This Content Downloaded From 185.192.71.183 On Wed, 08 Mar 2023 19:16:33 UTC

Propositions or Objects?
A Critique of Gail Fine on Knowledge and Belief in "Republic"

V
Author(s): Francisco J. Gonzalez
Source: Phronesis , 1996, Vol. 41, No. 3 (1996), pp. 245-275
Published by: Brill
Stable URL: https://www.jstor.org/stable/4182535
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Brill is collaborating with JSTOR to digitize, preserve and extend access to Phronesis
This content downloaded from

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
All use subject to https://about.jstor.org/terms
Propositions or Objects?
A Critique of Gail Fine on Knowledge and Belief
in Republic V
FRANCISCO J. GONZALEZ
According to most contemporary epistemologies, the distinction between

true belief and knowledge concerns, not objects, but propositions. What we
know or believe are propositions about objects, not these objects them-
selves.' Knowing something is thus rendered completely disanalogous to
seeing something. Whether or not I see something depends on the direct
relation I have to the object and on the character of the object, i. e., whether
or not it is something visible (as opposed, for example, to something only
audible). Knowledge, on the other hand, does not depend on any sort of
direct relation to the object, but instead on the justification or explanation of
propositions; furthermore, presumably anything about which we can formu-
late meaningful propositions can in theory - though not, of course, in fact -
be known. Many interpreters believe that Plato's conception of knowledge
does not fit this model. Plato appears to take very seriously the analogy
frequently found in his writings between knowing and seeing. A locus
classicus is the argument at the end of Republic V that, in apparently con-
cluding that one can know only intelligible forms and that nothing more
than belief is possible with regard to sensible objects, appears to understand
knowledge and belief as direct cognitive relations to objects and as restrict-
ed to certain kinds of objects.
This common reading of the argument, however, has been challenged by
Gail Fine in a couple of articles, the first appearing almost two decades
ago.' Fine rejects the "objects analysis" that interprets the argument as
correlating knowledge with certain kinds of objects. She defends instead a
"contents analysis" that interprets the argument as correlating knowledge
' Some modem philosophers of course recognize the existence of what is called "know-
ledge by acquaintance," but they do not consider this to be the kind of knowledge that is
contrasted with belief. See, for example, H. H. Price's claim that the distinction between
knowledge and belief is always a distinction between "knowledge that" and "belief that"
(Belief [London: George Allen & Unwin Ltd, 1969], 79).
2 "Knowledge and Belief in Republic V," Archiv fUr Geschichte der Philosophie 60
(1978): 121-39; "Knowledge and Belief in Republic V-VII," in Epistemology, ed. Ste-
phen Everson (Cambridge University Press, 1990), 85-115.
Phronesis 1996. Vol. XLI/3 (Accepted April 1996) 245

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
"with certain sorts of propositions, saying that there is some content of the
cognitive condition" (1978, 124), and thus as leaving open what objects
these propositions are about. The result is a radical departure from the
traditional interpretation of Republic V as supporting a "two worlds theo-
ry": the argument on Fine's reading does not rule out the possibility of
having knowledge of sensibles and only beliefs about forms.3 On the basis
of this interpretation, Fine in the most recent article can even argue, through
a reading of Republic VI-VII, that Plato, far from thinking that knowledge
involves "any special sort of vision or acquaintance" (1990, 87), was actual-
ly a coherentist and thus "surprisingly up to date" (115).4
Despite the extreme unorthodoxy of Fine's carefully argued reading, its
prominence as probably the most thorough account of the argument current-
ly available, and its momentous consequences for our understanding of
Plato, it surprisingly has not yet been seriously challenged.5 In the present
paper, however, I argue that it is untenable. I hope thereby to shed some
light on what the argument in Republic V is in fact trying to show.
I. The Argument (476e7-479e1)
In the following outline I list only the major steps in order to render clear
the structure of the argument. I also briefly describe the important transi-
tions.
Before the argument begins, Socrates has made a distinction between
philosophers and lovers of sights and sounds. The latter, unlike the former,
are unable to distinguish beauty itself from beautiful sensible objects and
therefore deny its existence. For this reason, Socrates claims, they possess
only belief and no knowledge. Acknowledging that they will fight this
claim, Socrates proposes to convince them by means of an argument. Ask-
3 The view that Republic V assigns knowledge and belief to "two worlds," the intelligi-
ble and the sensible, has also been rejected by J. C. Gosling ("Doxa and Dunamis in
Plato's Republic," Phronesis 13 [1968]: 119-30), on the basis of the same kind of
veridical reading defended by Fine and criticized below, and by Julia Annas (An In-
troduction to Plato's Republic [Oxford: Clarendon Press, 19811, 209-15), whose very
different reasons are also examined below. For a critique of Gosling's interpretation in
particular, see Nicholas P. White, Plato on Knowledge and Reality (Indianapolis: Hack-
ett, 1976), p. 106, n. 9.
4 For a critique of Fine's "coherentist" reading of Republic VI-VII see ch. 8 of my book,
Dialectic and Dialogue: Plato's Practice of Philosophical Inquiry, forthcoming.
' Fine's interpretation of Republic V is assumed both in Terence Irwin's Plato's Ethics
(New York: Oxford University Press, 1995), 267 and in Fine's own On Ideas (Oxford:
Clarendon Press, 1993): see, for example, 59-60 & 91-4.
246

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
ing Glaucon to speak on behalf of the lovers of sights and sounds, Socrates
first secures their agreement to the following four theses:
(1) Knowledge is of something (TtO, of what is (6v). (476e7-477al). It is therefore

"set over" (tnt) what is (477a9).
(2) Ignorance is of what is not (til 6v) (477a2-4), is "set over" what is not
(477a9- 10).
(3) If (dt) something both is and is not (obFTO IXr-L (bg Elvac E xaLi E IvaL),
then it lies between (REttcCvi) pure being and absolute not-being (Toi EtklX-
QLVWg 6VTOg xca' toi c' abt rba&tI 6vrog, 477a6-8).
(4) We must search for something between (ReTactv') knowledge (yv(ioLg) and
ignorance (&yvwoCta), if there is such a thing (Et IL TUyXdVCL 6V TOL01OTOV),
to be set over what is between being and not-being (477a9-bl).
Socrates then proceeds, with the following argument, to prove that there is
in fact something between knowledge and ignorance and that it is belief
(60'a).
(5) Different powers (b5vacLEt;) do different work (6 drEQydirTcat) and are set
over different objects (np' () fcrL, 477c6-d6).
(6) Knowledge and belief are different powers because they are infallible (dv-
acidQTryTov) and fallible, respectively (477e6-478a2).
(7) Therefore, belief and knowledge must do different work and be set over
different objects (478a3-b2), i.e., what is known cannot be the same as what is
believed (478bl-2).
(8) Since according to (1) knowledge is set over what is, and since according to
(7) what is known cannot be the same as what is believed, what is believed
cannot be what is (478alO-b5).
(9) Belief also cannot be set over what is not, since it cannot believe nothing
([L'6Ev) (478blO-c8).
(10) Since according to (2) ignorance is set over what is not, belief can no more be
identified with ignorance than with knowledge (478c8).
(11) Belief is darker than knowledge and brighter than ignorance (478clO-15).
(12) Therefore, belief must be between knowledge and ignorance, rather than
above knowledge or beneath ignorance (478d 1-4).
Now Socrates can appeal to (3) and (4) to conclude:
(13) Belief is set over what is between pure being and absolute not-being, i.e.,
what both is and is not (478dS-12).
Socrates at this point has yet to prove to the lovers of sights and sounds i)
that there is such a thing as "what is and is not" and ii) that the sights and
sounds which they love fall into this category. He needs one more premise
to prove this:
(14A) The beautiful things loved by the lovers of sights and sounds are also not
beautiful, the just things are also not just, the pious things are also not pious,
247

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
etc. In short, the objects of their love both are and are not and thus are
between being and not-being (479a5-d2).
(14B) Socrates equates this conclusion with a discovery that "the many conven-
tional opinions of the many concerning beauty and other things are tossing
about somewhere between not-being and pure being" (ta' T(OV [OXX(UV
noXX'a v61tua xaXou TE nrQL xaci TOV &XXWV [UatLv Wou XUXLV&LTaL
TO01 TE R' 6VTO Xa'L TOV 6vro0 dXLXQtV6g, 479d3-5).
(15) From (13) and (14), Socrates can conclude that the lovers of sights and
sounds, in contemplating (0ewp?voug) only things that both are and are not,
have no more than belief (479d7-e6) and that, conversely, those who con-
template the things themselves which are always and in every way the same
(acbT1aE txaoTCt ... xa' 6te1 xaMM TaOTl dbaocttowg 6vTcx) have knowledge
(479e7-9).
II. Fine's Interpretation
As already noted, Fine sees this argument as primarily addressing not the
objects of knowledge and belief but their propositional content. But in
assigning knowledge to "swhat is" and belief to "what is and is not," is not
the argument clearly assigning them to different objects? Not if, according
to Fine, we read the word "is" in the argument as neither existential nor
predicative in meaning, but veridical. On this reading, premise (1) above
must be understood as assigning knowledge not to what exists nor to what is
F (where F stands for some property), but to what is true. Similarly, conclu-
sion (13) assigns belief not to what exists and does not exist, nor to what is
F and not F, but to what is true and false (1978, 123-6; 1990, 87-9). Fine,
however, rejects the "degrees of truth" interpretation, according to which
each proposition contained in the content of belief is both true and false,
maintaining instead that the set of propositions included under belief is
disjunctively true and false, i. e., some are true and some are false (1978,
126; 1990, 89-90). Understood in this way, the argument does not restrict
knowledge or belief to specific kinds of objects, but simply assigns them to
different contents, namely, true propositions in the one case and the set
containing both true and false propositions in the other. Fine explains the
reason for such an assignment by interpreting the controversial steps (5)-(7)
as maintaining the following: because knowledge is infallible, it succeeds in
collecting only truths; because belief is fallible, it collects both truths and
falsehoods (1978, 128-9; 1990, 90-1). Thus, what the distinction between
knowledge and belief in the argument comes down to is simply the claim
that knowledge entails truth, while belief does not (1978, 139). Nothing is
said here about their objects.
Fine must, of course, acknowledge that at premise (14A) the word "is"
has a predicative, rather than veridical sense: in saying there that something
248

