2012 - Nicholas Smith - Plato On The Power of Ignorance

Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

Offprint from

OXFO R D S T U D I E S
IN A NC I E N T
PHILO S O P H Y
EDIT O R: B RAD IN W O O D

S U P P L E M EN TA RY VOLUME


V IRTUE AN D HAPPINESS:
ESSAYS IN HONOUR
OF JU LIA AN NAS

E D I T E D B Y R AC H AN A KAMTEKAR

3
  
P L A T O ON THE POWER
OF I G NORA NCE

NI CH O LAS D . S M I T H

. Introduction

O of the truisms of ancient Greek is chalepa ta kala: good things


are difficult. In the area of epistemology this truism is expressed
in the obvious observation that knowledge is difficult and rare,
whereas ignorance is easy and common. Absolutely anyone can
be ignorant, for ignorance does not require any special ability or
skill. Moreover, ignorance may be applied to any subject—one
may be ignorant of anything and everything. Only God cannot
achieve ignorance; the rest of us manage it all the time, in regard to
nearly everything. Plato, too, was well aware of this obvious truth,
and portrayed it in the figure of Socrates locked in what turned
out to be a fatal conflict with the ignorance he found everywhere
he turned. Socrates was the wisest of men only because he alone
recognized the extent of his ignorance (Ap.   –).
But in book  of the Republic Plato provides an analysis of ig-
norance that seems to violate these truisms. Starting at    he
has Socrates and Glaucon distinguish between three distinct cog-
nitive powers (dunameis): knowledge (epistēmē), belief (doxa), and
© Nicholas D. Smith 
I am indebted to many for their help with various earlier drafts of this paper. In
alphabetical order, these include: Hugh Benson, Lee Churchman, Rebecca Copen-
haver, Zina Giannopoulou, Brad Inwood, A. A. Long, Joel Martinez, Debra Nails,
Andrew Payne, Clerk Shaw, and David Wolfsdorf.

Throughout my discussion I will translate the Greek word δύναμις as ‘power’.
Others who have preferred this translation have included J. Adam (ed. and comm.),
The Republic of Plato, nd edn, intro. D. A. Rees,  vols. (Cambridge, ), i.
Introduction and Books I–V, , note on   ; A. Bloom (trans.), The Re-
public of Plato (New York, ); R. Larson (trans.), The Republic (Arlington
Heights, Ill., ); C. D. C. Reeve, Philosopher-Kings (Princeton, ). The Greek
word has also been translated as ‘faculty’ (e.g. by R. C. Cross and A. D. Woozley,
Plato’s Republic: A Philosophical Commentary [Commentary] (London and New
York, ), and P. Shorey in the Loeb Classical Library translation, Plato: The
Republic, nd edn. (Cambridge, Mass., )), ‘capacity’ (e.g. by J. Annas, An
Introduction to Plato’s Republic [Introduction] (Oxford, ),  ff., and T. Grif-
fith (trans.), Plato: The Republic (Cambridge )), or ‘ability’ (e.g. by R. W. Ster-
 Nicholas D. Smith
ignorance (agnōsia). Powers, Socrates goes on to explain, are dis-
tinguished in virtue of what they are related to (I will henceforth
call this ‘the relata condition’) and what they accomplish (which I
henceforth call ‘the accomplishment condition’: ἐφ᾿ ᾧ τε ἔστι καὶ ὃ
ἀπεργάζεται,   ). In this section of the dialogue the second of
these two differentiae is invoked again, but only briefly and not ex-
plicitly each time; instead, the distinctions Socrates makes here are
mostly made in terms of the relata of the different powers. Know-
ledge, we are told, is related to what is (to on); ignorance is related
to what is not (to mē on); belief is related to what both is and is not.
Later, we are told that ‘what is not’ is the same as ‘nothing at all’
(  ). In her excellent book on the Republic Julia Annas mostly
avoids the topic of what Plato does with ignorance in this passage,
and we can already perhaps see why: according to the truism with
which I began, it would appear that ignorance requires no ability,
yet here in book  it seems to have become an ability (or power)
after all! Also according to the truism, ignorance may be applied to
absolutely anything and everything. But here in the Republic Plato
tells us that it actually applies to ‘nothing at all’. Something seems
to have gone very wrong here!

ling and W. C. Scott (trans.), Plato: The Republic (New York, )). Aristotle later
famously uses this same Greek word for ‘potentiality’. As far as I can tell, nothing
important hangs on the choice of translation here, though my argument will make it
clear that we should not understand Plato’s δυνάμεις as cognitive states.

Scholars have been very troubled by the fact that Plato provides two differen-
tiae. See e.g. Annas, Introduction, ; M. Stokes, ‘Plato and the Sightlovers of the
Republic’ [‘Sightlovers’], in A. Barker and M. Warner (eds.), The Language of the
Cave (Edmonton, ), – at –. I propose to explain their various roles

herein. Annas, Introduction.

In saying this about her discussion of this passage, I do not intend to assume
some interpretative ‘higher ground’—after all, in my own earlier work on this very
passage (N. D. Smith, ‘Knowledge by Acquaintance and “Knowing What” in
Plato’s Republic’ [‘Acquaintance’], Dialogue: Canadian Philosophical Review, 
(), –) I provide an excellent example of shrugging off ignorance as not
likely to admit of analysis parallel to the other two cognitive powers (see especially
the closing words of the article at ). At least Annas never made such an absurd
claim! I am not, however, the only one to have made this mistake: a very recent
example may be found in F. Fronterotta, ‘Plato’s Republic in the Recent Debate’
‘[Recent’], Journal of the History of Philosophy,  (), – at : ‘Now, if
we recognize the fact that what is not, in that it is conceived as what does not exist
at all, or as pure nothing, really is not—and so does not constitute an autonomous
“existential” and still less “object” level to which we can opine or think anything—
this hierarchy ends up distinguishing just two degrees of existence and so are and
exist in two different ways.’ Fronterotta derives this result from his assumption of
an existential rendering of Plato’s being, about which see sect.  below.
Plato on the Power of Ignorance 
Like Annas, most scholars who have studied this passage have fo-
cused almost exclusively on the distinction between knowledge and
belief. Their focus is not entirely unreasonable, for it is clear that the
distinction between knowledge and belief is the primary one that
Plato wishes to explicate here, as it is in terms of this distinction
that the important difference between the philosopher-rulers and
ordinary rulers will be drawn. The distinctions between knowledge
and ignorance, and belief and ignorance, are only very briefly men-
tioned, and ignorance itself remains almost wholly unexplained. In-
deed, all that we are told directly about ignorance is that it is re-
lated to what is not (  ;   ;   ), which, again, Plato
tells us is ‘nothing at all’ (  ). Otherwise, all Plato manages to
have Socrates say about ignorance is by way of distinguishing belief
from ignorance: belief is neither knowledge nor ignorance (  ),
because belief is neither clearer than knowledge nor more obscure
than ignorance (  ), but rather darker than knowledge and
brighter than ignorance (  ) and so between the two ( 
). Therefore, the power related to what is between being and not
being would be neither knowledge nor ignorance, but between them
(  ).
Despite the extreme terseness of Plato’s remarks about ignor-
ance, however, I believe that careful thought about what little the
text does provide on the subject can help us to avoid errors scho-
lars have made in attempting to understand the other two cognitive
powers—knowledge and belief—which all agree are of more obvi-
ous importance to Plato’s goals at this stage of the dialogue. In this
sense, although my official topic is Plato’s treatment of ignorance,
my more general focus is the same as that of other scholars. We will
be able to understand the two more important cognitive powers, I
claim, if we manage to recognize well precisely how they are dis-
tinguished from ignorance, and how the ways in which Plato draws
these distinctions limit the ways in which we can understand his
epistemology more generally in this section of the Republic. Care-
ful attention to these distinctions, moreover, will allow us to re-
move entirely the apparent contradiction between Plato’s analysis
of the cognitive powers and their relata and accomplishments, and
the truisms about ignorance with which I began.
 Nicholas D. Smith

