Plato (Ancient Philosophies)

Plato
Ancient Philosophies
This series provides fresh and engaging new introductions to the

major schools of philosophy of antiquity. Designed for students of
philosophy and classics, the books offer clear and rigorous presenta-
tion of core ideas and lay the foundation for a thorough understand-
ing of their subjects. Primary texts are handled in translation and the
readers are provided with useful glossaries, chronologies and guides
to the primary source material.
Published
The Ancient Commentators Neoplatonism
on Plato and Aristotle Pauliina Remes
Miira Tuominen
Plato
Ancient Scepticism Andrew S. Mason
Harald Thorsrud
Presocratics
Cynics James Warren
William Desmond
Stoicism
Epicureanism John Sellars
Tim O’Keefe
Forthcoming
Classical Islamic Philosophy Indian Buddhist Philosophy
Deborah Black Amber Carpenter
Confucianism Socrates
Paul R. Goldin Mark McPherran
Plato
Andrew S. Mason
www.routledge.com
'JSTU1VCMJTIFECZ"DVNFO
1VCMJTIFECZ3PVUMFEHF
1BSL4RVBSF .JMUPO1BSL "CJOHEPO 0YPO093/
5IJSE"WFOVF /FX:PSL /: 64"
3PVUMFEHFJTBOJNQSJOUPGUIF5BZMPSBOE'SBODJT(SPVQ BOJOGPSNB
CVTJOFTT
© Andrew S. Mason, 2010
This book is copyright under the Berne Convention.

No reproduction without permission.
"MMSJHIUTSFTFSWFE/PQBSUPGUIJTCPPLNBZCFSFQSJOUFEPSSFQSPEVDFEPS
VUJMJTFEJOBOZGPSNPSCZBOZFMFDUSPOJD NFDIBOJDBM PSPUIFSNFBOT OPX
LOPXOPSIFSFBGUFSJOWFOUFE JODMVEJOHQIPUPDPQZJOHBOESFDPSEJOH PSJO
BOZJOGPSNBUJPOTUPSBHFPSSFUSJFWBMTZTUFN XJUIPVUQFSNJTTJPOJOXSJUJOH
GSPNUIFQVCMJTIFST
/PUJDFT
1SBDUJUJPOFSTBOESFTFBSDIFSTNVTUBMXBZTSFMZPOUIFJSPXOFYQFSJFODFBOE
LOPXMFEHFJOFWBMVBUJOHBOEVTJOHBOZJOGPSNBUJPO NFUIPET DPNQPVOET
PSFYQFSJNFOUTEFTDSJCFEIFSFJO*OVTJOHTVDIJOGPSNBUJPOPSNFUIPET
UIFZTIPVMECFNJOEGVMPGUIFJSPXOTBGFUZBOEUIFTBGFUZPGPUIFST
JODMVEJOHQBSUJFTGPSXIPNUIFZIBWFBQSPGFTTJPOBMSFTQPOTJCJMJUZ
5PUIFGVMMFTUFYUFOUPGUIFMBX OFJUIFSUIF1VCMJTIFSOPSUIFBVUIPST
DPOUSJCVUPST PSFEJUPST BTTVNFBOZMJBCJMJUZGPSBOZJOKVSZBOEPSEBNBHF
UPQFSTPOTPSQSPQFSUZBTBNBUUFSPGQSPEVDUTMJBCJMJUZ OFHMJHFODFPS
PUIFSXJTF PSGSPNBOZVTFPSPQFSBUJPOPGBOZNFUIPET QSPEVDUT
JOTUSVDUJPOT PSJEFBTDPOUBJOFEJOUIFNBUFSJBMIFSFJO
isbn: 978-1-84465-173-3 (hardcover)

isbn: 978-1-84465-174-0 (paperback)
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available
from the British Library.
Typeset in Minion.
Contents
Preface vii
1. Introduction 1
2. Plato’s development and Plato’s Socrates 15
3. Plato’s metaphysics: the “theory of Forms” 27
4. Knowledge 61
5. The soul 99
6. Politics 119
7. Ethics 135
8. God and nature 161
9. Aesthetics 181
Notes 201
Plato’s works 207
Further reading 209
Bibliography 219
Index 221
v
This page intentionally left blank
Preface
This book aims to present Plato’s thought to readers with no previous

knowledge of it. In line with the general aims of the Ancient Philoso-
phies series, it is written with senior undergraduates and first-year
graduate students particularly in mind, but I have tried to make it
accessible to a wider readership as well. No knowledge of philosophy
is assumed, but I have sometimes referred to contemporary philo-
sophical issues, so that readers who are familiar with them can see
how Plato’s thought relates to them.
No introduction can be a substitute for reading Plato’s own works.
This is true of every philosopher worth studying, but it is particu-
larly true of Plato, since his dialogues are not only expositions of
philosophical ideas, but also displays of philosophy in action; they
aim to draw the reader into a philosophical discussion. This book
concentrates on presenting major ideas that Plato puts forward in
his works. (That these ideas can reasonably be seen as Plato’s own is
argued in Chapter 1.) In doing so, however, it reveals only one aspect
of Plato; only by actually studying the dialogues can we appreciate
Plato’s thought as a whole.
Because of this focus on Plato’s central ideas, I have not said much
about those dialogues that seem not to reach definite conclusions,
but, rather, aim to depict the process of enquiry. The classic example
of such a dialogue, among those works that reflect Plato’s mature
vii
plato
thought, is the Theaetetus, and therefore it is only discussed briefly

here. This work very much repays reading, but it is not easy to extract
from it philosophical morals that can be presented in isolation.
Likewise, I have said little about the interaction between liter-
ary and philosophical elements in Plato’s dialogues, although this
is certainly an important aspect of his work. For this reason what
is perhaps Plato’s greatest work from a literary point of view, the
Symposium, is not covered at length; only a small part of that work
is actually devoted to the exposition of philosophical ideas, although
the whole of it certainly illuminates Plato’s philosophical thought in
complex ways.
Another book in this series will deal with Socrates, and therefore
I have not focused on Plato’s so-called “Socratic” dialogues: those
that are often seen as presenting the thought of Plato’s teacher rather
than his own. (That they are plausibly seen as Socratic is argued in
Chapter 2.) I discuss these dialogues, for the most part, only when
they can be used to illuminate Plato’s mature thought, either by simi-
larity or by contrast.
Of course, many aspects of the interpretation of Plato remain
controversial. My readings are certainly not the only ones possible,
and within the limits of this work I have not always been able to
set out alternative views and to defend my positions at length. The
Further Reading section should guide readers to other possible inter-
pretations.
While working on this book, I have benefited from discussing
Plato with colleagues and students at the University of Edinburgh
and at King’s College, London and members of the Archelogos group
at Edinburgh. Earlier versions of some of the ideas used in this book
have been presented to seminars at the universities of Edinburgh and
Glasgow, at Trinity College Dublin, and at the APA Pacific Division
meeting in Pasadena, 2004; I am grateful to all who took part in
those discussions. Thanks are due to three anonymous referees for
Acumen for their extremely helpful and challenging comments, and
to Steven Gerrard for being a supportive and patient editor. Finally, I
thank my parents, Kenneth and Barbara Mason, for all their support
over the years. This book is dedicated to them.
viii
one
Introduction
The figure of Plato
Plato is one of the most significant figures in the history of philoso-

phy. His work has inspired and fascinated philosophers in all ages.
Platonism was the dominant school of philosophy in the Mediter-
ranean region in late antiquity, and has had many revivals since
then; much philosophy has also been produced in reaction against
his thought. Plato had an additional impact on the development of
philosophy through his pupil Aristotle, who was both an admirer
and a critic, and who has an equal claim to be the most influential
philosopher of antiquity.
Plato stands at the beginning of many debates that have contin-
ued throughout the history of philosophy. Nevertheless he seems
alien to us in many ways. His central metaphysical position, the
“theory of Forms”, involves the claim that things in the world that
we see are images of eternal patterns; although this position has
some affinities with more recent views, it is presented in his works
in a way that makes it seem very unfamiliar. His dualist conception
of the mind, according to which it can exist independently of the
body, is unpopular in recent thought, and his conception of God is
at least alien to mainstream philosophy. His political views are also
now generally rejected; and while his scientific theories are in one
1
plato
way strikingly modern, as he sees the physical world as having a

mathematical basis, they are, of course, in detail wholly out of date.
However, his moral philosophy is certainly relevant to issues that are
still being discussed and, with its emphasis on virtue and its relation
to happiness, is perhaps becoming more so as these topics become
more central in ethics. Some aspects of his philosophy of mind, in
which he reflects on such topics as the conditions for responsibility
and the connection between freedom and rationality, are also rele-
vant to ongoing debates.
The study of Plato faces two dangers. One is to make him too
much “one of us”. By applying a principle of charity one can make
him sound more like modern philosophers than he really was;
alternatively, one may recognize the alien elements in his thought,
but engage in a process of “rational reconstruction” to present an
updated version of Platonism from which outlandish elements are
excluded. The other danger is to assume that, because Plato’s way of
thinking is so different from ours, he has nothing to say to us as a
philosopher, but is of purely historical interest.
I hope to show that it is possible to trace a path between these two
extremes, and to see Plato as a philosopher who writes from a very
different perspective from ours, and yet has things to say that are
relevant to us. Sometimes his ideas may actually be seen as making
a contribution to current debates; at other times they may draw our
attention to issues that are unfairly neglected.
The emphasis that Plato places on argument is important here; he
was committed to following an argument wherever it might lead. It
may be that he did not always live up to that commitment, but never-
theless his works are full of arguments; he is not just the spokesman
of a worldview, but presents reasons for his beliefs, which we can still
follow and either accept or argue against.
Plato’s thought has two dominating features, which may at first
sight seem to conflict with one another. On the one hand, he is often
seen as an “otherworldly” philosopher; this is certainly true by con-
trast with Aristotle, and even more so by contrast with the materialist
philosophers who were influential both before and after Plato’s time.
Indeed, he did not believe in a literal other world, wholly detached
2
introduction
from this one. But he did believe that there is more in the world than
we can be aware of through the senses, and that the part that is not
perceptible is the most significant.
One of his most distinctive beliefs was in what he calls “Forms”;
put very briefly Forms are universal essences, for instance the essence
of goodness, of beauty, of justice, and also of human being, of fire
and so on. They are grasped by reason, not by perception; they are
eternal and unchanging; and they do not depend for their existence
on particular instances of goodness, beauty, humanity or whatever
it might be. They may also be thought of as patterns, of which the
things we perceive in the world are imitations.
Plato believed in an immortal soul that is able to exist in sep-
aration from the body. He also believed in a God who designed and
made the world; although, like most Greeks of his time, he probably
believed in many gods, this single supreme God played a central part
in his thought. It seems likely that while in the earlier part of his
career the Forms and the immortal soul were central to his view of
the world, later the figure of God became more important to him.1
The other major feature of Plato’s thought is that he is an intensely
practical philosopher. By this I mean not that he is down to earth
and realistic – on the contrary he is extremely idealistic – but that
his philosophy is always directed towards action. Philosophy for him
is a way of life, and philosophical reasoning is important in guiding
the way we should live. While for Aristotle the practical life and the
life of philosophical contemplation are distinct, each with their own
virtues and their own kinds of happiness, Plato makes no such dis-
tinction; for him philosophical knowledge helps us make practical
decisions, and to live a good life one should either be a philosopher
oneself or be guided by someone who is. Only through philosophy
can we gain knowledge or understanding of the good, and we need
this knowledge if we are reliably to do what is good. This is rele-
vant both at the personal, ethical level – people need knowledge of
the good if they are to guide their own lives effectively – and at the
public, political level – rulers need knowledge of the good to guide
the states they rule effectively. While Plato’s writings deal with a
wide range of subjects – metaphysics, theory of knowledge, theories
3
plato
about the soul, God, nature and so on – these are always related to
practical concerns.
Equally, though, this practical orientation is always guided by
Plato’s metaphysical vision. Forms are ideals that we should strive to
be like, and knowledge of them will help us govern our lives; central
to the system of Forms is the Form of the good. The immortality of
the soul is important because we must take care of our souls and
prepare them for the afterlife. God is significant because he gives
purpose to things in the world, and in his later works Plato puts
forward an ideal of “becoming like God” (Tht. 176b; Ti. 90c).
Plato’s life
Plato was born in 427 bce to a wealthy upper-class family in Athens.

Athens was a city state, and at that time was, in the Ancient Greek
sense, a democracy; indeed, it was seen as the prime example of
democracy in Greece. This is not a democracy in the modern sense –
women were excluded from politics, and slavery was practised – but
all adult male citizens were allowed, and to some extent expected, to
take part in politics. All were allowed to take part in the assembly
that made major state decisions; many office-holders were selected
by lot. This, together with the small size of the state, meant that
politics played an important part in many people’s lives.
During his youth Plato became a friend and companion of the
philosopher Socrates. At that time Socrates was a public figure in
Athens; he was eccentric, and not universally popular (the comedian
Aristophanes wrote a play, the Clouds, largely devoted to satiriz-
ing him); but was on good terms with some influential people. He
did not write, and did not teach formally, but hung out in public
places, markets, gymnasia and so on, meeting people, asking ques-
tions and starting discussions. Socrates was immensely influential;
people came from all over the Greek world to see him. Not only
Plato but also many other philosophers were influenced by him;
later, most of the great schools of Greek philosophy would trace
back their lineage to him.
4
introduction
During much of Plato’s youth Athens was at war with Sparta,

a rival state with a contrasting, oligarchic political system. In 404,
when Plato was twenty-three, the war came to an end with a Spartan
victory. A commission of thirty members was set up, with Spartan
support, to reform the constitution on oligarchic lines; this included
two relations of Plato, Charmides and Critias, both of whom also had
links with Socrates. In practice the thirty took over the government
of the city. They rapidly alienated the people, behaving in an unscru-
pulous manner, and carrying out unlawful arrests and executions in
order to seize people’s wealth, and swiftly came to be known as the
thirty tyrants. Soon there was a popular uprising in which they were
overthrown, and Charmides and Critias were killed.
It seems likely that these events would have led to Plato becoming
disillusioned with the oligarchic politics to which his upbringing
might have attracted him; that he was so disillusioned is suggested
by a letter, ascribed to Plato (Seventh Letter 325a), although it is
uncertain whether it is really by him. If so, he was soon to become
disillusioned with democracy as well (325b–c). In 399, four years
after the fall of the thirty, Socrates was accused of “corrupting the
youth, and not recognising the gods whom the city recognises” (Ap.
24b), tried by a democratic court, found guilty and put to death.
Socrates was seventy years old, and had been practising philosophy
for thirty years or more, and had been at least tolerated up to that
point, so it is surprising that the Athenian people turned against
him; perhaps his links with Charmides and Critias, and the thought
that he had influenced them, contributed to this change in public
attitudes.
In any case, after Socrates’ death Plato, with other companions of
Socrates, left Athens. It is not known where he went. A tradition has
him spending some time in Egypt – this may have been inspired by
references to Egypt in his writings. A rather more secure tradition,
supported by his letters, if they are genuine, suggests that he went
to Syracuse in Sicily, at that time a Greek city, where he came into
conflict with Dionysius, the dictatorial ruler of Syracuse. About the
mid 380s he returned to Athens, and made it his base for the rest of
his life, but did not take part in the political life of the city.
5
plato
Plato founded a school at Athens; it was in the precincts of a

shrine dedicated to the hero Academus, and so became known as the
Academy: the precursor of all the schools, colleges and learned soci-
eties that have since been given that name. It was a school not only
of philosophy but also of mathematics; Plato himself does not seem
to have been an active mathematician, but his colleagues Eudoxus
and Theaetetus were, and Eudoxus was also an astronomer. Plato
thought mathematics was valuable because it encouraged the prac-
tice of abstract thought and turned us away from reliance on the
senses. Aristotle, the most important philosopher of the next gen-
eration, was a student, and perhaps later a teacher, in the Academy.
Teaching in the Academy seems to have been informal, and based
on discussion, rather in the manner of Socrates. Plato gave only one
public lecture, which was not very successful (Aristox. Elem. Harm.
II.30–31); most of his teaching was carried out in private.
Plato’s aim in founding the Academy was practical; he hoped that
the teaching would enable people to govern states more effectively;
young members of ruling families were encouraged to come there
to make them into better rulers. Having been disillusioned both
with elites based on wealth or military ability and with democracy,
Plato wanted a government formed by people who had understand-
ing of the aims of government: philosophers. His most famous
work, the Republic, is, among other things, a manifesto for philo-
sophical government and an explanation of what it might look like.
It is also possible that Plato remained involved in politics at Syra-
cuse throughout his life – although once again the evidence for this
depends on his letters – and that there, in particular, he tried to
establish a philosophical government by educating Dionysius the
Younger, son of the tyrant Dionysius, as a philosopher. If so, the
plan was unsuccessful.
Plato died in 347 bce. He had never married and had no children;
his nephew, Speusippus, succeeded him as head of the Academy.
6
introduction
Plato’s writings
We have handed down to us from antiquity thirty-five dialogues and

a number of letters under the name of Plato. It is disputed whether
the letters and a few of the dialogues are really Plato’s; but we know
that most of the dialogues are his from the testimony of his pupil
Aristotle, who often refers to them. So far as we can tell, nothing
that Plato wrote for publication has been lost; this is very unusual
among ancient writers.
It is not clear when Plato began writing dialogues. Possibly, as
some stories from antiquity suggest, it was during Socrates’ life; if
not, as is now more widely believed, it was shortly after his death.
The dialogue – what purports to be a record of a philosophical
discussion – was a new literary form in Plato’s time. In the Poet-
ics Aristotle refers to it specifically as the Socratic dialogue (Poet.
1447b10), suggesting that it developed among the followers of Soc-
rates, and was inspired by his method of philosophical discussion,
as if his style of philosophy could not be captured in a treatise. Many
disciples of Socrates wrote dialogues, although most survive only in
fragments; apart from Plato’s, only Xenophon’s still exist as complete
works.
Socrates is the chief speaker in most of Plato’s dialogues. In a few
late dialogues – the Timaeus, Critias, Sophist, Statesman and Laws –
he is replaced as chief speaker by someone else, although he appears
in all these dialogues except the Laws; while in the Parmenides he
shares the honours with Parmenides, a leading philosopher of the
previous generation. Most of the dialogues are called after a character
who appears in them, but there are a few exceptions. The Apology
(defence speech) has a name that expresses its form; the Symposium
is called after its setting (a drinking party); and the Republic, Sophist,
Statesman and Laws are called after their themes.
In general, each of Plato’s dialogues is an apparent record of a
philosophical conversation, although again there are a few excep-
tions: the Apology, a reconstruction of Socrates’ defence speech at
his trial, is a continuous speech; the Menexenus and Timaeus are
introduced by passages of dialogue, but the main portion of each of
7
plato
them is a single speech; and it seems that the Critias, an unfinished

sequel to the Timaeus, would have been similar.
Plato’s dialogues cover an immensely wide range of themes. They
include material that we would now class as moral and political phil-
osophy, aesthetics, metaphysics, theory of knowledge, logic, and
philosophy of mind, of science and of religion, and also material
that we would find hard to classify under any of these standard head-
ings, perhaps most notably his discussion of love. What is more, his
philosophy is a seamless web; discussion of one of these topics is
always bound up with discussion of the others, and while there is
generally in each dialogue a single theme that dominates the discus-
sion, simply stating that theme often does not indicate the complex-
ity of the work.
The topic that receives the most discussion in Plato’s work is eth-
ics. There is a large group of dialogues that are widely thought to
have come early in Plato’s career and to reflect the philosophy of
his teacher, Socrates, whose own interests were exclusively ethical.
(Questions about Plato’s development and his portrait of Socrates
will be discussed in Chapter 2.) These include two dialogues devoted
to the defence of Socrates, the Apology and the Crito, and many
dialogues in which he is shown pursuing his enquiry into the vir-
tues, including the Protagoras, Euthyphro, Charmides, Laches, Lysis,
Euthydemus and Gorgias. The Meno also includes an enquiry into
the nature of virtue in Socratic style, but adds to this a discussion of
knowledge and of philosophical method, and ventures into meta-
physics when it puts forward a hypothesis about the previous life
of the soul.
Plato’s best-known work, the Republic, is generally thought to have
been written about the middle of his career. It is a good illustration
of the way in which Plato interweaves many themes in one work.
One of its central claims is that there is a parallel between justice in
the individual soul and justice in the state; it is, therefore, a contri-
bution both to ethics and to politics, describing an ideal state and
asking how it might be brought about, while also discussing justice
in the individual and whether it is beneficial. Because Plato’s vision
of the ideal state involves rule by philosophers, it also includes a
8
introduction
discussion of metaphysics and the theory of knowledge, as well as

an extended contribution to aesthetics in a discussion of the place
of poetry in the state.
Other works often seen as belonging to the central period of
Plato’s career are the Phaedo – which, in the context of an account
of Socrates’ death, argues for the immortality of the soul – and the
Symposium and Phaedrus, both of which are discussions of love;
they approach it from rather different perspectives, but both see
it as something that stimulates us to philosophy. The Cratylus is a
discussion of language, the Theaetetus an enquiry into the nature
of knowledge, and in the Parmenides Plato subjects his best-known
metaphysical theory, that of Forms, to criticism.
Plato did not abandon his ethical and political interests in the
last phase of his life. Among works commonly thought to have been
written then, the Philebus is an account of pleasure from an ethical
perspective, while both the Statesman and the Laws have a political
theme; the Laws is another plan for an ideal state, less utopian and
so perhaps more realizable than that of the Republic. The Sophist dis-
cusses a logical problem concerning the possibility of negative state-
ments and of falsehood, while the Timaeus deals with the origin and
nature of the world, including both theology and physical science.
The dialogue form
One of the most striking features of Plato’s works is that they are
dialogues, not treatises. This both raises the question why he wrote
in this form and presents us with a puzzle as to whether the dialogues
can really be used as evidence for Plato’s own views.
We have seen that the dialogue form was a new one in Plato’s
time, and was created by the followers of Socrates. To them it must
have seemed a suitable way of capturing their master’s approach to
philosophy. Since he was concerned more with enquiring and ques-
tioning than with stating positive views, it was not possible simply to
expound his philosophy; it had to be demonstrated, by recording the
kind of conversation in which he engaged. Yet Plato continued to use
9
plato
the dialogue form in works where he seems clearly to be developing

his own ideas, and where the chief speaker expresses positive claims.
It is true that whereas in some dialogues a real enquiry or debate
seems to be taking place, others are more like treatises, with the chief
speaker dominating discussion and others mostly agreeing with him.
Yet Plato never wrote a book simply setting out his positions, but
always used the dialogue form. Why is this?
In the Phaedrus, Plato’s Socrates expresses the view that speech
is preferable to writing as a medium for philosophy (Phd. 275d ff.);
writings can be read without understanding, and if their meaning is
unclear they cannot be asked to explain it. The best way of learning
philosophy is through a face-to-face discussion. There is, of course,
a paradox here; it is through his writings that we learn that Plato
defended this position. But the paradox is not insurmountable; Soc-
rates is not saying that a philosopher never commits his own views
to writing, only that if they are so committed they cannot be a source
of knowledge. It seems likely that Plato did indeed believe something
like this, and that he did not intend his writings, by themselves, to
give his readers the kind of knowledge that a philosopher aims at;
this, for him, could only be achieved by actual discussion. “Phil-
osophy” for him is the name of a way of life or of a motive, love of
wisdom; the actual activity of philosophers he calls “dialectic”, the
art of discussion (although by a fruitful ambiguity it can also mean
the art of making distinctions). The point of his writings, in this case,
will be to give a demonstration of what philosophy is like, attracting
people to philosophy and stimulating philosophical thought.
This may make us uncertain whether we can actually see any
teachings of Plato in the dialogues; the views expressed are always,
strictly speaking, those of a character, not of the author, and we may
wonder why we should believe what that character says, if the work
is not meant to be a source of philosophical knowledge. Certainly,
we should be cautious in ascribing views to Plato on the basis of
what his characters say. He need not be committed to every point
that they make; the dialogue form allows him to put ideas forward
while standing at a remove from them. But it is generally clear which
character has most of Plato’s sympathy; Socrates, and the chief speak-
10
introduction
ers of the later dialogues, seem to be seen as philosophers with real

insight, and so it seems probable that the views they put forward are
ones Plato agrees with, at least provisionally. When the same ideas
turn up in the mouth of the chief speaker in several dialogues, it is
likely that they are fairly settled features of Plato’s thought.
One specific reason why we should be careful about attributing
the views of a chief speaker to Plato is that, if a dialogue depicts a
genuine ongoing enquiry, the views of the chief speaker may not
be stable throughout it. For instance, it is controversial whether
Plato sees the ideal state that he describes in the Republic as a real
possibility. When Socrates is asked this question he says that it is
possible, if philosophers should ever become rulers (Resp. 473c–e),
and in a couple of later passages he explicitly says that this is a real
possibility, although hard to achieve (502c, 540d); he even explains
how philosophers should put his proposal into action by sending
away from the city everyone over the age of ten, so that those who
remain can be trained in the kind of life that he proposes as an ideal
(541a).
On the other hand, he gives a lengthy description of the training
a philosopher needs to reach the insight required for government,
in a way that suggests it could only happen within the ideal state
(536d ff.),2 leaving it unclear how philosophers could bring that state
into being in the first place; and at the end of the main discussion of
the Republic, he accepts that this city does not exist on earth, but says
that “perhaps it is a pattern laid up in heaven, which he who wishes
can look upon, and looking upon it establish himself ” (592b). This
echoes an earlier passage (472b ff.), which says that the description
of an ideal is valuable even if it is not achievable; it can be read, in the
light of the claim that justice in the soul and in the state are parallel,
as meaning that a person can take the ideal state not as a political
proposal, but as a pattern for ordering his own life.
If we look for a consistent position for Socrates it is hard to find,
but it is possible to see his view as evolving within the dialogue: at
first he is trying to see the ideal state as a practical possibility, but
later he accepts that it is a pattern, not practicable but still useful as
an inspiration.
11
plato
There is another reason why it might be thought dangerous to

look in the dialogues for Plato’s thought; it is not simply that the
views put forward in them are put forward by characters rather than
by Plato himself, but that each dialogue is rooted in a particular situ-
ation, and its arguments are directed to a particular interlocutor. It is
reasonable to think that the chief speaker in each dialogue will tailor
his arguments to the person with whom he is debating, arguing in
a way that is likely to appeal to that person, and perhaps appealing
to his convictions (as Socrates does explicitly in some dialogues).
It is certainly true that the situation and characters are important
in each dialogue – although they are emphasized more in some dia-
logues than in others, where they tend to fade into the background
– and that for this reason we cannot simply identify what Socrates
says, or the assumptions he seems to make, with what Plato thinks.
Yet we should remember that although in the first instance Plato’s
arguments are addressed to the characters in the dialogues, Plato cre-
ated those characters (even if they are based on real people, as most
of them seem to be); he gave them the specific commitments they
are shown in the dialogues as having. His reason for creating them
in the form in which he did is presumably to enable him to make
philosophical points that will be relevant to his ultimate audience,
the readers of his work.
I have suggested that Plato’s aim is not simply to show a philoso-
pher in action but also to stimulate philosophical thought in his audi-
ence: not just to report debate, but to start a debate in which we,
the audience, can become participants. If we are to do this, we must
to some extent be able to abstract the positions and arguments put
forward in his works from the particular context in which he shows
them arising, although this will obviously affect the way they are pre-
sented; it must be possible to consider those arguments from our own
point of view, and ask whether they are philosophically convincing.
Certainly the fact that a dialogue deals with a specific situation
does not always prevent it having relevance beyond that situation. Of
Plato’s works, one of the most clearly rooted in a particular context
is the Crito, in which Socrates, facing the prospect of death, defends
his decision not to escape from prison. Yet the arguments of the Crito
12
introduction
are among the most “live” of Plato’s arguments even today, and are
still discussed in debates about the source and extent of our duty to
obey the law. Those who face these issues now are not in exactly the
same situation as Socrates, and this has to be taken into account in
considering whether his arguments are relevant; but this does not
stop us seeing those arguments as a contribution to the debate.
If we consider both that Plato’s chief speakers are, in general, fig-
ures whom he intends us to see sympathetically and that he wants us
to see his arguments as having relevance beyond the specific context
in which they are put forward, it is reasonable to think that we can
see Plato’s own thought lying behind what his characters say – as
indeed his readers have done since antiquity – although care must
always be taken in ascribing particular views to him.
The plan of this book
My aim in this work is to introduce some of Plato’s central ideas.