185.192.71.183 on ffff:ffff on Thu, 01 Jan 1976 12:34:56 UTC
"is and is not," Socrates means that it is and is not beautiful, just, pious, etc.
Yet Fine explains this shift her reading requires from the veridical to the
predicative "is" by pointing to the connection between (14A) and (14B) and
arguing that reliance on sensible properties that are F and not F issues only
in nomima that can be either true or false (1978, 136). In 1990 Fine clarifies
this point by explaining that while the lovers of sights and sounds may have
true beliefs about what things are beautiful, at least their beliefs about what
beauty is are false (1990, 93). Fine must also explain the undeniable fact
that before the argument begins and at its conclusion Socrates does in some
sense assign knowledge to forms and belief to sensibles. This is her expla-
nation: "To say that knowledge is set over forms is shorthand for the claim
that all knowledge requires knowledge of forms; to say that belief is set
over the many Fs is shorthand for the claim that if one is restricted to
sensibles, the most one can achieve is belief' (1990, 94). To summarize,
knowledge is set over true propositions that can be about either forms or
sensibles; however, its power of infallibly collecting only true propositions
depends on knowledge of forms. Belief is set over both true and false
propositions that can be about either forms or sensibles; the fallibility of its
power in collecting false as well as true propositions, however, is all that is
possible for someone restricted to sensibles.
IIl. The Dialectical Requirement
Assuming for the moment that Fine's "veridical" reading is consistent with
the text, what reasons do we have for accepting it over its rivals, the exist-
ential and predicative readings? Fine does not go so far as to claim that
these other readings are incompatible with the text. Instead, she argues that
they violate what in 1978 she calls the principle of noncontroversiality, and
in 1990 the dialectical requirement: i.e., the principle that an argument
should use only claims that are believed to be true and that the interlocutors
accept.6 Fine maintains that on the existential or predicative readings, So-
6 See 1978, 125 and 1990, 87. The "dialectical requirement" was already central to
Gosling's interpretation (121), as it is to that of Michael C. Stokes' ("Plato and the
Sightlovers of the Republic," Apeiron 25, n. 4 [1992], 109). However, Stokes takes the
principle one step further than Fine: if the premises of the argument are chosen accord-
ing to what is acceptable to the lovers of sights and sounds, why assume that Plato is
committed to them? Thus Stokes can argue, contra Fine, that these premises support a
"two worlds theory," while still agreeing with Fine that this theory "need not be accept-
ed as an item of Plato's own stock of beliefs on the basis of Republic 5 ..." (132).
Similarly, Theodor Ebert has maintained that the argument does not represent Plato's
249

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
crates' argument would begin with controversial assumptions that the lovers
of sights and sounds could not be expected to accept.
What are these controversial assumptions? Both readings, according to
Fine, make the argument assume a sharp separation between the objects of
knowledge and belief (1978, 125; 1990, 88). The predicative assumes that
what both is and is not F can only be the object of belief. Why would the
lovers of sights and sounds grant this? The existential reading "consigns the
objects of belief to the realm of 'half-existent"' (1978, 125); "But why
should the sightlovers agree at the outset that every object of belief only
half-exists?" (1990, 89). At the outset? As the above outline of the argu-
ment shows, the claim that the objects of belief both are and are not is the
conclusion of a complex argument (step [13], 478d5-12), not an initial
assumption. Fine, however, lists among the "opening premises" the claim
that "Belief is set over what is and is not" (1990, 87). This is what enables
her to maintain that the existential and predicative readings controversially
separate the objects of knowledge and belief and thus violate the dialectical
requirement. Yet the actual opening premises in the text, (1)-(4) above, far
from maintaining that belief is set over what is and is not, do not even
assume that there is anything either between knowledge and ignorance or
between being and not-being. All they assert is that if there is something
between being and not-being and if there is something between knowledge
and ignorance, the latter must be set over the former.7
If the text is read correctly, is there anything controversial in the opening
premises under the existential or predicative readings? Let us first consider
the existential reading:8
(le) Knowledge is of what exists.

(2e) Ignorance is of what does not exist.
(3e) If something both exists and does not exist, then it lies between pure existence
and absolute non-existence.
view, but is meant to reveal the deficiencies of the interlocutors (Meinung und Wissen in
der Philosophie Platons [Berlin: Walter de Gruyter, 1974], see especially 107-9, 121,
130). My reading of the argument does not rule out this dialectical interpretation ("Pla-
to" can be understood in this paper as shorthand for "what is defended without being
refuted in Plato's dialogues"), but it does reject the reasons that have been offered for
seeing the premises as unworthy of Plato (see, e.g., Stokes, 120-1, 123, 130, and notes
12, 24, 31, 36 and 50 below).
7In her statement of step (4) in 1978, Fine leaves out the crucial qualification: "if there
is such a thing" (123). This misreading only becomes more prominent in 1990 where
Fine is not sticking as closely to the text.
8 The existential reading is defended by, among others, R. C. Cross and A. D. Woozley,
Plato's Republic: A Philosophical Commentary (London: Macmillan & Co. Ltd., 1964),
and Michael C. Stokes, in the article cited above.
250

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
(4e) We must search for something between knowledge and ignorance, if there is
such a thing, to be set over what is between existence and non-existence.
In 1978 Fine does not argue that any of these premises need be contro-
versial; the only premise she finds controversial is the one that, as we have
seen, does not exist. In 1990, however, she sees a problem with (le):9 she
allows that it is plausible if "one takes knowledge to be some sort of ac-
quaintance," "But it is unclear why we should assume at the outset that
knowledge consists in or requires acquaintance with what is known" (89).
But why should we assume at the outset that the knowledge in question is
propositional?'? If we instead regard the structure of knowing as parallel to
that of perceiving, (le) becomes perfectly plausible: in this case it is natural
to assume that one can no more know what does not exist than one can see
what does not exist." This parallel is also needed to explain (2e), which
might otherwise appear blatantly false: for is it not obvious that I can be
ignorant of things that yet exist?'2 We need to recognize that Socrates is
here talking about what ignorance is "set over," what its object is. In being
totally ignorant of the form of beauty, the object of my ignorance is not the
form of beauty - it is not an object for me at all - but nothing. If knowl-
edge, like seeing, is a direct cognitive relation to some object, then igno-
rance, like blindness, is simply the absence of such a relation or, stated
differently, a relation to nothing.
Would, then, this parallel between knowing and perceiving be contro-
versial to the lovers of sights and sounds? The opposite is clearly the case.
At this point the lovers have no reason not to assume that "what is" in-
cludes, or is even restricted to, their sights and sounds and that knowing is
9 For the same objection, see Annas, 196, and Charles Kahn, "Some Philosophical Uses
of 'to be' in Plato," Phronesis 26 (1981): 112.
'1 J. Hintikka has claimed, I think rightly, that with the Ancient Greeks there was a
"tendency to think of knowledge in terms of some sort of direct acquaintance with the
objects of knowledge, e.g., in terms of seeing or of witnessing them" ("Time, Truth and
Knowledge in Ancient Greek Philosophy," American Philosophical Quarterly 4, no. 1
[1967]: 6).
" As Neil Cooper comments, "One could no more know something which did not exist
than meet someone who 'wasn't there"' ("Between Knowledge and Ignorance," Phrone-
sis 31, no. 3 [1986]: 234). At Theaetetus 188e-189b Socrates infers from the impossibil-
ity of seeing, hearing or touching what is not, the impossibility of believing what is not.
When Republic V 478b6-c6 also declares impossible believing what is not, we have
reason to assume that the same analogy made explicit in the Theaetetus is here implied.
12 This is Stokes' objection. In order to avoid this problem, he interprets Socrates as
meaning at 477a9-bl that "All non-existents are the objects of ignorance," rather than,
"All and only non-existents are the objects of ignorance" (114). Yet Stokes recognizes
that the stronger reading is the one found later in the argument (125).
251

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
nothing more than perceiving. What would be controversial to them is a
conception of knowledge as being primarily of propositions, since, as Glau-
con tells us, they never willingly concern themselves with logoi or pursue
that kind of activity (475d4-5).
Fine also objects in 1990 that the existential reading of the opening
premises assumes the "difficult notion of 'half-existence"' (89). Yet this is
again a misreading of the text: premise (3e) says only that if there is some-
thing that both exists and does not exist, it must lie between pure existence
and absolute non-existence. We are not yet shown or asked to assume that
there is such a thing; nor must we understand the meaning of the phrase
'between pure existence and absolute non-existence' in order to grant the
logical point that what is both x and y is in some sense between x and y.'3
That this objection also has no force against the existential reading of the
argument as a whole will be seen below.
Let us now turn to the predicative reading:"
(ip) Knowledge is of what is F.

(2p) Ignorance is of what is not F.
(3p) If something both is F and is not F, then it lies between pure F-ness and pure
not-F-ness.
(4p) We must search for something between knowledge and ignorance, if there is
such a thing, to be set over what is between F-ness and not-F-ness.
Fine's main objection to this reading is that it is by no means obvious or

noncontroversial that one cannot know that something is both F and not-F.'"
Why can I not know that a pencil is equal and not-equal (1990, 89) or that
an action is just and not-just (1978, 125; see also 135)? Yet it is possible
that the type of knowledge in question here is not knowledge that a thing is
F or not F, but rather knowledge of what F is. In this case, what the initial
'3 This is recognized by Stokes, who even claims that the antecedent of the conditional
(3e) must at this point appear a contradiction to the lovers of sights and sounds, and from
a contradiction anything can be granted to follow (1 13).
14 The predicative reading is supported by, among others, Vlastos, "Degrees of Reality
in Plato," in Platonic Studies (Princeton University Press, 1973), 63, and Annas, ch. 8.
15 Annas' defence of the predicative reading misses the point: "1 can only know of
something that it is long if it is long. If it isn't long, I can't know that it is long" (198).
But Fine's question is: If it isn't long, why can't I know that it isn't long? Cooper makes
the same objection (233), though he recognizes that the predicative reading is not im-
plausible if understood as compatible with the existential reading: "Of course, he [Plato]
could have meant that anybody who knows something, knows something which is F for
some F, since something which, for all F, was not F, could not be known. To be is to
have a property. Plato could well have meant something as complex as this and if he did,
we may regard the predicative and the existential interpretations as being compatible"
(233-4, my emphasis). This is a suggestion that will be defended below.
252