. Are the powers of or about their relata?

Most scholars have understood the relationships between the cog-


nitive powers and their distinct sets of relata (Plato’s ἐφ᾿ ᾧ τε ἔστι at
  ) as the representative or intentional relations between cog-
nitive states and what such states are of or about. So, for example,
we find Gail Fine using the expressions ‘belief is set over [Fine’s
translation of Plato’s epi] . . .’ and ‘knowledge is set over . . .’ in-
terchangeably with ‘belief is of . . .’ or ‘belief is about . . .’, and
‘knowledge is of . . .’ or ‘knowledge is about . . .’. For the sake
of brevity, let us call this understanding of Plato’s cognitive power-
epi-object relation the ‘subject-matter interpretation’. A moment’s
reflection on Plato’s power of ignorance, however, will show one of
the many problems with this assimilation. Plato assigns ‘what is not’
in relation to ignorance. As the difficulty with which I began makes
plain, however, the nature of this relation cannot be an articulation
of something like the subject-matter of ignorance. As I said at the
outset, ignorance may be of or about anything at all.
Part of the problem here, of course, may be an ambiguity in the
term ‘ignorance’ (in English or in Greek), for one may be ignorant
simply by lacking cognition in relation to a subject-matter, or one
may be ignorant in the sense of having cognitions, but ones properly
regarded as terribly faulty in terms of their content, generation, or
warrant. The former sense obviously works very differently from
knowledge and belief: one is a knower or a believer on the basis of
knowing or believing something. But one is not necessarily ignorant
on the basis of ignoring something—instead, one might have done
nothing at all to be ignorant. If Plato’s account of ignorance is to
be compared to our own conception, accordingly, it cannot be to


G. Fine, ‘Knowledge and Belief in Republic V–VII’ [‘Knowledge’], in S. Ever-
son (ed.), Epistemology (Companions to Ancient Thought, ; Cambridge, ),
– passim, esp. –. See also Annas, Introduction, – passim; Cross
and Woozley, Commentary, passim; L. Gerson, Knowing Persons: A Study in Plato
[Persons] (Oxford, ), – passim; D. Sedley, ‘Philosophy, the Forms, and the
Art of Ruling’ [‘Art’], in G. R. F. Ferrari (ed.), The Cambridge Companion to Plato’s
Republic (Cambridge, ), –, esp. ; N. P. White, A Companion to Plato’s
Republic (Indianapolis, ), . These examples are merely representative. In
fact, nearly every scholar who has worked on the epistemology of Republic  has
made this same assimilation.

See e.g. Annas, Introduction, , who explicitly makes this connection.

I owe this point to Rebecca Copenhaver.
Plato on the Power of Ignorance 
the sense in which ignorance is the absence of cognition. The ab-
sence of cognition seems more like the effect of not doing anything,
rather than a power by which we achieve some accomplishment. I
will later return to the specific requirements Plato applies to ignor-
ance, but for now it is enough to note that the universal range of
the subject-matter of ignorance seems enough for us to doubt that
Plato’s cognitive power-epi-object relation is the relation of cogni-
tion to subject-matter.
The error of the subject-matter interpretation probably derives
from the anachronistic application of modern epistemology’s focus
on what is often called either ‘propositional’ or ‘informational’
knowledge, where knowledge is understood as a cognitive state
that is of or about some proposition or information, or where the
proposition or information is counted as the content of the cog-
nitive state. In such analyses, knowledge is counted as a species
of belief—more particularly, warranted (which is usually taken
as justified or reliable) true belief. But again, the inapplicability
of this approach to what Plato is doing in Republic  should be
obvious from the outset: after all, Plato’s epistēmē is very obviously
not a species of belief (doxa), but an entirely distinct cognitive
power—indeed, that is the whole point of the discussion!
Cognitive states represent information, and so when we say that
knowledge, belief, and ignorance are about their subject-matter, it
is the information represented that we have in mind. Since, how-
ever, Plato did not have this intentional relation in mind in assign-
ing distinct ranges of objects to the cognitive powers, we should not
make the mistake of supposing that Plato’s cognitive powers are the
same as the cognitive states we typically discuss in contemporary
epistemology.


See e.g. N. Gulley, Plato’s Theory of Knowledge [Theory] (London, ), –:
‘just as . . . Plato implies that knowledge is “propositional”, so by what he says about
doxa and the “images” with which it deals he implies that doxa is propositional’. See
also Fine, ‘Knowledge’, who also treats Plato’s epistemology as propositional.

For a comprehensive recent discussion of different theories of warrant in con-
temporary epistemology see A. Plantinga, Warrant: The Current Debate (New York,
).

In the Meno Plato does seem to make knowledge a species of belief, which is one
of many reasons why I am sceptical about scholars’ attempts to assimilate the view
presented in the Republic to that given in the Meno (as, for example, Fine does). It is
not my purpose here, however, to discuss the epistemology in dialogues other than
the Republic.

I first insisted on the differences between powers and states in Smith, ‘Acquain-
 Nicholas D. Smith
Now, Plato’s own analyses do not at all invite the anachronism
just described. For one thing, the analogy he explicitly provides, by
which we are supposed to understand what he is doing with the cog-
nitive powers in this passage, is between cognitive and perceptual
powers: the first two examples of powers Plato explicitly mentions
as examples of the sort of thing he has in mind are sight (opsis) and
hearing (akoē) (  ). So when Plato says that he distinguishes
powers by what they are related to and what they accomplish, it is
easy enough to see (at least in general terms) how to apply these dif-
ferentiae to these models: sight is related to things that can be seen
and produces visual states, whereas hearing is related to things that
can be heard and produces auditory states.
So even if Plato’s model cases—sight and hearing—do not make
it simply obvious how we are supposed to apply the relata condi-
tion, the cases would seem to indicate that the intentional relations
inherent in what representations are ‘of’ or ‘about’ do not apply
to the powers themselves. Sight is not of or about things that can
be seen. Instead, when the power is working rightly, it is by sight
that we see them. So, too, hearing is not of or about things that
can be heard; rather, it is by this power that we hear such things.
But when we do see or hear things that can be seen or heard, the
result (which would seem to go into Plato’s accomplishment condi-
tion) generally is a representation (or a misrepresentation): I use my
power of sight to look into the tree, and see a red-winged blackbird.
tance’. In continuing to insist on this distinction, I diverge from the account of this
passage given in J. Szaif, ‘Doxa and Epistēmē as Modes of Acquaintance in Repub-
lic V’ [‘Doxa’], Études platoniciennes,  (), –, which in other ways presents
a view similar to the one I advance herein. See also nn. , , , and  below.