To gain a full appreciation of his thought, one needs to see it in the
context in which he presented it, within the specific dialogues; this
introduction cannot be a substitute for that. But there are continuing
themes in his work, which develop from one dialogue to another,
and it is useful, in understanding Plato, to get an overview of these
themes; hence, I have adopted a thematic structure.
After a short chapter on Plato’s development and his relation to
Socrates, I deal first with Plato’s principal contribution to metaphys-
ics, the “theory of Forms”, which, although rarely Plato’s central
theme, forms a necessary background to his thought in many areas,
and next with his theory of knowledge, which is intimately linked
with the Forms. I consider his view of the soul, which is both an
important theme in itself and a significant background to his thought
on practical issues, and then move on to his practical philosophy,
considering first his politics and then his ethics. Finally I deal with
two areas of Plato’s thought that have had an especially important
historical impact, not confined to academic philosophy: his theory
of God and nature, and his aesthetics.
13
plato
Many aspects of Plato are missing from this account. His view of
love, an intriguing topic in itself, will be mentioned here only in so
far as it illuminates other aspects of his thought; and his contribu-
tions to the theory of logic and language will also not receive much
attention. I hope I can nevertheless present enough of Plato’s thought
to give a picture of this fascinating philosopher, and inspire readers
to study his work further.
14
two
Plato’s development and
Plato’s Socrates
Two puzzles
Two major problems confront anyone who wants to study Plato’s

philosophy. First, Plato’s writing career lasted at least fifty years, and
it is reasonable to think that there were some changes in his views
during that time. But how radical were the changes, and to what
extent can we trace the development of his thought? Can we use
positions expressed in one dialogue in interpreting another?
This problem is particularly acute because we have no direct evi-
dence on when each of his works was written. Occasionally it may
be possible to date a work by a reference to a real world event, or
to find in one dialogue a reference to another, which enables us to
determine the order in which they were written. Many scholars have
believed that they can trace, by scientific means, the development of
Plato’s style, which allows them to work out the order of his works
(see Brandwood 1992), but it is highly controversial how much this
method can show. Often, therefore, hypotheses about the order of
the works depend on theories about how his thought might plausibly
have developed; but this obviously produces a danger of circularity.
Some interpreters stress the unity of Plato’s thought. Others focus
more on differences between the dialogues.1 One influential theory
sees his thought as falling into three periods. In the first, the theory
15
plato
claims, he was exclusively concerned with ethical issues, and his

work was exploratory in nature; in the second, the “middle period”,
he became more dogmatic, and, while still concerned with ethics,
also developed the metaphysical views for which he is famous, the
immortality of the soul and the theory of Forms. It is to this period
that what are probably his most famous works, the Phaedo, Sym-
posium, Republic and Phaedrus, belong. In the last period, which
includes the Parmenides, Theaetetus, Sophist and Laws, he was more
interested in questions to do with logic and knowledge, together with
practical issues. Some have even argued that during this period he
abandoned his most distinctive metaphysical theory, that of Forms;
others suggest that, at least, it became less important to him.2
The second problem is this: in most of Plato’s dialogues, the chief
speaker is his teacher, Socrates. Should we suppose that this is a
portrait of Socrates as he really was? Or is the Socrates who appears
in the dialogues a spokesman for Plato, expressing either his own
views, or at least views he wants his audience to consider? Since
antiquity, people have looked to Plato’s dialogues for the expression
of his own philosophy, but also for evidence about Socrates. But is it
really possible to find both of these in the dialogues at once?
The view that Plato’s Socrates represents the real Socrates in every
place where he appears is now generally rejected; in the dialogues
with the strongest metaphysical content, such as the Phaedo and the
Republic, he is expressing views that are distinctive of Plato. But one
group of dialogues is often seen as giving us evidence of the histor-
ical Socrates, and for this reason they are often called “Socratic dia-
logues”. This group includes two works, the Apology and the Crito,
that deal with Socrates’ trial and the events leading up to his death,
and a number of works in which he appears engaged in an investi-
gation of virtue and the virtues, such as the Protagoras (virtue as a
whole), Euthyphro (holiness), Charmides (temperance) and Laches
(courage).
If this is right, the question of Plato’s Socrates becomes closely
linked with that of his development, for the “Socratic” dialogues
are precisely those that are often seen as coming at the beginning
of Plato’s career. It is widely thought that, at the beginning of Plato’s
16
plato’s development and plato’s socrates
development, he would have been very much under the influence

of his master, and would have written works that closely reflected
Socrates’ thought, while later he may have found his own voice and
begun to express distinctive views of his own.
In what follows I shall consider to what extent we should follow
this view of Plato. I shall begin with the Socratic question, in the
hope that this will illuminate Plato’s development.
Plato’s picture of Socrates
The dialogues commonly called “Socratic” paint a distinctive picture

of Socrates. His primary concern is with ethics, and he does not obvi-
ously hold distinctive metaphysical views about Forms or the soul.
He denies that he has knowledge of moral subjects (Ap. 21b), and
his investigations rarely reach a definite conclusion. He is committed
to the testing and, where necessary, refutation of views: often these
are the views of others with whom he is engaged in debate, although
sometimes they are proposed for discussion by him. He is often
concerned with definition of ethical terms; as well as the Euthyphro,
Charmides and Laches, mentioned above, there is the Lysis (friend-
ship) and the Hippias Major (the fine or beautiful). He often uses
inductive methods of argument, in a broad sense of the term; that is,
he defends a general claim by looking at particular instances where
it is true. Finally, despite his profession of ignorance, he maintains
certain moral convictions. These include an “intellectualist” concep-
tion of virtue, involving the claims that virtue is knowledge3 and that
no one knowingly does wrong (Prt. 352b ff.); a belief that virtue is
sufficient for happiness (361b; Meno 89a); and a conviction that one
should not return evil for evil (Cri. 49b ff.).
This picture of Socrates seems to contrast in many respects –
although not all – with that found in the classic works commonly
assigned to Plato’s “middle” period. The Socrates of those works has
a much wider range of interests, holds metaphysical views about
Forms and the soul, does not emphasize his ignorance (although
occasionally he still professes it) and argues for definite conclusions.
17
plato
We should be careful in just how we draw the contrast between

these two groups of dialogues. Some of the contrasts that are per-
ceived simply relate to an element that is present in one group but
not in the other. Thus, in one group of dialogues there is a breadth
of interests that is absent in the other, and specific metaphysical
views that the other does not contain; but it is not clear that this
means that they represent either two different thinkers or two dif-
ferent phases in Plato’s thought; it might be explained simply by
Plato’s different purposes in writing them. We should not suppose
that he was obliged to mention in every dialogue every topic he was
interested in or every doctrine he held at the moment of writing
it. Likewise, the fact that some dialogues lack definite conclusions,
and seem more concerned with the testing and refutation of claims,
while others argue for definite positions, can be explained in terms
of a difference of purpose.
Nevertheless, there are at least two respects in which there seems
to be a real opposition – not just a difference of focus and approach
– between the two groups of dialogues. First, in the “Socratic” dia-
logues, perhaps most clearly in the Protagoras, Socrates defends the
famously paradoxical view that everyone desires what is good and
no one desires what is bad (Prt. 458d; Meno 77c ff.); hence if people
pursue what is actually bad, this must be explained by ignorance
of the good. This allows him to argue that virtue is knowledge, and
that akrasia or weakness of will, the phenomenon by which good
intentions, although we have them, are overcome by unruly desires,
is not in fact possible. By contrast, in the Republic Plato’s Socrates
argues that there are different sources of motivation, not only the
rational desire for the good but also desires for pleasure and for repu-
tation, which can oppose our rational desires (Resp. 439c ff.). If this
is so, it seems that akrasia is possible, that there are other sources of
wrongdoing besides ignorance, and that knowledge is not sufficient
for virtue, which also requires mastery of desire.
Secondly, in the “Socratic” dialogues Socrates is ready to dis-
cuss philosophy with everyone, and wants to encourage everyone
to examine their views in a philosophical way; this is his vocation,
as set out in the Apology, and we see him actually doing it in other
18
works. By contrast, in the Republic he is shown taking the view that

only a few are capable of philosophy, and that others should ideally
be subject to their care and government. One might think that there
is no direct opposition here, for the Socrates of the Apology could
be seen as thinking not that all are actually capable of philosophy,
but only that all should at least be given the chance to do it. But two
points of contrast stand out. First, in the Apology Socrates claims
that “the unexamined life is not worth living” (Ap. 38a), while in the
Republic he seems to be encouraging many people to live unexam-
ined lives. Secondly, in the Republic (539a ff.), encouraging the young
to do philosophy, when they are not adequately prepared for it, is
condemned as damaging, not just as unnecessary, thus seemingly
condemning what Socrates is shown doing in many dialogues (for
instance the Charmides, the Lysis and the Euthydemus).
It seems, then, that there is a real contrast in Plato’s works between
two points of view. It also seems that the first of these points of view –
the intellectualist and egalitarian one – is found mainly in works that
have the other features commonly seen as Socratic: an exclusively
ethical concern, an exploratory rather than dogmatic outlook and
the absence of metaphysical theories about Forms and the soul. This
suggests that these features are not just explained by a difference of
purpose, but reflect a different outlook from that found in classic Pla-
tonic works such as the Republic. It is reasonable, therefore, to divide
Plato’s works into two groups, reflecting different ways of thinking.
(This is not to say that the distinction is an absolutely rigid one; two
works in particular, the Gorgias and Meno, seem to have elements
that link them with both groups). What is not yet clear is whether
these just represent two stages in Plato’s development, or whether
one of them expresses the thought of the real Socrates.
Is this the real Socrates?
To answer this question we must look at what other authors have to

say. Aristotle, Plato’s pupil, frequently refers to Socrates; it is often
not clear whether he is talking about the historical figure or the
19
plato
character in Plato’s works. However, in two places in the Metaphys-

ics (A 987a9–b10; M 1078b9–32) he contrasts Socrates with Plato.4
In both he claims that Plato originated the idea of separated Forms,
although in doing so he was trying to cope with a problem raised
by Socrates’ thought; Socrates was concerned only with ethics, and
not with the nature of things as a whole. However, he did introduce
into philosophy the search for definitions and the use of an induc-
tive method. It seems that Aristotle is using some works of Plato
as evidence for Socrates, since his picture of Socrates fits what we
find in works such as the Charmides, Laches and Euthyphro, while
he takes other works, such as the Phaedo and Republic, to reveal the
thought of Plato.
Another work, the Magna Moralia (traditionally ascribed to Aris-
totle, but perhaps actually by a pupil of his) contrasts the ethical
views of Socrates and Plato (Mag. mor. I 1182a15–30); it says that for
Socrates the virtues are forms of knowledge, while Plato recognized
an irrational element in the soul that meant that virtue could not
simply be identified with knowledge. Once again, the views ascribed
to Socrates seem to be those that are found in the Protagoras and
related dialogues, while those ascribed to Plato can be found in the
Republic, as well as the Phaedrus and Timaeus.
How might Aristotle (and his pupils) have known which elements
in the dialogues reflected the thought of the real Socrates? We can-
not say with certainty, but nor can we rule out the possibility of their
knowing. Perhaps Plato was actually open about this, explaining to
his students that some elements in his works represented his own
thought while others came directly from Socrates. Or perhaps Aris-
totle was relying on the testimony of other pupils of Socrates who
were still active in Athens when he first came there; if there was a
consensus between them about some aspects of Socrates’ thought,
it would have been reasonable to ascribe at least these aspects to
Socrates himself.
In any case Aristotle must have had some reason for drawing the
distinction between Socrates and Plato in the way he did. If he had
nothing to go on but the dialogues themselves it is not clear how
he could have drawn that distinction; he might have thought either
20
that everything ascribed to Socrates in the dialogues represents his

real position, or that, as they are literary productions, nothing in
them can be seen as historical. That he did neither, but distinguished
between some elements that represent Socrates and some that do
not, suggests that he did have independent evidence to support that
distinction.
Some later authors seem to divide the dialogues in the same way.
Cicero, writing in the first century bce, contrasts Socrates with
Plato (Cic. Acad. I.4.16–17), claiming that Socrates was concerned
exclusively with ethics, denied that he knew anything and focused
on refuting others, while Plato began to work out a definite philo-
sophical system. He clearly is using some of Plato’s dialogues as a
source for Socrates, but he says that the writings of other followers
of Socrates confirm this way of seeing him.
The Stoics, a philosophical school that began to flourish two
generations after Plato, saw Socrates as a forerunner of their way
of thinking; their picture of him seems, once again, to have been
similar to that found in the “Socratic” dialogues. In particular the
Stoic teacher Epictetus, who worked in the late first and early second
century ce, describes the different tasks that God gives to philoso-
phers (Epict. Discourses III.21.18–19), and says that he gave Socrates
the task of elenchus – refutation or examination – as opposed to the
task of positive teaching, which he reserves for Zeno, the founder
of Stoicism.
All in all, it seems that these dialogues were widely seen as giving a
portrait of Socrates, and as distinct from those that introduced Plato’s
own philosophy, and that it is reasonable to see them as genuinely
Socratic.
Plato’s Socratic dialogues
This need not mean that Plato’s Socratic dialogues are actual reports
of conversations in which Socrates engaged; they can be seen as fic-
tions, but ones that seek to present Socrates as he was, with his real
attitudes and views, trying to show how he would have confronted
21
plato
a particular issue or situation. One work that raises particular puz-

zles is the Apology, a reconstruction of Socrates’ defence speech at
his trial (in Greek apologia means a defence, not an apology in the
modern sense). We need not see this as accurately representing what
Socrates actually said at his trial – many of Socrates’ pupils wrote
defences for him, not all of which can be accurate reports of what
was said – but as it is a defence of Socrates, it must reflect his life as
Plato saw it.
Why should Plato write Socratic dialogues? One reason would be
to honour his master and to defend him. This would obviously apply
to the Apology; and in some other dialogues we see him following
out the vocation, which he describes in the Apology, of testing the
views of others and their claims to wisdom. The Crito is another
dialogue that has the aim of defending Socrates; set in prison after
his condemnation, it seeks to justify his action in refusing to escape
(as he could easily have done) and remaining to face death. Another
reason for writing dialogues in a Socratic manner would simply be to
stimulate philosophical thought, as Socrates himself had tried to do.
What, then, can this tell us about Plato’s development? I suggest
that it tells us less than we might hope. For if Plato’s aim in these
dialogues is to give a portrait of Socrates, we cannot be sure that they
represent Plato’s own thought at the moment that he wrote them.
He must at least have had enough sympathy with Socrates’ position
to think it worth setting before his readers, but that does not mean
that he completely agreed with it; nor does it mean that he had no
views of his own on topics about which Socrates professes ignorance,
nor that he had no interest in topics beyond those which Socrates
is seen discussing.
Talk of a Socratic period in Plato’s work is dangerously ambigu-
ous. It may mean either a period during which Plato’s aim in his
dialogues was to give a portrait of Socrates, or a Socratic period in
Plato’s thought, in which he shared a Socratic point of view. Often,
it has been assumed that the dialogues of Plato’s early period are
Socratic in both these senses. Thus, Gregory Vlastos has claimed
both that these dialogues are a representation of the historical Soc-
rates and that in them Plato never ascribes to Socrates a view that
22
he does not share.5 But it may be thought unlikely that both these
things could be true at once; one wonders whether Plato would have
simply continued to follow his teacher’s thought over a long period
during which he produced many dialogues.
If these dialogues are intended to represent the thought of the
historical Socrates, it is possible that Plato’s own thought was already
developing away from Socrates at the time that he wrote them; he
may have developed metaphysical views about Forms and the soul,
and perhaps a psychological view different from Socrates’ intellec-
tualism, while he was recording Socrates’ position. Indeed, he may
even have been writing, at the same time, dialogues that presented
his own view: the difference between Socratic and Platonic dialogues
can be explained, not by development, but by a difference in inten-
tion. (Stylistic evidence strongly suggests that the Republic is later
than the bulk of the Socratic dialogues, but it seems not to do so for
the Symposium and Phaedo, other works that express Plato’s meta-
physical views; see Brandwood 1992.) If this is right, while Plato’s
thought no doubt did develop during the early part of his career, we
cannot trace that development with confidence.
Plato’s development: the later phase
However, even in those works that seem clearly to express Plato’s own
views there are variations in style and approach, and, apparently, in
doctrine. How are we to understand Plato’s later development?
On one influential view, there is a contrast between the central
group of works that express Plato’s most famous metaphysical posi-
tions – the Phaedo, Symposium, Republic and, perhaps, the Phaedrus
– and a later group of works, distinguished on grounds both of style
and content, including the Sophist, Statesman, Philebus, Laws and,
probably (although this is controversial), the Timaeus. While in the
earlier group Socrates is shown putting forward positive views, he
does so largely in a context of debate; in the later group ideas are
presented in a more didactic way, with the chief speaker (who is often
not Socrates) facing little opposition to his claims. The subjects are
23
plato
also more down to earth; they include logical questions (in the Soph-
ist), ethical and political issues (in the Statesman, Philebus and Laws),
and the physical world (in the Timaeus). While extravagant meta-
physical positions are not wholly absent, they are less emphasized,
except in the Timaeus, and even there they function as background
for Plato’s physical theory.
Between these two groups come two dialogues that seem to reveal
a more critical and questioning position: the Parmenides, in which
objections to Plato’s central metaphysical theory of Forms are con-
sidered; and the Theaetetus, which, in some ways recapturing the
spirit of the Socratic dialogues, investigates the concept of knowledge
without coming to a definite conclusion.
On grounds of both style and content I think something like this
picture of Plato’s development is correct. There is a group of dia-
logues that can be recognized as coming late in Plato’s career, includ-
ing the Timaeus, Sophist, Statesman, Philebus and Laws. As well as
a more didactic approach, there are several thematic links between
these works. They take the material world more seriously than earlier
works, and show a more optimistic view of it. They give the creator
God a central role; and they also show a more down-to-earth, prac-
tical approach to ethical and political issues, emphasizing the point
that abstract philosophical knowledge needs to be supplemented by
awareness of particular facts.
What is less clear is whether the change between the central works
and the later ones is a change of focus and method only, or also of
doctrine. It has been argued, most notably by G. E. L. Owen (1953),
that a radical shift in Plato’s views took place late in his career, and
that he abandoned his most distinctive metaphysical position, the
theory of Forms. Certainly arguments against that theory are put
forward in the Parmenides, and there is no explicit answer to them.
However, it does not follow from this that Plato actually means us
to abandon the theory; indeed Parmenides, the chief speaker of the
dialogue, himself suggests that Forms are necessary for all discourse
(Prm. 135b–c). (I shall discuss this in more detail in Chapter 3.) I
suggest, therefore, that while the theory may have changed in detail,
it cannot have been wholly abandoned.
24
What seems to be recognizably the same theory is present in dia-

logues normally dated later than the Parmenides. This is most clearly
true in the Timaeus, where Forms play a major role; Owen had to
argue that this dialogue comes earlier in Plato’s career than is com-
monly thought. However, not only does its style link it with later
works but also it shares some themes with them. The figure of God,
the creator, who is central to the Timaeus, also appears in other late
works; the theme of classification, of a world structured into genera
and species, which plays a large part in the Sophist, Statesman and
Philebus, can also be found in the Timaeus.6 It seems likely, therefore,
that Plato did continue to believe in Forms in the last phase of his
career, and that there was no truly radical shift in his doctrines; and
if we accept this, it is possible to see the theory of Forms at work in
other places too, including the Statesman and Philebus.
Continuity and change in Plato’s thought
This does not mean, however, that Plato’s thought was completely
static; he was open to changing his positions in response to argu-
ment. Indeed, some aspects of his thought about Forms seem to have
changed. In the Phaedo and Phaedrus he says that our knowledge of
Forms was gained by an experience that we underwent in a discarnate
state before birth, and in this life we are trying to recover knowledge
that we have lost. In the Republic and Timaeus, on the other hand, this
doctrine is not mentioned, and an account of the afterlife is given that
leaves no room for it; it seems possible to gain knowledge of Forms for
the first time in this life. (I discuss this at more length in Chapter 4).
Again, in the Timaeus, Forms are said to be eternal in the sense
of being timeless (Ti. 37e ff.); “was” and “will be” cannot be applied
to them. This view does not appear in earlier dialogues, where they
are simply said to exist for ever. It may well be that Plato actually
introduced the idea of timelessness in response to arguments put for-
ward in the Parmenides, which imply that whatever is in time must
be in change (Prm. 141a–d, 152a–e); hence if Forms are changeless,
as Plato thinks they must be, they must also be timeless.
25
plato
Plato’s view of the soul also seems to have changed. In the Phaedo
the soul is something simple and indivisible, and is identified with
the rational element; what we might naturally think of as mental
conflict is instead seen as a conflict between the soul and the body.
In the Republic, Phaedrus and Timaeus the soul has three elements,
the rational, the spirited and the appetitive (the part where desires
connected with the body are found), and conflict between them is
possible. In the Phaedrus the whole soul is immortal; the Repub-
lic and Timaeus compromise by making only the rational element
immortal.
Another example of a change in Plato’s view concerns the origin
of the soul. In the Phaedrus he insists that it is ungenerated (Phdr.
245d); he can also be seen as implying this in the Republic, since
there he claims that everything that is generated will be destroyed
(Resp. 546a), which means that if the soul were generated it could
not be immortal. However, in the Timaeus (34c ff.) and the Laws
(892a and elsewhere) he refers to the soul as being generated; and
in the Timaeus (41a ff.) he seems specifically to renounce the view
that everything generated will be destroyed, saying instead that while
everything generated is destructible, some things are preserved in
being by the will of God.
Not only are Plato’s views not unchanging, but it is not clear that
they develop in a linear way. It is quite possible that he sometimes
abandoned a position and later returned to it. Indeed, as I noted in
Chapter 1, the dialogue form means that Plato need not commit
himself to all the views that his chief speakers express, even if he
clearly intends to present them favourably.
Nevertheless, we can see some stability in Plato’s central ideas;
the existence of Forms and of a creator God, the immortality of the
soul and the centrality of virtue are convictions that remain with him
through most of his career. We must be cautious in assuming that
what is said in one dialogue represents Plato’s settled position; but
it is possible to see some unity in his ideas, and discuss his thought
as a whole.
26
three
Plato’s metaphysics: the
“theory of Forms”
One of the best-known aspects of Plato’s thought, and one of the

most pervasive, is his belief in “Forms”:1 that is, very roughly, in
essences shared by those things that (in our terms) belong to a sin-
gle kind or possess a single property. For instance, there is in his
view something, the essence or nature of goodness, that all good
things have in common; this is what Plato calls the Form of the good.
Likewise there is an essence that all large things have in common,
one that all living creatures have in common, and so on. Plato has
a distinctive conception of these essences, central to which are the
claims that they are eternal and unchanging, that they are grasped by
pure reason rather than by perception, and that they do not depend
for their existence on their perceptible instances.
Plato does not, in fact, always use the term “Form” to denote these
essences; his most common expression for them is “the so and so
itself ” (e.g. “the good itself ”, “the living creature itself ” and so on);
on some occasions he also uses abstract nouns to describe them
(e.g. “justice itself ”). However, he does sometimes refer to them by
two Greek words, eidos and idea, both of which can be translated by
the English word “form”; and this way of referring to them, which
is particularly useful when speaking of all the Forms as a class, has
become standard in later writers, beginning with Plato’s pupil Aris-
totle in his discussion and criticism of the theory. (While the word
27
plato
idea is indeed the ancestor of our modern English “idea”, it should

not be taken to imply, as the English word normally does, that these
things are in the mind; for this reason, while older books often refer
to “Platonic Ideas”, “Forms” is now the more usual term.)
Plato never devotes a dialogue to the subject of Forms, with the
partial exception of the Parmenides, the first part of which is taken
up by a consideration of some criticisms of the theory; more often, it
appears as part of the background to his thought on other subjects,
such as the immortality of the soul (in the Phaedo), love (in the
Symposium and Phaedrus), ethics and politics (in the Republic), and
cosmology (in the Timaeus). Plato never expounds his position on
the subject systematically, and, although I shall continue, for con-
venience, to speak of the theory of Forms, it is not clear that his
views were really systematic enough to be called a theory. It may be,
however, that they became more so in the course of Plato’s career:
in the dialogues where they first appear, Forms tend to be taken for
granted; in later works we see him considering criticisms of them
(in the Parmenides) and presenting his only explicit argument for
them (in the Timaeus).
Nevertheless, it is clear that Plato’s views about Forms were seen in
his time as distinctive and controversial; they go beyond the simple
thought, which might be seen as a platitude, that when many things
are (correctly) described by a single term, there is in some sense
something that they have in common. Plato thought that his position
was worth defending, and others thought it worth attacking. The
arguments discussed in the Parmenides seem to be based on those
used by contemporary opponents of the theory; and Aristotle, who
agreed with his teacher about many things (the existence of God, the
presence of purpose in the world, the connection between virtue and
happiness), nevertheless has a philosophy that differs radically from
Plato’s, largely because of his rejection of Plato’s account of Forms.
In this chapter I shall first try to clarify what Plato’s conception of
Forms is and what is so distinctive about it, and then look at some
specific problems that his account of Forms raises.
28
plato’s metaphysics: the “theory of forms”
What is a Form?
A Form, for Plato, is in the first place the nature or essence shared by
many things, when the same term is rightly applied to them; to know,
for instance, the Form of the good is to know what goodness is – to
know the essence that all good things have in common. Hence, a
Form is in some ways like what in modern philosophy is called a uni-
versal; indeed, Aristotle, who introduced the term “universal” into
philosophy, saw Plato’s theory as a kind of anticipation of his own.
Sometimes when Plato discusses Forms he is simply concerned with
this aspect of them: the idea of an essence shared by many things. In
the Socratic dialogues Socrates is often presented as searching for
an account of what something, for example virtue, is – an account
of the characteristic shared by all the people who are virtuous, and
manifested in particular virtuous acts and in all the more specific
kinds of virtue – and in two places, in the Euthyphro (6d–e) and the
Meno (72c–e), he calls what he is looking for a Form.
In the Phaedo Plato’s Socrates claims that Forms are causes or
explanations of things having a certain property (Phd. 100c ff.); for
example, the Form of large is the cause of things being large. This can
be read as meaning not that the Form actually acts on things to make
them large, but that by understanding what largeness is we can come
to understand why things are large. It is rather as we might say “This
is a square because it has four equal sides and four equal angles”;
knowing what a square is enables us to know why a particular object
is a square.2 This, again, echoes the Socratic dialogues, where Soc-
rates claims that it is because of a single Form – for instance, that of
piety – that individual pious acts are pious (Pl. Euthphr. 6d).
Plato clearly believes that these essences have a real and objective
existence that we can discover; this is parallel to what we would now
call a realist theory of universals. This is by no means uncontrover-
sial; many philosophers, called nominalists, have held that there is
nothing over and above the particular things that we touch and see.
But the realist view is nevertheless a widespread one; Plato’s accept-
ance over it does not in itself make his metaphysical views unusual.
However, his account of Forms has some more distinctive features.
29
plato
One of the most notable aspects of his theory is the so-called

separation of Forms. In the Parmenides he refers to Forms existing
“separately” (Prm. 130b ff.); elsewhere he speaks of a Form existing
“itself by itself ” (Phd. 65d; Symp. 211b; Prm. 129a), or of its not
being in anything other than itself (Ti. 52a). Exactly what this means
is rather obscure. It does not seem to mean that Forms are literally
in another place than their sensible instances; strictly speaking they
are not in a place at all. One thing that the language of separation
seems to imply is that Forms are not affected by what happens to
their instances. Another possible implication is that they do not
depend on their instances for their existence; there would be a Form
of beauty, for instance, even if nothing beautiful existed in the world.
That Plato did indeed believe this is implied by the Timaeus, where
we are told that Forms existed before the world was made (Ti. 52d),
and were used as a model by God as he brought things in the world
into existence (29a and elsewhere).
Another aspect of Forms that was clearly important to Plato is that
they are eternal and unchanging. The contrast between unchanging
Forms and the constant change of the sensible world is drawn in
the Phaedo, where it is argued that the soul is more akin to Forms
and is therefore likely to be immortal (Phd. 78c ff.); and again in the
Timaeus, where it is seen as a reason for thinking that there can be
knowledge, in the strict sense, of Forms and not of sensible things (Ti.
27d ff.). In the same dialogue Timaeus, the chief speaker, takes the
further step of saying that whatever is in time is in change, and hence
Forms, if they are truly changeless, must also be timeless (37e ff.).
A third important feature of Forms is that they are not perceptible,
and that it is through reason, not through the senses, that they are
known. This point is emphasized in the Phaedo, where it is argued
that the soul does not make use of the body in gaining knowledge
of the Forms, and is, in fact, better equipped to contemplate them in
separation from the body (Phd. 65d ff.), and in the Republic, where
a contrast is drawn between the realm revealed through the senses
and the realm (of Forms) grasped by the intellect, and we are urged
to practise mathematics as a way of reducing our reliance on the
senses (Resp. 523a ff.).
30
All these features of Forms are certainly controversial; but none

of them seems wholly unreasonable. The idea of separation – of uni-
versals that do not depend on their instances for their existence – is
unpopular in modern metaphysics, but it was held by philosophers
such as Bertrand Russell and G. E. Moore; and for some properties
it does not seem too implausible. Perhaps this is particularly true of
moral properties; it makes sense to say that there is such a thing as
justice even if no one is actually just. Likewise with some mathemati-
cal properties; one might say that there is such a thing as circularity
even though nothing is perfectly circular.
The changelessness of Forms is also quite a plausible claim; while
particular things may become, for instance, larger and smaller, what
largeness or smallness is is always the same, and does not change
when the instances change.
Finally, we may well think that Plato has underestimated the role
played by perception in our gaining knowledge of the natures of
things, if he thinks this can be achieved without any perception at
all; but it still seems right to say that we do not directly perceive many
properties of things, and that perception alone cannot give us know-
ledge of their natures, but rather reflection is required. This would be
true, again, of moral properties, and also of extremely abstract prop-
erties such as being, sameness or difference. (I shall explore this ques-
tion further in Chapter 4, in discussing Plato’s theory of knowledge.)
None of these features of the theory of Forms, then, explains
why the theory seems so striking, and why it has made so strong an
impression on some philosophers, and seemed so alien to others.
To understand this, we should look at another aspect of the theory,
the evaluative one: the way in which Plato sees the Forms as ideals
of which the sensible world falls short.
Plato sees Forms as having a special value, which other things do
not share; this reveals itself in two ways. First, it is good to understand
and to contemplate Forms. This is something to which we should
aspire and from which, when it is achieved, we will get pleasure; it
will also have a good effect on our soul, stimulating us to virtue.
This is perhaps made apparent most dramatically in two dialogues,
the Symposium and the Phaedrus, which have the theme of love. In
31
plato
both of these it is argued that the ultimate object of love is the Form
of beauty, and that love of beautiful people is valuable in so far as it
stimulates us to look for this Form. The Form, therefore, is an object
of emotion similar to, but stronger than, what we ordinarily feel
for our beloved (Symp. 210e ff.; Phdr. 250b ff.). A similar thought
is found in the Phaedo, where the philosopher is shown looking
forward to death because this will enable him to understand Forms
more clearly without the distraction of the body (Phd. 65d ff.), and
in the Republic, where the philosophers who are rulers of the ideal
state prefer to spend their lives in contemplation of Forms, and need
to be compelled to take part in government (Resp. 519c ff.).
Secondly, it is also good to resemble Forms. Forms are paradigms,
models or patterns that we may look to in shaping things in the
world, and a thing is in a better state the more closely it corresponds
to them. This is especially apparent in two places: in the Republic
philosopher rulers are described looking to the order and harmony
of the realm of Forms, and trying to produce a similar order in the
state they are governing (500b ff.); in the Timaeus God, in the same
way, looks to a Form as he makes the world, and seeks to make his
product as perfect as possible by making it resemble the Form as
closely as possible (Ti. 30c ff.).
This brings to light an important aspect of Forms; things can
instantiate them to a greater or lesser degree. In saying that there is
a Form of beauty, Plato means not simply that there is an essence
that all beautiful things share, but that there is an essence that things
share to the extent that they are beautiful; and as things can be more
or less beautiful, they can share in the Form to a greater or lesser
extent. But in fact Plato seems at points to go further than this. It
is not just that things can fall short of the Forms, but that things in
the sensible world necessarily do so: they can, by divine or human
efforts, be brought closer to the perfection of the Forms, but will
never fully attain it. This produces a contrast between the perfection
of the Forms and the imperfection of the sensible world, and gives
the Forms an importance they would not otherwise possess.
In a way, I suggest, it is not so much Plato’s actual view of Forms
that is so distinctive, but his view of the inadequacy of the material
32
world, which leads him to contrast it with Forms. It is this contrast