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
premises maintain on the predicative reading is the following: knowledge of
what F is must have as its object what is truly F, not what is not F nor what
is both F and not F. Would this claim be acceptable to the lovers of sights
and sounds? Could they, for example, agree that knowledge of what beauty
is must have as its object what is truly beautiful? Since they presumably
think that nothing is more beautiful than the sights and sounds of which
they are so enamoured, why would they object? In their minds (1p) would
simply state: one cannot know what beauty is by looking at sensible objects
that are not truly beautiful. The suggestion that knowledge is not possible in
relation to what is both F and not F need not trouble them at this point: they
have not been asked to assume that there in fact is such a thing; and though
they are eventually forced to grant that the objects they love are as ugly as
they are beautiful, this is clearly not something that would immediately
occur to them. In addition, the opening premises are very general and the
lovers of sights and sounds need not have only the predicative reading in
mind as distinct from the existential. As I show below, the two readings are
perfectly compatible and enforce one another if understood together.
Fine's argument for the superiority of the "contents analysis" of the ve-
ridical reading in large part depends on showing that the other two readings
violate the dialectical requirement, since she does not deny that they are
compatible with the text (see 1978, 122). Yet the only opening premise
identified by Fine that violates this requirement has been seen not to be an
opening premise at all, but a demonstrated conclusion. Even this conclusion
(13), one can add, need not be objectionable to the lovers of sights and
sounds since it does not yet show that there is a class of things that are and
are not and that beautiful sights and sounds fall into this class. Therefore, to
the extent that Fine's interpretation is supported by an appeal to the dialecti-
cal requirement - and it in large part is - it collapses.
IV. The Predicative Reading
Fine's criticism of the existential and predicative readings is not confined to

their supposed violation of the dialectical requirement. She also sees them
as not providing a philosophically coherent account of the argument as a
whole. Of course, as Fine recognizes, this alone does not prove these read-
ings to be incorrect, since the argument may in fact be incoherent. Never-
theless, before turning to the incompatibility of Fine's own reading with the
text, I wish to show that it is not the only philosophically coherent one.
Fine's major philosophical objection against the predicative reading has
already met with a response in the previous section: this reading need not
interpret the argument as absurdly maintaining that one cannot know that
253

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
something is both F and not F, but can see it as claiming that knowledge of
what F is cannot be "set over" what is both F and not F. Yet is even this last
claim defensible? Two questions need to be answered here: 1) What does
knowing what F is have to do with knowing what is truly and only F? 2) As
long as I know that something is F, why should the fact that it is also not F
prevent it from providing me with knowledge of what F is? The answer to
the first question can be found in Charles Kahn's extremely important ob-
servation that "To speak of what F is or of what is (truly) F for Plato, is to
speak of the same thing" ("Philosophical Uses," 109). Knowledge of what
is (truly) F is not simply, nor even primarily, knowledge that F belongs to
some x as a property, but rather knowledge of what F itself is. If we are to
speak of a predicative "is" here, we need to understand predication in the
way in which R. E. Allen has claimed Plato understands it: in the predi-
cation "x is F," what is primarily referred to and named is not the x, but the
F (so that F is not really a "predicate" in the modem sense); what the
predication expresses is the relation of the x to this F, whether it be a
relation of identity or a relation of approximation and imperfect instantia-
tion.'6 The predicative "is" in this case does not assign the F to the x but
rather the x to the F: it is the "property" that is here primary, while the
"subject" is secondary. The ontological presupposition of this conception of
predication is that a thing is not a substance distinct from and prior to its
"properties," but is constituted by these properties, either because it is iden-
tical with them ("the F itself') or because it is nothing but an imperfect
imitation of them: a sensible object is simply, as it were, a clouded mirror
that inadequately reflects what hardness is or what beauty is; more accurate-
ly, it is nothing but the dim reflection itself.'7
An answer to the second question is found in Book VII. At 524a, we are
told that a perception that is "set over" (epi) the hard is necessarily also "set
16 "'F' is a name, a name whose prime designate is a Form: 'F' names the F' (R. E.
Allen, "Participation and Predication in Plato's Middle Dialogues," in Plato: A Collec-
tion of Critical Essays, vol. 1, ed. Gregory Vlastos [Garden City, New York: Anchor
Books, Doubleday, 1971], 169). Allen concludes that Plato's is "a theory of predication
without predicates" (170).
'7 As Allen observes, particulars "have no independent ontological status; they are pure-
ly relational entities, entities which derive their whole character and existence from
Forms" (181); as "exemplifications" they are "not ... substances in which qualities inhere
but ... relational entities, entities in which resemblance and dependence so combine as to
destroy the possibility of substantiality" (183). In the dialogues, this point is most force-
fully stated at Timaeus 52c2-5. Rafael Ferber rightly characterizes as a misunderstanding
the view "daB Plato seine Sinnendinge wie Aristoteles als Dinge (ousiai aisthetai) kon-
zipiert hat," and chooses instead to speak of "Sinnesphanomenen" (Platos Idee des
Guten, 2d ed. [Sankt Augustin: Academia Verlag Richarz, 1989], 27).
254

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
over" (epi) the soft, with the result that the hard and the soft are confounded
(oruyxEX,u[va, 524c) and presented to the perplexed soul as being the

same (45 TatiTO6v, 524a3). Yet this again appears an odd claim. Can I not
perceive that one and the same thing is both hard and soft without confusing
hardness with softness? Is it not the case that my perception signifies only
the identity of the thing that possesses these two properties and not the
identity of the properties themselves? To understand the soul's perplexity
here we must recognize that, according to Socrates' description and in line
with the ontological presupposition described above, perception is "set
over" and "signifies" (aorIIaiveL) not the sensible object per se as a sub-
stance distinct from its properties, but the properties themselves.'8 In this
case, it is reasonable to claim that perception confounds or identifies these
properties in perceiving them as in the same sensible object without being
able to distinguish or abstract them from this object. Perception is perfectly
accurate in perceiving the hardness or softness of an object: what it mani-
fests deficiently (tv&6w bXkoi, 523e7) is what hardness and softness
themselves are. These observations show that Kahn's equation is confirmed
negatively by perception: perception of what is not truly F (i.e., is no more
F than not-F) is for Plato equivalent to confused and contradictory percep-
tion of what F is.'9
18 Socrates introduces this discussion by asking (523e3-6) if sight sees adequately

(Exavdog 6Q() largeness (T6 [EyrOoq) and smallness (f oIiLxoonTg) and if touch reveals
adequately thickness (3rdxog) and thinness (XentT6Tq), softness ([1aXax6TrJq) and hard-
ness ((JX11Q0T%). What he does not question is the ability of sight or touch to perceive
that an object is small or large, hard or soft; such an ability is presupposed by the
discussion. At 524a6-8 Socrates claims that perception of something as hard and soft
puts the soul in aporia concerning "TL 7tOTE OTU[LVCLEL aitbTfj 1 CdtkOJloL; T6 oxXQOQV
...." The word t6 axXlQ6v must here be equivalent in meaning to ox~QOJT1Jg: the
property of hardness, not the thing that has this property.
'9 In the claim that a sensible object is both F and not F, what are the possible values for
F? Socrates' examples in both Books V and VII are opposites: sensible objects are hard
and not hard, heavy and not heavy, just and not just. But then what about terms that have
no opposites? In Book VII (523c 1 -d6) Socrates himself gives us an example of such a
term: we do not perceive a finger as also not a finger; such a perception, therefore, does
not cause perplexity. But then can we have knowledge, as opposed to belief, about things
which, like fingers, have no opposites? This is the conclusion that Annas reaches and
that enables her to reject the "two worlds theory" (209-15). Fine apparently accepts
Annas's argument (1990, 88, n.8), though she of course wants to go further in maintain-
ing that any claims about sensibles, and not simply those that do not concern opposites,
can be known. Does Annas' argument, then, provide us with a way of rejecting a "two
worlds theory" without subscribing to Fine's more extreme position? Not if we heed
Nicholas P. White's observation: "Notice that Plato does not say here that 'finger'
applies to sensible objects without qualification. What he says is that it does not give
simultaneous appearances of applying and not applying. But it will still apply to a
255

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
Socrates' argument, on this reading, is that such perception can lead to no
more than doxa, where doxa is primarily to be understood, like the knowl-
edge with which it is contrasted, not as doxa that something is F, but as
doxa concerning what F itself is. As the dream analogy immediately preced-
ing the argument shows, doxa is characterized by a failure to distinguish
what F itself is from the things that only imitate and participate in F, and
thus by a failure to see clearly, with waking vision, what F is (476c2-8).
Here we touch upon an important controversy concerning the meaning of
doxa: some scholars have maintained that doxa is to be identified with
judgement, while others have insisted that doxa is more analogous to per-
ception, being essentially intuitive and even nonpropositional.20 In a sense
both sides are right. It is undeniable that Plato sees the person whose cogni-
tive state is doxa as believing that certain things are the case and thus as
affirming numerous true or false propositions; thus, at step (14B) of the
argument, the content of belief is identified with v6tZtLa. However, it is
nevertheless possible that when Socrates speaks of doxa he is referring
primarily not to the assertion that something is F, but rather to the confused
intuition of F itself that underlies and guides this assertion. This possibility
is supported by the apparent association of doxa with confused and contra-
dictory perception of what F is, a perception that can of course be the basis
for all kinds of true or false statements, but cannot be identified with any or
sensible object only with temporal qualification" (A Companion to Plato's Republic

[Indianapolis: Hackett, 1979], 198). As we know from the Symposium (21 1a3), one
possible meaning of the claim that a sensible object is both F and not-F is that it is at one
time F and at another time not-F. Therefore, there is no reason to restrict the argument of
Book V to opposites.
20 R. S. Bluck claims that "the Greek word 66ta is not necessarily to be associated with
propositions. It may refer simply to what a thing 'seems like' ... . There is no reason to
suppose that at the time of the Republic Plato associated 66ba any more than tnLrrLTJ
with propositions ..." ("Knowledge by Acquaintance in Plato's Theaetetus," Mind 72
[1963]: 259). N. R. Murphy also emphasizes the intuitive character of doxa (An In-
terpretation of Plato's Republic [Oxford: Clarendon Press, 1951], 121). Others, like
Norman Gulley, maintain that "doxa is 'propositional"' (Plato's Theory of Knowledge
[London: Methuen, 1962], 65). Behind this disagreement is the general debate concern-
ing whether doxa is 1) something analogous to perception (repr6sentation, Vorstellung)
or 2) a form of judgment (Meinung), a debate discussed and documented by Yvon
Lafrance, La Theorie Platonicienne de la Doxa (Montreal: Bellarmin; Paris: Les Belles
Lettres, 1981), especially p. 12, and J. Sprute, Der Begriff der Doxa in der Platonischen
Philosophie (Gottingen: Vandenhoeck & Ruprecht, 1962). Even though their critique of
Plato's argument depends on taking the objects of knowledge and belief to be simply
"that-clauses" (171-3), Cross and Woozley admit that "there are times here too when he
[Plato] seems to mean by belief some sort of immediate awareness ... " (176; see also
143-4).
256

