective qualities of experience exhaust Socrates' conception of hedonic and algesic appearances. At least, it is worth considering the relation between beliefs and appearances since doxastic language pervades Socrates' account of affective conditions. In the analogy of spatial positions he deploys at 584d1–585a7 to explain the irrational man's quasi-pleasure, Socrates says that one who moved from the bottom to the midpoint might mistake the midpoint for the top. Socrates repeatedly characterizes the mover's psychological condition in doxastic terms. The mover is said to ‘believe’ that he is being carried to the top (584d7, 8).Footnote ⁴⁴ Socrates speaks of such individuals as inexperienced in what is truly top, middle and bottom; and he concludes:

Likewise, earlier in the misperception argument, Socrates refers to what irrational men ‘say’ when they are ill (583c11, d3–4).Footnote ⁴⁵ I take it that the statements of the sick reflect their beliefs about their conditions.

Socrates' discussion of illusions in Republic 10 throws further light on the relation between affective appearances and beliefs in Republic 9. In the later book Socrates once again refers to shadow-painting, and he characterizes it as γοητεία (602d2). Recall that when Socrates says that there is ‘nothing healthy in the phantasms <of irrational men> relative to the truth of pleasure’, he says that these phantasms are a kind of γοητεία (584a10). In Republic 10, Socrates describes the magical effects of shadow-painting as creating on two-dimensional surfaces appearances of concavity and convexity (602c11–12). In the same passage he refers to non-artificial cases of illusion: an object at different distances from the observer appearing to be different sizes and a straight object in water appearing to be crooked (602d6–9). In these cases Socrates says that calculating, measuring and weighing can provide assistance so that the illusory ‘appearance’ does not ‘rule’ within us (602d6–e3). He says that calculation, measuring and weighing are functions of ‘the calculating part of the soul’ (602e1–2), and he continues (32):

Socrates does not specify which part of the soul is responsible for the illusory appearance, but he does say that it is one of the ‘base’ parts (603a7–8). Thus, it is either the appetitive or spirited part.Footnote ⁴⁶ On the basis of Timaeus I would argue that it is the appetitive part. But in either case, beliefs are attributed to both the part responsible for the appearance and the calculating part.

I suggest that the contents of (32) indicate that Socrates does not here draw a distinction between appearance and belief. To appreciate this, observe that Socrates' account is an argument for the view that two different parts of the soul hold contradictory beliefs; and consider the following slightly more formal description of Socrates' account:

1. (i) As a result of its measuring, the calculating part of the soul indicates that p.

2. (ii) Simultaneously not-p appears.

3. (iii) The same part of the soul cannot simultaneously believe that p and not-p.

4. (iv) Thus, the part of the soul that believes that not-p is not the same as the part of the soul that believes that p.

Implicit in the transition from (ii) to (iii) is a transition from ‘not-p appears <to a part of the soul > ’ to ‘ < the same> part of the soul believes that not-p’. The latter transition assumes that ‘p appears to s’ is equivalent to ‘s believes that p’. In other words, there is no indication here that beliefs are formed on the basis of appearances.Footnote ⁴⁷

If this interpretation is correct, then we should not reduce affective appearances to affective qualities of experience. Rather, we should hold that affective appearances are qualitative-cum-doxastic states. Indeed, prior to Plato's explicit distinction between bare or relatively bare αἴσθησις and δόξα at Theaetetus 184–6, there is no good reason to think that he recognizes a distinction between bare perception and perceptual belief. Thus, Socrates' conception of appearance in Republic 9 is akin to the conception of φαντασία that the Eleatic Stranger states at Sophist 264b2: a mixture of αἴσθησις and δόξα – with the crucial difference being that since Sophist is a post-Theaetetus dialogue, in composing the phrase ‘a mixture of αἴσθησις and δόξα’ in the later dialogue, Plato recognizes a distinction between bare or relatively bare perception and perceptual belief, whereas in Republic he does not.