that is the really striking aspect of his theory; but this turns as much
on his theory of material things as his theory of Forms. In the next
section I shall look more closely at his account of the material world,
and ask in what way it falls short.
What is wrong with the world?
How does the material world fall short of Forms? I want to suggest
that there are several ways in which it does so, and it is dangerous
to concentrate on one of these ways to the exclusion of others. In
particular, it may be that it falls short in different ways in relation
to different Forms.
According to a traditional reading of Plato, things may fall short
of Forms by possessing the properties of which they are the Forms in
an imperfect way. For instance, a person or a society may fall short
of the Form of justice by not being perfectly just; a diagram may fall
short of the Form of circularity by not being perfectly circular. This
view is not wholly wrong, but we should be cautious in using it to
interpret Plato.
Clearly, Plato thought that the material world does in many cases
fall short of Forms in this way; and it seems plausible that he thought
it necessarily true that it does so. At the end of the Republic Plato’s
Socrates seems to conclude that the ideal state is not actually achieva-
ble, but is “an example laid up in heaven” to which people can look in
guiding their own lives (Resp. 592b). Also in the Republic he suggests
that the movements of the heavens, because they are visible, cannot
be wholly unchanging and without deviation (530a–b; although it
is possible that he later changed his mind about this, as he came to
give more significance to the universe and its maker3).
However, this cannot be the only way in which the world falls
short of Forms, for with some Forms it is hard to make sense of
the idea of a thing possessing a property imperfectly. For instance,
consider largeness. It is in a sense true that nothing is perfectly
large, but this is not because we have a concept of perfect largeness
33
plato
and nothing happens to instantiate it; it is because “perfect large-

ness” does not make sense. “Large” is a comparative term, and for
anything we see as large, we can always imagine something larger.
When we call something large we normally have some comparison
in mind. For instance, a small elephant is one that is smaller than
other elephants; it will still be large in comparison with land animals
generally. Plato often makes use of relative terms such as “large”
in drawing attention to the contrast between Forms and things.
The point here is not that material things’ instantiation of Forms is
imperfect but rather that it is qualified; what is large in one way will
be small in another.
In the case of largeness, the obvious way in which a thing’s
instantiation of the Form may be qualified is by comparison: the
thing is large in comparison with one thing, small in comparison
with another. But there are other possibilities. For instance, in the
case of unity something might be unified in one respect, multiple
in another (by having many parts or many properties). In the Sym-
posium, Diotima (the chief speaker in the central portion of the
dialogue) lists a number of ways in which the beauty of something
may be qualified: it may be beautiful in one respect, ugly in another;
beautiful in comparison with one thing, ugly in comparison with
another; or beautiful in one context, ugly in another (Symp. 211a).
One might wonder why it should be seen as a defect of things that
they possess their properties in a qualified way. Certainly this could
be used to show that there is a distinction between Forms and things;
we can hardly identify largeness itself with the class of large things if
they are only qualifiedly large. It might also be used to show that we
do not get our knowledge of Forms from things (at least completely).
We cannot come to understand what largeness is just by looking at
an elephant, since the elephant is also in some respect small; we need
to use reflection, interpreting our experience in the light of concepts
that do not derive from experience. But why should it make things
less perfect, less valuable, than Forms?
I think that it does so precisely because it makes them unsuitable
as a source of knowledge. For Plato knowledge is something valuable,
and one reason that Forms are more valuable than sensible things is
34
that they are appropriate objects of knowledge; if we seek knowledge

we must not focus on the sensible world, but on Forms.
In this context it is worth noticing that in the Symposium Diotima
also refers – apparently as a gloss on the point about different contexts
– to the possibility that something might be beautiful to one person,
not to another. This is rather surprising, since Plato normally defends
objectivity in the realm of values; one would expect him to say that
if two people disagree about whether something is beautiful, one of
them is wrong, and so its beauty is not really qualified. One possibil-
ity is that he is thinking not just of a difference in taste but about two
people with different purposes: an object may be well made, and so
beautiful, when considered in relation to one purpose, but not when
considered in relation to another. But alternatively one might read
the passage as meaning that when people disagree about whether
something is beautiful then, even if it is truly beautiful, it is inad-
equate as an example of beauty because it is not an appropriate source
of knowledge. We cannot learn what beauty is from an example, if we
cannot even agree whether the example is really beautiful.
While in many cases it seems clear that Plato thinks of sensible
things falling short of Forms because their possession of properties is
qualified, this way of falling short will again not work for all Forms.
In particular, it will not work for the Forms of natural kinds, such as
human being and other animals, or fire and water. There is no obvi-
ous way in which our status as human beings is either imperfect, or
qualified. Is there, then, a way in which we fall short of the Form?
Once again the Symposium provides a clue, for it mentions a fourth
way in which things may fall short of perfect beauty, besides respect,
comparison and context. They may also be beautiful at one time, ugly
at another. Plato, then, seems to think of change as one of the ways
in which material things fall short of Forms. Things can of course
change in respect to beauty, and size, shape and other qualities that
correspond to Forms: but they can also, for Plato, change in respect
to their natural kind. The Timaeus makes it clear that the physical
“elements” can turn into one another (Ti. 49b ff.): water, for instance,
turns into air when it evaporates, and air may condense into water.
Something similar can be true of species of living creature. When I
35
plato
die, my body will turn to dust, while my soul, for Plato, will go on to
live another life, either out of the body or in another body, and may
no longer be human. Hence, being human is a temporary property of
me, and in this way I fall short of the Form of human being.
As we have already seen, Plato saw change as pervasive in the
material world, and thought of this as one of the central contrasts
between that world and the Forms. What is less clear is just what, in
his view, the extent of this prevalence of change was. In the Phaedo
he claims that sensible things “one might say, never remain the same”
(Phd. 78b). In saying this he seems to be accepting what is com-
monly called a theory of flux: a theory that in his time was associated
with the Presocratic philosopher Heraclitus, who is credited with
the famous sayings “Everything is flowing” and “It is impossible to
step into the same river twice”. Aristotle tells us that as a young man
Plato was influenced by the thought of Heraclitus, and that this is
what motivated him to believe in separated Forms. We cannot, he
thought, have knowledge of sensible things because they are in con-
stant change, so, if there are objects of knowledge they must be sepa-
rate from the sensible world (Arist. Metaph. A 987b1 ff.). Certainly,
as both the Phaedo and Timaeus imply, Plato accepted the view that
there was some kind of flux in sensible things, but just how extreme
was the kind of flux in which he believed?
In the Theaetetus Socrates discusses a very extreme theory of flux
(apparently more extreme than Heraclitus’ own) according to which
things are changing in all respects all the time, and never remain the
same in any respect (Pl. Tht. 179d ff.). This theory is rejected, and it
seems unlikely that Plato ever thought that it was true of the sensible
world. For one thing, it has very strange consequences: it implies
– as the Theaetetus shows – that we can never hope to speak the
truth about any sensible thing, because it will already have changed
while we are speaking. Clearly, Plato – even if he believed we have
no knowledge, strictly speaking, of the sensible world – thought we
could speak meaningfully about it. For another thing, in the Timaeus
the constant change of the sensible world is described as something
that we see: yet we clearly do not see things changing in all respects
all the time.
36
What then does Plato mean when he speaks of the world being
in constant change? There are two possibilities, both of which may
be aspects of his view. First, he may have thought that everything
is changing in some respect all the time, even if in other respects it
remains stable enough to be recognizably the same entity. This is a
possible reading of Phaedo 78b, and is also supported by the Sympo-
sium, where Diotima describes both our body and our soul as being
involved in constant change (Symp. 207d ff.). While it is not a directly
observed fact – many things we see have no visible change going on
in them – it is supported by many scientific theories; there are con-
stant processes of change maintaining the visible stability of things.
Secondly, everything is liable to change in all respects, and per-
haps will sooner or later change into something of another kind. This
seems to be the point being made in the Timaeus, where the elements
are described changing into one another: it is not that all fire is con-
stantly changing into air (although perhaps some fire, somewhere,
is always doing so) but that all fire is liable to change. Although this
passage is one of the most puzzling in Plato’s work, and there is much
dispute about its philosophical point, it can be read as meaning that
what we see is not essentially fire, since it is capable of changing into
something else: fieriness is just a quality that it has. If I point to my
desk and ask “What is that?”, it would not be appropriate to answer
“brown”, although it is brown, because brownness is just a quality it
has and might lose. Plato is suggesting that those predicates, such
as “fire”, that we might naturally think of as identifying the object
in question, are really similar to “brown”: they really only ascribe
qualities to things.
This is the aspect of change that is most directly relevant to the
question how things fall short of Forms. Ever since Aristotle it has
often been thought puzzling why the fact that something possesses
a property at one time but not at another should be taken to mean
that it possesses it less than perfectly. But Plato may have felt that if
something is at one time beautiful, at another not, and so beauty is
not essential to it, this means that it cannot be beautiful in the fullest
sense, because beauty is not part of what it is; it is, rather, something
it happens to possess because of its relation to something else.
37
plato
In any case, it seems clear that Plato did see the changeableness
of the sensible world as an important part of the contrast between
it and Forms, and as one of the ways in which it falls short of them,
however he worked this out in detail. He saw stability and reliabil-
ity as good qualities, and so was disturbed by the constant change
of the world around him. Because of this, the fact that Forms are
unchanging – that what justice, beauty and so on are is always the
same – not in itself a particularly surprising claim, took on for him
a special significance.
Forms, bodies and space
In the Timaeus, Plato gives a fuller account of the nature of bod-

ies and their relation to Forms. We have seen that the attributes
that physical things owe to Forms are treated there not as part of
their essence, but rather as qualities that might change; and these
attributes include not only things such as colour and shape, but also
attributes such as fieriness, which define bodies as the distinctive
sort of thing they are. This raises the question whether there is any
element in physical things that is not liable to change in this way. It
turns out that according to the view presented here there is: this is
a mysterious entity that is described by a number of metaphorical
terms, of which the best known is the “the receptacle of becoming”
(Ti. 49a). Plato clearly finds the receptacle puzzling and difficult to
explain; it is not understood through reason, as Forms are, but nei-
ther is it directly perceptible; nevertheless he thinks we have to admit
its existence. While bodies such as fire and water can properly only
be called “such and such”, not “this” – they are only qualities, not
things in their own right – the receptacle can be called “this”: it is
the real thing that we are pointing to when we seem to be indicat-
ing a body (49d–50a). While at first it is mysterious just what the
receptacle is, in the end it is identified with space (52a–b).
The passage of the Timaeus that deals with the receptacle is very
obscure, and almost anything that can be said about its interpretation
is controversial. But it is possible to see the receptacle as in some
38
way underlying and composing bodies, so that bodies come into

being when parts of the receptacle take on qualities derived from the
Forms. Thus, for instance, fire comes into existence when a region of
space becomes fiery. Sometimes the fire seems to be identified with
the qualified region of space (51c) and at other times it is identified
with the quality it takes on (49d ff., 52c), but the basic conception of
what happens is the same in either case.
An analogy that may illuminate the relation of bodies to the
receptacle is the relation of a wave to water. The wave comes into
being when water takes on a particular form, but the wave may move
through the water, so that different parts of the water take on the
form in turn. In the same way, when a body moves through space we
can interpret this as different regions of space taking on the quality,
for example fieriness, that is distinctive of that body.4
Why does Plato introduce the receptacle into his theory? One
reason may arise from his view of change. It is possible to argue
that in every change there must be a factor that remains unchanged.
Otherwise it is not really a change, that is, an object changing from
one state to another; rather, it is just one object being replaced by
another. For instance, if a man turns into a pig something must have
remained throughout the change, or else we cannot say that he has
turned into a pig, but only that a man has disappeared and a pig
appeared. What remains through the change cannot be in itself either
human or porcine; it has those qualities at different times, but they
cannot be its essential properties, in terms of which it is identified.
In ordinary changes, as of shape or colour, it is normally easy
to pick out the thing that persists through change, and the proper-
ties that identify it. But the most fundamental kind of change, that
between the physical “elements”, fire, water and so on, involves all
sensible qualities. So if there is something that persists through this
change, it must have no sensible properties of its own.
These thoughts have led some philosophers to posit something
that has no properties of its own at all. This theory, the theory of
“prime matter”, is sometimes ascribed to Aristotle, and was certainly
held by later Aristotelians. But it is an uncomfortable view. If the
underlying thing has no properties of its own, how do we identify it?
39
plato
Plato, by identifying the underlying thing with space, overcomes

this problem. Space has properties of its own – its three-dimensional
structure – but it has no sensible properties, such as colour, sound,
heat and so on. Hence it can be seen as what underlies sensible
change. When one object is transformed into another, this can be
seen as a change in space; the region of space persists through the
change, but is transformed from one quality to another.
However, there is another reason why Plato posits the receptacle;
this has to do with the status of bodies as images. At 52d, Timaeus
says that we ordinarily believe in the existence of space because we
think that everything that exists must be in something; this is not in
fact true of realities (Forms), which do not exist in anything other
than themselves, but is true of images. Because an image derives its
nature from something else, it must also exist in something else. The
point seems to be that images are relational entities. An image of
Socrates, for instance, is constituted by the relation of resemblance
something bears to Socrates. Relations of this kind cannot exist on
their own; there must be something that has the resemblance.
Is it generally true that images must exist in something other than
themselves? The passage is sometimes read as referring specifically
to mirror images, which clearly do exist in a medium distinct from
themselves, the mirror. But it can also be read as covering more sub-
stantial images, for example a bronze statue, if the image is identified
not with the figured bronze but with the figure in the bronze. An image
of Socrates exists in bronze; it exists because the lump of bronze bears
a relation to Socrates. In the same way bodies exist in space; they exist
because a region of space bears a relation to a Form. Forms and space
are the ultimate existents (at least in relation to the sensible world,
although a full account of what exists would also have to include God
and souls); bodies are constituted by the relation between them.
How many Forms?
One puzzle to which Plato’s theory gives rise is just which general
terms have Forms corresponding to them. Plato faces this problem
40
directly at one point, at the beginning of the Parmenides (130b ff.).

This dialogue is generally thought to have been written after those,
like the Phaedo and Republic, where the theory of Forms is first
developed at length. In it Socrates, as a young man, is shown pro-
posing a version of the theory of Forms, and facing criticism from
Parmenides, the greatest philosopher of the previous generation.
Parmenides raises the question what Forms Socrates believes in. His
answer is that he definitely believes in two classes of Forms: those
for extremely general predicates such as like and unlike, one and
many, rest and motion, and those for moral or evaluative predicates
such as just, beautiful and good. He is less certain whether there are
Forms for natural kinds, such as human being, fire and water, and
he definitely rejects Forms for “worthless” things such as hair, mud
and dirt.
Parmenides, however, replies that Socrates is saying this because
he is young, and that as he grows older and is more “gripped by
philosophy” he will see none of these things as beneath his notice.
It seems that Parmenides wants to expand the realm of Forms to
cover a wider range of predicates. Perhaps we should see Socrates as
too focused on the contrast between Forms and the sensible world,
thinking of Forms as something exalted of which sensible things
fall short; the emphasis on evaluative Forms encourages this way of
thinking. If he were to take more seriously the fact that the sensible
world is an image of Forms, he would recognize that the kinds of
thing we find in the sensible world also have Forms corresponding to
them. Certainly in the Timaeus Forms of natural kinds are present,
both the “elements”, such as fire and water (Ti. 51b–e), and kinds of
animal, of which human being is an example (39e–40a). While he
does not explicitly mention a Form of hair, he does discuss the place
of hair in the scheme of creation, making it something that might
appropriately have a Form (76b–d).
It seems, then, that Plato accepts at least three classes of Form:
those of extremely general properties, those of moral and evaluative
properties and those of natural kinds. The first two are more domi-
nant in the earlier works that feature Forms. The evaluative Forms
are Plato’s central concern, while the extremely general Forms such
41
plato
as unity, largeness and equality often feature in arguments intended

to reveal the contrast between Forms and sensible things. In later
works the extremely general Forms become more of an object of
interest in their own right, while the natural kind Forms also become
important.
Plato sometimes refers to a fourth class of Forms: those of arte-
facts, such as a weaver’s shuttle or a bed (Cra. 389a–c; Resp. 596b ff.).
He suggests that craftsmen, when they create these things, work with
the relevant Form in mind. Some have felt that in these cases Plato’s
language should not be taken seriously. But in fact it seems appropri-
ate, given Plato’s general outlook, that there should be such Forms,
for there is, for him, no fundamental difference between the works
of craft and those of nature. Both are produced by intelligence, in
the one case human intelligence, in the other divine, and the maker
of the universe, himself described as a craftsman, looks to a Form
in making the world, just as human craftsmen do in producing their
works.
It seems, then, that Plato accepts quite a wide range of Forms.
But is the range unlimited? Is there a Form for every meaningful
predicate? One line that suggests there is can be found in Book 10
of the Republic: “we usually posit a single Form in connection with
each of the many things to which we apply the same name” (Resp.
596a). That is, it seems, if there is a class of many things to which the
same name (i.e. the same general term, such as “good” or “shuttle”)
is applied, there is a Form. However, it is possible to read this with
the emphasis on “posit”, in which case it may mean not that there
actually is a Form in every such case, but that it is reasonable to start
by assuming that there is; however, this assumption may turn out
to be wrong.
In what circumstances might it be wrong? A clue is given by a pas-
sage in the Statesman, where the chief speaker, the Eleatic Stranger,
says that there is no Form corresponding to the term “barbarian”,
meaning a human being who is not Greek (Plt. 262d); this is because
it is not appropriate, in classifying human beings, to begin by divid-
ing them into Greeks and everyone else. As we shall see in Chapter
4, in several dialogues Plato discusses the classification of things in
42
the world, and distinguishes between natural and arbitrary ways of

classifying things, insisting on the importance of “dividing at the
joints” (Phdr. 265e). It seems possible, therefore, that he thinks there
is a Form for every predicate that corresponds to a natural, rather
than an arbitrary, way of classifying things.
Self-predication
One of the most puzzling aspects of Plato’s discussions of Forms is

that he often seems to think of them as instances of themselves; for
example, the Form of the good is described as something good (Resp.
532c), the Form of living creature as a living creature (Ti. 30c), and
so on. This raises many problems. For one, some Forms, if they are
instances of themselves, will turn out to have properties that are not
appropriate for Forms. An obvious example is the Form of change
(discussed in the late dialogue, the Sophist; e.g. Soph. 255d); if it is
an example of itself it must be in change, but it is central to Plato’s
conception of Forms that they do not change. Another example is
provided by Forms of artefacts, such as bed or shuttle. One might
well feel that it is essential to these that they were made by a crafts-
man, but it seems to be essential to Forms that they are ungenerated.
(In Book 10 of the Republic Socrates does actually suggest that the
Form of bed was made, by God [Resp. 597b], but this seems hard to
fit into his overall theory.)
Another problem is posed by Forms of relative properties such
as largeness. As we have seen, the contrast between these Forms
and the things that share in them turns largely on the thought that
nothing is large without qualification; a thing must always be large
in comparison with one thing, small in comparison with another.
If the Form is predicated of itself, then presumably it will be large
without qualification, but this seems not to make sense; the idea of
something large without qualification is not just one that happens to
have no instances in the world, but one that is incoherent.
Yet another problem is one of concreteness. If, for instance, the
Form of human being is a human being, it must have specific concrete
43
plato
qualities of the kind all human beings have: it must have hair of a
particular colour, a nose of a particular length, and so on. If this is
so, actual people will resemble the Form more closely, and so come
closer to perfection, the closer they are to having these particular
qualities. But this is implausible. It is more natural to think of the
ideal for human beings as something more abstract, which does not
fill in every detail; and certainly Plato gives no sign of thinking there
is a completely specific ideal for human beings, or for any other kind
of thing. (Another problem arising from self-predication, the “third
man” argument, will be discussed later.)
These problems have led some to suggest that Plato’s use of self-
predicative language should not be taken seriously. One suggestion
(Vlastos 1972) is that Plato is using the idiom found, for instance,
in Chapter 13 of Paul’s First Epistle to the Corinthians, where he says
“Love [or in some versions ‘Charity’] is kind and long-suffering”. This
clearly does not mean that love is itself kind, in the sense in which
people are kind, but rather that, necessarily, everyone who loves is
kind. As a limiting case of this, one might say that love is loving, since
it is true, although unexciting, that everyone who loves is loving.
Certainly some of Plato’s uses of self-predicative language might
be explained in this way; and the fact that this idiom exists may
be part of the reason why he found it natural to use this language.
However, it does not seem that this can account for all Plato’s uses
of this language; for in some places it seems to be important to him
that the Form itself has the property in question, and because of
this it explains the possession of the same property by its instances.
An alternative view, put forward by Michael Frede and others,5 is
that statements of the form “X is F” (e.g. “justice is virtuous”) can be
understood in two ways: one in which they mean that X has Fness
as a property; the other in which they mean that X has Fness as
part of its nature. For instance, we might say that justice is virtuous,
meaning that being virtuous is part of what it is to be just. Another
example of this kind of predication is something like “The porpoise
is a mammal”. This statement is not about an individual porpoise,
but about the species, and does not mean that it literally is a mam-
mal – that it has warm blood, suckles its young, and so on – but
44
rather that being a mammal is part of what it is to be a porpoise. It

is this second kind of predication that is relevant to Plato’s account
of Forms. When he says something like “justice is virtuous”, this can
be taken to mean that being virtuous is part of what it is to be just,
and, once again, “justice is just” can be read as a limiting case of this;
being just is the whole of what it is to be just. (In other cases Plato
can be seen as using the first, more ordinary kind of predication in
connection with Forms. When he says, for instance, that Forms are
eternal, this means simply that they have eternity as a property.)
This account makes sense of much of what Plato says, and may be
seen as implicit in much of his discussion of Forms. If we now want
to defend something like a theory of Forms, this would seem to be a
good path to follow. Nevertheless, I am doubtful whether Plato ever
himself clearly recognized this distinction between kinds of predica-
tion. There are some things in his work that do seem to require that
Forms should be seen as literally instances of themselves, possess-
ing themselves as a property. In the Symposium, it is central to the
argument that the Form of beauty is itself something beautiful, an
appropriate object of love, the contemplation of which is worth aspir-
ing to and gives satisfaction. In the Republic the Form of the good
is described as the best of realities (Resp. 532c). Moreover, in many
places the relation between Form and instance is described as a rela-
tion of likeness, and the Form is described as a model (paradeigma;
e.g. Ti. 29a), of which the instances are copies; this is most simply
understood as meaning that the Form itself has the same property
that the instances possess, and they have the property by resembling
it. Likewise, the claim that instances fall short of the Form is most
straightforwardly understood as meaning that they do resemble it,
having a property in common, but not perfectly.
An interesting way of making sense of much of this language
without accepting the stranger consequences of self-predication has
been proposed by I. M. Crombie (1962–3: vol. II, 274–5), who sug-
gests that a Form should be seen as something like a design: the
relation of instances to Forms is like the relation of, say, a house to
the design on which it is built. (Of course, the comparison should
not be pressed too far; the design was made by someone for the sake
45
plato
of the house, while a Form is ungenerated and does not exist for the
sake of material things.) The design does not literally possess the
properties of the house, for example being made of bricks, or being
three storeys high; but these properties are contained in the design.
We can talk of the actual house as resembling the design, and indeed
as resembling it more or less closely, as falling short of it. Much of
the language Plato uses about Forms can be understood in the same
way. Moreover, it is reasonable to think that the design of something
beautiful will itself be beautiful. (The special puzzles raised by the
Form of the good will be discussed in Chapter 7.)
I do not want to suggest, though, that Plato had a well-worked-out
theory along these lines, but only that this will help to make sense of
much of what he says. The distinction between this view and literal
self-predication is quite easy to miss unless one is actually looking
for it, and Plato may not have fully distinguished them. It certainly
seems that in some places he sees Forms as standards, of which
particular things may fall short; the Form of justice, say, is not just
the quality common to all just people or states, but the standard of
justice, to which people and states may approximate more or less
closely. But “standard” can mean either an abstract standard – a prin-
ciple specifying what a truly just person would be like – or a standard
object – something that is itself truly just. It is the first notion that
the idea of a design seeks to capture, but Plato may not have clearly
separated it from the second. He wants to affirm that the Form of
justice represents perfect justice, contrasting with imperfectly just
individuals; but then it is easy to fall into thinking of the Form as a
perfect example of justice. However, I hope I have shown at least that
something that is recognizably a theory of Forms can survive without
this implication. Certainly, if we emphasize this aspect of Forms we
will end up with a deceptive picture of the theory; the realm of Forms
will come to seem too much like a literal other world – just like ours,
only more perfect. It seems to have been this aspect of Forms that
repelled Aristotle. He could not see any point in the theory if it just
duplicated our world (e.g. Ar. Metaph. B 997b5 ff.). It is better to
see the realm of Forms not as a literal other world, but as a pattern,
providing standards against which our world can be judged.
46
Forms and reality
Another striking aspect of the theory of Forms is that Plato often

refers to Forms as having “being”, by contrast with sensible things,
which are sometimes described as “becoming” (e.g. Ti. 27d–8a),
while in the Republic they are referred to as “between being and not
being”, in that they “are” in one way but “are not” in another (Resp.
478d ff.). What does “being” mean here? It seems unlikely that it
means existence; Plato is not saying that the sensible world does
not exist, that it is an illusion. It seems plausible, rather, that Plato
is talking of the predicative sense of “being”: the sense in which it
is used in “being large”, “being beautiful” and so on. Indeed, this is
borne out by the Republic, which spells out the claim that sensible
things are “between being and not being” by saying that the many
beautiful things will in a way be ugly, the many just things in a way
unjust, and so on. One may think, then, that Plato is using “being” as
shorthand for “being so and so”; to say that only Forms have being
in the fullest sense is to say that only they possess properties in an
unqualified way.
Yet this does not seem to do justice to everything Plato says about
the being or reality of the Forms. For instance, in the Phaedo Socrates
warns us not to see sensible things as “being”, although the pleasure
and pain they produce may force us to do so (Phd. 83b ff.). This can
hardly mean that we should not think they exist; how, if they did not
exist, could they produce pleasure and pain? But nor does it plausibly
mean that we should not think they have properties, such as size,
equality and so on, in an unqualified way. Why should pleasure and
pain lead us to think they have being in this sense?
There are a number of things that Plato may have in mind when
he speaks of the Forms as “being” in a way in which other things
are not, and I think it would be wrong to choose one of them to
the exclusion of others. First, as Vlastos (1965) has pointed out, the
words for “being” (on, ousia) can also be translated as “real”, or “real-
ity”. “Reality” can be used in a way equivalent to “existence”, as when
we contrast what is real with what is imaginary, but it can also be
used in another sense, of a genuine example of some kind of thing,
47
plato
as when we contrast real with imitation pearls, or like the Velveteen

Rabbit in the story by Margery Williams, who wanted to be real, that
is to be a real, living animal. “Real” can also sometimes have evalu-
ative implications; we speak, for instance, of a car that we admire as
“a real car” in a way that suggests that a defective example of a kind
is not a real example.
If we ascribe reality in this sense to Forms, it is indeed a kind of
predicative being, but not any old predicative being, but one in con-
nection with an especially important predicate: the one by reference
to which something is identified as the kind of thing it is. In saying
that something is real, we do not mean just that it has some property
without qualification, but that it is, genuinely, the sort of thing that it
claims to be or by reference to which it is defined. So, if we are look-
ing for real beauty, for instance, we should look at the Form rather
than at sensible beautiful things; this may be important both from a
cognitive point of view, if we want to know what beauty is, and from
an evaluative one, if we see beauty as something valuable that we aim
to possess. In this case sensible beautiful things will indeed lack real
beauty because their beauty is qualified in some way. But the mistake
Plato is warning against is not simply thinking that their beauty is
unqualified, but rather thinking that they are the most genuine, most
fundamental, examples of beauty.
Secondly, Plato sometimes contrasts being with becoming (e.g.
Resp. 509b); he may think, therefore, that things in the world lack
being because they are involved in constant change, and so are always
becoming something new. This can be seen as contrasting with the
constant being of Forms: “being” can here be taken both in an exist-
ential sense, since Forms exist forever, and in a predicative sense,
since they always retain the same properties.
Finally, there is a possible sense of “real” in which the most basic
constituents of the world are real, and the way they manifest them-
selves is less real (even though it has objective existence, and is not
just an illusion). As we have seen, Plato sometimes suggests that
sensible things depend on Forms, along with the receptacle, for their
existence;6 Forms may therefore be seen as “more real” in the sense
that they are ontologically more fundamental.
48
The way in which Plato emphasizes the “being” of Forms and con-
trasts it with the “becoming” of sensible things should not be exag-
gerated. He often does speak of sensible things as “being” so and so.
And while in places “being” is used of something permanent, con-
trasting with what is constantly becoming something new, in other
places the relation between being and becoming is understood in
the natural way, with becoming interpreted as coming into being.
However, Plato does have a distinctive way of using the concepts of
being and becoming in some special contexts, and this is what we
have been concerned with here.
Why believe in Forms?
It is surprising that – although Plato clearly knew that his belief in