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
all of these statements. Furthermore, if we consider Socrates' critique of the
lovers of sights and sounds prior to the present argument, we see that his
reason for attributing only doxa to them is not that they affirm false or
unjustified propositions about what things are beautiful, but that they fail to
see (6Qav: 476b7, blO, dl) what beauty itself is. This is not to say that the
lovers of sights and sounds have no perception of beauty whatsoever - no
more than the passage from Book VII intends to say that perception pro-
vides no awareness of heaviness or hardness _,21 but rather that their percep-
tion of beauty, as confined to sensible objects that are as ugly as they are
beautiful, is equally "set over" ugliness and therefore cannot clearly sep-
arate beauty from its opposite.
Not only do the above observations suggest that the confused awareness
of F that characterizes doxa is perceptual rather than propositional,22 but the
predicative reading as I have interpreted it still assigns knowledge and
belief to objects: knowledge of what F is is not knowledge of some uni-
versal definition, but knowledge of what is truly F; belief about what F is is
"set over" objects that are both F and not-F, not propositions about F.
Therefore, the conception of knowledge required by the predicative reading
not only does not contradict, but complements the one required by the
existential reading: namely, the conception of knowledge as some sort of
direct acquaintance with objects analogous to perception. The connection
can be seen if we characterize knowledge and belief as follows: we know
what F is through direct acquaintance with what is truly F; we can only
"believe" (as opposed to knowing) "what F is" (this is bad English, but
apparently good Greek)23 through perception of sensible objects that are
21 At 505e1 -2 the soul is described as perplexed (deuoQEiv) about what the good is (rt
7roT' ot?iv) in language that parallels the later description of the soul as perplexed about
what hardness is (524a7-8). At 505e 1 Socrates makes clear that this perplexity does not
prevent the soul from dno0JiavTEvo[v1q TL ctvatL.
22 Though the relation between doxa and perception is not made clear in the text, it must
be a very close one, since the characteristic of being both F and not-F that makes
sensibles objects of doxa is the same characteristic they have as objects of perception
(compare 479b6-7 and 524al-0). In the Timaeus we are told that what becomes is b66n
3 ET' Ctto Eoetos 3EQLXr1t6V (52a7).
3 Stokes points to an apparent problem with characterizing doxa as a dunamis assigned
to objects, namely, that "the Lexicon suggests the absence of any external object for the
verb 'opine' (doxazein) in normal Classical Greek. One can know, but apparently not
'opine' something externally.... This might seem a solid obstacle to taking the things
opinion stands 'over against' in this argument as external rather than internal objects"
(124). Yet Stokes himself recognizes an unambiguous exception: Theaetetus 209b2-3,
where at 6bo~taov means "I had a belief about you." Pace Stokes, there is no in-
dication that Socrates' Greek is here unnatural or forced (see also 190d4 & d7).
257

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
both F and not-F. This characterization of knowledge and belief also has the
advantage, as will be seen more fully below, of preserving common ground
between them while assigning them to different objects: what knowledge
understands clearly and distinctly through acquaintance with its object is the
same as what belief understands confusedly through acquaintance with its
objects: what F is. In short, knowledge is here understanding and acquaint-
ance, understanding achievable only in direct acquaintance with certain
objects; doxa is the failure to achieve such understanding on account of the
character of the objects with which it is acquainted.24
If understood in this way, the predicative reading does not make Socrates
argue for the absurd view that, in the words of Fine, I can only believe, but
not know that something is both equal and unequal. This "knowledge-that"
is simply not the knowledge at issue here. This defense of the predicative
reading, however, need not be made at the cost of the existential reading. As
the next section will show, the existential reading can itself be made coher-
ent; furthermore, it does not exclude, but needs the predicative reading.
V. The Compatibility of the Existential with the Predicative Reading
Fine makes two related philosophical objections against the existential read-
ing. The first is that it requires a notion of degrees of existence that makes
24 In a very important paper, Nicholas D. Smith argues that "we must construe Plato's
concept [of episteme] to be a blend of knowledge by acquaintance and knowledge what"
("Knowledge by Acquaintance and 'Knowing What' in Plato's Republic," Dialogue 18
[1979]: 283). Platonic episteme, according to Smith, involves acquaintance because it
has a direct object (282; he sees the word epi as an indication of this); it also involves
"knowledge what" because it requires having certain truths that are essentially related to
the nature of the object (282). Though Smith seeks to distinguish this latter "knowledge
what" from propositional "knowledge that," the only real difference appears to be that
"knowledge what" is restricted to propositions about a thing's essence. Smith's account
therefore leaves unclear how exactly this propositional "knowledge what" is "blended"
with acquaintance. My own suspicion is that even what Smith calls "knowledge what"
can be nonpropositional and that therefore what he characterizes as a "blend" is more of
an identity. In defending a "predicative" against an "existential" reading, Annas charac-
terizes episterne solely as propositional "knowledge what" (192) or understanding (193,
200, 212) achieved, not through an intuition of objects nor through a justification that
provides certainty, but through setting the propositions one affirms in the wider context
of other propositions one affirms and their mutual explanatory relationships (200). In
thus maintaining that one's beliefs become knowledge through systematic mutual expla-
nation, Annas comes very close to Fine's view that Plato was a "coherentist." For a
critique of this view, see chapters 7 & 8 of my Dialectic and Dialogue: Plato's Practice
of Philosophical Inquiry (forthcoming).
258

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
no sense.25 Something either exists or does not exist; to claim that it "half-
exists" is, we are told, to talk nonsense. A thing's properties can vary in
degree, e. g., it can be more or less beautiful, cold, red, hard, etc., but a
thing's existence cannot in the same way be more or less. Fine's second
objection also assumes this distinction between a thing's properties and its
existence: the existential reading, she claims, would have to see the argu-
ment as inferring, from the claim that a sensible object both is and is not F
(14A), that it both exists and does not exist. But such an inference is clearly
invalid since "is-p does not carry existential import in this way. One cannot
infer from the fact that x is not F that x does not (fully) exist. The paper on
which I am now writing is not green, nonetheless, it exists" (1978, 134).
Julia Annas for the same reason sees the existential reading as giving Plato
"a very silly argument" (197).26
The argument is "silly," however, only on the assumption of a sharp
distinction between the existential and predicative senses of the word "is."
Yet, as the above discussion of predication has indicated, Plato does not
appear to distinguish a thing's "substance" from its "properties" or its exist-
ence from its essence in the way required by such a distinction. In addition,
Charles Kahn has convincingly shown the anachronism of reading this dis-
tinction back into the Ancient Greeks.27 One of the important reasons he
25 Annas asserts that "the notion of degrees of existence does not make sense" (196).
Vlastos' well-known argument that "degrees of reality" in Plato cannot be "degrees of
existence" is based primarily on the same assumption: "As we commonly use the word
'existence,' degrees of it (as distinct from degrees of perfection of things in existence)
make no sense whatsoever; the idea of one individual existing more, or less, than another
would be a rank absurdity. It would take strong, unambiguous evidence to establish that
Plato had any such thought in mind when he spoke of some things as being more, or less
real, than others" ("Degrees of Reality," 65). But why assume that Plato's understanding
of "existence" was the same as the one common today? And given this assumption, what
could possibly count as "strong, unambiguous evidence" of Plato's commitment to de-
grees of existence? The studies by Charles Kahn discussed below undermine Vlastos'
approach in showing that the Greeks did not possess our concept of existence. (In his
presidential address to the American Philosophical Association, Vlastos goes so far as to
claim, in a leap far beyond any of the evidence he presents, that the Greek language
itself "rules out as a monstrosity a tertium quid between existence and non-existence":
"A Metaphysical Paradox," in Platonic Studies [Princeton University Press, 1973], 49).
Ferber, in defending the notion of "Existenzgrade" (25-6), rightly criticizes the Vlasto-
sian objection as fixated on contemporary usage. Vlastos' "common sense" objection
that Plato would not want "to undermine our faith in the existence of the beds we sleep
on, buy and sell, etc...." ("Degrees of Reality," 65) is besides the point. The meta-
physical question of whether or not a bed exists as fully as its form is irrelevant to our
getting a good night's sleep.
26 See also Kahn, "Philosophical Uses," 113.
27 See "The Greek Verb 'To Be' and the Concept of Being," Phronesis 26 (1981):
259