The significance of this analogy is noted in Stokes, ‘Sightlovers’, , but
then dismissed as a dialectical assumption, intended to secure the agreement of the
sightlovers to the rest of the argument. See also next note. A far more plausible
understanding of the analogies to sight and hearing is offered in F. J. Gonzalez,
‘Propositions or Objects? A Critique of Gail Fine on Knowledge and Belief in Re-
public V’ [‘Propositions’], Phronesis,  (), –. Szaif, ‘Doxa’, rightly em-
phasizes the element of acquaintance in Plato’s analysis, on which see also Smith,
‘Acquaintance’.

It has been suggested to me that, following the echoes of Parmenides more
evident in this passage than perhaps anywhere else in Plato, we should understand
Plato’s ἐπί as operating more or less directionally, as in Parmenides’ different ‘ways’.
By this analysis, then, sight goes out to things that can be seen and hearing goes out to
things that can be heard. It is not just a matter of aiming, because one can aim and
miss; in the cases of the powers, it is about the causal interaction between the power
and what it goes to. I am grateful to A. A. Long and Brad Inwood for calling my
attention to this issue.
Plato on the Power of Ignorance 
My power of sight is not of or about the red-winged blackbird; but
my visual experience is, because it is a visual representation of the
bird in the tree. As I said, it could also be a misrepresentation: Py-
thagoras thought he heard the voice of a deceased friend in the cries
of a dog being beaten (D.L. . ). There is intentionality in both
perception and cognition, then, obviously—but the intentionality
may be found in the representational states produced by the per-
ceptual or cognitive powers, and not in the causal relation between
the powers and their relata.
Several important results flow from these simple observations.
For one thing, we may now understand that it cannot be a correct
formulation of whatever it is that Plato is trying to tell us, here at
the end of book , that there can never be knowledge, belief, and
ignorance of or about the same subject-matter. If Plato were telling
us that, we would have good reasons for finding his claims deeply
paradoxical, at best. Instead, we will get this result only if the ana-
lysis Plato provides actually entails that the cognitive products of
the various powers at work cannot have the same subject-matter
or be cognitions that are of or about the same subjects. And we
already have at least some reason to think that this will not turn out
to be the case.

. Relata, accomplishment, and propositional content

Scholars have also offered three distinct interpretations as to what


Plato may have had in mind as the bearers of the relata condition:
what is, for knowledge; what is not, for ignorance; and what both is
and is not, for belief. To understand what notion of being Plato may

Annas notes that this result has been regarded as ‘scandalous’ by many of Plato’s
critics (Annas, Introduction, ). One can well understand Annas’s own obvious
dissatisfaction with Plato’s analysis here, once she has dismissed the analogy with
perceptual powers as ‘play[ing] no part in the argument’ and begins by understand-
ing the relations of the powers to the objects in terms of representation and subject-
matter (see Annas, Introduction, –).

Worries about this problem have come to be called Plato’s ‘Two Worlds’ prob-
lem, according to which the distinct and separate worlds of Forms and sensibles were
supposed to entail that there could be no beliefs about Forms and no knowledge of
the sensible world—which would obviously defeat Plato’s claim that the philosopher-
rulers would be superior rulers (in the Cave/sensible world, obviously), on the basis
of their knowledge. For further discussion of this problem see N. D. Smith, ‘Plato
on Knowledge as a Power’ [‘Power’], Journal of the History of Philosophy,  (),
–.
 Nicholas D. Smith
have had in mind in this passage, scholars have proposed veridical,
existential, and predicative interpretations. In the veridical in-
terpretation ‘what is’ means ‘what is true’; ‘what is not’ means ‘what
is not true’ (or ‘what is false’), from which it is taken to follow that
‘what is’ and ‘what is not’ refer to propositions. Accordingly ‘what
both is and is not’ means ‘what both is true and is not true’, which
is generally then explained either as ‘every proposition that can be
believed is both true and false’ or as ‘the set of propositions that
can be believed includes some truths and some falsehoods’. Those
who favour the veridical reading argue that this understanding ac-
cords well with the contemporary view that what is known must
always be true, but what is believed can be either true or false. To
complete this analysis, then, we would understand the relata of ig-
norance to be only false propositions.
Neither of the versions of the veridical reading makes good sense
of the text, however. For one thing, when Plato characterizes what
he has in mind as the relata of belief, he has Socrates specify that
they are the sorts of things that both are and are not at the same time
(ἅμα,   ). Then, to give examples of what he has in mind, Soc-
rates mentions things that are both fair and ugly (  –), things
that are both just and unjust (  ), holy and unholy (  –),
double and half (  –), big and little (  ), light and heavy
(  ). Plainly, neither are the things that ‘both are and are not’
here propositions, nor is their being true and false at the same time
what is at issue here.

Examples include J. C. B. Gosling, ‘Doxa and Dunamis in Plato’s Republic’
[‘Doxa’], Phronesis,  (), –, and Plato (London, ); and Fine, ‘Know-
ledge’. Szaif, ‘Doxa’, closely links the veridical to the predicative interpretations (see
esp. ).

Examples include F. C. White, Plato’s Theory of Particulars (New York, ),
 n. , and also ‘The Scope of Knowledge in Republic V’, Australasian Journal
of Philosophy,  (), –; Stokes, ‘Sightlovers’, esp. ; Fronterotta ‘Re-
cent’, .

Examples include Annas, Introduction, esp. ; Cross and Woozley, Commen-
tary, ; G. Vlastos, ‘Degrees of Reality in Plato’ [‘Degrees’], in id., Platonic Stu-
dies, nd edn. (Princeton, ), –, esp. ; Gerson, Persons, esp. –; Sed-
ley, ‘Art’, ; Smith, ‘Power’; and Szaif, ‘Doxa’ (but see n.  above).

This is how Fine, ‘Knowledge’, renders Gosling, ‘Doxa’, and Gosling, Plato.

This is Fine’s own version of the veridical reading in ‘Knowledge’.

See Fine ‘Knowledge’, –.