Forms was controversial – there is very little explicit argument for
it in his works. In fact, in only one place, in the Timaeus (51b–d),
does he present a direct argument for the existence of Forms, and
the argument there is rather cryptic, and clearly presupposes some
knowledge of Platonic philosophy. Timaeus claims that unless Forms
exist as something distinct from material things, there is no dif-
ference between knowledge and true belief. He goes on to provide
reasons for thinking that there is a distinction between knowledge
and true belief, but the crucial premise, that there cannot be such
a distinction unless there are Forms, is left unsupported. Clearly,
someone who has never heard of Forms will have no reason to accept
this premise; rather, it is meant to appeal to someone who already
has some understanding of Forms and the role they play in Plato’s
system. The argument does not provide a reason to introduce a new
theory; it is intended to defend an existing theory, by reminding us
of the function, that of a secure ground for knowledge, that Forms
have within that theory, and suggesting that nothing else can fulfil
this function.
There are two other places in Plato’s work where he puts forward
considerations that could be used as part of an argument for Forms,
but in neither passage does that actually seem to be his main aim.
49
plato
One is in the Phaedo (74b ff.), where Socrates is arguing for the pos-
ition known as the doctrine of recollection: that we had knowledge of
Forms in a disembodied state before our birth, and in our present life
are sometimes reminded of them; for instance, when we perceive two
equal sticks this reminds us of the Form of equality. In the course of
this argument, Socrates puts forward an argument that supports the
claim that the Form of equality is different from the equal sticks that
remind us of it. But his main aim does not seem to be to argue for
this difference; in fact his respondent, Simmias, has already agreed
this at the beginning of the discussion. Rather, Socrates’ aim is to
show that we do not get our knowledge of equality from perceived
equal things such as sticks, but must already have some knowledge
of it if we are to recognize them as equal.
The other passage is from the Republic (476e ff.), where Socrates
argues that only things with the distinctive characteristics of Forms
can be objects of knowledge, because only they possess properties
without qualification; sensible things possess their distinctive prop-
erties – largeness, beauty and so on – in a qualified way. It is easy to
see how this could be used as part of an argument for the existence
of Forms. If we add the premise that knowledge does in fact exist,
we can conclude that, as only Forms can be objects of knowledge,
Forms must exist. In fact, this argument could be combined with
that from the Timaeus – which does give reasons for thinking that
knowledge exists as something distinct from true belief – to pro-
duce a stronger argument for the theory of Forms. But within the
Republic this is not Plato’s aim; rather, he is trying to show that as
only Forms are true objects of knowledge, it is only those who pur-
sue knowledge of them who are rightly called lovers of knowledge,
and so entitled to the name “philosophers”. He is seeking to rule
out the suggestion that dilettantes, lovers of sights and sounds, have
a right to that name; the real philosopher is someone who knows
something of significance.
We know from Aristotle (Metaph. A 990b9 ff.) that arguments for
belief in Forms were used in Plato’s school, the Academy. Clearly he
did think that argument in favour of the theory was possible, and
sometimes useful. But it is in fact difficult to find arguments that
50
could, by themselves, provide a motive for introducing the theory

of Forms.
Some arguments seem too broad, and it is not clear why they
should motivate the introduction of Forms rather than of some other
entity. For instance, it is sometimes suggested that Plato introduced
the notion of Forms because he thought that the objects of know-
ledge should be unchanging, and that, as things in the material world
are clearly in change, there must be something else that plays this
role. In a way, this is certainly right: Plato did think that objects of
knowledge must be unchanging, and that Forms were able to play
this role. And this is indeed an advantage for the theory of Forms
over rival views on which everything changes. But it is not clear why
this line of thought should have led Plato to posit Forms specifically,
rather than some other kind of unchanging reality, for instance God,
or atoms, the ultimate constituents of the physical world.
Other arguments seem to motivate the introduction only of a
particular class of Forms, not of the wide range of Forms in which
Plato seems to have believed. For instance, one might argue that we
cannot gain a clear conception of some properties, such as large and
small, just by looking at examples of them, since what is large in com-
parison with one thing will be small in comparison with another;
hence, what largeness is must be something distinct from visible
large things, something not perceived through the senses. But this
argument will not extend in any straightforward way to other prop-
erties, such as being a human being; when we look at a human being,
it seems, we can see that she is human, and nothing in our immediate
experience shows us a respect in which she is not human. Indeed,
Plato himself has Socrates make a point along these lines in Republic
Book 7 (523a ff.): some properties, such as largeness and unity, are
useful for reducing our reliance on the senses; others, such as being
a finger, are not, since when we look at a finger our senses do not
show us any aspect under which it is not a finger. Yet Socrates does
not say that there are no Forms for properties such as being a finger,
or that the senses do give a completely adequate account of those
properties, but only that our immediate experience does not give us
a reason to think they do not. And elsewhere, as we have seen, he
51
plato
does introduce Forms for natural kinds, such as human being, and
artificial kinds, such as bed, which do not raise the same problems
as properties such as largeness. If Plato first introduced the theory
of Forms in response to these problems, as some think, why would
he extend it to other properties that do not raise them?
I suggest that these puzzles may be solved if we suppose that Plato
did not introduce Forms in response to an argument; rather, he had
a basic intuitive conviction, perhaps inherited from Socrates, that
something like Forms existed. When he began to work out his view
of Forms in detail and to apply it, he found that it could help to solve
various problems, and this gave it an advantage over rival views, and
so was an additional reason for accepting it. This explains why argu-
ing for Forms is not a priority with him: arguments for Forms arise
incidentally in the discussion of other issues. It also explains why
Plato feels able to use arguments that are limited in the ways I have
discussed. He is not introducing a theory in response to problems;
rather, he is claiming it as an advantage for an existing theory that
it can solve certain problems. Thus it can be seen as an advantage of
Forms that they perform a certain role, even if something else could
perform that role, and it will also be an advantage of the theory if
one group of Forms helps to solve a particular problem, even if other
Forms are not relevant to it. The main reason for adopting the theory
lies not in any one specific argument, but in its coherence and its
general usefulness.
There is a clear shift between the Socratic and Platonic dialogues
in their attitude to Forms, but this can be seen as a change not so
much of doctrine as of perspective. Plato is not introducing new
entities; as we have seen, Socrates does believe in universal essences
such as virtue itself, holiness itself and so on, and even occasionally
calls them Forms. Nor is Plato exactly ascribing new characteristics
to Forms; it is quite likely that, if Socrates had been made to reflect
on the Forms in which he believed, he would have agreed with many
of Plato’s distinctive views about them. It is quite reasonable to say,
for instance, in the light of the way he discusses virtue, that for him
the nature of virtue does not change, and that we know it through
reasoning rather than through the senses. Even the aspect of Forms
52
that is often seen as most distinctive of Plato, separation, can be

seen as implicit in Socrates’ thought; he might well have agreed that
virtue itself does not depend for its existence on particular virtuous
people, especially as he was doubtful about whether there were any
really virtuous people.
But in the Socratic dialogues these aspects of Forms are never
emphasized. Socrates’ concern is not with the general claim that
Forms exist, and their status, but always with the particular Form
that he is investigating on each occasion, and its ethical implications.
Plato, by contrast, reflects on Forms as such, and the way they con-
trast with sensible things, and begins to think systematically about
them. For him it is important, for instance, that they are unchanging,
that they are objects of knowledge and that they have no imperfec-
tion about them: all respects in which they contrast with the material
world.
Problems for Forms
In the first part of the Parmenides, as we have seen, Plato shows Soc-
rates, as a young man, proposing a version of the theory of Forms; he
then faces criticisms of the theory from Parmenides. These criticisms
are not answered within the dialogue, although it does not follow
that Plato thought them unanswerable. Here I shall discuss two of
the problems that Parmenides raises.
The first of these is commonly called the “third man” argument
(Prm. 132a–b). It seems to have been a well-known argument in
philosophical circles at the time; Aristotle refers to it in his Metaphys-
ics (A 990b15), and Alexander in his commentary on the Metaphysics
makes it clear that this was the same argument which we know from
the Parmenides (Alex. Aphr. in Metaph. 83.34 ff.). It is to them that
we owe the name “third man”, since they discuss a version of the
argument that uses as an example the Form of man. The Parmenides,
however, uses instead the Form of large.
Parmenides begins by getting Socrates’ agreement that although
there are many things that we call by one name, “large”, there is one
53
plato
thing, the Form, in virtue of which they are all large. He then argues
that if we consider the Form and the other large things together,
there must be another Form in virtue of which they are all large. (So
if an ordinary large thing is the “first large”, and the original Form
the “second large”, this will be the “third large”.) However, there is
no reason to stop there; if we look at this “third large” together with
the original Form and the other large things, there must be a further
Form, a “fourth large”, in virtue of which all these are large. And so
ad infinitum: an infinite series of Forms is produced.
This result is obviously disturbing, not only because it produces an
infinite number of Forms for each predicate, when Socrates’ original
proposal is that there is one Form in virtue of which, for instance, all
large things are large, but also because the intention is to explain why
large things are large by their relation to the Form. If we then have
to posit another Form to explain why that Form is large, yet another
Form to explain why that is large, and so on, we have an infinite
regress of explanation, and it is often felt that this is not genuinely
explanatory: it leaves the fact with which we started basically unex-
plained. So, if the theory of Forms leads to this conclusion, it may
seem best to abandon the theory.
However, are we forced, just by believing the basic principles of
the theory of Forms, to accept Parmenides’ argument? In fact, it
looks as if his argument rests on some assumptions that could be
questioned. One we are already familiar with is the self-predication
assumption: that the Form of large is itself something large, and
likewise for other Forms. Parmenides says that we should consider
the Form of large and other large things together, and then we will
discover another Form in virtue of which they are large. If the Form
in virtue of which things are large is not itself large no problem arises;
there is no need to introduce anything else to explain its largeness.
However, this is not the only assumption we need to make the
argument work. Why should we not say that the Form of large is large
in virtue of itself, although other large things are large by relation
to it? To block this move the argument needs some further assump-
tion. The assumption that fulfils this function is normally called a
non-identity assumption. There has been some dispute about just
54
how the non-identity assumption is best formulated, but the sim-

plest formulation is one proposed by Gail Fine (1993: 206): nothing
is large in virtue of itself (and likewise for other predicates, at least
those that have Forms corresponding to them). Parmenides seems to
be making some such assumption as this when he claims that there
must be another Form in virtue of which the original Form and its
instances are all large.
These assumptions are not obviously unreasonable; it is plausi-
ble that someone who believes in Forms might accept them. As we
have seen, Plato’s language often suggests that he believes in self-
predication, and it is essential to the theory that, at least in ordinary
cases, something like non-identity is true: ordinary large things are
not large in virtue of themselves, but through a relation to the Form.
However, nor is it obvious that either assumption should be accepted.
We have seen that some have interpreted Plato in such a way that he
is not committed to self-predication, while there may also be reasons
to reject non-identity when this is applied to the Form itself. Fine
(1993: 226–8) has argued that ordinary large things cannot be large
in virtue of themselves precisely because their largeness is imperfect
or qualified; the same need not be true of the Form, which is large
in an unqualified way.
It seems possible to give up one or other of the assumptions. If we
maintain both we face a damaging regress, but either on its own does
not pose a problem (or at least this particular problem). Another
possibility, however, is to see some kind of ambiguity in the terms
in which the problem is stated. If this is right, both assumptions
might be true in a way, but on different interpretations of the terms.
One way of achieving this result has been suggested by Constance
Meinwald (1992). It turns on the distinction between two kinds of
predication mentioned above (§ “Self-predication”). On ordinary
predication, where an object possesses a property, self-predication
will be false (the Form of largeness does not have the property of
being large) but non-identity will be true (each thing that has the
property of being large has it in virtue of something else, the Form).
On the special kind of predication where an object has a property as
the whole or part of its nature, self-predication will be true (the Form
55
plato
of largeness is large in that largeness is its nature) but non-identity

will be false (something that has largeness as its nature need not have
it in virtue of something else).
Another possible solution, which also turns on ambiguity, can be
found if we compare Forms with standard weights and measures.7
The Form of large can be seen as a standard of largeness in the same
way that something can be a standard for a particular weight or
measure, for example the (former) standard metre as a standard for
the property of being a metre long; other things are large by resem-
bling the Form. This raises a problem over how we are to use the
word “large”. On the one hand, we might well suppose that if other
large things resemble the Form, then it has the same property that
they have, so it is large; in this case self-predication will be true, non-
identity false. On the other hand, in saying things are large we are
comparing them with the Form, and the Form cannot be compared
with itself; it does not literally resemble itself, so it seems wrong to
call it large. Both views seem to have some intuitive power, but we
cannot maintain both of them at once without paradox. However,
which one we accept – how we actually use the word “large” – may
in the end be just a matter of choice. The problem seems to be one
about how the word is to be used rather than about what Forms are
really like.
Interestingly, a parallel puzzle exists in modern philosophy about
standard measures, such as the standard metre that at one time
existed in Paris. The metre was defined in such a way that other
things were a metre long if their length was the same as that of the
standard. Some philosophers, such as Ludwig Wittgenstein (1953:
50), have argued that the standard should not itself be described as
a metre long, since in saying things are a metre long we are compar-
ing them with the standard, and it cannot be compared with itself.
Others, such as Saul Kripke (1981: 54), hold that, as the standard
has the same length as other things that are called a metre long, it
must be right to say that it, too, is a metre long. One thing that we
can be certain of, though, is that the standard existed; the fact that
there was a puzzle about how to speak of it does not mean that we
need to deny its existence, or in general to think that weights and
56
measures cannot be defined by reference to standards. Something

similar may be true of Forms.
Before leaving Parmenides’ arguments, we should note that while
the third man argument is the one that has stimulated most discus-
sion, it is not the one that Parmenides himself thinks is the greatest
difficulty for Forms. He gives this title to an argument that seeks to
show that knowledge of Forms is impossible (Prm. 133b ff.).
This argument turns on the claim that Forms are related, in the
distinctive way that their names imply, not to ordinary things, but
to other Forms; conversely, ordinary things are related to other
ordinary things, and not to Forms. Thus, for instance, human slaves
are slaves of human masters, not of the Form of master, and human
masters are masters of human slaves, not of the Form of slave; the
Forms of master and slave, however, are what they are in relation
to one another. If this reasoning is extended to knowledge, Parme-
nides claims, it will follow that only the Form, knowledge, can be
related to truth itself, and the Forms of specific knowledges (i.e.
branches of knowledge) to specific realities, that is, Forms. Our
knowledge, on the other hand, will be related to ordinary realities,
not to Forms.
There is no space here to explore this argument in detail. A full
treatment would have to show not only how it might be answered,
but also why Parmenides sees it as so intractable. It is certainly the
case that Forms do not typically stand to us in the relation that their
names indicate: the Form of master is not our master, the Form of
statesman not our ruler, the Form of bed not something we could lie
on, and so on. It does not obviously follow from this that we cannot
stand in any relation with Forms. Knowledge may be an exception;
so, as we have seen, may love. Perhaps the problem is one of giving
a principled criterion to distinguish the relations that we can stand
in with Forms from those that we cannot. One may also suspect
that the puzzle arises from an excessively literal application of the
thought that the Forms constitute a distinct world. One might think
that if they do not stand to us in relations like master this is because
they are not part of our world but are located elsewhere; but if this
is true it is hard to see how we could have any access to them. But
57
plato
as I have suggested, this may not in any case be the best way to look
at the realm of Forms.
Plato’s response to these problems
What was Plato’s own response to these problems? Some have

thought that they were so devastating that they led him to abandon
the theory of Forms altogether. But the language of the passage does
not suggest that. As we have seen, Parmenides calls the problem of
the unknowability of Forms “the greatest difficulty”, yet he insists
that it can be answered, although only an exceptionally gifted person
would be able to follow the answer (Prm. 133b, 135a–b). Later, he
says that if we simply refuse to accept that Forms exist, we make all
discussion impossible (135b–c). In addition, dialogues that most
scholars see as later than the Parmenides – most notably the Timaeus
– include discussions of Forms, and not just as common essences
that things share, but as paradigms that they resemble, while neces-
sarily falling short of them (e.g. Ti. 37d): one of the clearest state-
ments of the classic conception of Forms.
Did Plato, then, have definite answers to the problems in mind?
No such answers are stated in the Parmenides, but perhaps we are
being challenged to work them out for ourselves. One possible view
is that we are meant to see Socrates as holding an inadequate ver-
sion of the theory of Forms (perhaps implying, also, that Plato in
earlier works had expressed an inadequate version of the theory).
Parmenides’ arguments expose difficulties, and we have to revise the
theory to overcome them, although it will remain the same theory
in essentials. For instance, we can see the third man argument as
depending on the assumptions of self-predication (understood in a
fairly literal way) and non-identity. Perhaps the moral is that one of
these assumptions must be given up.
Alternatively, we may see the problem as lying not in the actual
theory Socrates holds, but in the way he expresses it. Unclarities in
the way he expounds the theory give rise to misunderstandings,
which allow apparent problems to be raised. In this case the moral
58
is simply that we have to clarify the way we speak about Forms in

order to show that these problems do not really arise.
However, there is another possibility: that Plato did not have any
specific solution to the difficulties in mind, but nevertheless believed
that one was possible, and set out the difficulties as a stimulus to
future research. One might think that it would be wrong to cling to
a theory in the face of such problems if one could not oneself see
any answer to them. This might be right in the case of a scientific
theory, put forward as a solution to a problem; if it raises difficulties
that cannot be overcome, that is a reason for thinking it is not a good
solution. However, I have suggested that Plato’s belief in Forms was
not like that. Rather, he had a deep intuitive conviction that such
things as unity itself, justice itself, beauty itself, and so on exist. It
might be reasonable to cling to such a conviction even in the face of
problems that one cannot immediately solve. The philosopher Zeno
of Elea, a friend of Parmenides, who makes a brief appearance at the
beginning of the dialogue, was famous for putting forward some
notorious paradoxical arguments that seem to show that there is no
such thing as plurality or motion (the most famous of his arguments
is probably that which turns on the paradox of Achilles and the tor-
toise). Some of these arguments have proved very hard to answer,
but clearly the right response is not to stop believing in plurality or
motion – we know, by experience, that these things exist – but to
go on believing in them while seeking a solution to the paradoxes.
Plato may have seen it as right, in the same way, to go on believing
in Forms, and to look for a solution to the paradoxes that they raise.
59
four
Knowledge
Our discussion of the Forms has already shown that knowledge plays
a central role in Plato’s thought. It is important to him that they are
suitable objects for knowledge, and he sees knowledge of them as
something to aspire to. In this chapter we shall look more closely at
this aspect of his philosophy. I shall begin with a brief look at the one
work, the Theaetetus, where Plato confronts the topic of knowledge,
as it were, head on, asking what knowledge is. After this I shall look
at a number of other aspects of Plato’s view of knowledge, and in
particular of the relation between knowledge and Forms.
The Theaetetus: Plato on knowledge
Although the Theaetetus is generally thought to have been written

relatively late in Plato’s career, in many ways it seeks to recapture
the style and method of the “Socratic” dialogues. It presents Soc-
rates, as he appeared in many of those dialogues and as he may
have been in real life, examining a number of proposals rather than
stating a worked-out view, and reaching no definite conclusion.
Near the beginning of the dialogue Socrates introduces the famous
image (which may go back to the historical Socrates) of himself as a
midwife, trying to help his interlocutors to give birth to ideas (Tht.
61
plato
150b ff.). In the course of the work he enables the young mathemati-
cian Theaetetus to come up with three proposals for the definition of
knowledge, but in the end all are rejected. However, the discussion
of these proposals is used as a peg on which to hang various further
discussions, which go some way beyond the dialogue’s official aim of
defining knowledge. Among the topics discussed are: the Heraclitean
theory that the world is in radical flux; the relativism of the Sophist
Protagoras, according to whom “a human being is the measure of
all things” (Tht. 152a) and whatever seems true to me is true for me;
the possibility of false belief; and the relation between elements and
composite wholes.
The Theaetetus is one of Plato’s most fascinating works, but it
is valuable as an example of philosophy in action rather than as a
source of philosophical teaching. It is far from clear what moral, if
any, we are meant to take away from the reading of it. What is certain,
though, is that it does not leave us with a definition of knowledge.
In what follows I shall look briefly at the three proposed definitions
and why they are rejected.
Theaetetus’ first proposal is that knowledge is perception (Tht.
151e). This is rejected on the grounds that while our awareness of
some properties such as colour, heat, musical tone and so on comes
through the senses, there are others such as sameness and difference,
likeness and unlikeness, number and, especially, being that we grasp
not with the senses but with the intellect. It is therefore not percep-
tion, but belief, that enables us to grasp these things, and Socrates
says that if we do not grasp being we do not grasp truth, and so do
not attain knowledge (184b–6e). It is not entirely clear what Socrates
means by this but, at any rate, what he says does seem to give us a
reason for rejecting the definition. It would seem that, at the very
least, properties such as being can be objects of knowledge – that we
can know that something exists, that two things are different and so
on – and if these properties are not grasped by the senses this is a
reason not to identify knowledge with perception.
The second proposal discussed, following on from this, is that
knowledge is true belief (187b). This is rejected on the grounds that
it is possible to persuade people, rhetorically, of something that
62
knowledge
happens to be true, although in such a case they will not have know-
ledge; an orator, for instance, might persuade a jury that a certain per-
son committed a crime, when he did indeed do so, but the members
of the jury do not know this, as is shown by the fact that the orator
could equally have persuaded them of the opposite had he so chosen
(201a–c). It is somewhat obscure what Socrates thinks would be suf-
ficient to produce knowledge in such a case. At one point he suggests
it might be produced by “teaching”, that is, presumably by going over
the evidence in a rational way; the problem is that court proceedings
do not allow enough time for this. Later, however, he seems to say that
only an eyewitness could know who committed a crime. But in any
case, it seems right to say that rhetorical persuasion can produce true
belief without knowledge, so Socrates is right to reject this proposal.
Theaetetus’ third proposal (201d) is that knowledge is true belief
with an account (logos). At first sight this is much more plausible
than the first two. It seems to resemble the view, much discussed
in recent philosophy, that knowledge is justified true belief, since
a justification for a belief might well be called an account of it. But
it emerges that this is not exactly what Theaetetus has in mind. He
means, rather, that to know a thing is to have true belief about the
thing along with an account of the thing. Socrates then brings a
number of criticisms against this theory, based on different inter-
pretations of the term “account”. One of his criticisms, however,
although directed against a particular reading of “account” – that
it means a distinguishing mark that enables us to pick something
out from other objects – is of wider relevance, and poses a problem
for many theories of knowledge. It turns on the question whether,
when we say that true belief must be accompanied by an account, we
mean true belief about the account or knowledge of the account. If
the first, then it seems that the requirement that we have an account
adds nothing to what is already involved in true belief; to think about
something at all we must have a way of picking it out. If the second,
our definition becomes circular; it uses knowledge to define know-
ledge (209d–10b).
This problem can be generalized: it is an objection to any theory of
knowledge that sees it as true belief together with some further piece
63
plato
of information. If we require only true belief about this new piece of

information, it is not clear why this should by itself turn our belief
into knowledge. If one piece of true belief does not constitute know-
ledge, why should two pieces of true belief do so? (Both, after all,
might have been produced by rhetorical persuasion.) But if know-
ledge of the new information is required, the definition becomes
circular. This is not to rule out the possibility that acquiring a new
piece of information might in fact enable us to turn our true beliefs
into knowledge. But if it does so, it is because it brings about some
change in our mental state, not because its presence is, by defini-
tion, enough to constitute knowledge. Thus in what is probably a
much earlier dialogue, the Meno, Socrates claims that true belief can
by turned into knowledge by working out the explanation (Meno
98a). But this need not mean that knowledge is simply true belief
together with possession of an explanation, for this would raise the
same problem: true belief about an explanation, or knowledge of it?
Working out the explanation brings about the transformation into
knowledge, but this does not in itself say exactly what the transfor-
mation involves. Plato never, in the Theaetetus or elsewhere, achieves
a satisfactory account of what knowledge is.
Knowledge and Forms
We now turn to look at other aspects of Plato’s theory of knowledge,

concerned not so much with what knowledge is but with what it is
knowledge of and how it is gained, and in particular with the con-
nection between knowledge and Forms. There are three aspects of
the theory of knowledge that especially deserve discussion. First, I
shall look at general questions about the knowledge of Forms, what is
involved in knowing Forms, and in particular the question whether
Forms are the only thing we can know. Secondly, I shall consider the
theory of recollection, the proposal that we gain knowledge through
remembering something we have learned, outside the body, before
our birth into the present life. Finally, I shall look at Plato’s view of
the structure of knowledge and of philosophical enquiry, and of the
64
knowledge
special part played in it by the Form of the good; Plato’s account of

these issues is given in the central part of the Republic, by means of
the three famous analogies of the sun, the line and the cave.
Forms and definition
What is it to know a Form? Essentially, I suggest, it is to know what

some kind of thing is: to know the Form of good is to know what
goodness is; to know the Form of large is to know what largeness is;
and so on. It is, therefore, the sort of knowledge that would naturally
be expressed in a definition. This is not surprising, when we remem-
ber that the theory of Forms arose out of Socrates’ quest for defini-
tions. Many (although by no means all) of the Socratic dialogues are
concerned with the search for definitions, and in two of them, the
Meno (72c) and Euthyphro (6d–e), Socrates calls what he is looking
for a Form. That quest is not abandoned in those dialogues where
the theory of Forms is present in a more developed state, although,
because Plato thinks this knowledge is hard to attain, he rarely gives
actual examples of definition. However, one of the central aims of the
Republic is to find a definition of justice. The definition actually given
is provisional, but it is suggested that an investigation of Forms, link-
ing it with the Form of the good, is what we need to achieve a more
certain result (Resp. 504b–5a). This interest in definition remains
in Plato’s later works: the Theaetetus, as we have seen, is a quest,
although an unsuccessful one, for a definition of knowledge, and
Plato searches, apparently successfully, for definitions of the sophist
and the statesman in the dialogues of those names. (A passage at
the beginning of the Sophist [217a ff.] also suggests that Plato at one
point planned a third dialogue in the group, which would have been
concerned with the definition of the philosopher.)
In many places Plato uses the language of perception in describ-
ing our knowledge of Forms – he speaks of seeing or looking at the
Forms, or of touching them – and this has sometimes been taken to
mean that the knowledge he is looking for is not propositional. But
Plato’s language of perception can be seen as a metaphor, like “seeing
65
plato
the point”, “grasping the truth” and so on. Knowledge of Forms is a

kind of insight, and is clearly presented as a striking experience of a
kind that can be life-transforming, but this may still be knowledge
that can in principle be stated. A Form is a really existing object, and
so when we know a Form there is some object that we know, but that
does not mean that knowledge of Forms is acquaintance as opposed
to propositional knowledge; it can be seen, rather, as knowledge of
what they are.
The combination of Forms
However, definitions are not the only thing Plato thinks we should
be looking for in our investigation of Forms. In later works, espe-
cially the Sophist, he emphasizes the importance of relations between
Forms. The chief speaker of the Sophist, the Eleatic Stranger, draws
attention to the way in which Forms can “combine” or “blend” with
one another, and, in describing dialectic, the distinctive activity of
philosophers, claims that they should investigate which Forms can
combine and which cannot (Soph. 253b ff.).
Just what is meant by “combination” is somewhat unclear; it may
be that Plato is using the term in a rather broad sense, so that it
covers what we would think of as several different relations. He cer-
tainly seems to think that two Forms combine when one of them is
predicated of the other. Thus, for instance, the Forms of motion and
rest both exist or are something. This means that the Form of being
is predicated of both of them, or that they have being as a property;
and this is seen as a combination between these Forms and being
(254d). On the other hand, the Stranger also claims that all discourse
depends on the combination of Forms (259e), yet it is clearly not the
case that every assertion predicates one Form of another; “Socrates
is wise”, for instance, does not. It may be that Plato sees combination
as including not only cases where one Form is predicated of another,
but also cases where two Forms are predicated of the same object.
“Socrates is wise” could than be taken to imply the combination of,
for instance, the Forms of human and wise. Thus, in investigating
66
knowledge
the combination of Forms, we would be enquiring into which Forms

are compatible, in that they could belong to the same objects, and
which are not.
The idea of combination of Forms is not a wholly new one in the
later works. There is in fact a reference to Forms combining in the
Republic (476a). In the Parmenides, indeed, there is a passage where
the young Socrates is sometimes read as saying that different Forms
do not combine (Prm. 129d–e), but in fact he need only be taken to
mean that opposite Forms, such as likeness and unlikeness, do not
combine. But while the basic idea is not wholly new, the focus on it
certainly is. Plato’s philosophy is taking a rather different direction
here from that found in the classic works written earlier in his career.
In the Sophist, the Stranger undertakes an investigation of the
relations between five very great or very important Forms – being,
sameness, difference, motion (or change) and rest (or stability) –
enquiring which of these can, and which cannot, be predicated of
one another, but also emphasizing that they are all distinct from one
another (Soph. 254c–7a). (These are sometimes referred to as the
greatest Forms, but the Stranger does not say that they are the great-
est; others, such as unity, may be equally important.) The method
demonstrated here could no doubt be applied to other Forms as
well, including natural kinds and evaluative Forms, for we have no
reason to think that Plato has abandoned interest in these. Yet it is
significant that Forms of extreme generality are here singled out
as among the most important. It may be that Plato thinks that an
understanding of them is necessary to any investigation of relations
between Forms.1
In this passage the most notable features ascribed to Forms in ear-
lier dialogues – their eternity and changelessness, their being grasped
by reason and not through the senses and so on – are not mentioned
(although changelessness is discussed in an earlier passage of the
Sophist [248a ff.]), and the tone is for the most part drily logical,
without the mystical elements found in many earlier accounts of
Forms. This may lead us to doubt whether the Forms discussed here
are the same as those found in earlier works. Yet even in the Sophist
the Stranger at one point shows an attitude of reverence for the Forms
67
plato
he is discussing. Referring specifically to the Form of being, he calls it

divine, and says that it is because of the brilliance of this Form, which
blinds us when we try to look at it, that it is hard to discern the nature
of the philosopher, who is concerned with being (Soph. 254a–b).
In fact it is possible to see the contrast between the Sophist and
earlier works as resulting, not from a change of doctrine, but from a
difference in purpose. The theory of Forms, as expounded in works
such as the Phaedo, Symposium, Phaedrus, Republic and Timaeus,
is primarily a theory about the relation between Forms and things:
sometimes the contrast between Forms and things, sometimes the
way in which Forms underlie things. It is also a theory about our
relation with Forms: how we come to know them, and how we can
move towards contemplation of them. For this reason it focuses on
the realm of Forms as a whole, and on the characteristics that they
share. But in the Republic Socrates describes the true method of dia-
lectic as “proceeding from Forms to Forms, and ending with Forms”
(Resp. 511b–c); it is not concerned with the relation of Forms to us
or to the world, but with their own nature and their relations with
one another. Someone doing dialectic in this way would focus not on
the ideal qualities of the Forms, which set them apart from sensible
things, but on the individual nature of each Form and what makes
it what it is. It does not follow that these Forms should not be seen
as having those same ideal qualities. The discussion in the Sophist
of these very important Forms and their relations may in fact be the
clearest example in Plato of the method of dialectic described in the
Republic, but not actually practised there. Much of what Plato writes
is practical in aim, directed either towards the guiding of our lives in
general or towards moving us to do philosophy, but in this passage
of the Sophist it is possible to see him doing philosophy itself, from
a purely theoretical perspective.
Collection and division
One aspect of the relations between Forms in which Plato takes a

special interest is the way they may be related as genus and species.
68
knowledge
Sometimes, a generic Form may have a number of more specific