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
gives is that "for Aristotle, as for Plato, existence is always ctvCL Tt, being
something or other, being something definite. There is no concept of exist-
ence as such, for subjects of an indeterminate nature" ("Existence in Greek
Philosophy," 333).28 A thing's existence is not for Plato sharply distinguish-
able from its being what it is, e.g., beautiful. To say that it "is" in the sense
of "exists" is always at the same time to say that it "is such and such."
Conversely, to say that it "is such and such" is at the same time to say that it
"exists," since, as Lesley Brown has rightly pointed out, Plato "nowhere
allows that X is F does not entail X is but is consistent with X is not."29 But
105-34 (especially 247-9 & 259) and "Why Existence Does Not Emerge as a Distinct
Concept in Greek Philosophy," Archiv Pfir Geschichte der Philosophie 58 (1976): 324-
334. In addition to the existential and predicative meanings, Kahn also recognizes a
veridical meaning that, unlike the one appealed to by Fine, understands truth objectively,
not simply as the property of propositions ("The Concept of Being," 250; "Existence,"
328; "Philosophical Uses," 106, 115, 126; Kahn contrasts his own veridical reading with
Fine's on p. 130, n. 18; for his recognition, however, that Plato does not always keep
distinct the two senses of truth, see "Philosophical Uses," 126, and "The Concept of
Being," 252). The objective veridical meaning of being is, according to Kahn's own
account, compatible with the existential and predicative meanings: in fact, he sees the
latter two as only special cases of the veridical, which he considers to be the funda-
mental meaning of being for the Greeks ("The Concept of Being," 251-2). Ferber also
defends what he calls an "alethische" reading of the "is" (he intentionally avoids the
term "veridical"; see p. 281, n. 8), while denying that the truth in question here is the
truth of propositions: "Offensichtlich ist mit dieser Wahrheit keine Aussage- oder Ur-
teils-, sondern eine ontische Wahrheit gemeint: Ihr Sitz ist nicht der Mensch, sondern
das Sein" (22). He identifies this truth with Echtheit. An account of how an objective
sense of truth restricts knowledge to the forms is to be found in Sprute, Der Begriffder
Doxa, 78. It can be argued, therefore, that Fine's "veridical" reading does not even do
justice to Plato's conception of truth.
28 Elsewhere Kahn makes this point by asserting that "Every truth for Plato can properly
be expressed in the copula form 'X is Y.' Even the existential proposition can be so
expressed ... . 'Justice exists' is expressed as 'Justice is something (ti)" ("Linguistic
Relativism and the Greek Project of Ontology," in The Question of Being [University
Park: Pennsylvania State University Press, 1978], 39). After pointing out that "the cop-
ula proposition in turn is to be interpreted ontologically in terms of participation: 'X is Y'
is true only if and because X participates in Y-ness or in the Y' (39-40), Kahn concludes:
"In the last analysis, I suggest, Plato's concept of being is the being-of-a-Form, or the
being-related-to-a-Form by way of participation" (40), so that, Kahn implies, there is no
concept of being as existence distinct from either.
29 "Being in the Sophist: A Syntactical Enquiry," Oxford Studies in Ancient Philosophy,
vol. 4 (1986): 56. Brown criticizes G. E. L. Owen's sharp distinction between a com-
plete use of the word "is" (as in "X is," i.e., exists) and an incomplete use (as in "X is
F'), a distinction upon which Owen rests his thesis that only the incomplete use is to be
found in the Sophist ("Plato on Not-Being," in Plato: A Collection of Critical Essays,
vol 1., ed. Gregory Viastos [Garden City, New York: Anchor Books, Doubleday, 197 1 1,
223-4). Brown also criticizes Vlastos' attribution to Plato of a parallel sharp distinction
260

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
if to exist is to exist as a determinate something, and if a sensible object is
no more F than not-F, then this indeterminacy of its F-ness prevents it from
fully existing. On the other hand, to be fully F is to exist fully. On this view,
what is absurd is not the notion of degrees of existence, but the modem
notion that a sensible object can be imperfectly beautiful and yet perfectly
exist, that its beauty and existence can be kept so distinct that the imperfec-
tion of the one does not affect the other.
The sharp, exclusive distinction between predicative and existential
meanings of the word "is" with which Fine and others approach the text
therefore depends on philosophical assumptions that Plato need not, and
evidently did not, share.30 That both meanings can be found in the text is a
result not of Plato's "confusion,"'" but of the fact that for him they are
between existential and predicative ("real") senses of the word "is" ("Metaphysical
Paradox," 46-9; "Degrees of Reality," 59-66; for the criticisms see Brown, 56, 64).
Brown provides a detailed and clear account of the distinction between complete and
incomplete uses of the "is" on pp. 52-7. What enables her to reject the sharp distinction
made by Owen and Vlastos is the description of a complete use that, while having no
complement, allows a complement (53). This allows Brown to claim, in addition to what
is cited above, that Plato does not draw a distinction "between the non-existent and that
which isn't anything at all" (62; see also 60, 61, n.16). Brown's general conclusion is
that the complete and incomplete uses are related as follows: "X is (complete use) entails
X is something and X is F entails X is. X is not (complete use) entails X is not anything at
air' (69). Ferber appears to be getting at a similar point when he characterizes the
modem use of the existential "is" as "nonindexical" and Plato's in contrast as indexical
(25). He specifically argues that Plato's existential "is" stands in relation to "Dauer" and
"Wahrheit bzw. Echtheit": i.e., that which "is more" exists more genuinely and longer
than that which "is less" (25).
30 Even Kahn, who opposes reading the distinction back into the Ancient Greeks (see n.
28 above), appears to take it for granted when he rejects the existential reading of
Republic V in favor of the veridical and predicative readings ("Philosophical Uses,"
112-3). Stokes assumes the same dichotomy in defending an existential reading, with the
consequence that he must explain the "leap" from the predicative "is" of step (14) to the
existential "is" of the rest of the argument (129). Yet not only is his explanation inade-
quate (as he appears to recognize, since he makes the lovers of sights and sounds
responsible for step [14], 130), but it is unnecessary, since there is no leap here. Ferber,
on the other hand, who also defends an existential reading, argues that the existential
"is" is not ruled out, but rather presupposed, by the predicative or "copulative" is (21-3).
Though Gerasimos Santas, like Ferber and unlike Kahn or Stokes, claims that the formu-
la "is and is not" can include both predicative and existential claims ("Knowledge and
Belief in Plato's Republic," in Greek Studies in the Philosophy and History of Science,
ed. Pantelis Nicolacopoulos, Boston Studies in the Philosophy of Science, vol. 121
[Dordrecht: Kluwer Academic Publishers, 1990], 52), his account of the forrns depends
on keeping these claims sharply distinct (52-4).
31 Vlastos accuses Plato of "defective analysis" in failing to explain the difference
between the predicative "is," which admits of degrees, and the existential "is," which
does not ("Metaphysical Paradox," 56).
261

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
simply different aspects of the same meaning: the existential "is" does not
mean existence as distinct from, and completely indeterminate with regard
to, essence, but rather determinate existence as such and such; the predi-
cative "is" does not assign F as a predicate to some x that may or may not
exist, but assigns x to that F by which it exists. The existential and predi-
cative readings are therefore reconcilable. Knowledge is assigned to "what
is" in the sense that to know what F is is to have direct acquaintance with
what truly exists as truly F. For example, I know what beauty is through
direct acquaintance with what determinately and completely exists as deter-
minately and completely beautiful. Belief is set over "what is and is not" in
the sense that merely to "believe what F is" is to be confined to acquaint-
ance with what does not fully exist by being no more F than not F. For
example, I can only have belief about what beauty is in perceiving things
that only imperfectly exist by being only imperfectly beautiful.
If the predicative and existential readings do not violate the dialectical
requirement, and if they, along with the two conceptions of knowledge they
require, i.e., "knowledge what" and "acquaintance," are compatible and
complementary, then what reason remains for rejecting these readings in
favor of Fine's? The lack of such a reason, of course, does not in itself show
that Fine's reading is incorrect. We must still examine this reading's own
merits when measured up against the text.
Vl. The Incompatibility of Fine's Reading with the Text
The existential and predicative readings as characterized above all belong

within what Fine calls an "objects analysis" and are thus incompatible with
her own veridical reading understood as a "contents analysis." While I have
argued that versions of the existential and predicative readings are compat-
ible with the text and with each other, Fine's propositional reading, I wish
to show, is not compatible with the text. The four following objections are
not to be given equal weight: the first two, though important, are not deci-
sive, while each of the last two is by itself sufficient to render Fine's
interpretation untenable.
i) The first problem is that Fine's propositional reading is at least hard to
square with some of the language of the argument that is more suggestive of
objects. Specifically, the argument begins with the claim that knowledge
must be not only of what is, but of something (ti). In the course of the later
argument that belief cannot be "set over" what is not, Socrates asks, "Is it
not the case that not-being would be most properly addressed not as some
one thing (otOx lv TL), but as nothing (iin&v)?" (478bl2-cl). These claims
certainly appear to assign knowledge to what exists and ignorance to what
262

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
does not exist. How are we to explain them on Fine's reading? If by "what
is" Socrates is referring to true propositions, is he suggesting at 476e7-
477al that only true propositions are something? If by "what is not" he is
referring to false propositions, is he suggesting at 478bl2-cl that false
propositions are nothing?32
ii) Another problem is the difficulty of fitting Fine's propositional read-
ing into the broader context of the argument. As Michael C. Stokes has
argued, and as I have also suggested above, the dispute with the lovers of
sights and sounds concerns the knowledge of beauty itself and the existence
of beauty itself, not the truth of propositions ( 10-1). As we have seen, the
lovers of sights and sounds are described as absolutely unconcerned with
propositions. Their dispute with the philosophers is about what objects can
be known. They think that in perceiving beautiful sights and sounds they
know exactly what beauty is; the philosophers tell them that such objects do
not provide knowledge. Given this context, it is no wonder that, as Santas
points out, in the argument itself "we hear nothing about logos."33 This is,
however, extremely puzzling if, as Fine maintains, the argument is in fact
primarily about logoi and not objects.
iii) The first major difficulty is reconciling Fine's interpretation with
what the present argument is in large part meant to prove: that knowledge,
belief and ignorance are "set over" different "things" (this is Fine's word
and I use it so as not to assume what Fine denies, i.e., that these things are
objects rather than contents). Far from seeing any difficulty here, Fine as-
serts that her interpretation succeeds where the two worlds interpretation
(under both existential and predicative readings) fails: i.e., in explaining the
argument at steps (5)-(7) that knowledge and belief, being different powers
doing different work, must also be set over different things. Fine joins
others in objecting: "For why should knowledge and belief not be different
capacities with different work on the same things?" (1978, 128; see also
1990, 90).
It first must be pointed out that Fine is wrong in suggesting that the
existential and predicative readings cannot answer this question. The an-
swer lies ready to hand if we take Fine's own dialectical requirement seri-
ously and ask: what are the lovers of sights and sounds likely to consider
paradigm dunameis? Seeing and hearing, of course. But the two conditions
Socrates claims must be met by different dunameis - i.e., that they must do
32 These claims are listed by Fine only in the 1978 paper (123 & 130). Fine on pp.
130-131 appears to recognize the problems these claims present. For a criticism of her
attempted solution, see below.
3 "Knowledge and Belief," 49.
263