See Gerson, Persons, . Szaif, ‘Doxa’, does not explicitly notice this quali-
fication, but takes the use of ‘being’ in Plato to include both predicative and also
veridical elements in such a way as to indicate that he thinks something can actually
be true and false at the same time: ‘Take the example of an instance of beauty. [Plato’s
Plato on the Power of Ignorance 
No doubt a significant factor in why the veridical interpretation
ever suggested itself is what Plato has Socrates say about the distinc-
tion between knowledge and belief at   –, according to which
knowledge is infallible and belief fallible. Scholars have offered dif-
ferent understandings of this claim, however. The weaker interpre-
tation would have it that ‘knowledge but not belief entails truth’.
But the weaker interpretation is plainly too weak; for as Lloyd Ger-
son puts it, ‘“Infallible” cannot be simply equivalent to “unmis-
taken” or “not false” if one kind of belief is true belief. If that were
what it meant, true belief is no less unmistaken than is knowledge.’
Now, scholars have understood this difference between the
powers as involving something propositional—after all, one ob-
vious sense we can make of a cognitive power being mistaken is
that it assents to something that is false, rather than true. By this
reasoning, then, it is a short step to supposing that what Plato
is doing in this entire passage is talking about the ways in which
knowledge, belief, and ignorance are related to true and false pro-
positions. This short step, however, is a misstep: it misinterprets
Plato’s relata condition, for as we have seen, Plato can hardly have
true and false propositions in mind as the relata of the powers, or as
what he means to refer to when he distinguishes knowledge, belief,
and ignorance by relating them to ‘what is’, ‘what is and is not’ (at
the same time), and ‘what is not’, respectively. Instead, then, we
should understand the fallibility and infallibility of the powers as
having something to do with their accomplishments.
But even this is not enough. In an earlier paper on this sub-
ject, I argued—partly in order to make sense of Plato’s association
of knowledge with infallibility and belief with fallibility—that we
should understand the accomplishments of the cognitive powers
to be cognitive states. Hence, the product of knowledge at work
would be some cognitive state (an example of what contempor-
ary epistemologists might call ‘knowledge’), whose content would

discussion] tells us that this is an example of something which  and  , because
it is and is not beautiful—or because it is both true and false to say of it that it is beau-
tiful’ (Szaif, ‘Doxa’, , emphases original). I will later argue that the ambiguities
of what is and is not, in Plato’s sense, do not extend to their sentential truth (and
hence, would not require Plato to have in mind anything other than a two-valued
sentential logic).
 
Fine ‘Knowledge’, . Gerson, Persons, .

Smith, ‘Power’. It now seems to me that, on this issue, I was closer to the truth
in Smith, ‘Acquaintance’.
 Nicholas D. Smith
be given in a proposition, what is known. In standard analyses of
knowledge, the content appears as the variable identifying the pro-
position: the ‘p’ in ‘S knows that p’. I now think my earlier ana-
lysis is mistaken on this point. Instead, I think the accomplish-
ments of the cognitive powers are something like what we now call
‘conceptions’. So, for example, with regard to F-ness, knowledge
produces a very vivid, clear, and accurate conception of F-ness; be-
lief produces a much less clear, vaguer, and less accurate conception
of F-ness.
Again, the case of ignorance is helpful here. What goes wrong in
the case of ignorance may not simply be assent to falsehood—after
all, this sometimes occurs with belief as well. There are other ways
that cognition can fail than just by affirmation of falsehood. One
might, instead, suffer from misconception. Although at the heart
of misconception is some error, misconception might serve as the
ground for some other true belief: one who has a misconception
about lions (supposing they were native to Australia) would not go
wrong in thinking that lions are carnivores. In my earlier work I
supposed that Plato conceived of the accomplishments of the power
of ignorance as cognitive states whose propositional content was al-
ways false. I now think that the products of the power of ignorance
are not propositional states at all.
Now cognitive states are of or about things: my belief that
Coriolanus is the same person as Gaius Marcius is a belief about
Coriolanus. Moreover, if I know that Coriolanus is the same person
as Gaius Marcius, then this is knowledge of Coriolanus, know-

On this point I take it that I agree entirely with Szaif, ‘Doxa’ (but see also n. 
above, and nn. , , and  below).

This is not the place to review the complications of de dicto and de re senses
of ‘belief’, according to which the misconceiver’s thought, ‘lions are carnivores’, is
belief in the de dicto sense, whereas the de re sense of that belief is that the mis-
conceiver is believing that tigers are carnivores. Whether or not Plato had a way to
make precisely this sort of distinction, he could certainly distinguish cases in which
one might express a thought in a way with which others who knew what they were
talking about might agree, but whose expression of that thought was the product of
confusion or misconception. Socrates’ interlocutors, for example, often said things
that were true; but even when they managed to get things right, Socrates’ interro-
gations eventually revealed the gross confusions and misconceptions grounding the
interlocutor’s claims. Even the Thrasymachus who speaks in book , for example,
might affirm that Plato’s philosopher-rulers were just, if they could be shown to go-
vern in such a way as to maximize their self-interest, relative to the options available
to them. I think a proper reading of the Republic would support such a claim, in fact,
but that reading would not square with Thrasymachus’s own conceptions of either
self-interest or justice.
Plato on the Power of Ignorance 
ledge about him. Others—those who have not read Plutarch or
Shakespeare—may be ignorant of this identity. To say, accord-
ingly, that knowledge, belief, and ignorance can all be of or about
the same objects is one of epistemology’s obvious truisms. In
the case of knowledge and belief, the truism is actually one of
analyticity—it follows directly from the fact that knowledge (as a
cognitive state) has belief (as a cognitive state) as one of its neces-
sary conditions. So, strictly speaking, every case of knowledge will
be a case in which there is both knowledge and belief about the
same object. This, to put it rather bluntly, is Epistemology .
Another of the basic claims one learns in Epistemology  is
that knowledge does—and belief does not—have truth as a neces-
sary condition. But as we have seen, when Plato says that knowledge
is infallible, it cannot simply mean only this. We can, however, re-
cognize that nothing in what Plato says in book  requires any de-
viation from what we learn in Epistemology  in understanding
the intentional relations between cognitions and the objects repre-
sented in them (between, for example, my belief about Coriolanus
and Coriolanus as the object the belief is about), or in understand-
ing that one cannot have false knowledge, though one can have false
belief.
But for the reasons I have already given, these truisms—though
entirely consistent with and even entailed by Plato’s discussion at
the end of book —should not be confused with what Plato has in
mind when he relates the cognitive powers to different sets of ob-
jects, nor do they explain at all how Plato would suppose that in
achieving the relevant cognitive relation to the appropriate set of
objects, there will also be differences in what the cognitive powers
produce in us. To see how all of this works, we need to return to
the question of what the relata are for each power. Once we have
understood this adequately, we can then ask how the relationships
between the powers and their relata have different cognitive effects.