Forms falling under it; for instance, the Forms of land animal, water
animal and so on may be seen as falling under the Form of living
creature (see Ti. 30c, 39e ff.); Forms such as justice and temperance
may be seen as falling under the Form of virtue. It is worth remem-
bering here that the most common Greek word for form, eidos, can
also be translated “kind”. This need not mean that Plato is thinking
of Forms as kinds in the sense of classes of objects, but he does
sometimes think of each Form as defining a kind of thing and set-
ting it apart from other kinds of thing. The more specific Forms are
sometimes referred to as parts of the more general one. This language
should probably not be taken too literally, since it is hard to fit it
together in detail with the idea of Forms as ideal patterns (although
Plato does sometimes use both kinds of language together, notably
in the Timaeus). The point is rather that the “parts” can be seen as
more specific versions of the “whole”: the Form of living creature
determines, in a rather broad way, the requirements that something
must satisfy to be a living creature; the Form of land animal deter-
mines a specific way of satisfying these requirements. There may
in turn be more specific versions of the Form of land animal. By
investigating them we can draw a classificatory tree, going down to
the individual species. Investigation of these relations may move in
either direction: it may start with a number of species and look for
the genus within which they fall, or it may start with a genus and
seek to articulate its various species.
Once again this theme goes back to an early stage in Plato’s
thought. The idea that Forms can stand in part–whole relations is
found as early as the Euthyphro, where holiness is called a Form,
but is also described as a part of justice (Euthphr. 6d–e, 12d). How-
ever, the Phaedrus is probably the first work to focus explicitly on
the investigation of these relations. There Socrates, having made a
speech in praise of love, begins to reflect more systematically on
what he did in that speech, and points out that he did two things
in particular (Phdr. 265d ff.). One was to collect or bring together a
number of seemingly disparate phenomena under one name, love,
and give a definition of love that fits the class as a whole. The second
69
plato
was to make a distinction or division within the general field of

madness. Love is seen as a kind of madness, but it is important to
distinguish between a bad kind of madness and a good, or divinely
inspired, kind. The divine kind is then subdivided into four varieties,
found in mystical initiation, prophecy, poetry and love. Socrates here
introduces the famous metaphor of “dividing at the joints”: there is a
natural way in which to classify things, and we should not split them
up in an arbitrary way.
Both these operations, collection and division, are connected
with the part–whole relation. The first begins with a multitude of
phenomena and seeks the whole of which they are parts; the sec-
ond begins with a whole and distinguishes the parts within it. Nei-
ther operation is presented as wholly new, for Socrates says that
he has always seen them as typical of dialectic; what is new is the
possibility of applying them to rhetoric. Indeed, the first operation,
collection, can be seen as a development of the practice of seeing
the one in the many, and searching for what all examples of some-
thing, for example virtue, have in common, which is typical of the
Socratic dialogues. As for division, in those same dialogues Soc-
rates sometimes shows an interest in classification and in making
distinctions, notably in the Gorgias, where he classifies a number
of arts or skills, and also a number of knacks, based on experience
rather than knowledge, which he contrasts with these skills (Grg.
464b ff.). This aspect of Socratic dialectic may go back to the histor-
ical Socrates, since Xenophon also ascribes to him an interest in
“making distinctions according to forms (or kinds)” (Xen. Mem.
IV.5.12). It exploits the way in which the name of the art can mean
not only “art of conversation” but also “art of making distinctions”.
But it is the Phaedrus that makes explicit this double tendency of
dialectic; it both seeks unity, discerning what many things have
in common, and draws distinctions, recognizing many varieties
within an encompassing whole. It is the Sophist, however, within
Plato’s work, that explicitly links these operations with Forms; the
Stranger says that it is characteristic of dialectic to “divide accord-
ing to Forms”, and this is linked with enquiry into the combination
of Forms (Soph. 253d).
70
knowledge
The investigation of the genus–species relation between Forms

might serve several different purposes. One is simply to give a map
of some area of enquiry, setting out all the species that fall within
some important genus, thus improving our understanding of the
whole area and of the particular species within it. This aim is par-
ticularly prominent in the Philebus, another late dialogue connected
in a number of ways with the Sophist and Statesman. The passage
there in which Socrates introduces this method is extremely cryptic
(Phlb. 16c ff.), but the general point being made seems to be that, in
order fully to understand a field of enquiry falling under some gen-
eral Form, we should not simply point to the wide variety of things
that fall under it, but rather seek to classify the various objects in
a structured way, looking first for species, then for subspecies and
so on (19a–b). The examples given here include the classification of
vocal sounds, of the sort represented by letters; these are divided first
into vowels, voiced consonants and unvoiced consonants, and then
into more specific kinds (18b–d). Socrates claims that we cannot gain
knowledge of any one of the items in this scheme of classification
without knowledge of the others. Later in the dialogue this method
is applied to the classification of kinds of pleasure and knowledge,
since the value of pleasure and knowledge, and the contribution they
make to the good life, is the overall theme of the work.
Another aim of division may be to draw attention to philosophi-
cally significant differences. When two species belong to the same
genus, this means that they have something in common, but also that
there is some significant difference between them; it is important not
to think that because one of them has a particular property the other
must do likewise. Thus in the Phaedrus, as we have seen, two kinds
of madness are distinguished; in both we are not wholly under the
control of reason, yet one is harmful, the other beneficial.
However, in the Sophist and Statesman Plato links his interest in
classification with his interest in definition. It is possible to define a
term by showing where it fits into a scheme of classification. Thus,
starting with a very general Form such as production or acquisition,
one can divide it into species, then subdivide one of the species and
so on, until a very specific Form, that of the sophist or the statesman,
71
plato
has been identified. The definition, when achieved, will show just
how the Form being defined fits into a larger pattern of Forms. The
search for definitions of the sophist and the statesman follows this
pattern. Indeed, in the Sophist, the definition given at the end of the
dialogue, according to which the sophist is a producer of deceptive
images, lists all the Forms under which he falls, from the most spe-
cific to the most general, production (Soph. 268c–d), although the
same pattern is not followed in the Statesman.
The Sophist and Statesman present a rather different aspect of
division from that which we found in the Phaedrus. Whereas there
it was concerned with making a few philosophically significant dis-
tinctions, in the later works it aims to be comprehensive, and seems
more like the scientific schemes of classification associated with Lin-
naeus. There is, in fact, evidence that biological classification was
practised in the Academy; in a fragment of Greek comedy some of
Plato’s pupils are shown attempting to classify a pumpkin (Epicrates,
quoted by Ath. Deipnosophists II.59). Indeed, in the Statesman, the
Stranger, in the course of trying to classify arts that care for animals,
makes some remarks on the classification of animals themselves (Plt.
262a ff.). This, however, does not mean there is a radical difference in
the underlying idea; the method of collection and division is equally
applicable to philosophically significant Forms, and to more mun-
dane ones.
Can only Forms be known?
In places in Plato’s work, both in the Republic and, perhaps more

explicitly, in the Timaeus, the suggestion is made that only Forms
can be known. This is certainly a puzzling claim; we have to consider
both why Plato puts it forward and just what he means by it. In the
Republic, it is linked with the claim that only Forms have unquali-
fied “being”. The argument (Resp. 476e ff.) is a very puzzling one. At
the beginning, Socrates asserts that we can have knowledge only of
what is, and indeed of what “completely is”: things that are “between
being and not being” can only be objects of true belief, which lies
72
knowledge
“between knowledge and ignorance”. The opening claim would be

plausible if “is” were taken to mean “exists”; it is not an unreasonable
suggestion that we can have knowledge only of things that exist. (It
is perhaps true that we can know about, say, Sherlock Holmes, who
does not exist, but one would not normally speak simply of know-
ing Sherlock Holmes). This, however, raises problems with the idea
of “being completely”. This suggests there are degrees of being, and
if “being” means “existence”, this must mean degrees of existence:
a puzzling idea. (Alternatively, “is” might mean “is true” or “is the
case”, in which case we have a reasonable claim that we can only
know what is true, but the idea of degrees of truth is once again puz-
zling, and at any rate the way the passage then develops suggests we
are discussing knowledge of things, not of facts.)
At the end of the argument, however, Socrates explains that the
many objects of ordinary people’s attention – by contrast with the
Forms – are “between being and not being” in the sense that in one
way they are large, heavy, beautiful and so on, and in another way
they are not (Resp. 479a–b). For this reason they cannot be objects
of knowledge, but only of true belief. It seems, therefore, that they
are “between being and not being” in the sense of predicative being,
the kind of being involved in being large or being beautiful, although
one would not have guessed this from the way the topic is introduced
at the beginning of the argument.
As I suggested in Chapter 3, it is possible that when Plato speaks
of ordinary things lacking being he means that they lack reality, in
the sense of really being the sort of thing that they claim to be: the
way in which imitation pearls, or the Velveteen Rabbit, lack reality.
However, one may wonder why this should mean that they cannot be
objects of knowledge. Can we not have knowledge of velveteen rab-
bits? It is true that they would not be a good place to look if we were
seeking knowledge of rabbits – live ones – and by the same token,
it may be that sensible beautiful things are not a good source for
knowledge of the Form, beauty itself. But it is not obvious that they
cannot themselves be known. It may be, then, that Plato has made
a false inference by slipping between the various possible senses of
“being”; he has moved from the plausible claim that we cannot know
73
plato
what does not exist, or what is not true, to the less plausible claim
that we cannot know what lacks “reality” in this distinctive sense.
In the Timaeus, we are again told that Forms are objects of know-
ledge, and sensible things only of perception and belief, but here
this is linked with the claim that the sensible world is in change (Ti.
27d ff.). As I suggested in Chapter 3, this need not mean that, for
Plato, the world is in such constant change that nothing stays the
same long enough to be observed. Rather, his point is simply that
everything is liable to change. The idea that we cannot have know-
ledge of things that are liable to change clearly implies an extremely
high standard that something will have to meet in order to count as
knowledge; but it is not completely implausible. Plato clearly sees
knowledge as something that is itself secure, and cannot be shaken.
But any belief based on things that change, one might think, is liable
to be lost. A belief about particular changeable facts, based on obser-
vation – for example, that my desk is brown – will not remain true
forever, and when it is no longer true, we will no longer be able to
establish, by observation, that it used to be true. A general belief
about the changeable – for example, that all swans are white – is
always vulnerable to the possibility of an example turning up that
falsifies it. Hence, we might argue, the only beliefs that are completely
unshakeable, and so can count as knowledge, are beliefs about nec-
essary truths, beliefs about what is unchanging, and these beliefs,
of which mathematical beliefs are perhaps the clearest example, are
gained by reasoning, not by perception. For Plato, they will be beliefs
about the eternal natures of things, Forms.
Does Plato really want to restrict the scope of knowledge so radi-
cally? In a number of places, in order to illustrate some point about
knowledge, he does refer to knowledge of mundane matters: in the
Meno to knowing who Meno is (Meno 71b), or knowing the road to
Larisa (97a); in the Theaetetus to knowing who committed a crime
(Tht. 201a–c). It is possible to see him, in these passages, as using
“know” in a more relaxed way, for something he does not think
should count as knowledge in the strictest sense. However, it is clear
at least that he is not proposing a general scepticism, a policy of radi-
cal doubt: he does think some beliefs about the sensible world are
74
knowledge
perfectly reasonable and well founded, or else he would not be able

to describe them as knowledge even in a relaxed sense. Still, they do
not have the complete security that our knowledge of Forms has.
Sometimes, this reading of Plato is objected to on the grounds
that Plato clearly sees knowledge as having practical relevance, but
our actions are concerned with things in the sensible world; if we
have no knowledge of things in the sensible world, how can our
knowledge guide our actions? (See e.g. Annas 1981: 214.) In fact,
however, it seems that knowledge of abstract matters can be relevant
to action. We can, for instance, know what justice is, and this will
be relevant as we seek to act justly. It is true that we will not strictly
speaking know the particular situations in which we try to act justly,
but we can perceive and form beliefs about them, and then act in
them in the light of what we do know. An example of this is found
at Republic 501b–c, where the philosophers who rule the ideal state
are shown looking alternately at the ideal – the Forms – and at the
community that they are trying to mould as an image of the Forms.
Their knowledge of the ideal guides what they do in the actual world.
In the Meno, Socrates claims that true belief is converted into
knowledge by working out the explanation (Meno 98a); and Plato
often uses a word for knowledge, epistēmē, which can also be trans-
lated “understanding”. This may suggest that what Plato is really
interested in is understanding, rather than simply knowledge or
certainty, and it may seem more reasonable to restrict understanding
to general truths, grasped by reason, than to restrict knowledge as
such. With mathematical or logical truths, for instance, it is plausible
that we come to grasp them by seeing why they are true, while for
particular matters of fact this need not be the case. I might come
to know that there is a dog in the next room by hearing it barking
without having any idea of why it is there.
I think the kind of knowledge Plato has in mind is indeed one
that involves understanding: that in coming to know Forms we come
to see why they are as they are, not just that they are as they are.
But this does not mean that all he is saying is that we do not have
understanding of particular matters of fact. In the same passage of
the Meno he claims that knowledge is “tied down”, made secure, by
75
plato
explanation; explanation is important because it gives beliefs a stabil-

ity they would not otherwise have. This is presumably because the
kind of explanation he has in mind shows why these things must be
true; and once we have seen that something must be true, there is no
evidence that can count against it. There can, of course, be explana-
tions of particular matters of fact, but they do not show in the same
way that something must be true. For this reason our beliefs about
the sensible world remain fallible.
Recollection: the Meno
One of the most striking claims about knowledge put forward in

Plato’s works is that we acquire it by remembering something we
have learned, in a discarnate state, before we were born into our
present life: this is commonly called the theory of recollection.
Plato introduces this theory in a number of works, from rather
different perspectives. Probably its first appearance is in the Meno
(80d ff.). There, Socrates and Meno have been discussing the ques-
tion “What is virtue?”, and have found that they are completely
stuck. Socrates insists that he is not deliberately trying to confuse
Meno, but, just like him, does not know at all what virtue is. This
prompts Meno to raise the famous paradox (sometimes called
“Meno’s paradox” or “the paradox of enquiry”): if you do not know
at all what a thing is, how can you look for it, and how will you rec-
ognize it if you find it?
Clearly in some cases this is not a serious problem. Often the
knowledge we need to understand a question, and see what evi-
dence might be relevant to answering it, is quite different from the
knowledge we need to give the answer; consider, for instance, the
question “What was the date of the Battle of Bannockburn?” But
with an abstract question such as “What is virtue?” there does seem
to be more of a puzzle. While Socrates and Meno are unable to give
a definition of virtue, they seem to have some awareness of what it
is, which enables them to understand what they are searching for
and to see what evidence might be relevant, but it is not clear how
76
knowledge
they would explain what they know about it, or how the knowledge
that enables them to start the enquiry differs from the knowledge
they are seeking.
Socrates, in answering this question, appeals to a religious trad-
ition, according to which we have lived many lives, seen things here
and things in Hades (i.e. the world of the dead), and so have learned
everything. Having once known what virtue is, and so on, we are able
to recover that knowledge. It is not made entirely clear how this helps
to solve the paradox. How does the fact that we once knew something
help us to enquire about it now, if we have lost the knowledge? The
thought is presumably that while we no longer have explicit know-
ledge of what we once learned, we do have some kind of implicit
awareness of it, which enables us to direct our search, to accept some
claims about virtue and reject others when they are presented to us,
and so to move towards an explicit account.
Socrates then gives a demonstration of this with a slave of Meno’s,
guiding him, by careful questioning, towards the solution of a geo-
metrical problem. At the end the slave is able to give a correct answer
to the problem, although at the beginning he did not know it. Socra-
tes claims not to be “teaching” the slave, that is, actually telling him
the right answers; rather, the slave is finding out the answers for him-
self. It is often felt that, in fact, Socrates is giving the slave too many
clues, by leading questions, so that the slave could have reached the
right answer just by following Socrates’ suggestions. But it seems that
what Socrates is trying to illustrate here is at least possible, although
perhaps it really needs a longer demonstration than Plato has space
for within the Meno. When asked the right questions we can see for
ourselves that some answers are right, others wrong, although we did
not know this before; we need guidance to make sure that we go in
the right direction. But we do not accept the answers just because the
teacher gives us them; we are working them out for ourselves. This
seems to be true of both mathematical and philosophical learning.
It contrasts with other areas such as history, where we do have to
accept what people tell us.
Socrates claims that, within the dialogue, the slave has only
achieved true belief about the answer to his problem, but that with
77
plato
more questioning he will be able to achieve knowledge (presumably

because he will not only be able to give the right answer, but also
explain why it is the right one). He will have found this knowledge
within himself and, Socrates claims, recovering knowledge from
within oneself is recollection (Meno 85d). Thus, the theory of recol-
lection is established.
However, Socrates’ argument here is questionable. It may be true
that the slave is finding knowledge within himself in the sense that he
is not actually acquiring it from anyone else, but this may mean not
that it was actually present in him, but merely that he has a capacity
to work things out for himself, to see how concepts are connected
to one another, and so on. Alternatively, the knowledge may have
been present in him in some sense, but not because he had actually
learned it in the past or once known it consciously; he may just have
innate, implicit knowledge. In this case it would be odd to describe
what he was doing as “recollecting”.
One puzzle that this argument raises is just what we are supposed
to be remembering. Explicitly, we are told that we are remembering
things that we have learned in previous lives. But it seems that the
emphasis is on the discovery of abstract truths, including mathemat-
ical truths (which are what the demonstration with the slave focuses
on) and philosophical truths such as the nature of virtue (which is
the theme of the dialogue as a whole). It is knowledge of this kind,
not knowledge of empirical matters, that we can reasonably be said
to recover from within ourselves. But it seems implausible that we
could have learned these truths by ordinary experiences in previous
lives, here or in Hades, given that we do not learn them by experience
in this life. The Phaedo is going to make it clear that what we are
recollecting are Forms, and that we gained the original knowledge
by contact with Forms while in a discarnate state. Although this is
not explicit in the Meno, it is compatible with what is said there: the
context is an enquiry into the question “What is virtue?”, and virtue
is described earlier in the dialogue as a Form (Meno 72c). So we can
see Socrates and Meno as having had some knowledge of the Form
of virtue before their birth.
78
knowledge
Recollection: the Phaedo
In the Phaedo (73a ff.), Socrates discusses recollection again, but

presents it in a rather different way. In the Meno recollection is
prompted by questioning, and what it produces, when complete,
is explicit knowledge, involving the ability to give an explanation.
In the Phaedo it is prompted by perception, and (although this is
controversial) it seems that what it enables us to do is to recognize
examples of a Form, to acquire a general concept or see the one in
the many. For instance, by looking at pairs of sticks that are equal (in
length) to one another, we may acquire a concept of equality; and
this, Socrates argues, means that we are remembering the Form of
equality, which we knew before birth.
Socrates’ argument is that when we see sticks or stones that are
equal to one another, this makes us think of equality itself, and this
means that we are being reminded of equality; this, in turn, implies
that we have known it before. Why should we think that this is a
case of being reminded? To begin with, Socrates argues simply that
the equal sticks and stones are distinct from equality itself, and that
when, perceiving one thing, we are made to think of something else,
this is a case of being reminded. This, as a general claim, does not
seem plausible; perceiving one thing might stimulate us to invent
something new; for instance (in an example suggested by David Bos-
tock [1986: 63]) James Watt, seeing steam coming out a kettle, was
stimulated to come up with an improved design for a steam engine.
However, Socrates goes on to say not merely that equal sticks
and stones are distinct from equality itself, but that they fall short
of it (Phd. 74d), and that when we see them we recognize this. This
seems a possible basis for a rather stronger argument for recollection.
When, say, I see a picture of Socrates, and this not only makes me
think of Socrates, but also question to what extent it is like him, this
does imply that I already have some knowledge of Socrates and am
being reminded of him; I could not judge that the picture fell short
without some prior knowledge of what it falls short of.
What is meant by saying that sensible equal things fall short of
the Form? As we saw in Chapter 3, this could mean either that their
79
plato
equality is imperfect – they are not, in fact, precisely equal, only

approximately so – or that it is qualified – they are equal in one way
but not in another (e.g. equal in length but not in thickness). In
the case of equality both these ways of falling short may be in play.
In the case of other Forms mentioned in this passage, one may be
more relevant than the other: for instance, with largeness the point
may be that things are large in a qualified way (since it does not
makes sense to think of anything as perfectly large), but with justice
the main point may be that people and societies are imperfectly
just.
Although this point is never made wholly explicit, one can see
Socrates as arguing that we cannot simply get our notion of equal-
ity from sensible equal things; our concept of equality cannot just
be that of “what these sticks have”. Rather, we bring a conception of
equality to the sensible equal things, although we do not have that
conception consciously until we encounter the sensible equals. This
is why we can speak of the sensible equals reminding us of equality.
This can be supported by reference either to imperfect possession
of properties, or to qualified possession of them. If we recognize
sensible things as approximating to, but falling short of, equality,
we must already have some idea of what equality is. Likewise, if
their equality is qualified we must already have some conception of
equality, to distinguish the respect in which they are equal from the
respect in which they are not.
It also seems (although here there is some dispute about the trans-
lation) that Socrates refers to the possibility of disagreement about
whether two sensible things are equal (Phd. 74b). This can be seen as
supporting the same point; we cannot be getting our conception of
equality just from the sticks if we disagree about whether the sticks
are equal, but agree on what equality is, and can distinguish it from
its opposite.
All these points tend to support the view that our concept of
equality is not derived just from perception of its instances; rather,
we already have that concept, and bring it to perception. However,
once again, it does not follow that we acquired it by some past experi-
ence, before birth. Socrates, in drawing this conclusion, is going
80
knowledge
beyond what his argument warrants. It might be innate in us, or we

might have acquired it by some unconscious process during this life.
Recollection: the Phaedrus
A third dialogue in which recollection features is the Phaedrus. Once

again, as in the Phaedo, it is made clear that the objects of recollec-
tion are Forms. However, it introduces some interesting new ideas
concerning the question what recollection enables us to do.
Recollection in the Phaedrus seems to have two aspects. On the
one hand, we are told that a soul cannot be incarnated as a human
being (as opposed to a lower animal) unless it has, before birth, a
vision of the Forms, since human beings have to “understand the
language of Forms, proceeding from a multitude of perceptions to
a unity gathered together by reasoning” (Phdr. 249b–c). This seems,
as in the Phaedo, to refer to an ability that all human beings exer-
cise (although this has been disputed). But about just what ability
this is, Plato’s language is ambiguous. It could refer, once again, to
seeing the one in the many, forming a general concept. We have
many individual perceptions of trees, and we come to see them all
as belonging to a single kind, but we could not do this unless we had
some inner, unspoken, awareness of the Form. But it could also refer
to the way we become aware of even one tree: from many individual
acts of perception we come to see it as a single unified thing. Plato
may here be suggesting that we could not do this unless we had an
inner awareness of the kind of thing it is.
However, the principal use that the Phaedrus makes of recollec-
tion is different. The theme of the central speech in the Phaedrus is
the praise of love, and in it Socrates describes the experience of fall-
ing in love: the result produced in us by the vision of beauty (Phdr.
251a ff.). He says that when we see a beautiful person this reminds
us of the Form of beauty, which we perceived before our birth, and
wakens in us a desire to regain that heavenly vision; the metaphor-
ical wings of our soul, lost at birth, begin to grow again. What is
distinctive of this account of recollection is that in it we recollect
81
plato
the experience; we are not just enabled to give an account of Forms,

or to recognize instances of Forms, but rather we remember what it
was like to contemplate the Forms, and so we are moved to want to
recapture that experience. This gives recollection a mystical aspect
that it does not have in other places where it occurs. Beauty plays a
special role here because its instances – unlike the instances of other
central Forms such as goodness and justice – are visible; but once
visible beauty has stimulated our desire for the vision of the Forms,
we can pursue the other Forms as well. Hence, beauty plays a central
role in moving us to philosophy.
The significance of recollection
One important question that arises in connection with recollection

is how widespread it is. On a traditional view, which I have followed,
in the Phaedo and the Phaedrus it is very widespread indeed, being
something we all do whenever we deploy concepts. In the Meno,
however, it is more restricted, and happens only when we engage in
a conscious process of learning. The kind of learning in which the
slave is shown engaging is indeed very common; many of us must
have experienced recollection while studying mathematics. But it is
plausible that recollection of things such as the nature of virtue is
much less common, and indeed is confined to philosophers. In the
light of this it has been suggested (see Scott 1987, 2003: chs 1–2) that
the Phaedo and Phaedrus too should be interpreted so as to make
recollection more restricted, and that they refer not to the acquisi-
tion of concepts, but to the achievement of conscious knowledge of
Forms.
It is true that in the Meno the name “recollection” is restricted to
what happens at the end of the process, the actual achievement of
conscious knowledge, but our implicit awareness of the nature of
virtue and similar matters seems to be at work before that. It is what
enables us, first to ask the question “What is virtue?”, then to see
what evidence might be relevant to the answer. Socrates does not
suggest that the search for an answer should begin by ignoring our
82
knowledge
ordinary beliefs about virtue. On the contrary, he makes use of such

beliefs, both general beliefs such as “virtue is beneficial” (Meno 87e),
and beliefs about examples, such as that justice and temperance are
virtues (73d–4a). Indeed, until we achieve knowledge of what virtue
is, such beliefs must be provisional (this will be discussed later in the
chapter, when we look at the method of hypothesis), but they can-
not be abandoned entirely; they do guide us in our search, and it is
by reflection on them that we are led to the truth. This makes most
sense if they are themselves derived from our prenatal awareness of
the Forms, which is present, although only in an implicit way, when
we begin to form the concept of virtue.
In the light of this, it is not implausible that in the Phaedo recol-
lection is seen as being involved in the actual formation of concepts.
This is not inconsistent with the account given in the Meno; it may
be that we have a latent knowledge of Forms, acquired before birth,
that enables us first to recognize examples, then, when stimulated by
questions, to move towards a more articulate account. But if that is
right, “recollection” is being used in the two dialogues as a name for
different stages in the process (see Crombie 1962–3: vol. II, 143–4).
In the Phaedrus, “recollection” refers both to the formation of con-
cepts, and to the process, stimulated by the vision of beauty, that
brings us to conscious knowledge of the Forms.
The claim that all of us have some implicit awareness of the Forms,
and manifest this in our daily life, might be taken to support the
view that all are capable of philosophy, as Socrates in the Socratic
dialogues seems to have believed. However, I am not sure how close
the connection between these views in fact is. Certainly the Phae-
drus, although it insists that all human beings must have had the
prenatal vision of Forms, still implies that only some are capable of
philosophy, which suggests that even if we all have latent knowledge
of Forms, some extra capacity is needed to make this knowledge
explicit.
A second significant question about recollection is how important
it is to the theory of Forms. In the Phaedo, they are clearly presented
as inseparable (Phd. 77a); Socrates claims that the existence of Forms
stands or falls with the existence of the soul before birth, apparently
83
plato
because it implies recollection. However, it is not obvious that they

must be connected so closely. The theory of recollection seems to
have two aspects. First, our knowledge of Forms in some sense comes
from inside us; we do not gain it straightforwardly from percep-
tion, or from a teacher, although both perception and teaching may
stimulate us to achieve active knowledge. Secondly, this knowledge
was acquired by an experience in a discarnate state before birth,
and is now being remembered. The first element seems essential to
Plato’s thought, but the second is less so; might we not simply have
an innate capacity to grasp the nature of Forms when appropriately
stimulated?
Plato’s doctrine of recollection has often been compared to the
modern concept of a priori knowledge: knowledge which is gained
by reasoning rather than by experience. In a way this seems right;
he is pointing to a kind of knowledge that does not depend simply
on sense-experience. But he may nevertheless be making it too like
empirical knowledge if he thinks it was gained by an experience
before birth; he writes as if we must actually have met the Forms
to have knowledge of them, and this must have happened before
birth, since we clearly do not directly perceive them in this life. But
if our knowledge of the Forms rests on an intellectual grasp, not a
literal seeing, it ought to be possible to gain that knowledge in this
life as well.
It is, in fact, possible that Plato moved away from the doctrine
of recollection. In two dialogues, the Republic and Timaeus, we
are given accounts of the afterlife that do not include the vision of
Forms that is needed for recollection. What is more, the Republic
urges us to pursue wisdom in this life in order to be able to make
correct choices in the afterlife (Resp. 619d–e) and the Timaeus gives
an account of the origin of the soul that points to a different expla-
nation of our capacity to know Forms; the soul is made by God in
such a way that its nature is similar to that of Forms, and so is able
to know them on the principle that like knows like (Ti. 35a ff.). It
seems, then, that recollection is not essential to the theory of Forms,
and we should not assume that Plato believed in it throughout his
career.
84
knowledge
Philosophical enquiry
Our last topic in this chapter is the nature of philosophical enquiry.