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
different work and be set over different things - are fulfilled by the different
dunameis of perceiving, as Michael C. Stokes has observed:
"Now, sight, hearing and the other senses are capacities each distinguished from
the others by both province and effect. Notoriously you cannot see sounds, smell
colours, hear tastes, touch smells; each sense has in that way its own province ....
Of course their effects are different too: to see is not the same as to hear, and so on.
Once we set Socrates' point in its original context of an argument against the
sightlovers, it becomes clear" (120).
This is why when Socrates introduces the notion of dunamis his examples
are none other than seeing and hearing (477c3). Would, then, Socrates'
characterization of the dunameis of believing and knowing as analogous in
structure to the dunameis of seeing and hearing be acceptable to the lovers
of sights and sounds? This question answers itself. Furthermore, Socrates'
assumption has legitimacy beyond its plausibility for the lovers if knowl-
edge and belief are understood to involve some sort of direct acquaintance
with objects.34 The important conclusion in the present context is the fol-
lowing: an "objects analysis," which sees belief and knowledge as assigned,
like seeing and hearing, to objects and not to propositions, can at least make
as much sense of the argument in steps (5)-(7) as Fine's "contents analy-
SiS."35
However, I wish to go further and argue that Fine's "contents analysis"
cannot make sense of the present passage nor of the argument's general
thesis that knowledge, belief and ignorance are "set over" different things.
As we have seen, Fine explains steps (5)-(7) as follows: the work of knowl-
edge does not err and therefore succeeds in collecting only truths; belief errs
in its work and therefore collects both truths and falsehoods (1978, 128).
34 Stokes does not attribute to Plato himself the view that the powers of sight and
hearing are analogous to those of belief and knowledge, but asserts that "what is true for
Plato is irrelevant in an argument designed to convince the sightlovers" (123). Michael
L. Morgan likewise sees the "dunamis-object" principle as appealing to the preoccupa-
tion with seeing and hearing on the part of the lovers of sights and sounds and as "not
even intended as a rule Plato is committed to" ("Belief, Knowledge, and Learning in
Plato's Middle Dialogues," Canadian Journal of Philosophy, Supplementary Vol. 9
[19831: 97).
3S It may appear that the analogy between knowing and perceiving is incompatible with
the characterization of knowledge as infallible. Thus Stokes sees the switch from gnosis
to episteme in the course of the argument as necessary to introduce infallibility. "'Gnow-
ledge' (sic) is discernment, realisation, recognition, and there is nothing obviously in-
fallible about any of these" (I 15). Yet Plato might have considered the "seeing" of the
forms infallible in the way waking is infallible in comparison with dreaming. There is no
reason to assume that the infallibility in question here is the infallibility of propositions.
I see no evidence for Vlastos' claim that "logical certainty" is a "less inflated, more
exact" term for what Plato calls "infallibility" ("Degrees of Reality," 73).
264

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
This "contents analysis," Fine claims, avoids any objection against So-
crates' association of doing different work with being set over different
things. We must recall, however, that Fine rejects the "degrees of truth"
view according to which each proposition included in belief would be both
true and false:36 instead, she claims that belief includes some propositions
that are completely true and others that are completely false. But in this case
belief loses any province peculiar to itself: the true propositions over which
it is set are the province of knowledge; the false propositions over which it
is set are the province of ignorance. Fine in 1978 creates the appearance of
a difference between the things over which knowledge and belief are set by
calling the former "pieces of knowledge" and the latter "beliefs" (1978,
129). This is only an appearance because on Fine's view the same true
propositions can be either known or believed: what makes a true proposi-
tion a "piece of knowledge" rather than a "belief' is not anything intrinsic
to it, but simply the fact that it is known, as Fine acknowledges: "A given
proposition is a belief when it is believed, and a piece of knowledge when it
is known ... ." (1978, 129-30). Of course, certain propositions, namely, false
ones, cannot be known, but only believed; however, these propositions are
apparently not the peculiar domain of belief but belong also to ignorance
(though, as will be seen below, the text ultimately forces Fine to give a
different account of the relation between belief and ignorance).
Fine's account in 1990, though less detailed, clarifies her position by
dropping all talk of "pieces of knowledge" and characterizing the difference
between knowledge and belief as follows (1990, 91):
1) Knowledge is of the set of only true propositions

2) Belief is of the set of true and false propositions.
Though Fine grants that these sets overlap, she maintains that they differ in
not being co-extensive. Yet it is clear that Fine does not mean that their
contents could not be co-extensive, since she later grants that some believ-
ers could conceivably have only true beliefs, while others could conceiv-
ably have only false beliefs (1990, 93). The difference this interpretation
allows in that over which knowledge, belief and ignorance are set is not
36 In 1978 Fine rejects this view because she sees it as having the "unintuitive" result
that knowledge and belief have irreducibly different contents, knowledge ranging over
truths, and belief over partial truths (1978, 126; see also 137-8). In 1990, Fine steps back
from this position, admitting that it might be possible to make sense of the notion of
partial truth; now, however, she claims that such a notion is controversial and thus
violates the dialectical requirement (1990, 89). Gosling, on the other hand, characterizes
belief as being of "partial truth" (130).
265

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
even one between propositions, but only one between logical sets.37 This
purely formal difference does not give the "epi" any meaning distinct from
the work that the different powers do,38 since the difference between the sets
of knowledge and belief is reducible to the fact that knowledge, being
infallible, collects only true propositions, while belief, being fallible, may
collect false as well as true propositions. A knower and a believer could
affirm exactly the same propositions (all true): what would allow us in this
case to describe the province of the former as "the set of only true proposi-
tions" and the province of the latter "as the set of true and false proposi-
tions" is not the content of what they affirm (the propositions are the same),
but the infallibility or fallibility with which they affirm it. With this in-
terpretation, the difference between what is believed and what is known is
reduced to a difference between the property of being believed and the
property of being known.39
37 Already in 1978 Fine in a note distinguishes between the token contents of belief and
knowledge, on the one hand, and the sets of belief contents and of knowledge contents,
on the other (1978, 130, n. 14). The token contents of belief and knowledge can be the
same, but "the set of belief contents is not the same as the set of knowledge contents,
since it has one different property: containing true and false members, and not only true
members" (1978, 130, n. 14). Therefore, even here the difference between what is
believed and what is known is a difference between sets, and not between specific
contents.
38 Gosling's reading also in effect eliminates the tnC by reducing it to a difference in
work (124). J. Hintikka has seen the argument as doing the opposite: as reducing the
difference in work or function to a difference in objects ("Knowledge and its Objects in
Plato," in Patterns in Plato's Thought, ed. J.M.E. Moravcsik [Dordrecht / Boston: D.
Reidel, 1973], 7-18; also in Knowledge and the Known: Historical Perspectives in
Epistemology [Dordecht / Boston: D. Reidel, 1974]). Hintikka sees this reduction as
explaining Plato's "tendency to view knowledge as some sort of direct relation between
the knower and the objects of knowledge" (18), or "knowledge by acquaintance."
Against both interpretations, Santas has shown that the text clearly distinguishes the
objects of belief and knowledge from their functions ("Hintikka on Knowledge and its
Objects in Plato," in Patterns in Plato's Thought, ed. J. M. E. Moravcsik [Dordecht,
Holland: 1973], 36-42). Santas' own solution (see "Hintikka on Knowledge," 46, 48-9)
nevertheless depends on Hintikka's claim that knowledge for Plato is a direct relation to
objects (as Santas explicitly acknowledges in "Belief and Knowledge," 55).
39 Fine explicitly maintains that in claiming that what is known and what is believed
cannot be the same, "Plato might only mean that the properties of being known and
being believed are different properties" (1990, 91). However, she distinguishes this
interpretation from the one that attributes to Plato the supposedly stronger claim that
"the set of propositions one can believe is not co-extensive with the set of propositions
one can know" (91). As I try to show, however, there is no real difference here. Fine
appears to acknowledge this in deed, since she makes no use of the distinction between
the two interpretations nor defends one against the other.
266

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
Can this reading plausibly be made consistent with the text? The most
obvious difficulty is simply the word "epi."40 Fine claims that "'epi' can
range over contents as well as objects" (1978, 129). This claim would
provide an adequate defence of the "degrees of truth" reading, since that
reading indeed sets knowledge and belief over different contents: truths, in
one case, and partial truths, in the other. Fine, however, rejects this reading
and in doing so undermines the possibility of having "epi" range over
different contents. On her interpretation, as we have seen, knowledge and
belief can have exactly the same propositional content. What then is left for
"epi" to range over? Only formal sets. Though belief and knowledge may in
any actual case have the same content, belief is set over the possibility of
including false propositions, while knowledge is set over the impossibility
of including false propositions. Yet, not only is there no textual basis for
such a use of "epi," but, as already noted, this interpretation reduces the
"epi" to the difference between the fallibility and infallibility of the work
belief and knowledge do and thus eliminates it as a distinct criterion. Such
an elimination is not compatible with the text. An adequate interpretation of
(5)-(7) must explain both 1) how the two criteria for distinguishing between
powers are related and 2) why they are introduced as distinct criteria. The
interpretation offered above and based on an analogy between knowledge
and perception can meet both requirements; Fine's interpretation cannot.
Yet even if we ignore the "epi" altogether, there are still claims made in
the text that cannot plausibly be made consistent with Fine's interpretation.
One of them is Socrates' claim that what is known is "what is" and that
therefore "4what is believed must be something other than what is" (?kio TL
&iv 6o80aoTv fl To ov E'L, 478b3-4). To see the inconsistency, we need
only note how Fine's interpretation, in providing a veridical reading of this
claim, would apparently contradict itself: if we substitute "what is true" for
"what is," Socrates is maintaining that what is true cannot be believed; yet
Fine maintains not only that what is true can be believed, but also, in 1990,
that some people may believe only what is true. Fine apparently attempts to
avoid the inconsistency by taking Socrates' claim to mean that what is
believed is not only what is true but also something else in addition (i.e.,
what is false) (see 1978, 130). Yet we certainly require more than the needs
of Fine's interpretation to convince us that the phrase "other than" (&kXo i)
in this passage is to be understood as "not only, but also ...
The argument at 478a8-e6 is most plausibly read as showing that the
province of belief, "what is and is not," is not simply a lumping together of
the provinces of knowledge and ignorance, but a third thing distinct from
40 Stokes appears to think that the word epi itself rules out Fine's reading (116-7).
267