. Being and not-being

We have now seen why it is a mistake to suppose that Plato under-


stood the relata of the cognitive powers to be true and false pro-
positions, as the veridical interpretation has it. So what else, then,
might Plato mean by ‘what is’, ‘what is and is not’, and ‘what is
 Nicholas D. Smith
not’, as the relata belonging to knowledge, belief, and ignorance,
respectively? As I said earlier, two other interpretations have been
offered to explain this, the ‘existential’ and ‘predicative’ readings
of Plato’s ‘being’. Each of these readings, however, appears to en-
counter a problem. The existential reading seems to fail at the same
place the veridical reading fails—in Plato’s claim that ‘what is and
is not’ manages to be both at the same time (ἅμα,   ). Existence
does not seem to be a matter of degree: that which does not exist
does not also at the same time exist, and what does exist does not
also at the same time fail to exist. No sense can be made of some-
thing managing both existence and non-existence at the same time.
On this topic, the predicative reading offers obvious advantages,
for Plato’s actual examples of things that both are and are not are
obviously cases of things that take on both of a pair of opposite pre-
dicates at the same time.
The trouble with the predicative reading seems to come from
the power of ignorance: what could Plato have in mind as the set
of objects that ‘in no way is’ (  –,  )? What Plato has to
say here seems emphatically to endorse the existential reading of
‘what is not’: ‘But surely that which is not could not be designated
as some one thing, but most rightly as nothing at all’ (  –
, trans. Shorey). The predicative reading, however, appears to re-
quire the identification of some subjects and some predicates where
those predicates could ‘in no way’ apply to those subjects. One can
imagine cases of this sort, of course. Ordinary failures of reference,
but also category mistakes and analytic falsehoods, are obvious ex-
amples: ‘present king of France’ is an example of the first, ‘sleeping
mineral’ is an example of the second; ‘married bachelor’ is an ex-
ample of the last. All such expressions refer to nothing at all.
So one way the predicative reading might be able to accommo-
date the apparently existential commitment in Plato’s identification
of ‘not-being’ with ‘nothing at all’ would be by understanding ig-
norant errors in terms of failures of reference. Things that do not or
cannot exist are neither tall nor short, and are neither one of these
even more or less. Nor are non-existents just or unjust, fair or ugly,
or anything else or its opposite, to any degree whatsoever. Perhaps it

This understanding, indeed, is so obviously compelling that some translators
manage to put explicit references to predication into the text itself. So Shorey, in the
Loeb translation, renders   –: ‘And likewise of the great and the small things,
the light and the heavy things—will they admit these predicates any more than their
opposites?’
Plato on the Power of Ignorance 
is precisely because nothing at all can be predicated of nothing at all
that P. T. Heath concludes, in his classic article on ‘Nothing’ in The
Encyclopedia of Philosophy, that with respect to nothing ‘altogether
the less said about it the better’. Without access to predication,
there is nothing to be said!
But understanding Plato’s ‘nothing at all’ as indicating the
ground of ignorance in terms of the subject place in propositional
content does not seem to be the right sort of failure of reference
for the analysis we find in Republic . As I noted earlier, the ac-
tual examples Plato gives of ‘being’ appear to be predicative. But
this would seem to require that the failure we should expect, in
what is accomplished by the power of ignorance, is a failure in
the predicate place in propositional content. In other words, the
error inevitably made where ignorance is at work, for Plato, must
be a misconception of the predicate F (a misconception of F-ness,
in other words), rather than an adequate conception of F-ness
misapplied to something that does not exist. A case of the latter
sort would be the error of supposing the present king of France is
just, for example, where one has some appropriate conception of
what justice is. The sort of case Plato seems to require, however,
must be something like supposing that the laws of Athens are just,
when one wholly misconceives what justice is.
Plato’s account of how knowers can predicate justice of things
and have the result be infallibly correct derives from the way in
which knowers attain their conception of justice. By applying the
power of knowledge to the Form, the conception of justice they
achieve is clearest (see Plato’s contrast between saphēneia and
asapheia at   ). Belief, Plato tells us, is ‘darker than know-

‘Nothing is an awe-inspiring yet essentially undigested concept, highly es-
teemed by writers of a mystical or existentialist tendency, but by most others
regarded with anxiety, nausea, or panic. Nobody seems to know how to deal with
it (he would, of course), and plain persons generally are reported to have little
difficulty in saying, seeing, hearing, and doing nothing. Philosophers, however,
have never felt easy on the matter. Ever since Parmenides laid it down that it
is impossible to speak of what is not, broke his own rule in the act of stating it,
and deduced himself into a world where all that ever happened was nothing, the
impression has persisted that the narrow path between sense and nonsense on this
subject is a difficult one to tread and that altogether the less said about it the better’
(P. L. Heath, ‘Nothing’, in P. Edwards (ed.), The Encyclopedia of Philosophy, 
vols. (New York, ), v. – at ).

I note that in her analysis Annas seems to me to get this exactly right: ‘Some-
thing is an object of my ignorance if it totally lacks the quality I predicate of it’
(Annas, Introduction, ).
 Nicholas D. Smith
ledge but brighter than ignorance’ (  –), by which he seems
to think that, for example, the conception of justice that believers
apply when they predicate justice of things is less clear—because
it derives only from empirical and equivocal examples of justice,
rather than the purity (eilikrineia,   ;   ;   –) of
Justice Itself. One who is ignorant, then, is one who either has no
conception of justice at all or, perhaps more likely, a conception
that is wholly confused and inapplicable to making judgements
involving justice, for ignorance engenders conceptions via a pro-
cess that derives from what is not-just—what is, as far as justice
is concerned, ‘nothing at all’. Those who have confused or inad-
equate conceptions have only wholly unreliable ways to employ the
predicate and no idea of what they are talking about, if we find them
actually applying that predicate in their speech. It does not follow
that everything they say must be sententially false—after all, they
might mimic and repeat verbally the judgements of those who have
a better idea of what justice is. But even as they say something true,
they have no grasp at all of the truth they pronounce. Whatever
they may have in mind, then, when they talk about justice, has
‘nothing at all’ as its source, and thus has nothing at all to do with
justice.
Now, as we have seen, Plato has far too little to say about what ex-
actly he has in mind here. On the one hand, he actually tells us that
‘not being’ refers to what is the opposite of what is. Recall the ex-
amples he gives of things that ‘are and are not (at the same time)’:
things that are fair and ugly (  –), things that are both just
and unjust (  ), holy and unholy (  –), double and half

Szaif first treats ignorance as ‘no cognition at all’, but then seems immediately
to modify this claim by defining the state of ignorance as being ‘of x and x  
(anything at all)’ (Szaif, ‘Doxa’, , emphases original).