How does Plato think we should go about it?
This question is addressed most fully in Book 6 of the Republic,
in the passage that introduces the famous simile of the divided line
(Resp. 509d–11e). The passage is complex, and seems to be making
many points at once, but its principal aim is to introduce a view
of philosophical method, and in particular the difference between
philosophical and mathematical enquiry. Socrates asks us to imagine
a line, apparently vertical, divided into two unequal parts, each of
which is then subdivided in the same ratio as the whole. Thus, when
the division is completed, the line will have four sections.
When the line is first introduced we are told that the different
sections represent different kinds of object: the two sections making
up the lower line represent the realm of visible things – ordinary
sensible objects – while the two sections making up the upper line
represent intelligible things, a class that at least includes Forms. Of
the two sections that make up the lower line, the lowest represents
images, such as shadows and reflections, in the visible world and
the other represents concrete objects, such as animals, plants and
manufactured things. However, when we reach the two sections of
the upper line, we are not told that they represent different objects;
rather, they seem to represent two kinds of enquiry, the lower of the
two sections standing for the kind of enquiry typical of mathematics,
while the other, the highest section of the line, stands for that typical
of philosophy or dialectic.
It is odd that while in the lower line we are introduced to two
kinds of object, in the upper line we are told only of two kinds of
enquiry. Many readers of the passage have supposed that the two
sections of the upper line should indeed be seen as standing for
different kinds of object, with which the two kinds of enquiry are
concerned: dialectic will be concerned with Forms, mathematics
with some other kind of object. There is evidence from Aristotle that
Plato did at some point in his career believe in special objects with
which mathematics deals (Ar. Metaph. A 987b14–18). However, here
85
plato
Socrates seems to say that mathematical reasoning is concerned with

Forms: “the square itself and the diagonal itself ” (Resp. 510d–e). The
intended correspondence between parts of the line should be seen
as relating not so much to objects as to states of awareness: math-
ematical reasoning is related to dialectical reasoning in the same way
that perception of images is related to perception of concrete things.
The underlying thought seems to be that while mathematicians are
thinking about Forms they are doing so indirectly, just as someone
looking at a reflection of trees in water is perceiving the trees but
indirectly. Mathematicians do not have the direct grasp of Forms that
philosophers have, and which is implied by the term “knowledge”.
Why is mathematicians’ grasp of Forms inadequate? Socrates
gives two reasons. One is simple: they use visible diagrams. Although
they are indeed thinking of “the square itself ”, not of the particular
square that they have drawn (since they intend their conclusions
to be universal), they cannot grasp the Form by pure thought, as
it should be grasped, but need to rely on visible diagrams to get a
clear idea of it.
The other difference between mathematicians and philosophers
is more obscure; it is that mathematicians rely on hypotheses. This
means that they have concepts associated with certain basic math-
ematical terms – the odd and the even, kinds of shape or of angle
– of which they cannot “give an account”; this may mean both that
they cannot explain further what these concepts amount to – that
they cannot give a definition of them – and that they cannot give
reasons why these concepts are correct. Later, Socrates is to compare
the state of mind of someone who cannot give an account of some
term with dreaming (Resp. 533b), and he thinks of dreaming as mis-
taking an image for the reality (476c). The mathematicians have
in their minds a conception – an image, in an extended sense – of
what, for instance, a square is, and because they take this conception
for granted and cannot explain or justify it, they do not have a real
grasp of the Form of square, even though that is in some sense what
they are thinking about. Hence, once again, their state of mind is
like that of someone looking at shadows or reflections rather than
concrete things.
86
knowledge
How, then, is the philosopher to move beyond this state of mind,

and come to know Forms as they are? To answer this we must look
more closely at Plato’s use of the concept of hypothesis, which he has
already introduced in earlier works.
The method of hypothesis
In two places, in the Meno (86e ff.) and Phaedo (100a ff.), Socrates
recommends a method of enquiry – derived from mathematics –
that starts from hypotheses. By this he seems here to mean state-
ments that are plausible, so that it is reasonable to accept them as
a starting-point for enquiry, even though they are not known to be
true. If enquiry had to start with something that we knew for cer-
tain, it is not clear that it could ever get off the ground. Thus in the
Meno Socrates implies that we cannot know whether virtue has any
particular quality – specifically, whether it is teachable, the official
topic of discussion in the dialogue – unless we first know what it is
(Meno 71a). But how can we discover what it is without making ref-
erence to its qualities? The concept of hypothesis allows us to break
this circle; starting from a plausible claim about its nature we can
make inferences about its qualities, even without knowledge. There
is some dispute about just how Socrates is using the term “hypoth-
esis” in this passage, but on what seems the most plausible reading
he adopts the hypothesis that virtue is knowledge, and infers from
that that it is indeed teachable; the claim that virtue is knowledge
is itself supported by reference to a further hypothesis, that virtue
is good (i.e. advantageous). Likewise in the Phaedo the existence of
Forms is taken as a hypothesis, and used as a basis for arguing to an
account of the explanation of becoming and perishing, and finally
to the immortality of the soul.
A hypothesis is not, on this account, primarily a claim put forward
to be tested; it is something that the participants in a discussion find
plausible and are therefore able to use as a basis for further discus-
sion. But, since hypotheses are not known to be true, they may be
challenged. In fact, in the Meno the hypothesis that Socrates and
87
plato
Meno first agree on, that virtue is knowledge, is later challenged

(Meno 89d ff.); in the Phaedo the actual hypothesis used is not chal-
lenged, but the possibility is explicitly mentioned (Phd. 101d). If this
happens, we may either abandon the hypothesis, or find some more
basic claim, on which agreement can be reached, on whose basis it
can be defended.
However, this form of reasoning is widely seen as unsatisfactory;
even if we do in fact manage to achieve agreement, the method will
only expand our stock of plausible beliefs; it will not give us know-
ledge, since if the starting-points of reasoning are not known, what
is derived from them will not be known either.
Hypothesis in the Republic
In the Republic, as we have seen, Socrates criticizes mathemati-

cians for using hypotheses as starting-points and failing to give an
account (logos) of them (Resp. 510c ff.). It is not wholly clear that the
kind of hypotheses he is discussing here are the same as those dis-
cussed in earlier works, for in them hypotheses were propositions,
while here he refers to mathematicians as hypothesizing things: the
odd and the even, figures and kinds of angles. Yet it seems plau-
sible that when Socrates speaks of mathematicians hypothesizing
these things he means, at least in part, that they assume the truth
of some claim about them: either that they exist, or an account of
their nature. In this way the account of hypotheses here will har-
monize with that found in earlier works. In saying that they fail to
give an account of the things they hypothesize, he may mean that
they fail to give either a proof, or an explanation of why these things
should be so, but for Plato these two ideas will be closely linked, if
he thinks, as is suggested in the Meno, that we attain knowledge of
something by working out the explanation. It may well be that Soc-
rates also means that mathematicians cannot clearly state what they
mean by these central terms. If this is right then any belief they hold
about what the square and so on are will be held only implicitly. But
it seems that we should see them as holding some belief about the
88
knowledge
nature of these things, which might be true or false, but which lacks
justification.
The problem with the mathematicians is not that they use hypoth-
eses, but that they see no need for anything further; they fail to rec-
ognize that a method that starts from hypotheses, giving no proof
or explanation of them, cannot give us knowledge. By contrast the
method of dialectic, which Socrates recommends, seeks for a more
secure starting-point on the basis of which our hypotheses can
be established (Resp. 511b ff.). No doubt some hypotheses will be
rejected in the course of this enquiry; others may remain but, having
been established on a firmer basis, will no longer be hypothetical.
As we have seen, the Meno and Phaedo accepted that hypoth-
eses would sometimes be challenged, and would then need to be
grounded on a firmer basis. In the Republic, however, this is not
treated just as something that may happen occasionally in the course
of discussion, but as an essential part of philosophical method; a
philosopher should not remain content with hypotheses but should
seek for a further ground on which they can be established. What is
more, this process will end not simply with another hypothesis with
which our partners in discussion will agree, but with an “unhypo-
thetical starting-point”: something that is not merely hypothesized,
but known, and so is able to serve as a ground for further knowledge.
It seems, moreover, that there is a single first principle that is able to
serve as the starting-point for all knowledge.
How is this search for a first principle meant to work? It seems
likely that the process is gradual; we do not look, at once, for a first
principle that will ground all our beliefs, but rather move by stages
towards it, first finding propositions that will support our initial
hypotheses, then propositions that will support them, and so on.
But what motivates our acceptance of a proposition? If we are right
in thinking that a foundational claim gives the explanation of what
follows from it, this will help to structure our search; we will look for
propositions with explanatory power. If, moreover, the ultimate end
of our search is a single first principle for all branches of knowledge,
it is likely that as we progress towards it we will find more general
claims that enable us to unify various areas of enquiry. Indeed, later
89
plato
in the Republic Socrates says that the dialectician is someone able

to take a synoptic view (Resp. 537c). The first principle, when we
discover it, will be a claim of great explanatory power over a wide
range of branches of knowledge.
This, however, is not to say that our acceptance of the first prin-
ciple is grounded simply in its explanatory power. Since the claims
we are setting out to explain are not initially known, it is not clear
that we could come to know a principle just through its power to
explain them; it seems possible that there might be different systems
of propositions, each coherent and each containing a first principle
that explains the others, but only one of them can be right. It is likely
that Plato sees the first principle as something that, although it is
hard to discover, once we have found it will be self-evident; some-
thing that, once it is understood, cannot be rejected.
The distinctive feature of this account of philosophical enquiry is
the twofold process that it envisages; first the way up, moving from
our accepted beliefs to the first principle; then the way down, return-
ing from the first principle to more specific areas of enquiry, of which,
in the light of the principle, we can now gain knowledge. (While I
have written, in line with the usual way of thinking today, of the
first principle as a foundation and as grounding other claims, Plato
actually sees it as “higher” than other claims, which are envisaged as
hanging from it, and this metaphor guides the way he describes the
process of enquiry.) On the way down we will be going again over
material we had already traversed on the search for the first prin-
ciple, but now with knowledge, not just provisionally; knowledge of
the first principle may also enable us to make new discoveries. This
twofold process was clearly important to Plato; Aristotle tells us that
he often used to ask his pupils to clarify whether they were moving
to or from first principles (Ar. Eth. Nic. 1095a30–b1).
This allows us to begin our enquiry with common beliefs, but not
to accept them uncritically; the starting-points for enquiry are not
the same as the starting-points for demonstrative reasoning. It dif-
fers, therefore, from the traditional understanding of foundational-
ist views, which require us to begin our enquiries with things that
are indubitable; these are often identified with facts known through
90
knowledge
perception. Nevertheless, Plato’s view does posit a foundation; it

does not suggest that beliefs are validated simply by the way they
hang together. But for Plato the foundation – since it is the starting-
point not for enquiry, but only for demonstration – may be some-
thing quite abstract, the discovery of which requires much thought.
A similar view is found in Aristotle, who distinguished between
“things better known to us”, which are the starting-points of enquiry,
and “things better known in themselves”, which are harder to find
but, when found, can serve as a foundation for knowledge (Ar. An.
post. 71b33–72a5). But for Aristotle there are different principles for
the different sciences, and a number of principles in each of them;
for Plato there is just one first principle for all the sciences.
A real-life analogy for the kind of structure that Plato has in mind
may be provided by the axiomatic geometry that was worked out, a
couple of generations after his time, by Euclid. Euclid starts with a
small number of postulates from which he derives the various theo-
rems of geometry; but of course he did not actually begin his enquiry
with these postulates; rather, already being aware of various geomet-
ric proofs, he searched for the axiomatic structure that would enable
them to be brought into the simplest and most coherent system. But
while the structure here is similar to what Plato had in mind, he
would not have accepted Euclid’s postulates as first principles, but
would have insisted on the search for a more fundamental principle
that would explain them.
Knowledge and the good
One question remains; what is the unhypothetical first principle?

We are never told in so many words, but the context strongly sug-
gests that it is the Form of the good or perhaps, strictly, a statement
of the nature of that Form. Immediately before the image of the
divided line, in which Socrates contrasts mathematicians with phil-
osophers, he introduced the image of the sun, in which the Form of
the good is presented as the source of knowledge and reality (Resp.
507a ff.). Socrates says that as the sun both promotes the generation
91
plato
and growth of sensible things and enables us to see them, so the good
is responsible both for the being or reality of Forms, and for our
knowledge of them. The dialectical method is concerned with Forms,
so it is reasonable to see the good, which gives us our knowledge of
Forms, as the first principle in that method.
However, the claim that both the reality of Forms and our know-
ledge of them derive from the good is clearly an obscure one, and
Socrates says little to make it clearer. Indeed, it is reasonable to see
the two claims as going together; if the Forms are what they are
because of the good, it is plausible that knowledge of the nature of
the good will enable us to understand the other Forms as well. But
how, exactly, can it be responsible for their reality?
In the Phaedo, Socrates refers to the view that things are as they
are because it is best that they be so. He thinks that if this could
be shown, it would be the basis of a good explanation of things,
although he does not think that any explanation of this kind has in
fact been achieved (Phd. 97d ff.). Might Plato be thinking along the
same lines in the Republic, now applying this thought not to sensible
things but to Forms?
Perhaps Plato is suggesting that Forms exist, and have the nature
they have, because it is best that they should do so. In this way they
could be said to owe their reality to the good, and we could gain
knowledge of them by understanding what is good. Socrates does
indeed describe the system of Forms, at Republic 500c, as ordered
and divine, free from injustice and so on, so he does seem to see it
as in some sense good. Alternatively one might also suppose that
instances of the Forms are good; justice, for example, is a Form,
and it is good that instances of justice, just souls and communities,
exist in the world. One might, therefore, explain what justice is not
by reference to the way in which the existence of the Form itself is
good, but rather to the way in which its instances are good. Or one
could combine these two lines of thought, suggesting that the way
the Forms themselves are ordered is good, but also that things are
good to the extent that they instantiate the Forms, and that both
these relations to the good may be involved in explaining the nature
of a Form.
92
knowledge
Working out what this would amount to in detail is still problem-

atic. It is true that for some Forms in which Plato believes, a con-
nection with the good is reasonably clear. These include the moral
or more generally evaluative Forms, for example justice and beauty;
we can see justice as either a kind of goodness or a state of character
that promotes the good, so we can come to grasp its nature through
understanding the good.
With other Forms the link may be through teleology, the idea that
things exist for a purpose: we come to grasp the nature of things by
grasping the purpose that they serve, how they promote the good.
This applies most straightforwardly to artefact Forms such as that of
bed. But if we believe, as Plato did, that natural things – for instance
the elements, or kinds of animal – are in the world for a purpose,
it can also apply to them. This, however, leaves unexplained other
Forms that Plato certainly recognized. These include the mathemati-
cal Forms, and also the extremely general ones such as being, same-
ness and difference.
Can Plato have thought that these Forms are in some sense good?
Certainly some have felt that beauty or appropriateness is to be found
in mathematical systems, and Plato may well have shared this atti-
tude. He does, for instance, refer to the sphere as the most perfect
of shapes (Ti. 33b), and he thinks that certain numbers stand in
relations of harmony with one another, independently of the sounds
they produce when applied to strings of the lyre (Resp. 531e). We
may find this implausible. We may hold that if the relations between
mathematical concepts are logically necessary, goodness cannot be
relevant to them. But Plato may not have shared this attitude.
One other possibility deserves mention: that the good should be
connected, not with each individual Form, but with the harmonious
system of Forms. Some have even suggested that the good is that sys-
tem, but it may be more plausible to see it as a property that the sys-
tem possesses, although other things may possess it as well. Certainly
Plato’s language at Republic 500 implies that the system of Forms, as a
whole, is something good, something harmonious. It can be seen as
fitting together somewhat in the way that we might describe a beau-
tiful or elegant mathematical or scientific theory as doing. In this
93
plato
case, we may suppose that to understand the nature of each Form we

must understand how it fits into the system, and to understand how
the system fits together we need to know the nature of the good that
it instantiates. In this case not all individual Forms need be seen as
contributing to the good as such; it is as part of a system that they do
so. One advantage of this view is that it explains how, despite linking
Forms with the good, Plato occasionally refers to Forms of bad things
such as injustice (e.g. Resp. 476a). We can say not that injustice itself
contributes to the good, but that it is good that there is a system of
classification in which things such as injustice have their place.
These are just a few attempts to explain something about which
Plato’s words are, as he recognizes, extremely cryptic. What is clear
is that he places the good at the centre not only of his ethics, but of
his metaphysics and epistemology. For him, therefore, philosophy
cannot be divided; the most abstract metaphysical studies are rele-
vant to the way we govern our lives and the state.
The image of the cave
Before we leave this section of the Republic, we should look at the

third important image that is introduced here, in one of the most
famous passages in Plato: that of the prisoners in the cave (Resp.
514a–17a). The passage goes over much of the same ground as the
images of sun and line, in a more graphic form, but it also introduces
some significant new points about the practical consequences of the
theory outlined there.
Socrates first asks us to imagine a group of prisoners in a cave,
chained in such a way that they cannot turn, and can see only the
wall of the cave in front of them. Behind them is a fire; between them
and the fire are people carrying statues, the people hidden behind a
wall, while the statues protrude above it. The prisoners can see only
the shadows cast by the statues, in the light of the fire, on the wall
in front of them. The people carrying the statues sometimes speak,
and the prisoners hear the echo of their voices from the wall. Since
the prisoners have no knowledge of what is behind them, they think
94
knowledge
that the shadows on the wall are real people and things, and that the
noises coming from the wall are their voices.
Socrates next describes one of the prisoners being freed, and
turned around to face the fire, so that he sees the actual statues;
such a person would be dazzled by the light of the fire and would
have difficulty adapting to this new way of seeing things. If he then
were led out of the cave into the sunlight, he would have even more
difficulty adjusting, and would at first not be able to look at real
things in the world outside, but would have to start by looking at
shadows and reflections. Later, however, he would be able to look
at real things in the outside world, and in due course would look at
heavenly bodies, and last of all at the sun.
Finally, we are asked to imagine such a person returning to the
cave, in the hope of liberating the prisoners who still live there.
Whereas at first he had difficulty adjusting to the light, he would
now find it equally difficult to adjust to the darkness. As a result,
those in the cave will think that his eyesight has been ruined; they
will not want to leave the cave, and if he tries to liberate them they
will kill him.
This story is often seen as expressing Plato’s theory of the sensible
world, according to which everything we perceive is an illusion, like
the shadows on the walls of the cave. But in context it seems that
Plato’s aim in introducing this story is epistemological and politi-
cal, not metaphysical; in the initial description of the prisoners he
is describing a particular state of mind and society, not the sensible
world as such. It is true that Socrates says that the cave represents the
world revealed by sight, and the fire within the cave the sun. But even
within the cave it is possible to turn towards the fire and see solid
objects; although the things we see in the sensible world are images
of something further, the Forms, this does not mean that they are
illusory. As for the escape from the cave, it does not mean a literal
escape from the sensible world. It is something we can achieve in this
life, while still literally living among sensible things: it is a change of
the focus of our attention, from the visible to the intelligible.
If the story is read in this way, some parts of it are quite easy to
interpret (see Resp. 517b ff.). Those in the cave are those whose minds
95
plato
are concerned wholly with the sensible world; the escape from the
cave represents the turn from these concerns to intelligible realities.
The person who, on first emerging from the cave, looks at shadows
and reflections in the outside world, represents the mathematician,
whose awareness of Forms is indirect. Socrates is about to go on to
recommend mathematics as a preparation for philosophy, and in
particular as a way of turning our attention away from the senses
(521c ff.). The person looking at real objects in the outside world
represents the philosopher contemplating Forms, while the vision
of the sun represents the contemplation of the Form of the good.
What is harder to interpret is the distinction of levels within the
cave. There are two different situations within the cave; that of the
prisoners, who see only shadows, and that of those who have turned
towards the fire, who can see the actual objects of which they are
shadows. If we insist on a precise correspondence between this image
and that of the line, it seems that the first should represent looking
at (literal) shadows and reflections in the sensible world, the second
looking at concrete objects. Yet this is hardly plausible; the prisoners
are described as “like us” (515a) but we do not spend our lives look-
ing at literal shadows and reflections. The condition of the prisoners
is clearly meant to represent something philosophically significant;
and the shadows they look at are said at one point to include “shad-
ows of justice” (Resp. 517d), showing that literal shadows cannot be
all that is meant.
It seems that Plato wants to distinguish, among those who are
concerned wholly with the sensible world, between two groups: those
whose lives are governed by illusions, and those who have some rea-
sonably reliable true beliefs (perhaps because they are guided by
philosophers). The second group, perhaps, although they have no
real understanding of the nature of justice, might generally be able
to recognize examples of it, while the first cannot get beyond appear-
ances, supposing, for instance, that people are just if they are rich
and so able to repay debts.2 It is reasonable that Plato should want to
make this distinction, since it seems that, by the time he wrote the
Republic, he did not think that everyone was capable of philosophy,
yet he does not seem to have thought that most people were con-
96
knowledge
demned to a life of radical illusion. But if he is drawing the distinc-

tion in this way, he nowhere makes it explicit, so the image remains
hard for us to interpret.
At one point it seems that the cave represents a corrupt state, such
as Plato thought existed in his time. When the philosopher, return-
ing to the cave to liberate its inhabitants, is put to death, this clearly
recalls what happened to Socrates at Athens. Later, however, we are
told that even in the ideal state the cave is still present: when philoso-
phers in that state are required to leave the life of contemplation to
take part in government, they are described as returning to the cave
(Resp. 519d). Here the cave seems simply to be the world of practical
concerns. Perhaps we can associate the two levels in the cave with
these two aspects of it: the prisoners described at the beginning are
the deluded inhabitants of a corrupt state; those who have been freed
but are still within the cave are like the lower classes in Plato’s ideal
state, and have reached the highest level of which ordinary people,
without intellectual interests, are capable.
Two morals in particular emerge from the story of the cave.
One we have already seen: while it is hard for the person who has
been brought up thinking of mundane matters to adjust to abstract
thought, it is equally hard for the person devoted to abstract thought
to adjust to the mundane again. This is why philosophers are often
failures at practical politics (see Resp. 517d ff.). They have difficulty
applying their knowledge, at least at first, until they have become
used to mundane affairs; but it does not follow that their knowledge
is not relevant to practical matters.
The other moral is that a conversion is necessary if we are to gain
knowledge of the intelligible realm. The prisoners must be turned
round and led up to the outer world before they can perceive the
things there, and in particular the sun; in the same way the soul must
be turned towards the intelligible realm and especially the good (see
Resp. 518b ff.). Socrates says that education is not putting knowledge
into the soul. In context, the point is not to deny that it is putting
information into the soul (although Plato would indeed reject this)
but to deny that it is putting the power of knowledge into the soul.
One cannot put the power of sight into the eye, but one can turn
97
plato
the eye to the light. In the same way we already have the power of
knowledge, but we must be turned towards the good. But, Socrates
goes on, the eye cannot be turned unless the whole body is turned,
and in the same way the intellect cannot be turned unless the whole
soul is turned. We must be motivated to pursue truth, and only if
we have the right motives shall we discover it.
98
five
The soul
Plato’s conception of the soul
The Greek word that is commonly translated “soul”, psuchē, means

the principle of life in a living thing. It does not necessarily imply
dualism: the view that the soul is something distinct from and inde-
pendent of the body. Nevertheless, Plato’s conception of the soul was
dualist; he regularly distinguishes and opposes soul and body. This
seems to have been quite normal in his time. Even philosophers
whose general view of the world was materialistic often thought of
the soul as a special kind of body, rather than a state or aspect of
ordinary bodies. Plato’s Socrates does at one point, in the Phaedo
(85e–6d), discuss a theory that is closer to materialism in a modern
sense, according to which the soul is the “harmony” of the body, the
way in which the constituent elements in it are combined, but this
theory is quite swiftly rejected (91c ff.). Most of the time dualism is
taken for granted. However, Plato does not see dualism as immedi-
ately implying that the soul is immortal. This is a further claim that
for him needs argument.
Plato’s position on what functions and capacities belong to the
soul varies from one work to another. In the Phaedo he assigns only
rational activities to the soul; perception, irrational desire and even
belief belong to the body, and for this reason there can be a conflict
99
plato
between soul and body (94b). This is one of his reasons for rejecting
the claim that the soul is the harmony of the body. In other works,
however, including the Phaedrus, Republic, Timaeus and Laws, this
view is abandoned; the soul also includes non-rational elements, in
particular the elements of spirit (anger, daring and ambition) and
appetite (bodily desire – that is, desire for things connected with
the body) (Phdr. 246a ff.; Resp. 435e ff.; Ti. 69c ff.; Leg. 896e–7a). In
some dialogues these elements still arise in some way from the body,
either being derived from it, as in the Republic (611b ff.), or being
created by the gods, but in order to cope with bodily needs, as in the
Timaeus (69c–d). Elsewhere, however, most clearly in the Phaedrus,
they seem to exist independently of the body.1
In any case, the capacities that Plato assigns to the soul are in
general ones that we would now class as mental: thinking, perceiv-
ing, feeling, desiring and so on. He does not, like Aristotle, think of
biological functions such as breathing and digestion as belonging
to the soul.
In what follows, I shall look in more detail at three aspects of
Plato’s view of the soul: his definition of the soul as a motion that
moves itself; his belief in immortality; and his conception of the soul
as divided into three parts, reason, spirit and appetite.
Self-motion
In two places in Plato’s work, in the Phaedrus (245e) and in the Laws
(895e–6a), a definition of soul is offered; it is a motion that moves
itself.2 Given that Plato is generally reluctant to commit himself to
definitions, the fact that he returns to this one is quite striking. In
the Phaedrus the definition is used in an argument for immortality,
and in the Laws in an argument for the existence of gods, and it is
most often discussed in the context of these arguments. Yet it seems
to have a wider relevance. In both places the idea of soul as self-
motion is accepted as if it were already well established, suggesting
that it formed a fairly stable part of Plato’s view of the soul; it is only
the specific uses to which it is put that are presented as new. (This,
100
the soul
of course, does not mean that Plato held it throughout his career; it
is not clear, for instance, if it can be reconciled with the view of the
soul adopted in the Phaedo, where its unchanging nature is empha-
sized.) In addition, Plato makes use of the concept of self-motion in
the Timaeus, although it is not there presented as a definition (Ti.
37b, 46d–e, 77b–c, 88e–9a); there it is linked with rationality, giving
it a wider relevance in Plato’s thought.
In adopting this definition of soul, Plato seems to be making use
of a widespread view in ancient Greece that it is characteristic of
living things to move themselves; it is notable that in the Laws this
information is volunteered by Cleinias, the (not particularly philo-
sophical) respondent, not by the Athenian Stranger, the chief speaker
in that dialogue (Leg. 895c), although it is the Athenian who argues
that this can be seen as a definition. Moreover, both Cleinias and the
Athenian agree that we see that living creatures move themselves; it
is because we see a thing being moved from within that we recognize
it as alive. However, for Plato, as a dualist, it is strictly speaking the
soul that moves itself; it then sets the body in motion, so that living
bodies can be described as moved from within, but not, properly, as
moving themselves. (However the living thing as a whole, body and
soul, can be described as moving itself.)
The term translated “motion” (kinēsis) can be applied to change
more generally, although Plato does seem to think of motion in space
as in some way primary; it is the fact that living things originate their
own spatial motions that makes it plausible to call them self-movers.
Plato does sometimes use language that implies that the soul has a
position in space and performs spatial motions, but even if we do
not accept the spatial conception of the soul, it is possible to think
of it as moving itself as it performs its mental functions of thinking,
feeling and so on.
What precisely does Plato mean by the claim that the soul moves
itself? What feature of soul or of living things is this meant to cap-
ture? On one view, soul for him is something like energy or motive
force: the power of moving without being pushed. However, this
seems too weak to express what Plato means. Many things that are
not alive at least seem to move without being pushed – rising flames
101
plato
and falling stones, rivers moving towards the sea, or magnets – but
Plato does not ascribe soul to them. (He does ascribe it to the stars,
but this, I shall argue, is not merely meant to bear witness to the fact
that they move without being pushed, but rather to account for the
apparent rationality of their motions.) Hence, when we recognize
living things as moved from within, we cannot simply be recognizing
that nothing pushes them.
While this view of self-motion is too weak, there is another that,
I suggest, is too strong. On this view something is not self-moved
if its motion is in any way dependent on anything beyond itself;
hence what is self-moved must be ungenerated. Once again, how-
ever, this does not seem to harmonize with the claim that we recog-
nize things as living when we see them being moved from within.
We certainly do not see the ungenerated status of soul, and indeed
prima facie the soul or principle of life in a living thing is generated,
by its parents. It is true that – since the soul is not just something
that happens to move, but something that is essentially in motion
– what generates it can be seen as setting it in motion, so that if it
is generated it is in one respect moved from without. But it seems
possible that it should be moved from without in one respect but
move itself in another, if it sustains itself in motion, so self-motion
need not imply being ungenerated.
What, then, is central to the concept of self-motion? I suggest that
the core of the concept is the power that animals have to control their
own movements, to take initiative, by contrast with lifeless bodies,
which simply move as they are caused to move by external forces.
This is indeed a feature that we can recognize in animals and that
distinguishes them from other things. Self-motion, so understood,
will imply at least some rudimentary directedness or purposiveness,
although it will not by itself imply rationality or choice in the full
sense.
For Plato, of course, it will not strictly speaking be the visible
movements of the animal that are examples of self-motion, but the
inner movements of its soul, which cause the visible movements;
but the fact that animals control their own movements can be seen
as giving evidence of the self-motion of soul.
102
the soul
The significance of self-motion
What follows from Plato’s conception of the soul as something that

moves itself? First, it is essentially something active; its primary
function is that of a source of motivation, and its other activities,
thinking, perceiving and so on, are significant largely because of the
contribution they make to motivation. When Plato divides the soul
into parts, this is done on the basis of the different motivations it can
have. Secondly, the operation of the soul is never purely mechanical.
External things do indeed move us to action, but they do so in virtue
of the way we think of them or perceive them. We are responding to
them, rather than just being caused to act by them.
As we have seen, Plato’s speakers draw specific consequences from
the definition of the soul as self-motion in a number of places. We
shall consider the Phaedrus later in this chapter, and the Laws in
Chapter 8, when we consider Plato’s cosmology. Here we shall look
briefly at the use made of the concept in the Timaeus. In that work,
at several points, the idea of self-motion is linked with that of rea-
son. The passage that makes the point of this connection clearest
is 46d–e, which argues that only soul is capable of intelligence or
of acting for a purpose, because, implicitly, it moves itself, whereas
material things are moved from outside, and pass on the motions
they receive of necessity. Material things, because they are moved by
external causes, cannot act, in any particular situation, in any way
other than that in which they do act. Hence, they will act as they
do whether or not there is a reason for it: whether or not it serves
some purpose or promotes some good. The soul, on the other hand,
because it controls its own motions, can respond to reasons, and so
can act for the sake of the good.
Plato’s thought here – that only what is free from determination
by physical causes can act for a purpose – stands at the beginning
of a tradition linking freedom and rationality, which was to be very
important in the history of philosophy. Versions of the view are found
in Descartes, Kant and Hegel. This kind of freedom is not exactly
the same as free will as we normally understand it, since this is often
taken to imply that we are free at the moment of choice, and that
103
plato
our actions are not determined by our beliefs, desires and thoughts;
for Plato action can indeed be determined by beliefs, desires and
thoughts, but these are themselves not wholly formed by external
causes. Moreover, this freedom is not a freedom to choose between
right and wrong; Plato holds that if we are truly rational we will
always choose the good. Wrong actions are produced when reason
is obstructed in some way, either by ignorance – a failure of reason-
ing – or by some other power, such as anger or appetite, opposing
reason. (We shall look at this in more detail later in the chapter when
we consider the division of the soul.)
At several points in Plato’s later works we find the claim that nous
(reason or intelligence) is found only in soul (Ti. 30b, 46d; Soph.
249a; Phlb. 30c). This, I suggest, should not be taken to mean sim-
ply that the actual activity of thinking happens only in the soul, but
rather that only the soul can act for a reason or pursue some purpose.
Of course, bodies can serve a purpose, but only when they are under
the control of soul (directly in the case of living bodies, indirectly
for others). In taking this view, Plato’s vision is very different from
that of Aristotle, for whom purpose was found in the (unthinking)
processes of nature as well as in the mind. Plato’s view is much closer
to an early modern one, in which bodily nature acts in a mechanical
way, and needs to be brought under the control of the mind in order
to serve a purpose. We shall see some of the use Plato makes of this
idea in Chapter 8 when we look at his concept of God.
Immortality
Probably the best-known aspect of Plato’s view of the soul is his belief
in immortality; this was clearly important to him, as he returns to it
in many works. The Phaedo is largely taken up with arguments for
immortality, and they can also be found in the Meno, Republic and
Phaedrus. He also makes use of the idea, although without arguing
for it, in the Gorgias (524a ff.), Timaeus (41a ff.) and Laws (903d ff.).
In only one work does Plato seem to show doubts about immor-
tality: the Symposium. In the central speech of that work (given by a
104
the soul
priestess, Diotima, although recounted by Socrates) it is argued that