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
either. A decisive passage is 478e 1-2 which describes that over which belief
is set as participating (peTEXov) in both being and not-being. Since a major
point of the argument against the lovers of sights and sounds is that a thing
is not identical with that in which it participates (476c9-d3), the claim that
"what is and is not" participates in both being and not-being implies that it
differs from both and therefore cannot be their mere combination. This
characterization of that over which belief is set as a third thing is of course
incompatible with Fine's account. If "what is" means "what is true" and
"what is not" means "what is false," what is this third thing that only
participates in both and is therefore distinct from both? The answer cannot
be "partial truth," because Fine rejects this notion. The mention of partici-
pation makes as clear as possible what belief is being set over here: the
same things that are described as [iFTFXOVTa before the argument begins,
namely, sensible objects. This identification, however, undermines Fine's
veridical reading. If the "is" has veridical meaning at 478b3-c6, where
Socrates claims that belief is neither of what is nor of what is not, and if it
has veridical meaning at 478d5-12, where Socrates recalls the opening
premises (3)-(4), it cannot suddenly switch to an existential or predicative
meaning at 478el-3, where Socrates sums up the preceding argument with
the characterization of the object of belief as participating in both being and
not-being while not properly called either.4' In short, the strained reading of
478a8-e6 required by Fine's interpretation is at the outer limits of plausi-
bility.
Yet there is worse to come. So far we have pursued only Fine's account
of the relation between belief and knowledge, but the relation between
belief and ignorance is even more problematic for her reading. At 478b6-c6
(step [9] in the outline above), Socrates says not simply that what is be-
lieved is other than what is not, but that it is impossible (dt6UvaTov, 478b6,
478b9) to believe what is not, since this would be equivalent to believing
nothing. This claim places Fine in a terrible bind: if she reads "what is not"
as meaning "what is false," as she must for the sake of consistency, then
Socrates is claiming that it is impossible to believe what is false, i.e., that all
beliefs are true. But if all beliefs are necessarily true, then the only dis-
4' Fine appears to shy away from a veridical reading of 478el-2. She recognizes that
this passage is problematic for her reading (1978, 132), but brushes it aside with the
assertion that it "claims quite generally that whatever is between being and not being is
what is believed" (133). Fine cannot be allowed to leave the terms "being," "not being"
and "between" here vague and general (what about participation?) when she has forced
a very definite meaning on the rest of the argument.
42 Annas makes a similar objection to the veridical reading (198), though she does not
acknowledge Fine's distinction between "false" and "very false" beliefs.
268

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
tinction Fine recognizes between belief and knowledge is collapsed.42 Fine
attempts to get out of this bind by distinguishing between false beliefs and
"totally or very false beliefs" ("justice is a vegetable," to use her example)
and assigning only the latter to ignorance while leaving the former to belief
(1978, 131).43 What therefore prevents the believer with only false beliefs
(and according to Fine in 1990, as we have seen, there can be such believ-
ers) from being ignorant is that while he may think that it is just to murder
others and steal their property, he at least recognizes that justice is not a
type of squash. This interpretation need only be cited in order to reveal
itself as an act of desperation. Not only is there not the slightest evidence
for attributing to Plato a distinction between "very false" and "moderately
false" beliefs, but this distinction is surely not available to someone who
rejects the "degrees of truth"' theory." Furthermore, Fine's interpretation
requires us to give entirely different readings of Socrates' denial that we
believe "what is" and his denial that we believe "what is not," even though
both denials occur in the same sentence (478c6): when Socrates denies that
we believe "what is," he is supposed to mean that we believe not only "what
is," but also something else (i.e., not only what is true, but also what is
false); however, when he denies that we believe "what is not," he is sup-
posed to mean that it is altogether impossible to believe "what is not"
understood in one sense (as "what is very false"), while allowing that we
can believe "what is not" in another sense (as "what is moderately false").
Finally, what sense are we to make of 478el-2 on this reading? How do
moderately false beliefs participate in "very false beliefs"? In short, faced
with the above texts, Fine's interpretation cannot stand.
This interpretation only fares worse if judged not by the criterion of
inherent plausibility, but by the dialectical requirement. Can we really be-
lieve that when Socrates speaks of "pure being" and "absolute not-being,"
the lovers of sights and sounds would naturally take him to mean "the set of
only true propositions" and "the set of very false propositions"? As already
43 In 1990 Fine does not discuss the assignment of ignorance to "what is not," but only
refers us to the 1978 paper (1990, p. 87, n. 6). Therefore, the later paper provides no new
way of explaining the distinction between belief and ignorance.
44 Fine asserts: "Nor is this line of argument DT [the degrees of truth interpretation].
Plato's claim is now only that totally false beliefs are assigned to ignorance, and not that
all false beliefs are" (1978, 131). Yet I simply cannot see how this interpretation can
avoid assuming degrees of truth. What can prevent a belief from being totally false and
therefore assigned to ignorance except its partial truth? In general, if Fine accepted the
"degrees of truth theory" she could still avoid the "two worlds theory" (as she recog-
nizes, 1978, pp. 138-9, n. 23), while having a much more coherent interpretation, though
one still vulnerable to my other criticisms and plagued with new problems attending a
belief in "degrees of truth."
269

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
noted, it is highly implausible that they would be inclined to associate
knowledge with propositions at all, rather than with perception. When So-
crates asserts that what is believed is other than what is, could the lovers of
sights and sounds understand him to mean, "what is believed is not only
what is true but also what is false"? When Socrates argues that it is impos-
sible to believe "what is not," would they be likely to realize that he means
only that we cannot believe what is very false, though we can believe what
is moderately false? Only an audience of modem analytical philosophers, if
anyone at all, could read into what Socrates says all that Fine's interpreta-
tion requires.
iv) That Fine cannot abide by her own dialectical requirement is revealed
in another major difficulty: her interpretation of the argument can work
only by reading into it two assumptions that clearly violate this requirement,
as she virtually acknowledges (1990, 92-3, n. 13 & n. 14). These assump-
tions are needed to resolve the following problem: if, as Fine claims, the
argument does not assign knowledge to forms and belief to sensible objects,
but assigns both to propositions, thus allowing the possibility of knowing
true propositions about sensible objects and believing true or false proposi-
tions about the forms, then how can the argument reach what Fine takes to
be its conclusion: i.e., that all knowledge requires, though is not restricted
to, knowledge of forms and that one can achieve no more than belief if one
is confined to sensible objects? In short, if the argument does not rule out
knowledge of sensible objects, how can the lovers of sights and sounds be
persuaded that such knowledge requires knowledge of a form, something
that they claim does not even exist? According to Fine, the argument reac-
hes this conclusion by assuming that 1) "there is just one property, the F,
the same in all cases, in virtue of which all and only F things are F' and that
2) we cannot know anything about F things unless we can define what F is
(1990, 92-3).45 These assumptions must be so obvious to the lovers of sights
and sounds that Socrates does not need to defend or even explicitly artic-
ulate them in the argument.
Since, however, it is hard to imagine assumptions that would be more
controversial than these to the lovers of sights and sounds, the argument on
Fine's reading can reach its conclusion only through a violation of the
dialectical requirement more serious than the one with which she charges
45 Fine also requires the perhaps less problematic assumption that "knowledge is pos-
sible" (1990, 93). Since, however, the knowledge in question here is defined quite
strictly and distinguished from perception, there appears to be no reason why the lovers
of sights and sounds should grant that such knowledge exists.
270

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
rival interpretations.46 Would the same people who deny that beauty itself
exists grant that all beautiful things are beautiful in virtue of one property,
beauty, that remains the same in all cases? Would the same people who
shun all activity concerned with logoi grant that they must first define what
beauty itself is before they can know that the sights and sounds they love
are beautiful? By making the argument depend on these assumptions, Fine's
interpretation is refuted on its own terms.
VII. The "Two Worlds Theory"
Fine's interpretation, however, is at least in part motivated by legitimate

concerns about the "two worlds theory" in the extreme form in which it is
often presented. Fine's explicit reasons for seeking to avoid this theory are
the following (1978, 122; 1990, 86):47
i) It leads to the patently absurd consequence that "No one can know, for exam-
ple, what actions are just or good; no one can know even such mundane facts
as that they're now seeing a tomato, or sitting at a table" (1990, 86).
ii) It contradicts Socrates' claim to have only doxa concerning the form of the
good (506c-e)
iii) It contradicts the description of the philosopher who returns to the Cave as
knowing the sensible things there (520c).
Can these objections be met without recourse to Fine's untenable interpreta-

tion? Specifically, can the existential/predicative reading described and de-
fended above provide a solution?
The first objection is easily dismissed, since this consequence is absurd
only according to our conception of knowledge as justified true belief that
something is the case. If episteme is, as suggested above, nonpropositional
knowledge of whatness (e.g., of what justice is or of what goodness is,
through acquaintance with what is truly just or what is truly good), then
there is no absurdity in claiming that we do not "know" in this sense that a
particular action is good or that we are seeing a tomato; these facts are
simply not objects relevant to episteme. Nothing Socrates says rules out the
possibility of our having justified true belief that we are seeing a tomato;
this simply is not what he means here by episteme.48
46 In conclusion Fine recommends her account of the argument as one "on which at least
its opening premises satisfy DR" (1990, 94). I do not see why an account on which later
but absolutely crucial premises violate the dialectical requirement is to be preferred over
an account on which the opening premises violate this requirement. For a similar crit-
icism, see Stokes, 123.
47 Similar objections against a "two worlds theory" are to be found in Annas, 194.
48 It is important to recognize that Plato sometimes uses episteme in a weaker sense that
corresponds more to our "justified true belief." As Cooper points out, in Book 4
271

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
To resolve the contradictions stated in ii) and iii), we need to recognize
that the relation between forms and sensibles is not one between two com-
pletely distinct worlds, since sensibles do not exist independently of the
forms, but are only their images or imperfect instantiations. Therefore,
forms and sensibles are not completely separate, only extrinsically related
objects: the being of the sensible object is exhausted by its participation in
the form, it exists and is what it is only as intrinsically related to the form.
On account of this intrinsic relation, in recognizing beautiful bodies as
beautiful I can be said to have some awareness of the form of beauty they
imitate, but this is an awareness that fails to distinguish the form from the
imitations and sees it only as reflected in them. It is therefore a dream-like
awareness that confounds reality with what only resembles it. Does So-
crates' argument prevent us from calling this awareness doxa concerning
the form of beauty? It cannot and need not. It cannot, because what is
deficiently grasped in the doxa of the lovers of sights and sounds are not the
sensible objects themselves - the lovers can probably perceive and describe
them with perfect accuracy-, but the form reflected in them.49 In this sense,
and in this sense only, doxa must be about the form. Yet what Socrates'
argument maintains is that the deficiency of doxa, what makes it inferior to,
and not simply different from, episteme, is its failure to distinguish the form
imitated by the sensibles from these sensibles themselves, with the result
that its cognition is not set over the form - since the form is not even a
distinct object for it -, but is confined to the sensibles. The characterization
of doxa as a dream-state unable to distinguish the forms from their likeness-
es (476c2-8) both confines doxa to these likenesses and allows it to pre-
serve an indirect relation to the forms - its deficiency is precisely the
indirectness. Therefore, when Socrates claims to have only doxa concerning
(438c6-8) Socrates talks about "common or garden tAtTnjctLaL of such contingent mat-
ters as house-building and health (240). (This passage, incidentally, does not help Fine,
since she could not claim that house-building and health prove that there can be episteme
of sensible objects without having to assume that all house-builders and doctors have
knowledge of the forms). Sprute distinguishes between Erfahrungs-Erkenntnis and Ei-
gentliche-Erkenntnis (62).
49 Ebert has rightly drawn our attention to the need to reconcile the characterization of
belief and knowledge as two distinct dunameis with the characterization of belief at
476b-d as a deficiency (Mangel) of the power constitutive of knowledge (124). Unlike
Ebert, I take such a reconciliation to be possible. My interpretation here rejects Adriana
Bertozzi's claim that in the Republic doxa is "non piu un genere imperfetto di conoscen-
za, ma una conoscenza perfetta dell' imperfetto" ("11 termine 66ea nei Dialoghi di
Platone," Giornale di Metafisica 3 [19481: 41). Bertozzi's interpretation supports the
kind of extreme "two worlds theory" that would be vulnerable to Fine's criticisms ii) and
iii) above.
272