Here, too, I break with Szaif on the question of whether we should understand
the relative superiority of knowledge and faultiness of belief as involving some blur-
riness in sentential truth (in the case of belief), which would then force Plato’s ana-
lysis to require all ignorant judgements to be false. See Szaif, ‘Doxa’, : ‘The clos-
ing arguments [of book  use] the “argument from opposites” (which could also be
called an “argument from context-relativity”) in a very questionable way by arguing
from the co-presence of opposites in the case of natural and social instantiations of a
Form to the conclusion that such instantiations don’t even allow for an unqualifiedly
true judgement and thus cannot be object of knowledge’ (emphasis original). On the
contrary, I argue that Plato’s argument is not ‘questionable’ on these grounds after
all, because in my view the truth or falsity of judgements is not what is at issue in
this passage at all, but rather the adequacy of the conceptions on the basis of which
one might form true or false judgements. See also nn. , , , and  above, and
n.  below.
Plato on the Power of Ignorance 
(  –), big and little (  ), light and heavy (  ). So
‘what is not’, in regard to ‘being F ’, is not simply the non-F, but
the anti-F: where F is Justice, ‘what is not’ is thus injustice. But we
should worry about what exactly this commits Plato to. He would
probably not be willing to have negative Forms be the relata of ig-
norance, for we should expect all Forms (even negative ones) to be
the relata of knowledge instead. But what, then, could the injustice
that we should identify as the not-being of justice be? So perhaps,
instead, we should assume that he had something like the priva-
tion of F-ness in mind as the not-F. So, for example, with regard
to justice, the ignoramus might confusedly observe something that
has no justice in it and create a conception of justice on the basis of
some other quality he thereby observed. Plato certainly does some-
times talk as if there are things in which no justice can be found—
perhaps, for example, the ‘incurable souls’ in the myth at the end of
the Gorgias, in whose souls there is ‘nothing healthy’ (οὐδὲν ὑγιὲς ὂν
τῆς ψυχῆς,   –). Perhaps a complete privation of justice could
also appear in non-moral things as well. Plato’s identification of
‘what wholly is not’ with ‘nothing at all’ might help here: because
‘nothing at all’ is neither just nor unjust, perhaps Plato was thinking
that the required failure of predication was the result of an existential
failure. In regard to ‘nothing at all’, we would expect only priva-
tion to apply. But there are different problems with thinking that
Plato had privations of F-ness in mind as the relata of the power of
ignorance. We might wonder how conceptions that had absolutely
nothing to do with F-ness could really be F-conceptions at all.

We should be careful here, however. If, as the account of justice he gives in
the Republic strongly suggests, justice can be found in proper functioning, or per-
haps proper functioning in complex entities in which each part functions properly,
it might be that we could find images of justice in non-human animals, plants, and
many natural or artificial things. But perhaps Plato would think that some non-moral
things were sufficiently chaotic to have no trace of justice in them—mud or faeces,
for example.

One who argues for this position is Fronterotta, ‘Recent’. Fronterotta fails to
note that Plato’s actual examples of ‘what is not’ are predicative, and not privative.
In any case, it should be clear from what I have already argued that even if Plato
does identify ‘what is not’ with non-existence, it does not follow that his analysis of
‘what is’ and ‘what is and is not at the same time’ must be understood existentially,
as Fronterotta does.

With gratitude to Hugh Benson for supplying the example (an application of
one given in another context by Donald Davidson), consider the case of someone
who looks into his refrigerator and sees some oranges in there, but on the basis of
what he sees in his refrigerator reports his beliefs that hippopotamuses are round,
orange, about the size of baseballs, and sweet and tart in flavour. Should we really
 Nicholas D. Smith
It is perhaps possible that Plato had no distinct metaphysics in
mind for not-being in this passage, but was instead thinking only
that an ignoramus might form a conception of justice on the basis
of the injustice of something that was unjust—we might perhaps
think of some admirer of Thrasymachus or Callicles, or some other
terribly unjust person, conceiving of justice in terms of the very
qualities that made the person unjust, instead. But even these er-
rors seem to be of a sort we should associate with the processes of
belief, rather than the extreme of ignorance at work.
To return to the basic picture, then, Plato’s analysis requires the
cognitive powers to produce results of a certain kind, and in the case
of the other cognitive powers, it now seems that these results are
conceptions. Now conceptions, notice, are also intentional—they
are of or about a thing conceived. The problem about ignorance
with which we began, recall, was that ignorance could be of or about
anything at all, whereas Plato relates ignorance to nothing at all.
Earlier, I noted that ‘ignorance’ in English is ambiguous between
the privation of any relevant cognitive condition and the presence
of a cognitive condition we regard as profoundly defective. For
the reasons I have given, Plato’s power of agnōsia must be that by
which we produce defective conceptions—misconceptions—rather
than simple lack of cognition. Moreover, we are now in a position to
see just how defective the conceptions produced by ignorance must
be. By parity of reasoning, if the difference between what is pro-
duced by ignorance and what is produced by belief must be a differ-
ence of the same kind and degree as is the difference between what
is produced by belief and what is produced by knowledge, then it
follows that what is produced by ignorance must be something at
the furthest end of defect, relative to what grounds the infallibi-
lity of knowledge and the fallibility of belief. It must, accordingly,
always be a failure of some sort. As I said earlier, failures of refer-
ence, category mistakes, and analytic falsehoods would all qualify
as candidates for such failures, but because Plato’s focus is on un-
equivocal and equivocal applications of predicates, I concluded that
say that this person has any conception of what a hippopotamus is, rather than say
that he is confused about how to use the word ‘hippopotamus’?

I am tempted to think that it was because of his thinking about this question
that Plato was led never again to return to the epistemology he provides in the Re-
public. Instead, we find a very different account suggested in the Theaetetus, as well
as a very different account of not-being in the Sophist. But this is admittedly pure
speculation.
Plato on the Power of Ignorance 
the cases Plato must have in mind are cases in which one predicates
F-ness of something in a way that reflects one’s complete confusion
or misconception about what F-ness is. Those who have mere opi-
nions about F-ness will sometimes succeed in attributing F-ness to
F things; those who know what F-ness is will always succeed in at-
tributing F-ness to F things. But for one who is ignorant of F, even
if he manages to achieve some occasional success in attributing F-
ness to F things, such ‘success’ will be purely verbal, as it were, for
lying behind the ignoramus’s successes, no less than his failures, in
making such attributions will be a complete misconception of what
F-ness is. The fault in the cognitive states produced by ignorance,
then, is not that the propositional content of the cognition must be
false, but rather that the one in such a state has misconceived what
serves as the predicate in that propositional content. To return to
the case of justice, then, whatever the ignoramus has in mind when
he pronounces that something is just will be a cognitive failure be-
cause he misconceives justice. The ignoramus forms judgements
without having any idea of the grounds of such judgements. His
judgements, then, are not merely fallible; they are the kind of non-
sense that derives from misconception.
We can also now appreciate just how deficient the ignoramus will
be from a practical point of view. The problem with what is pro-
duced by ignorance is that it is wholly indiscriminate and thus fails
to be a conception in any normal sense. The whole purpose of con-
ceptions is to provide the grounds for discriminating things that
are different from one another. If a conception is so confused as to
ground no reliable discriminations, however, it has failed in a very
specific and critical way: it has failed to achieve what conceptions
are supposed to achieve, by their very nature. In using such utter
failures in making his judgements, then, the ignoramus is one who
is truly useless in the practical domain.