the only way in which human beings can attain a kind of immortality
is through their offspring (Symp. 207d ff.); these may either be physi-
cal offspring, our actual children, or spiritual offspring, if through
education, legislation or the creation of works of art we are able to
influence the minds of others. At the end of the speech it is possible,
on one reading, to see a hope of personal immortality being held out
to the philosopher, but not to people in general (212a).
Elsewhere, however, Plato seems to believe that the very soul that
lives in us now will live forever. Yet this belief seems to have caused
him some difficulty. He tried numerous arguments for it, none of
which were wholly successful, and in his later works he may be seen
as abandoning argument and rather believing in immortality on the
basis of faith.
Before looking at some of Plato’s arguments we should notice that
it is in general the rational element in us that he sees as immortal,
even in those works where he allows a non-rational element into
the soul (although the Phaedrus is an exception to this). However,
even if only the rational element is immortal, it does not follow that
only it survives death; other elements may live on for a while after
separation from the body, although not forever.
Arguments for immortality: the Phaedo
In the Meno (86b) Socrates suggests that the theory of recollec-

tion is a basis for belief in immortality; and when the theory is first
introduced in the Phaedo, it seems that he intends to use it in the
same way. According to this theory (discussed in Chapter 4) the
soul acquired some important kind of knowledge in a discarnate
state before this life; this implies that it can live in separation from
the body. However, it is pointed out by Simmias, one of Socrates’
interlocutors in the Phaedo, that, even if the theory is true, this is
all it shows; the soul has existed in separation from the body and so
could do so again, but we cannot tell that it will live forever (Phd.
77b).
105
plato
In response to this, Socrates produces two other arguments for

immortality that are worth looking at. The first (78b ff.) turns on the
similarity between the soul and the Forms. The argument is complex,
but the central point seems to be that both soul and Forms are ever-
lasting because they are simple or incomposite, that is, they do not
have parts. It is not unreasonable to think that what is composite is
likely to perish and what is incomposite is likely to last for ever, for
destruction can be seen as the dissolution of a thing into its compo-
nent parts; this, rather than sheer disappearance into nothingness,
is what destruction normally comes to in our experience.
This conception of destruction, with the consequence that what
is composite is destructible, does indeed seem to have been a per-
sistent feature of Plato’s thought. But it can be used to support the
soul’s immortality only if the soul is seen as simple, and it seems
that Plato did not persist with this way of thinking. In the Republic
the soul is seen as composite, since it has non-rational parts, but the
rational soul still seems to be simple; however in the Timaeus even
the rational element in the soul is composite, having one part that
contemplates Forms, another that makes perceptual judgements (Ti.
36d ff.).
The other argument in the Phaedo, which comes towards the end
of the dialogue, is an extremely complex one, involving the theory
of Forms and the nature of explanation, but at the heart of it there
seems to be a simple point: the soul is what produces life in us, hence
it should itself be seen as essentially alive, something that cannot die.3
This depends on a conception of causation that has a long history
in philosophy, but is often seen as problematic: a thing produces a
quality in something else by passing on that quality to it, and there-
fore must itself have that quality. This clearly fits some examples of
causation better than others. The example that most clearly illustrates
it – and which Socrates does indeed mention in this passage – is
that of heat. Fire is essentially something hot, and is also the cause
of heat in other things. It does not fit other cases so well: the stone
that breaks a window need not itself be broken.
But in any case, even if we accept this argument in relation to the
soul, it does not give us the conclusion that Socrates wants. Fire is
106
the soul
essentially hot in that it cannot continue to exist while being cold.

If soul is a parallel case, it is essentially alive in that it cannot con-
tinue to exist while being dead. This claim, if we can make sense of
it, is not implausible; it is hard to imagine a lifeless soul, as we can
a lifeless body. But still fire can be quenched, and so it would seem
that soul might still be extinguished. Socrates seems on the point of
recognizing this, but in the end insists that whatever is deathless (or
immortal) is also indestructible (Phd. 106d). This is true if “death-
less” has its normal sense; but within the context of this argument it
means only something that is essentially alive, that cannot exist while
being dead. Hence this argument, like the previous one, cannot be
seen as an effective proof of immortality.
Arguments for immortality: the Phaedrus and Republic
In the Phaedrus (245c), Socrates identifies the soul with “what moves
itself ”, and claims that what moves itself, “because it never lets go
of itself, never ceases to move”. It is not wholly clear whether this
is meant as an argument for immortality or just as an explanation
of it. But it can be read as an argument, similar in spirit to the last
one in the Phaedo, but perhaps somewhat stronger. While in the
Phaedo the soul was seen as essentially living and a source of life,
here it is seen more specifically as essentially moving and a source
of motion. The thought seems to be that bodily things will cease to
move when they lose their source of motion; but the soul, being its
own source of motion, has no reason to cease to move. This is not,
of course, conclusive; even if it is its own source of motion some-
thing might extinguish it and so bring its motion to an end. But it
has some inclining force; the obvious reason for death in the case of
bodies is not present in the case of the soul, leaving it unclear why
it should perish.
There is another, more complex, argument for immortality in
the same passage of the Phaedrus, also turning on the idea of self-
motion (Phdr. 245c–e). Socrates begins by affirming that what moves
itself is the source of all motion. This assumption is in fact open to
107
plato
question. It seems true that if we can trace a series of motions to a

source, that must be a self-mover. Clearly something that is (in all
respects) moved from outside cannot be the source of a new series
of motions; and, on Plato’s assumptions, an unmoving thing cannot
cause motion. But it is not obvious that every series of motions must
have a source. Could it not just go back infinitely into the past?
In any case Socrates, having affirmed that what moves itself is
the source of all motion, draws from this the consequence that it is
ungenerated; for if it were generated, the thing that generates, not it,
would be the source of its motion. From this, in turn, he concludes
that it is immortal, since if the source of all motion were to perish it
could not be generated again, and the universe would come to a stop.
These arguments have some force; but the most they can show is that
soul, the genus, is ungenerated and immortal. As we have seen, an
individual soul might move itself in one respect while being moved
from outside in another; hence it might be generated by something
else, although once it exists it sustains and directs its own motion.
But the genus soul, if it is the source of all motion, cannot be moved
from outside in any way; hence indeed it cannot be generated by
anything else. Likewise, an individual soul could perish, and might
be replaced, since there would be other souls still available to act as
generators; but the genus soul, being the source of all motion, could
never be replaced if it perished as a whole. If this is (as it seems to be
in context) an argument for the immortality of the individual soul,
then, like other such arguments, it fails.
Yet another argument for immortality is found in the Republic
(609a ff.). Most things are destroyed by their own defects: the human
body by disease, wood by rot, metal by rust and so on. But the defects
of the soul – ignorance and vice – do not seem to destroy it, so there
seems to be no reason why it should perish. It is not clear why bodily
defects should destroy the soul, except perhaps by causing defects
in the soul, which they do not appear to do. (Plato is, of course,
assuming that the soul can exist independently of the body; if this
is so, destruction of the body will not of itself cause the soul to per-
ish.) Once again this argument has some inclining force; it suggests
that there is no reason to think that the soul will be destroyed, since
108
the soul
the obvious reasons for perishing are not present in its case. But it
is not conclusive.
It is striking that in works that are thought to belong to Plato’s
later years, he moves away from arguments for immortality. In the
Timaeus there is an episode where God, the creator, says that every-
thing that is generated (which in the Timaeus includes the soul) can
be destroyed, but in fact things made directly by him (which include
at least the rational part of the human soul) will not perish, but be
preserved in being by his will (Ti. 41a ff.). While in the immediate
context this refers to the stars, it can also be applied to the human
soul, suggesting that Plato now sees the immortality of the soul not
as part of its nature, which can be demonstrated by reflection on its
nature, but as a divine gift. In the Laws, although he makes use of
the idea of immortality in defending the goodness of the gods, he
does not argue for it (although he has argued for the existence of
the gods and the fact that they are concerned about us); the part of
the dialogue in which immortality figures is called a “charm” rather
than a logical argument (Leg. 903a–b). Finally, in a letter ascribed to
Plato that, if genuine, must come late in his life, we are told that we
should accept immortality on the basis of a sacred tradition (Seventh
Letter 335a).
It seems likely that Plato was never wholly satisfied with his argu-
ments for immortality; it is interesting that in the Phaedo he makes
Socrates say that he is advancing these arguments to convince him-
self when faced with the prospect of death (Phd. 91a). While belief
in immortality was a more or less constant feature of his thought,
he never found a truly satisfactory argument for it, and it is not sur-
prising that later in his life it became for him more a matter of faith.
Why immortality matters
Plato’s position on why immortality is important also seems to shift.

In the Phaedo and Phaedrus the soul’s true destiny is the contempla-
tion of Forms, and this is best achieved outside the body because of
the distracting effects of perception and bodily desire; being in the
109
plato
body is a misfortune (Phd. 64d ff.). In the Phaedrus it is compared to

being in a tomb or in prison (Phdr. 250c), and the tomb metaphor
is also found in the Gorgias (Grg. 493a). We should, according to
these works, look forward to death; in the Phaedo philosophy is seen
as practice for death (Phd. 64a, 67c). What a philosopher achieves
after death is not simply a reward for a virtuous life, but the actual
fulfilment of the purpose for which he has been living.
In other dialogues, however, our present life is seen in a more
positive way. In the Republic the principal benefit of virtue, includ-
ing contemplation of Forms, can be achieved in this life. Only in
the last book, when Socrates goes on to deal with the additional
rewards of virtue, which it gets if it is recognized both by human
beings and by the gods, does he introduce the topic of immortal-
ity (see esp. Resp. 612b–c). Part of the significance of the afterlife
is that in it we make choices for our next life on earth (619b ff.).
It is true that the prison metaphor is used in the Republic, in the
famous analogy of the cave (514a ff.), but here escape from the cave
does not represent death but only conversion to a philosophical
life; the philosophical ruler is envisaged as returning to the cave
– that is, taking part in the day-to-day business of government –
after contemplating the realities in the world outside. Likewise in
the Timaeus the soul has a purpose within this world, helping to
complete the scheme of creation.
Even in these dialogues immortality remains significant, but for
a different reason; it helps to vindicate the claim that the world is
well ordered, and so defend the goodness of its rulers, the gods.
This has two aspects. On the one hand, it is simply, as the Timaeus
makes clear, bad for something well ordered to perish (Ti. 41b). On
the other hand, life after death enables the virtuous to be rewarded
and the guilty punished, which does not always happen in this life.
While the truest reward of virtue is an internal one, it is desirable
that the virtuous should also receive external rewards, if the gods
are just. This emphasis on reward and punishment is found in both
the Republic and the Laws (Resp. 612e ff.; Leg. 903b ff.).
110
the soul
The divided soul
Another important aspect of Plato’s thought about the soul is his

account of the division of the soul into three parts or elements, com-
monly referred to as the rational, the spirited and the appetitive or
desiring parts. This division is not to be found in every work of his
where the soul is considered – as we have seen, in the Phaedo non-
rational elements are assigned to the body – but it seems to have
become a fairly constant element in his thought, found in the Phae-
drus (246a ff.), Republic (435c ff.) and Timaeus (69c ff.) (although it
is not clear that his conception of the three parts is exactly the same
in all these works).
Plato seems to be motivated to divide the soul in this way by his
aim, in the Republic, of linking ethics and politics; he wants there
to be parts of the soul that are parallel to the classes he identifies in
the ideal state, so that the harmony between these parts can con-
stitute virtue in the soul in the same way that harmony between
classes constitutes virtue in the state. The three classes in the state
are rulers (later identified as philosophers), warriors and working
people (farmers and craftsmen); reason, spirit and appetite are the
corresponding part of the soul. However, it is possible that Plato’s
division of both state and soul was inspired by an earlier distinction
– thought to have been current before Plato’s time, and sometimes
ascribed to Pythagoras – between three kinds of life that people can
live, governed respectively by the love of gain, the love of honour
(ambition) and the love of wisdom (philosophy) (see Resp. 581c).
The rational element in the soul includes the actual power of rea-
soning; it is sometimes called the calculative part (439d). This, as
the Timaeus makes clear, includes both philosophical thought about
Forms and more mundane reasoning about perceptible things (Ti.
36d ff.; cf. 41d ff.). It also includes rational desires. At first, when this
part is introduced in the Republic, this means the desire for what is
good for the person as a whole, which it is the job of reason to discern
(Resp. 441e), and this aspect of it is not abandoned. Near the end of
the work we are told that the lower parts, those that aim at honour
and money, will also achieve their proper pleasures most effectively if
111
plato
guided by reason (586d–e). But later this part also comes to include
the philosophical desire for knowledge or truth (581b).
The “appetitive” part is said at one point to be extremely various
(580d), and is compared to a many-headed monster (588c ff.), but it
is regarded as primarily the seat of bodily desires, such as those for
food and drink, and sexual desire. It is also seen as the seat of the
love of money, which Socrates claims is sought primarily in order
to gratify these desires (581a). In this way this part can be linked
with the life dominated by love of gain, and with the money-making
class in the city.
The third part, the “spirited”, which is seen as intermediate between
these two, is more puzzling. The word translated “spirit”, thumos, can
also mean “anger”, and when Socrates first distinguishes this element
from reason and appetite it is indeed of anger that he is thinking
(439e ff.). However, it can also be the seat of pride in one’s own good
deeds (553d) and shame at one’s bad ones (Phdr. 254a ff.),4 of respect
and admiration for the achievements of others (Resp. 553d), of the
love of honour (ambition) and the love of victory (581a–b), and it is
characterized by a natural aggressiveness that, when it is governed
by reason, also makes it the seat of courage (442b–c). It might per-
haps be described as the self-assertive element; it brings together a
number of mental activities that, while not exactly rational – they
are certainly found in people not much given to reasoning – also do
not arise from simple desires, but rather from a conception of the
self. It is seen as dominant in the life of ambition, in the warriors
of the ideal state, and also in actual states, such as Sparta in ancient
Greece, which are governed by a military ideal.
The rationale of the division
It will be clear at once that this is not, in the first instance, a divi-
sion of the soul into faculties, of the kind that has been common in
philosophy ever since Aristotle; the rational part is not the faculty
of reasoning, nor the appetitive part the faculty of desiring. Rather,
it is in the first place a division into sources of motivation, which
112
the soul
can oppose one another; it is introduced, in the Republic, to make

sense of the phenomenon of mental conflict. Hence the distinction
of reason and appetite is nothing like Hume’s distinction of reason
and passion; what Plato calls reason has passions of its own. (It is
sometimes suggested that the second function of Plato’s rational part,
pursuing what is good for the person as a whole, is similar to the
Humean picture of reason [see Annas 1981: 133], but this is not
clearly so; pursuing the interest of the whole need not, for Plato, be
reducible to working out how to satisfy the desires of the various
parts; it can still be a source of motivation in its own right.)
On the other hand, we should not suppose that the division cov-
ers only kinds of motivation; other mental states are also assigned
to the various parts. Most obviously, the actual process of reasoning
belongs to the rational part, while sensations of pleasure and pain are
assigned to the appetitive part (Ti. 77b). On the other hand, it is not
clear where Plato would locate all mental states that he recognizes;
emotional states, in particular, cause a problem. While many emo-
tions, such as anger, pride, shame and respect, are naturally located
in the spirited part, this is less clear with others, such as pity and fear.
What is meant by the expression “part” of the soul? It certainly
means more than just an aspect. When Socrates argues for the divi-
sion of the soul (Resp. 436a ff.), he means more than that it can
engage in different activities, for he accepts this at the beginning of
the argument, but then goes to some lengths to show that it performs
these different activities with different parts. The main ground for
calling them different parts is that they can oppose one another;
reason can oppose appetite, as can anger, and reason can also restrain
anger. Socrates argues that one unified thing cannot possess opposite
attributes at the same time, so if there are two conflicting motives
in the soul it must be divided. One may question whether this argu-
ment strictly works in logical terms, in particular whether conflicting
motives really are opposite properties in quite the way that is needed
to make the argument work. But we may think that talk of division
makes sense simply as a metaphor for the possibility of conflict;
when different sources of motivation in the soul conflict, they are
in a way like different people engaged in a dispute.
113
plato
On the other hand, it is also possible that Plato understood the

concept of parts in a more literal way. Nowadays we tend to take
it for granted that if the soul is not bodily it is also not located in
space, so cannot have parts in the way that a material object has.
Plato, however, quite often speaks of souls as having spatial loca-
tion and performing spatial motions;5 it seems he sees the distinc-
tion between body and soul as depending on whether something
is perceptible, rather than on whether it is located in space. In the
Timaeus the parts of the soul are given locations, the rational part
being in the head, the spirited in the chest, and the appetitive in the
belly (Ti. 69d–70a).
One may wonder why, if the point of the division is to allow for
mental conflict, Plato recognizes only three parts. Is not conflict also
possible within these three broadly defined parts of the soul: between
two bodily appetites, say, or between anger and pride? Plato seems
to recognize this at least in the case of appetite; the later parts of the
Republic, which describe degenerate souls, do show conflict taking
place between the appetites. The image of the appetitive part as a
many-headed monster may be seen as pointing to this, the various
heads representing the possibly conflicting motives within the appe-
titive part. In addition, in the Republic Socrates does at one point
suggest that there may be more than three parts to the soul when he
compares harmony in the soul to a musical harmony between “high,
middle and low notes and however many there may be in between”
(Resp. 443d, emphasis added). It may be that, while Plato is confident
of the three parts he names as constant features of human nature, he
does not really want to rule out the possibility of further division.
Mental conflict
The point of the divided soul is to allow for the possibility of mental
conflict. Indeed, even in the Phaedo, where the soul is not divided,
what we would call mental conflict is accepted but is treated as con-
flict between soul and body. In other dialogues, perhaps more intui-
tively, the conflicting elements are all located within the soul.
114
the soul
This view seems to represent a step away from that found in the
Socratic dialogues. According to Socrates there, all desire is for the
good; hence, if I pursue something it must be because I judge it to be
good. It is not clear that this way of thinking can really accommodate
the phenomenon of psychological conflict. The later dialogues, by
contrast, allow different springs of motivation, not directed to the
good as such, but to such things as honour and pleasure, so making
conflict possible.
This also helps us to deal with the puzzling phenomenon of
akrasia (variously translated “incontinence”, “weakness of will” and
“lack of self control”): of acting in a way that is contrary to one’s
better judgement, overcome by desire, anger or the like. Socrates’
view as found in the Socratic dialogues seems to rule this out, and
indeed in the Protagoras he does deny the possibility (Prt. 352b ff.).
If all desire is for the good, we can only pursue something that is
not actually good if we have made a mistake about the good. The
explanation of wrongdoing, therefore, is ignorance; knowledge of
the good secures good action, and can be identified with virtue. The
theory of the Republic, by contrast, allows akrasia; there are springs
of action other than reason, and we can be overcome by them. (In
the Laws [863a ff.], we are explicitly told that there are three causes
of wrongdoing: ignorance, thumos – anger or spirit – and pleasure.)
However, while Plato certainly seems to believe that one can act
contrary to a specific rational judgement, it does not follow that for
him a truly wise person could act wrongly; it may be that, for him,
to become truly wise we need a desire for knowledge or truth that is
able to overcome distracting desires. In this case, wisdom can still
be seen as bound up with the other virtues.6
Problems of the divided soul
Plato’s account of the divided soul leaves us with two problems. First,
while it clearly accounts for mental conflict, can it really make sense
of mental harmony? Plato wants us to think that a virtuous soul is
in a state of harmony,7 with the parts in agreement under the rule of
115
plato
reason; but if their aims are so different, can they really agree? Here,
there seems to be a difference between the two lower parts of the
soul. Plato sees the spirited element as a natural ally of reason (Resp.
440b, 441a), and this seems plausible. Our feelings of pride, shame,
admiration, indignation and so on are often based on judgements
about what is right or wrong, good or bad, so are indeed responsive
to reasoning, although they do not always respond immediately, so
that conflict is still possible. The appetitive element, on the other
hand, seems to have objects of its own – bodily pleasures – that
do not depend on reasoning, and that it has such aims seems to be
essential to its identity as a part. So can it really be in agreement
with reason?
It may be that all that Plato can really defend here is not genu-
ine agreement, but only coincidence of aim; we might educate our
appetites so that they aim at things that reason judges are good for
us – things necessary for survival, and harmless pleasures – although
they are not aiming at those things because they are good. If this is
right, it seems that Plato’s analogy between a virtuous soul and a
well-ordered state is not perfect, since for him the lower classes in
a well-ordered state do have the power of reasoning and are able
actually to agree that the decisions of their rulers are good for them
(see Resp. 431d–e).
The other problem posed by Plato’s theory of the divided soul is
whether it can make sense of the idea of choice, and of an action that
is genuinely one’s own rather than that of some element within one.
The picture suggested by Plato’s account is that when we act ration-
ally, our reason is in control; when we act irrationally it is overcome
by one of the other elements. There seems no place for an act by
which we decide between the rational and irrational aims. In what
part of the soul might that decision be located?
Plato does sometimes write as if the self were something distinct
from the various elements, which chooses between them (see e.g.
Resp. 588e ff.). But often he seems prepared to accept the implications
of the theory of the divided soul. For many ancient thinkers, the true
self can be identified with the reason, and what we choose is what
seems good to us; if we pursue something that is actually bad, this
116
the soul
is either because we mistakenly think it is good, or because we have

been overcome by something, for instance appetite or anger, which
prevents us acting in accordance with our choice. Neither of these is
ultimately under our control. It seems that, much of the time, Plato
shared this way of thinking. Even in the Timaeus we are told that “no
one is willingly bad” (Ti. 86d–e), and in the Laws that “no one acts
unjustly except against his will” (Leg. 860d). These statements echo
the claim found in the Socratic dialogues that no one does wrong
willingly, but their significance has changed; for Socrates wrongdo-
ing was always caused by ignorance, whereas for the later Plato it
may be caused either by ignorance or by akrasia, being overcome by
anger or appetite. What is still excluded is the conscious villain, the
person who chooses to do what is bad, knowing it is bad. But while
this idea has some intuitive appeal, it is indeed hard to make sense
of, and Plato is not alone in the ancient world in rejecting it.
But even if, for Plato, we are not ultimately responsible for our
bad deeds, which are caused by something other than our true self, it
does not follow that our decisions are of no significance or that there
is no point in giving ethical advice (as Plato’s speakers, of course,
frequently do). This would be so only if our actions were caused
by something else in a way that bypassed our decisions. But this
is not so: our decisions do have an effect on action. In the case of
wrongdoing caused by ignorance, it is clear that our decision leads
to action; the problem is that we do not possess the knowledge that
would enable us to make the right decision, and here ethical advice
can directly help us. In the case of wrongdoing caused by anger or
appetite, our rational part has no immediate control over our action;
but our desires can be educated, so that their strength and direction
depends on decisions made by our rational part. Hence, this kind
of wrongdoing may in part be due to decisions made by the rational
part in the past. In any case, reason has work to do in educating our
desires, and ethical advice can be relevant here.
117
six
Politics
Plato’s views on practical issues form a central part of his thought.

They embrace both questions about the good of the individual – eth-
ics – and about that of the community – politics; and any attempt to
separate these is rather artificial. His most significant work in this
field, the Republic, is concerned with both, and they are closely inter-
twined. Nevertheless, I shall focus on his political views first, before
going on to discuss his ethical views in Chapter 7. I take them in this
order simply because, in the Republic, Plato’s Socrates expounds his
views on the ideal state first, before turning his attention to the indi-
vidual, and it seems appropriate to follow this order of exposition.
This chapter will be concerned primarily with views put forward
in the Republic, but will go on to consider Plato’s later works in the
same field, the Statesman (sometimes called the Politicus, from the
Greek word for statesman) and the Laws.
Plato’s political views are perhaps the most controversial aspect
of his thought. His picture of the ideal state has struck readers, from
Aristotle onwards, as unattractive; few would want to live there. This
is partly because of some features it has in common with other Greek
city states; indeed, in some respects, notably in connection with the
position of women, Plato does seem to progress beyond the ordinary
attitudes of his day. But other unattractive features of the ideal state
are distinctive of Plato. Nevertheless it is worth looking to see how
119
plato
Plato reaches his conception of the ideal state, and what kind of
principles he uses to justify it.
We saw in Chapter 1 that there is some uncertainty about the
extent to which Plato sees his ideal state as a practical proposal. But
he certainly sees it as a genuine ideal: a state that would be worth
living in if only it could be achieved, not just as a way of illustrat-
ing moral truths at the individual level (although it is that as well).
Later, in the Laws, where he has quite definitely given up any hope
of achieving the ideal state, and is more concerned with what can
be achieved in practice, a state resembling that of the Republic is still
treated as an ideal (Leg. 739c ff.); the state actually being planned in
the Laws is seen as second best. We must therefore take the proposals
of the Republic seriously as a reflection of Plato’s views (although,
as always, this does not mean that he must be seen as committed to
all of them in detail).
The plan of the Republic
As the Republic will be our main source for Plato’s ideas in the next
two chapters, we should begin with a brief outline of its arguments.
They are concerned primarily with two questions: what justice
is, and whether and in what way justice is beneficial to the just
person.
The first book, which is written in the style of the Socratic dia-
logues, shows Socrates discussing the nature of justice with two other
speakers: Polemarchus, who defends a conventional view of justice
(Resp. 331e ff.), and Thrasymachus, who holds that justice is simply
the result of laws imposed by rulers, and that it is neither advanta-
geous nor a virtue (336b ff.). Socrates succeeds in arguing Thrasy-
machus into a corner. However, in the next book two new speakers,
Glaucon and Adeimantus (Plato’s brothers), claim that Socrates’
arguments have not produced real conviction, and ask for a more
careful consideration of the issues, challenging Socrates to show that
justice is worth having, for its own sake, not just for the sake of
the rewards and reputation it brings (357a ff.). Socrates accepts this
120
politics
challenge, but, in a way typical of the Socratic dialogues, begins the

enquiry by considering what justice is.
Socrates points out that justice can be found both in a state or
community, and in an individual soul, and suggests that justice in
the state may be easier to discern (368d ff.); this leads into his well-
known sketch of an ideal state, which gives the work its name.1 This
lasts from the middle of Book 2 to that of Book 4. Having determined
what justice, and the other virtues, are when they are manifested
in the state (427e ff.), he goes on in the last part of Book 4 to give a
parallel account of the virtues in the soul (434d ff.). Book 4 ends with
the discovery of the answer to the first question, the nature of justice.
At this point we enter what is initially presented as a digression,
although in fact it contains much material relevant to the main
questions. Beginning with a consideration of the place of women in
the ideal state (449c ff.), Socrates moves on to consider whether the
ideal state is really achievable. The answer turns out to be that it is,
but only if philosophers are entrusted with the task of government
(473c ff.). This leads to a discussion of the nature of philosophy, and
of philosophical education. (Some aspects of this discussion have
been considered above in Chapter 3 and 4.) This whole discussion
occupies Books 5–7.
In Book 8 the main line of argument is resumed. Having described
a just society and a parallel just soul in Book 4, Socrates now describes
various forms of unjust societies and parallel unjust souls. Finally, in
the last part of Book 9, he compares just and unjust lives and gives
reasons to think that the just life is happier (576c ff.). Book 10 is
something of an appendix. It deals with two main themes, the place
of poetry in the ideal state (to be discussed in Chapter 9), and the
external rewards of justice, including those achieved after death, and
includes a discussion of the immortality of the soul.
The ideal state
As was usual in ancient Greece, this community is envisaged as a

city state. The central principle on which Socrates’ account of it turns
121
plato
is that each person should perform a distinct function. This is first