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
the good, he means that he is in some sense confined to sensible images
(such as the sun) in his understanding of the good, that the good is not a
direct, explicit object of his cognition.50 Socrates' claim is therefore perfect-
ly consistent with Book V's claim that doxa is set over sensible objects.
Given the above clarifications, it is not difficult to explain the descent
into the Cave. The philosopher who knows forms must certainly know their
sensible imitations in a way in which the prisoner ignorant of the forms
does not. But what is the nature of this knowledge of sensibles? What
Socrates says in his speech to persuade philosophers to descend back down
into the Cave is informative: "For once accustomed [to the darkness] you
will see with infinitely greater clarity ([VUQ6 WXtLov 6lEFOE) than those
down there, and you will know what each of the images (Et&wXa) is and
what it is an image of, because you have seen the truth (Tr&krOf-) concern-
ing the beautiful, the just, and the good" (520c3-6). The philosopher knows
sensibles precisely as imperfect images of forms that transcend them, while
the prisoners clearly lack such knowledge. Yet it is inaccurate to say that
the objects of this knowledge are the sensibles per se; it is precisely because
this knowledge is not of sensibles, but of the forms, that it can reveal
sensibles for what they are: nothing but deficient imitations of these forms.
An example might help to clarify this point. Imagine two people looking at
a portrait: person A knows the man of whom it is a portrait, person B does
not; in fact, person B does not even recognize that it is a portrait of anyone,
but believes it to be a mere product of artistic fantasy. In this case, we can
indeed say that person A knows what the portrait is, while person B does
5o This is not to say that Socrates is on the same level as the lovers of sights and sounds.
While he does not know what the good is and can talk about it only in images, he
recognizes that the good exists and is distinct from any of the sensible objects that
participate in it. The importance of the Divided Line lies precisely in its introduction of a
level of cognition between the doxa of the lovers of sights and sounds and the highest
form of knowledge attainable by the philosopher: this intermediate cognitive state is
dianoia, which recognizes that its objects are not sensible, but must rely on sensible
images in talking about them (5 10d-51 lal). On this point see the excellent discussion
in Morgan, 84-92. There are important parallels between dianoia and doxa: just as doxa
is said to be something in between ignorance and knowledge, so is dianoia said to be
"something in between" ([LETaCtv. TL) doxa and nous (51 1d4-5); just as doxa is said to be
brighter than ignorance but darker than knowledge, so is dianoia said to be brighter than
doxa but darker than knowledge (?tLiJT'LTI; 533d5-6); dianoia, like doxa, is compared
to dreaming (533b6-c3). Though Socrates' dianoia is brighter and more awake than the
doxa of the lovers of sights and sounds, it is still dream-like and dark; though he
recognizes that the forms are distinct from their sensible images and has more to say
about them than what these images reveal (see 506d8-e3), he nevertheless can no more
dispense with these images than the mathematician can dispense with his diagrams. See
Fine 1990, 105.
273

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
not. Yet the knowledge that A has and B lacks is not a knowledge of the
portrait per se as a sensible object: A and B can be imagined to perceive the
sensible properties of the portrait with equal accuracy and vividness. The
knowledge that distinguishes the two is a knowledge of the original of
which the portrait is an imitation. Likewise, the philosopher in one sense
does not know the shadows on the wall any better than the prisoner does; in
fact, at first he will be less able to predict their movements (516e9-517a4,
517d4-e2). What allows him, unlike the prisoner, to see the shadows for
what they truly are is not a knowledge of the shadows themselves, but a
knowledge of the originals that cast them. Even the description of the de-
scent, therefore, preserves the assignment of knowledge to the forms: it
only shows how this knowledge of the forms affects our understanding of
their imitations. The characterization of episteme as a waking-state that
succeeds in distinguishing the reality from its images (476c9-d4) both as-
signs episteme to the reality and preserves its indirect relation to the images
- where this indirectness is its virtue.
Fine attempts to bridge the two-worlds dichotomy by making proposi-
tions, which can indifferently be about either sensibles or forms, the imme-
diate intentional content of both knowledge and belief. Such a bridge, how-
ever, is provided by the objects themselves: in assigning knowledge to
forms and belief to sensibles, Socrates does not divorce them, but enables
them to be related as their objects are related: just as sensibles imitate
imperfectly what the forms themselves are perfectly, so does the belief
assigned to sensibles intuit confusedly what the knowledge assigned to
forms intuits clearly.5" Belief and knowledge are related as dreaming to
waking, because their objects are related as image to original. The exist-
ential/predicative reading can explain this relation as follows: though belief
is assigned to sensible objects that exist and do not exist by being F and
not-F, while knowledge is assigned to forms that truly exist as truly and
only F, what is imperfectly imitated and perceived in the objects of belief is
the same as what is perfectly exemplified and known in the objects of
knowledge: what F is. Problems arise only if one reads back into Plato's
dialogues the modem notion that knowledge and belief are propositional:52
one must then wonder why the propositional content of a belief concerning
a certain object could not be converted into knowledge about the same
object.
Fine does everything possible to make the text fit this modem conception
of knowledge. Indeed, her interpretation appears to be part of a more gener-
s' The proportion upon which the Divided Line is based is the following: "4; To
8o0aYT&v Q6Og To yVW(TOT6V, OiTiO TtO 6poLwOAiwV nQ6O T6 @ b)ROL6OY)" (5 lOa9- 10).
274

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC
al view that all knowledge in Plato's dialogues is propositional.53 Yet this
interpretation fails. It is not recommended by any advantages over other
readings. It itself violates the dialectical requirement which it wrongly ac-
cuses rival interpretations of violating. The problems it seeks to solve by
altogether rejecting a "two worlds theory" do not require anything so dras-
tic. It rests on unwarranted assumptions concerning the nature of knowledge
and the meaning of "to be." Finally, and most importantly, it cannot yield a
plausible reading of the text. The failure of this otherwise ingenious in-
terpretation has the positive result of making clearer than ever how far
Plato's episteme is from "knowledge" as generally understood today. Plato
is, for good or bad, not "up to date."54
Skidmore College, Saratoga Springs, NY
52 It is their view that the objects of belief and knowledge must be "that-clauses" that
leads Cross and Woozley to find the argument of Republic V incoherent (169-75).
5 Elsewhere Fine writes that "even if Plato's primary concern is knowledge of objects,
this concern can readily be phrased in the modern idiom as knowledge that a particular
proposition is true" ("Knowledge and Logos in the Theaetetus," Philosophical Review
88 [Jan. 1979]: 366-7). This is precisely the way in which Fine attempts to rephrase
Republic V, against all appearances and, as I argue, unsuccessfully. For a general de-
fence of the thesis that knowledge of the forms is nonpropositional, see my "Nonpropo-
sitional Knowledge in Plato," forthcoming.
54 I wish to thank the editor and an anonymous reader for helping make this a better
paper.
275

185.192.71.183 on Wed, 08 Mar 2023 19:16:33 UTC

This Content Downloaded From 185.192.71.183 On Wed, 08 Mar 2023 19:16:33 UTC

Uploaded by

Copyright:

Available Formats

This Content Downloaded From 185.192.71.183 On Wed, 08 Mar 2023 19:16:33 UTC

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

This Content Downloaded From 185.192.71.183 On Wed, 08 Mar 2023 19:16:33 UTC

Uploaded by

Copyright:

Available Formats

Propositions or Objects?

A Critique of Gail Fine on Knowledge and Belief in "Republic"

Stable URL: https://www.jstor.org/stable/4182535

This content downloaded from

According to most contemporary epistemologies, the distinction between

Phronesis 1996. Vol. XLI/3 (Accepted April 1996) 245

This content downloaded from

I. The Argument (476e7-479e1)

This content downloaded from

(1) Knowledge is of something (TtO, of what is (6v). (476e7-477al). It is therefore

Now Socrates can appeal to (3) and (4) to conclude:

This content downloaded from

II. Fine's Interpretation

This content downloaded from

IIl. The Dialectical Requirement

This content downloaded from

(le) Knowledge is of what exists.

This content downloaded from

This content downloaded from

(ip) Knowledge is of what is F.

Fine's main objection to this reading is that it is by no means obvious or

This content downloaded from

IV. The Predicative Reading

Fine's criticism of the existential and predicative readings is not confined to

This content downloaded from

This content downloaded from

(oruyxEX,u[va, 524c) and presented to the perplexed soul as being the

18 Socrates introduces this discussion by asking (523e3-6) if sight sees adequately

This content downloaded from

sensible object only with temporal qualification" (A Companion to Plato's Republic

This content downloaded from

This content downloaded from

V. The Compatibility of the Existential with the Predicative Reading

This content downloaded from

This content downloaded from

This content downloaded from

This content downloaded from

Vl. The Incompatibility of Fine's Reading with the Text

The existential and predicative readings as characterized above all belong

This content downloaded from

This content downloaded from

This content downloaded from

1) Knowledge is of the set of only true propositions

This content downloaded from

This content downloaded from

This content downloaded from

This content downloaded from

This content downloaded from

This content downloaded from

VII. The "Two Worlds Theory"

Fine's interpretation, however, is at least in part motivated by legitimate

Can these objections be met without recourse to Fine's untenable interpreta-

This content downloaded from

This content downloaded from

This content downloaded from

This content downloaded from

Skidmore College, Saratoga Springs, NY

This content downloaded from

You might also like