. Can there be knowledge, belief, and


ignorance of the same subjects, for Plato?

Scholars have hotly debated the question of whether Plato thought


there could be knowledge of particulars. This debate, however,

I am indebted to Rebecca Copenhaver for her help in formulating what I say in
this paragraph.

Those who think there can be knowledge of particulars include Gosling, ‘Doxa’;
 Nicholas D. Smith
has been profoundly confused by the assimilation of the objects of
cognitive states (the objects such states are of or about) to the relata
of Plato’s powers. Once we take care not to confuse things in this
way, however, we can make perfectly good sense of Plato’s epis-
temology in a way that neither defeats his philosophical agenda in
promoting the idea of philosophers as rulers (on the sole ground
of their cognitive advantage in knowledge) nor equivocates on his
divisions of cognitive powers by distinct sets of relata and accom-
plishments.
Knowledge, as all but those attracted to the veridical account of
‘what is’ acknowledge, is related to the Forms. When Plato refers to
‘what is’ or ‘what purely is’ (e.g. at   ), he refers to the Forms.
The power of knowledge, then, is actualized when (and only when)
an appropriate connection is made, in the production of a judge-
ment (a cognitive state), with a Form. The product of such a con-
nection may be a judgement whose content is of or about the Form:
‘Beauty is not ugly’. I have argued, however, that the differences
between the powers should not be conceived in terms of differences
of subject-matter. One reason to resist this result is to make possible
the sort of case Plato explicitly puts into Socrates’ mouth at  
– , where it is plain that whatever the latter has to say about the
Form of the Good, it will be only belief and not knowledge that
provides the basis of his judgement. But just as Socrates can have
beliefs about the Form of the Good, so his imaginary (doxastic) in-
terlocutors at the end of book  will prove able to answer Socrates’
important questions about things that appear to be both just and
unjust, beautiful and ugly, at the same time. These ‘belief-lovers’
(  ) might find themselves quite willing to agree that ‘beauty
is not ugly’, while continuing to reason in a way that conceives of
beauty in wholly empirical ways. The propositional content of their
belief, then, is the same as that which we find in the judgement
of one who produced the judgement on the basis of a conception
of beauty gained through the power of knowledge; but the way in
Reeve, Philosopher-Kings; Fine ‘Knowledge’; Stokes, ‘Sightlovers’; Gonzalez, ‘Pro-
positions’; and Smith, ‘Power’. Those who think there cannot be knowledge of par-
ticulars include Cross and Woozley, Commentary; Gerson, Persons; Gulley, Theory;
G. X. Santas, ‘Hintikka on Knowledge and its Objects in Plato’, in J. Moravcsik
(ed.), Patterns in Plato’s Thought (Dordrecht, ), –; Sedley, ‘Art’; and Vlas-
tos, ‘Degrees’.

For arguments as to why the denial of knowledge of particulars would have this
effect see Fine, ‘Knowledge’, and Smith, ‘Power’.
Plato on the Power of Ignorance 
which Plato’s believers produce judgements is quite different from
the way in which knowers do. The difference is not to be found in
subject-matter or content, but in the power applied, the objects to
which the power is applied, and the conceptions that are thereby
produced. In the case of ignorance, the ignoramus might also have
‘beauty is not ugly’ as the propositional content of some judgement,
but would have a completely wrong idea of what beauty is. Judge-
ment requires the application of some conception, but in the case
of ignorance, only utter confusion or misconception lies behind the
application. Notice, however, that in each case the way the cognitive
powers and their products become involved is through the process
of forming a judgement (again, here is where we will find cognitive
states in Plato’s analysis): each judgement will require an applica-
tion of the cognizer’s conceptions. So, strictly speaking, we should
not characterize the judgements made by knowers, believers, and
ignoramuses as examples of ‘knowledge of p’, ‘belief about p’, or
‘ignorance of p’. Each case, rather, is a case of judgement that p
(for example, let ‘p’ = ‘x is F ’) on the basis of some conception of the
predicate-term (F) in ‘p’. Because of the differences in the quali-
ties of their conceptions of F-ness, knowers are the best judges of F-
things—indeed, qua knowers, they are infallible judges of whether

Though Plato does occasionally talk this way, as at   , where he promises
his future rulers that when they return to the Cave, they will know (γνώσεσθε,  
) what the images there are, and of what they are images.

This provides my own answer to the conundrum posed by Szaif, who com-
plains that here in book  Plato makes knowledge and belief different powers, but
later claims that ‘the rational faculty (the “eye of the soul”) is one and that it can
achieve only doxa-type competence as long as it remains focused on the physical and
social world, but will produce epistēmē once it has been “turned around” and refo-
cused toward the realm of intellectual objects through the efforts of dialectic’ (Szaif,
‘Doxa’, –, emphasis original). In my view, the cognitive powers produce con-
ceptions, and the unified ‘power in the soul’ (δύναμις ἐν τῇ ψυχῇ,   ) is what
produces judgements—that is, cognitive states—in which these conceptions are put
to use, which explains why the power of judgement functions so differently when it
is employed under the influence of different cognitive powers. This also allows my
own view to avoid the uneasiness of Sedley’s suggestion that even though the philo-
sophers ‘cannot, even in principle, “know” that a given policy is just, [their] ability
to arrive at the (temporarily) correct “opinion” that in current circumstances such a
policy is the most just will depend upon prior knowledge, namely, . . . knowledge of
what justice itself is’ (‘Art’, ). I agree with Sedley that it is ‘knowledge of what
justice itself is’ that makes Plato’s rulers far better than any alternatives, but do not
agree that the ruler’s judgements should be regarded as opinions. They are, rather,
judgements made on the basis of the conceptual products of knowledge.

It does not follow that knowers can never err; only that they can never err as a re-
sult of knowledge. The dissolution of Callipolis will be the result of the philosopher-
 Nicholas D. Smith
something is or is not F, and to what degree; believers are unreli-
able judges of the F-ness of things; ignoramuses are simply con-
fused about what is or is not F, because their conception of F-ness
is wholly defective.
Similarly, an account can now be readily given of how knowledge
would apply to the judgement of particulars. (Again, this, strictly
speaking, should not be thought of as knowledge of or about par-
ticulars, but rather an explanation of how the power of knowledge
will provide the basis of the most reliable judgements in regard to
particulars.) If a belief-lover were to form the opinion that the rule
of philosophers would be a just one, then according to Plato, the be-
lief formed would be true. But for this to be a belief-judgement—a
judgement we would associate with the operations of the power of
belief, that is—it would have to be the product of the believer’s hav-
ing observed some equivocal particular cases of justice and formed
a conception of justice on the basis of these observations, and then
applied this conception in judging that philosophers would instan-
tiate that concept. Plato’s philosopher-rulers must also judge that
they will be just rulers, so once again it seems that the difference
between knower and believer cannot be made in terms of either
subject-matter or content of cognition. Instead, in applying the
power of knowledge, philosophers would make an appropriate con-
nection to one or more Forms (in this case, at least to Justice Itself),
which would result in their having the clearest possible concept of
justice, which they could then apply in the judgement of instances.
Seeing the ‘resemblance’ of Justice in the rule of philosophers, the
knower would judge such a rule to be an instance of justice. The
ignoramus, by contrast, having only misconceived what justice is,
might ape the words of his cognitive betters, but would use such
words without comprehension. Misconceiving justice, whatever the
ignoramus may say or think about just things is simply irrelevant
to whether or to what degree they may actually be just.

rulers’ error in judging the birth number. Presumably, then, an account of this error
must involve some reliance, by the rulers, on perceptual or cognitive powers other
than knowledge.

This is obviously not the place to engage the notorious problem of the philoso-
phers’ reluctance to rule. I give my own assessment of this problem in N. D. Smith,
‘Return to the Cave’, in M. McPherran (ed.), Plato’s Republic: A Critical Guide
(Cambridge, ), –.
Plato on the Power of Ignorance 

. Summary and conclusion

In this paper I have tried to explain how and why the epistemology
of Republic  can accommodate the ordinary truisms about what we
can and cannot be ignorant of, have beliefs about, or know. These
truisms, I have claimed, all apply (in Plato, as well as in contempor-
ary epistemology) to cognitive states. In the account Plato provides,
cognitive states are generated when people apply in judgement the
conceptions they have come to have as a result of the activities of
their distinct cognitive powers, whose relata are also distinct. The
relative quality of the concepts produced by the powers is what ex-
plains differences in fallibility. Plato regards the judgements made
by those who apply concepts produced by knowledge as infallible,
on the ground that cognitive connection to perfect entities prevents
the possibility of error in judgements made in the light of such en-
tities. The judgements made as a result of conceptions produced
by belief are fallible because the objects to which belief naturally
relates provide contradictory appearances, and so the conceptions
produced as a result of cognitive connection with such objects may
suffer defects that derive from the contradictory appearances, and
thus be only more or less accurate, as conceptions. Hence, appli-
cation of such conceptions in the formation of judgements will be
fallible—it may result in misjudgement. The power of ignorance
produces misconceptions that are the product of no appropriate
connection at all to what the conception is supposed to be a con-
ception of. The application of misconception in a judgement, ob-
viously, will be worse than simple fallibility. Instead, whatever the
ignoramus judges to be true is simply the product of a misconcep-
tion.
In achieving these results, then, I hope I have attributed a view
to Plato that avoids the ‘scandalous’ errors Annas reported others
finding in the epistemology of book . I would also argue that
the account provided herein satisfies what Fine has called the ‘dia-
lectical requirement’. It does so because it so readily accounts for
what I have called the epistemological truisms, which even the least
sophisticated interlocutor would accept. Even so, the way in which
Plato can derive what I have called the epistemological truisms may

Annas, Introduction, .

Fine, ‘Knowledge’, .
 Nicholas D. Smith
rightly be admired for its considerable sophistication. In this way,
then, Plato deftly explains how it is that ignorance relates to nothing
at all, but can take absolutely everything as its subject-matter. Here
again we find, I think, an epistemological truism that is the para-
doxical nature of ignorance, whose subject-matter is everything but
whose competence is nil.
Lewis & Clark College

BIB L IO G RA PHY
Adam, J. (ed. and comm.), The Republic of Plato, nd edn, intro. D. A.
Rees,  vols. (Cambridge, ).
Annas, J., An Introduction to Plato’s Republic [Introduction] (Oxford,
).
Bloom, A. (trans.), The Republic of Plato (New York, ).
Cross, R. C., and Woozley, A. D., Plato’s Republic: A Philosophical Com-
mentary [Commentary] (London and New York, ).
Fine, G., ‘Knowledge and Belief in Republic V–VII’ [‘Knowledge’], in S.
Everson (ed.), Epistemology (Companions to Ancient Thought, ; Cam-
bridge, ), –.
Fronterotta, F., ‘Plato’s Republic in the Recent Debate’ [‘Recent’], Journal
of the History of Philosophy,  (), –.
Gerson, L., Knowing Persons: A Study in Plato [Persons] (Oxford, ).
Gonzalez, F. J., ‘Propositions or Objects? A Critique of Gail Fine on
Knowledge and Belief in Republic V’ [‘Propositions’], Phronesis, 
(), –.
Gosling, J. C. B., ‘Doxa and Dunamis in Plato’s Republic’ [‘Doxa’],
Phronesis,  (), –.
Plato (London, ).
Griffith, T. (trans.), Plato: The Republic (Cambridge, ).
Gulley, N., Plato’s Theory of Knowledge [Theory] (London, ).
Heath, P. L., ‘Nothing’, in P. Edwards (ed.), The Encyclopedia of Philo-
sophy,  vols. (New York, ), v. –.
Larson, R. (trans.), The Republic (Arlington Heights, Ill., ).
Plantinga, A., Warrant: The Current Debate (New York, ).
Reeve, C. D. C., Philosopher-Kings (Princeton, ).
Santas, G. X., ‘Hintikka on Knowledge and its Objects in Plato’, in J.
Moravcsik (ed.), Patterns in Plato’s Thought (Dordrecht, ), –.
Sedley, D., ‘Philosophy, the Forms, and the Art of Ruling’ [‘Art’], in
G. R. F. Ferrari (ed.), The Cambridge Companion to Plato’s Republic
(Cambridge, ), –.
Plato on the Power of Ignorance 
Shorey, P. (ed. and trans.), Plato: The Republic, nd edn. (Cambridge,
Mass., ).
Smith, N. D., ‘Knowledge by Acquaintance and “Knowing What” in
Plato’s Republic’ [‘Acquaintance’], Dialogue: Canadian Philosophical
Review,  (), –.
‘Plato on Knowledge as a Power’ [‘Power’], Journal of the History of
Philosophy,  (), –.
‘Return to the Cave’, in M. McPherran (ed.), Plato’s Republic: A
Critical Guide (Cambridge, ), –.
Sterling, R. W., and Scott, W. C. (trans.), Plato: The Republic (New York,
).
Stokes, M., ‘Plato and the Sightlovers of the Republic’ [‘Sightlovers’], in
A. Barker and M. Warner (eds.), The Language of the Cave (Edmonton,
), –.
Szaif, J., ‘Doxa and Epistēmē as Modes of Acquaintance in Republic V’
[‘Doxa’], Études platoniciennes,  (), –.
Vlastos, G., ‘Degrees of Reality in Plato’ [‘Degrees’], in id., Platonic Stu-
dies, nd edn. (Princeton, ), –.
White, F. C., Plato’s Theory of Particulars (New York, ).
‘The Scope of Knowledge in Republic V’, Australasian Journal of Phi-
losophy,  (), –.
White, N. P., A Companion to Plato’s Republic (Indianapolis, ).

You might also like