introduced at 369 ff. as something we need in order to have a state or
community at all; rather than each person being self-sufficient, peo-
ple have different jobs and exchange goods and services, depending
on one another. But Socrates quickly turns this into a normative prin-
ciple. First, he insists that each person do the job for which he is best
qualified (370c). Then he applies it in ways that would not normally
be accepted in ancient Greece: first to the army (374a ff.), and then
to rulers (412b ff.). Normally in Greece all citizens were expected
to do military service. As for ruling, this was indeed restricted in
some states: whereas in a democracy all (free male) citizens took
part in government, in an oligarchy the opportunity to do so would
be restricted on the basis of wealth, and in the Spartan system on
the basis of completion of a programme of military training. (The
Spartan system was commonly seen as an oligarchy – which it cer-
tainly was in the simple sense of government by a restricted group
– but Plato regards it as something separate from the wealth-based
oligarchies, giving it the distinctive name “timocracy” [545b].) In
any case, Socrates establishes what is probably a much smaller class
of rulers than any actual Greek state had; and, quite exceptionally,
he sees it as a job to be done on the basis of expertise.
Socrates refers to the army as “guardians” of the state (374e), and
since the governing class is chosen from among the army, the term
also applies to them. Sometimes, however, he restricts the term
“guardian” to the rulers, referring to the soldiers as “auxiliaries”
(414b). Many of his provisions relate specifically to these two classes.
At 375 Socrates says the guardians should be both spirited and
philosophical – although “philosophical” is here used in a rather
attenuated sense, referring to a sensitivity that enables them to dis-
tinguish friends and enemies. He then gives a lengthy description
of the education – in gymnastics and “music” (including poetry)
– which is suitable for the guardians (376d ff.). (Part of this will be
discussed in Chapter 9.)
Socrates proposes that the division into classes be justified to the
citizens by means of a myth – the so-called “noble lie”2 – according
to which the people of the ideal state originally rose out of the earth,
122
politics
making the land their mother, and so are all kinsfolk, but nevertheless
are of fundamentally different natures, the rulers having gold in their
bodies, the warriors silver, and the craftsmen and farmers bronze and
iron (414d ff.). This may well give the impression that the qualities
that determine what class one belongs to are inherited, and indeed
Socrates says that this is normally the case. However, he insists that
it is not so invariably; a child fitted for one class may be born to par-
ents of another (415b–c). Members of the classes are shown as being
selected on the basis of ability (for the warriors, 374e ff., 423c–d; for
the rulers, 412b ff., 537a ff.). Warriors would need to be recruited at
an early age, as their education is strictly regulated, but the charac-
teristics desired in them, being spirited and “philosophical”, swift and
strong, are perhaps visible at an early age. The qualities required in
rulers – knowledge, aptitude for learning and love of the city – do not
become apparent so early, but this is not a problem as the training of
rulers does not begin until they are twenty.
Socrates argues that the guardians should not have private prop-
erty, or even private dwellings, but should live like soldiers in a camp
and be supported by the rest of the population (who, as farmers
and craftsmen, sometimes referred to as “money-makers”, do have
property of their own) (416d ff.).
Later, returning to the subject in Book 5, he argues – very surpris-
ingly in ancient Greece – that some women should be guardians,
since everyone should do the job for which they are best fitted, and
some women will have the skills needed for fighting and government
(451c ff.). Although this is one of Plato’s most striking proposals –
and one of the most controversial in his time – we should note that
it is very restricted in scope; it applies only to the two upper classes,
and women of the farmer and artisan class are not given equality in
the same way. Moreover, Plato’s speakers often take a rather negative
view of women; even in this passage, while insisting that individual
women may be better qualified than some men, and so should not
be excluded from fighting and government, Socrates is still made to
claim that women are on average less able.
He also introduces the notorious “community of women and chil-
dren”: the proposal that there should be no permanent marriages
123
plato
(among the guardians), but arranged pairings at special festivals, and

that children should be brought up in state nurseries, not recognizing
any particular people as their parents (457c ff.).
Unity in the state
In all this there is an emphasis on the unity of the state. The principle
of specialization is taken to promote this (Resp. 423d), presumably
because it makes all parts of the state dependent on one another.
Extremes of wealth and poverty must be avoided, as they tend to
divide the city, and the city must not get too large, as it will then
break up into separate groups (421d ff.).
Socrates also especially emphasizes the unity of the guardian class,
which he thinks will promote unity in the state as a whole (465b). The
absence of private property is thought to promote this, by removing
causes of quarrelling (418a–b), and the “community of women and
children” has the same effect, by making people see one another as
relations (462b ff.). Socrates sees unity as the greatest good for the
state, and indeed as necessary if it is to be a state at all. A state divided
against itself is not one state but a collection of states (422e–3b).
Socrates aims at a state where everyone is concerned about each
other, and sees their interests as coinciding; just as the whole body
is pained when one finger is injured, so the whole state should be
pained when one of its members is injured (462c–d). The rulers
should be those who most deeply identify their interest with that of
the community (412d–e).
Adeimantus asks whether the guardians, living in a camp without
private property, will be happy (419a). Socrates answers that what
matters is not whether they are happy but whether they contribute
to the happiness of the community. But later he says that they will
be happy, because they are honoured and supported by the rest of
the community, who owe their safety to them (465d ff.).
The claim that it is the happiness of the whole state that matters
has sometimes been taken to show a lack of concern for the hap-
piness of individuals. It may seem that Socrates is here treating the
124
politics
state as an organic entity with a happiness of its own, distinct from

that of its members, a thought encouraged by the fact that he draws
an analogy between the state and the individual soul. However, he
need not be read this way; he may mean only that it is the happiness
of all the citizens, not just of one group, which matters, and it is not
unreasonable to ask one group to suffer a reduction of happiness if
others are benefited by it.3
It is not wholly absurd to see the system Socrates proposes as tend-
ing to make all the citizens happy, although one has to make some
rather strong assumptions about human nature to justify this view.
The ordinary people are happy because they are wisely governed
and effectively protected; the rulers and warriors are happy because
they are supported by the people and honoured for their role in
governing and protecting them. Each group gains satisfaction both
from the benefits that other groups give them and from performing
their own part well.
Philosopher rulers
One final provision does not emerge in Socrates’ initial descrip-

tion of the ideal state, but only later, in Book 5, when the speakers
are considering whether the ideal state is truly possible. It is that the
rulers of the state must be philosophers (Resp. 473d ff.), not simply
in the sense in which it was earlier argued that the guardians must
be “philosophical”, having the sensitivity that enables them to distin-
guish friends from enemies, but in the sense of having philosophical
knowledge and training, including, as becomes clear in due course,
knowledge of Forms. Initially this is mentioned as a condition for the
ideal state to come into being in the first place; but later we find that
the philosopher rulers are to be a permanent presence in the state,
with much of the later part of the work given up to a description of
their training (535a ff.).
That rulers should be philosophers follows naturally from two
points. The first is that rulers should be experts: that ruling is a skill,
to be assigned to those best qualified for it, those with the relevant
125
plato
knowledge. This is already implicit in the earlier description of the

ideal state, from the point where the principle of specialization was
made to apply to rulers. The other is that the knowledge relevant to
ruling is philosophical knowledge; and this is a reasonable claim if
we think that the central object of philosophical knowledge is the
good, and that things in general – including states – are in a better
condition the more closely they approximate to an ideal pattern. A
central passage for understanding the work of the philosopher rulers
is 500b ff., where they are described looking at the pattern of Forms
and trying to produce an image of it in the state.
Virtue in the state
At the end of his first description of the ideal state, Socrates asks
where in the state he has described its virtue may be located. He
argues that, as a well-ordered state, it must possess all the virtues, and
assumes that the virtues are those included in what was to become
a traditional list, later called the cardinal virtues: wisdom, courage,
temperance and justice (Resp. 427e). (The Republic seems to be the
first place where this list is used in a serious context, although the
Symposium, probably earlier, uses it in a humorous context in the
speech of the poet Agathon [Symp. 196b ff.]. However, the choice of
topics for the Socratic dialogues show that these virtues were central
concerns of philosophers at the time; the Protagoras lists these four
together with holiness [Prt. 329c–d], but the Euthyphro suggests that
holiness is a part of justice [Euthphr. 12d].)
It seems clear that the wisdom of a state is located in its rulers
(Resp. 428d), the courage in its army (429b). This does not mean
that no one but a ruler can be wise, no one but a soldier brave, but it
is on account of the rulers that we say that the state is wise, and on
account of the army that we say the state is brave.
In his account of temperance (431e ff.), Socrates relies on the
widespread (although not universal) identification of temperance
with self-control. He points out that this term is actually paradoxical;
it does not make sense to speak of a person (as a whole) controlling
126
politics
himself (as a whole). What it means is that the better – the rational
– element in us controls the worse elements, the unruly desires.
Likewise a state, he suggests, can be called temperate when rational
people are in control, and the rest of the population agree that they
should rule. (This reference to agreement is important; it means that
the population is not kept down by force; and it allows the virtue of
temperance to be grounded in the lives of the whole population, not,
as with the two previous virtues, just one class.)
Finally we come to justice. This, rather surprisingly, turns out
to be identical with the principle with which we began: that each
citizen should “perform his own function”, meaning not just that
there should be a division of labour, but that everyone should do
the work for which he is best suited (433a ff.). The reasons given for
this identification are, at first sight, rather weak. The main one seems
to be simply that it is a virtue, as it makes an important contribu-
tion to the benefit of the state, and as it is not identical with any of
the other virtues it must be justice, which is the only one left over.
This obviously turns on the acceptance of the list of cardinal virtues,
which one might think is by no means obvious, especially if we are
engaged in giving revisionary accounts of just what the individual
virtues are. Socrates also says that, in ordinary speech, we often say
that justice is doing one’s own work and not intruding in that of oth-
ers, but here the phrase seems to have a rather different meaning: it
does not mean doing the work for which you are best qualified, but
rather sticking to matters that concern you or have been assigned
to you – in the modern idiom “minding one’s own business”. In any
case, whether or not the observance of this principle deserves the
name of justice, it is identified as an important virtue, central to the
preservation of the state.
Problems of the ideal state
Probably the feature of the ideal state that will first strike most
modern readers as problematic is the exclusion of large numbers of
citizens from the opportunity to take part in government. However,
127
plato
while Socrates in the Republic takes this to extremes, it was quite a

common feature of Greek states; Plato is by no means alone in his
rejection of democracy. Moreover, the ordinary people of Socrates’
state – the farmers and craftsmen – seem to live what would gener-
ally have been seen as quite a normal life, with little government
interference on a day-to-day basis; they are allowed to own prop-
erty (although extremes of wealth and poverty must be avoided), to
marry and have children of their own.
Much more striking are the restrictions placed on the life of the
guardian class. It is notable that it is in connection with them, rather
than the farmers and craftsmen, that Adeimantus worries whether
they will be happy. First, there is the fact that they cannot own prop-
erty (including homes) of their own, but are wholly dependent on
the rest of the state for support. Secondly there is the “community of
women and children”; quite apart from the unreasonable restrictions
this places on personal life, Aristotle was surely right to say that it
would not achieve its intended purpose (Ar. Pol. 1261a10 ff.). If we
regard everyone in our community as a relation, the ties of kinship
will be weakened; we will not come to regard everyone in the state
as we currently regard our parents, children or siblings, and the aim
of promoting unity will not be achieved.
Perhaps the most disturbing feature of Socrates’ ideal state, how-
ever, is simply the absence of personal choice. There are three pos-
sible ways of life, to which citizens are assigned on the basis of their
aptitudes, and then expected to live the kind of life appropriate to
their class, which, at least in the case of the ruler and warrior classes,
is quite rigidly determined. There is an assumption here that if some-
thing is beneficial to the state, government has the right to require it.
We should be careful here; to say that Socrates does not leave
room for personal choice is not to say that he favours coercion. On
the contrary, it is clear that he sees the citizens of the ideal state as
consenting to their position. This is shown both in the description
of temperance, where it is emphasized that rulers and ruled are in
agreement about who should rule, and in various passages dealing
with the unity of the city, and the way the various elements in it
rely on one another (see e.g Resp. 462b ff., 465d ff.). Socrates does
128
politics
sometimes use the language of compulsion,4 but this need not mean
that he actually imagines coercive measures being taken; it may sim-
ply mean that certain things are required of the citizens, with the
assumption that they will in fact agree. Indeed, he will later speak of
the rulers themselves being compelled to take part in government
(520a), but clearly they could not be physically forced to do so, and
in fact it is made clear that they consent. Socrates sees tyranny, which
does involve coercion, as the very opposite of the kind of state he
is describing.
We are likely, in imagining the ideal state, to think of what would
be necessary to achieve such a system of government in the actual
world; and this would no doubt require rigorous policing, since real
people are not, on the whole, like the citizens of Socrates’ state, and
would have to be compelled to fit into it. But according to the account
given by Socrates, if such policing were necessary the state would
already have failed. The state as he describes it probably involves less
coercion than most states of the time. It is striking that judicial mat-
ters and punishment are hardly mentioned in the sketch of the ideal
state (by contrast with the later, more down-to-earth Laws). Plato
recognizes, of course, that most actual people would not accept life
in such a state; if people are going to do so they must be educated
for it. Hence we have the strange suggestion that the ideal state can
be made real only if we start by sending away from the city every-
one over the age of ten (540e–41a). This may be read as an effective
admission that the state is not a practical possibility. But if we could
achieve a community of truly rational people, Socrates seems to be
saying, this is how they would choose to live.
This approach to the ideal state seems to be based on an incredible
faith in the power of reason; both the reason of the rulers, in discern-
ing the best solutions, and the reason of others, in recognizing that
they are the best and consenting to them. What is problematic is
not coercion as such, but the lack of recognition that different ways
of life, different ways of seeing the world, may be valuable. Socrates’
proposals may be seen as making sense, if we suppose not only that
there is a definite, well-defined good to be sought in government, and
that for each of us there is a best way of contributing to that good,
129
plato
but also that people are in a position to recognize this. But all these
assumptions are, of course, open to question.
It is very common in Plato’s dialogues for government to be
treated as an expertise in the same way that, for instance, medicine
or navigation is an expertise. In such cases, there is an agreed goal:
health in medicine, safety at sea in navigation and so on. What dis-
tinguishes experts is their knowledge of how to achieve this goal.
But is there a similar goal in politics? If there is an objective good
for people and states, then in principle there can be; we all want to
attain the good, and some people might be specially skilled at finding
a way to achieve it. If there is not, then the business of government
becomes something different, not finding the way to an agreed goal,
but rather deciding on what goals to aim at. But even if there is an
objective good, it does not follow at once that there are people quali-
fied to find it, and able to convince others that they have done so,
and in the absence of such people government will, in practice, have
to involve deciding on goals, choosing between different conception
of the good.5
Two features of the ideal state are particularly disturbing. One is
the censorship of literature, and of the arts more generally, at first
presented as part of the guardians’ education, but later seen as apply-
ing to the state as a whole. This will be discussed in Chapter 9. The
other is the way in which rulers are allowed to deceive the people. At
389b–d Socrates says that although in general lying is to be avoided,
the rulers in the ideal state may be allowed to lie in the same way that
doctors can. This picks up an earlier passage where we are told that
lies are “a useful drug” that can be used to stop our friends doing
something bad through madness or ignorance (382c). This – even
supposing it is true of doctors – will apply to rulers only if they are
seen as experts taking care of the people from a position of superior
knowledge.
This manifests itself in two places. One, which we have already
commented on, is the “noble lie” used to give citizens a sense of
identity. However, Socrates says that he hopes the rulers themselves
will one day come to believe it. We might, on the contrary, think it
hard to suppose that anyone will believe it; but in that case, one might
130
politics
think that it could serve its purpose just as well if it were regarded
simply as a useful fiction. The other place where lying is authorized,
even more disturbingly, is the lies told to the warrior class as part
of the programme of breeding; they believe that partners are being
assigned to them by lot, but in fact they are being selected in such
a way as to produce the best children (459c ff.). We would in any
case find such a programme disturbing, but more so if it is brought
about by deception. Once again, we have the assumption that rul-
ers, if they are genuine experts, are entitled to do whatever is in the
interest of the state.
The Statesman: a new approach
Later in his career, Plato returns to political themes in the States-

man. While the approach taken there has much in common with the
Republic, it also introduces some distinctive new features. In particu-
lar, it places a new emphasis on the concept of law. The Republic has
little to say on this subject, suggesting at one point that the details
of legislation are unimportant, the wisdom of the rulers being what
matters (Resp. 425c ff.). The Statesman gives a new importance to
the subject.
Government is treated, as in the Republic, as a skill like medicine
or navigation; but while in the Republic the expertise of the rulers
seemed to rest primarily on their knowledge of unchanging Forms,
in the Statesman emphasis is placed on their ability to make decisions
for particular people in particular circumstances (Plt. 293b ff., esp.
294b). Hence, it is argued, a true statesman should not be limited by
law, but left free to make what judgements he finds appropriate in
each situation. Interestingly, it is also claimed that the expert ruler
does not need the people’s consent. However, in the absence of a
true statesman – and it seems to be implied that they are in general
absent from actual states – law is an appropriate second best, acting
as a rough guide to what a true statesman would judge (295a ff.).
Hence while it would be wrong to subject the expert in government
to the requirements of law, it is even worse if a ruler who is not an
131
plato
expert ignores the law in order to benefit himself or do favours to

others (300a–b). It follows that of the various forms of government
that exist in the actual world, law-governed forms are better than
lawless ones; the first group include kingship, aristocracy (perhaps
here meaning the Spartan system) and constitutional democracy,
while the latter include unconstitutional forms of democracy, oli-
garchy and tyranny.
It is hard to say whether the Statesman gives law a higher or a
lower place than it has in the Republic. On the one hand, it seems
that the rulers of the Republic govern according to unchanging prin-
ciples that could well be called law-like, while here the flexibility of
the statesman and his ability to deal with particular circumstances
is emphasized. On the other hand, written law had no importance
in the Republic, since the wisdom of the rulers was the guarantee of
good government. Here, although written law is a second best, it is
taken seriously as being often the best of which we are capable, and
as superior to lawlessness.
Both these developments, however, can be seen as examples of the
tendency, which appears in many of Plato’s later works, to take the
sensible world and ordinary life more seriously. This leads him on
the one hand to emphasize the importance of particular judgement
as well as universal knowledge; this is also found in the Timaeus,
where the rational element in the soul includes not only the part that
contemplates Forms but also the parts that make perceptual judge-
ments (Ti. 37a–c); and in the Philebus, where it is pointed out that
our understanding of universal Forms must be supplemented by the
ability to recognize particular examples of them (Phlb. 62a–c). On
the other hand, it also leads him to consider the needs of actual states,
where no expert ruler is present, and to recognize the importance
of law in those states.
The Laws: politics in the actual world
Plato returns to the subject of politics in what is probably his last

work, and his longest, the Laws. The dialogue deals with the found-
132
politics
ing of a hypothetical new city in Crete, and the core of the work is
a proposed legal code for the new city, prompting Aristotle’s com-
ment that “The greater part of the Laws consists of laws” (Ar. Pol.
1265a1–2); however, there is much else as well, including histor-
ical material, discussions of the nature and purpose of law, religion,
education and, rather surprisingly, the benefits of drinking-parties.
The work is much more practically oriented than the Republic. The
new constitution is explicitly presented as a second best. Although
the communism of the Republic, including the “community of
women and children”, is still seen as an ideal, it is not practicable in
the real world (Leg. 739c ff.), so the Laws provides for the citizens
to own property and to have conventional marriages and families.
Consent is still considered important, and provision is made for
every law to have a preface in which the legislator tries to persuade
the people of the benefits of the law; but it is recognized that this
will not always succeed, so punishment also features significantly
throughout the code (718a ff.).
This more down-to-earth approach may be interpreted in more
than one way. On the one hand it can be seen as representing a
growth of pessimism in Plato. Having at one time made proposals for
a radically new kind of state, he has come to recognize that this is not
possible, and is offering a second best instead. But alternatively, we
may suppose that the ideal state of the Republic was never conceived
as a real possibility. In this case, the fact that he is now ready to offer
a set of practical proposals, even if they fall short of the ideal, can be
seen a showing more optimism than his earlier position.
Two features of the Laws deserve special attention. The first is
the view taken by the chief speaker, the Athenian Stranger, of the
purpose of law. In fact he makes two suggestions about its purpose.
One is that it exists to promote peace within the state (628c ff.); this
can be seen as echoing the claims of the Republic about unity. The
other is that the purpose of law is to promote virtue. The Athenian
is inspired here by the Spartan constitution and that of some Cretan
cities; these were normally seen as aimed at promoting courage in
war, but he argues that they should really be seen as aimed at virtue
as a whole (630e). However, he accepts that they are in fact more
133
plato
effective at promoting courage than other virtues such as temperance

(634a ff.). Again this echoes some elements in the Republic, notably
in the proposals for education, which aimed at encouraging virtue,
but the Laws places more emphasis on virtue as good in itself, not
just as something that ensures the good of the state (631b ff.). This
idea of virtue as the aim of law was also taken up by Aristotle. The
idea that it is the function of the state to promote morality is now an
unpopular one, but one can see how it arises, given two assumptions,
both Platonic: that it is possible to discern, objectively, what virtue
is; and that virtue is a great good to those who possess it. If virtue is
the greatest good we might expect the state to promote it, just as if
we take, for instance, wealth or freedom to be the greatest good, we
may see it as the business of the state to promote these.
The other central claim of the Laws, however, would be much
more widely accepted even now: that law itself should be sovereign in
the state (612d ff.). Here again the Athenian praises the Spartan and
Cretan systems because they do not commit power to any one person
or group, but, by a system of checks and balances, ensure that law
itself should be supreme. For this reason he thinks they are especially
deserving of the name politeia, “constitution”, which thus comes to be
used as the name of a particular system of government, as well as a
term for systems of government in general. Here law seems to rise to
a much higher status than it had in either the Republic or Statesman;
it is described as the “dispensation of reason” and seen as something
divine. Of course, law is made by human beings; it is called divine
because it is an expression of reason, the divine element in us. But
if we commit power to one person or group we are committing it
to them as complete persons, including both reason and appetites,
while if we commit power to the law we are committing it to some-
thing that can be recognized as rational. Checks and balances, in
turn, ensure that the law is observed, while a single ruler or group
of rulers could ignore it if this were in their interest. This idea of the
sovereignty of law, and of the division of power as a way of ensuring
it, has remained influential down to modern times.
134
seven
Ethics
Ethics – the philosophical consideration of the question how we

should live – is central to Plato’s thought throughout his career. His
thought, like that of many ancient philosophers, focuses especially
on the virtues – that is, on morally admirable states of soul or char-
acter – rather than on the rightness or wrongness of specific actions,
although he does, of course, sometimes consider the way the virtues
are expressed in action.
While Plato devoted many dialogues to ethical questions, most
of them are normally seen as Socratic dialogues, expressing the
thought of Socrates rather than of Plato himself. The work in which
the mature Plato presents his ethical ideas most fully is the Republic.
Its official theme is justice, but as justice is seen as bound up with
the other virtues, it does in fact present Plato’s thought about the
virtues more generally.
Plato continued to think about ethical issues later in his career;
as we have seen, two later works, the Statesman and Laws, deal pri-
marily with political topics, but these are, of course, bound up with
ethical ones in Plato’s thought, especially as he sees the promotion
of virtue as a major part of the work of government. He also wrote
one more work that deals directly with ethical questions, the Phile-
bus (the last of his dialogues to feature Socrates as chief speaker);
this addresses, rather more directly than the Republic, issues about
135
plato
the good or happy life for human beings (rather than about virtue),
and considers the place of pleasure and knowledge in the good life.
Much of the argument of the Philebus is hard to interpret, but it does
give significant insight into Plato’s ethical thought. Here, however, I
shall focus on the Republic, as his central presentation of his ethical
position.
The investigation of justice
As we saw in Chapter 6, the Republic is an enquiry into the nature

of justice both in the state and in the individual. At the individual
level, it addresses two central questions: what justice is and why
it is beneficial. These questions will form the main themes of this
chapter.
The word translated “justice”, dikaiosunē, is one of the most
important terms in ancient Greek ethics, and has a rather wider
application than our modern word. It has many of the connotations
of “justice”, being linked with law, with fairness, with punishment
and rectification, and so on, but it also shares some connotations of
“honesty”. Behaviour generally seen as characteristic of a just person
includes telling the truth, and refraining from stealing, cheating and
so on (see Resp. 331b ff., 442e ff., 485c ff.). We tend to use the word
“just” (in connection with people, as opposed to societies, states of
affairs and so on) only of people who are in a position to make judge-
ments – judges, teachers assigning marks, and the like – whereas the
ancient Greeks might use the corresponding term of anyone who
acted rightly towards others.
Socrates, however, sets out in the Republic to find a more precise
account of what justice is. In doing so he makes a striking assump-
tion. He takes it that justice can be found both in a community and
in an individual, and he supposes that justice is the same property in
both cases; a just individual will be like a just state (368e ff.). Socra-
tes does not blindly assume that this must be so; after finding what
justice in the state is, he allows that justice in the soul may turn out
to be something different (434d). But from the start he takes it to be
136
ethics
likely that justice in the state and in the soul will be parallel. This is
why he structures his enquiry in the way he does.
Since Socrates holds – plausibly – that justice in the state is a mat-
ter of the structure of that state, it follows that justice in the soul is a
matter of the structure of the soul, not, at least in the first instance, of
its attitude to other people or to the community. Certainly “justice”
is used in both contexts, of a state and of an individual, and this is no
simple ambiguity; the uses are connected. But one might well think
that a just individual was one who contributed to a just society, rather
than one who resembled one.
Socrates’ point cannot just be that a word, when used in a way that
is not simply ambiguous, must always stand for the same property,
for “just” is used of actions as well as people and communities, and of
course Socrates does not hold that a just action is one that resembles
a just person; rather, it is one that promotes and preserves the just
condition of the soul (443e). Instead, we can see Socrates as relying
on the assumption that, in general, qualities of a community derive
from similar qualities of the individuals who make it up.1 Thus, a just
person is indeed one who contributes to a just society, but he does
so by passing on to it a quality that he himself possesses.
The general principle is plausible in many cases. Take, for instance,
a case that Socrates considers at 435e: the love of learning. A studi-
ous community will be one that consists of studious people, and it
seems that we are using “studious” in a single sense when we say this.
However, it is less clear that this will be true of structural properties.
If the justice of a state is found in the relations between its parts, then
there will indeed be some property of the individual citizens that
contributes to the justice of the state, but there is no obvious reason
for thinking that this will be like the justice of the state – that it will
itself be a structural property.
Virtue in the soul
In searching for an account of virtue in the soul, Socrates makes use

of his account of virtue in the state, discussed in the previous chapter.
137
plato
Since the virtues of the state are found in the relations between its
parts, it follows that the soul must also have parts, whose relations
can constitute its virtues (Resp. 435b–c). It is at this point, therefore,
that the theory of the divided soul, outlined in Chapter 5, is intro-
duced. The three most significant parts of the soul are the rational,
corresponding to the rulers, the spirited, corresponding to the war-
riors, and the appetitive, corresponding to the money-makers (farm-
ers and craftsmen). Wisdom in the soul is found in its rational part,
courage in its spirited part. Temperance is the agreement between
the parts that reason should rule (although, as noted in Chapter 6, it
is a puzzle whether parts of the soul can really be seen as agreeing),
and justice turns out to be the condition in which each part of the
soul does its own work (441c ff.).
One might ask what this can mean; since parts of the soul are
defined by the kind of activities typical of them, how can they fail
to do their own work? Presumably, however, the point is not just
that they do the kind of thing typical of them, but that they do it in
a way that serves their proper purpose, for the good of the whole.
For instance, in Book 8 we read of a person dominated by the love
of money, whose rational part is devoted to working out how to get
money, and whose spirited part admires those who have it (553d);
while these parts are in a sense performing their own function (cal-
culating and admiring), they are not doing so in a way that serves
their overall purpose, as they do in a just soul. It follows that it is not
enough that my soul be governed by reason in the sense of reason-
ing: that I act on the basis of thought-out decisions, rather than of
impulse. It must be right reason, directed to right ends, that governs
my soul if I am to be virtuous.
The view of virtue taken here is interestingly different from that
ascribed to Socrates in the Socratic dialogues. For him, as we saw
in Chapter 5, all desire is for the good; if we know what is good,
we will choose it; hence the only source of wrongdoing is igno-
rance. Thus he was able to identify virtue with knowledge. In the
Republic, by contrast, there are parts of the soul that can conflict;
the rational part of the soul judges what is good, and pursues what
it sees as good, but we can be motivated, by spirit or appetite, to
138
ethics
pursue things we do not see as good. Hence virtue involves not only
right judgement but also the power of that judgement to govern the
other elements in the soul. It seems that judgement can be over-
come by anger or desire, so we are only virtuous if our judgement
remains in control.
In the Socratic dialogues Socrates is sometimes seen as having
an intellectualist concept of virtue; for him, virtue is simply a state
of the intellectual faculties. On the view presented in the Republic,
by contrast, it involves other parts of the soul as well. But it remains
fundamentally rationalist: it is the rational thought that something is
good, and desire for the good, which should motivate us. It is better
when our other desires and feelings harmonize with this, but they
are not in themselves central to virtue.
Virtue, knowledge and true belief
Socrates’ account of virtue raises a puzzle to do with the nature of

reason. As we saw in Chapter 5, reason has two aspects; on the one
hand it has a distinctive aim of its own, the aim of knowledge, which
is most fully expressed in the philosophical life; on the other hand
there is the aim of promoting the good of the whole self, which is
guided by knowledge. When Socrates speaks of the virtuous soul
being governed by reason, which of these aspects of it has he in
mind? They are not, of course, wholly separable; it is by philosophy
that we can attain the knowledge of the good that enables us to guide
our lives effectively. But does this mean that no one but philosophers
can be virtuous?
Certainly later in the Republic, in some of the arguments for the
connection of justice and happiness, the just life is identified with
the philosophical life (Resp. 581b ff.). However, it seems strange that
the lower classes in the ideal state cannot be seen as just. Indeed, it
is clear that Socrates sees them as contributing to the justice of the
state by performing their own function; and it is plausible that this
is because they have just souls, in which each part performs it own
function.2 Moreover, Socrates certainly wants the whole community
139
plato
to be happy (420b ff.). Can it really, for him, be happy if it is not

virtuous?
It may well be that Plato wants us to see only philosophers as
virtuous in the fullest sense; but he may still leave room for a lesser
kind of virtue possessed by others. In the Meno, Socrates argued
that virtue need not be constituted by knowledge, but could instead
merely be true belief (Meno 98c ff.). While in the Republic he no
longer holds that virtue is a purely intellectual state, so that it cannot
simply be either knowledge or true belief by itself, it seems reason-
able to say that it might involve true belief rather than knowledge;
one is virtuous if one has true beliefs about the good, and one’s feel-
ings and desires harmonize with them. So in the Republic the courage
of the warrior class is defined as “persistence in true belief about the
things which are and which are not to be feared” (Resp. 429c). This
true belief can be achieved by following the guidance of the phil-
osopher rulers; although the lower classes do not have philosophical
knowledge, they have sufficient rational powers to be able to agree
that what their rulers command is best.
In later works, particularly the Laws, Plato’s speakers suggest that
people may achieve virtue by being guided by law;3 but of course not
any law will do; it must be one whose own inspiration is philosoph-
ical. What is harder to see is whether any non-philosophers in the
actual world can be virtuous (or even philosophers in the actual
world, since it is not clear that they have actually attained know-
ledge). While the Meno seems right to suggest that true belief can
be a basis for virtue, it gives a less plausible answer to how people in
the world can attain this true belief, suggesting that it is a divine gift
(Meno 99c). This is often seen as ironic, casting doubt on whether
anyone is really virtuous. Yet surely it must at least be true that some
people approach closer to virtue than others, and one may wonder
how this can be explained. It does not seem that Plato ever finds a
wholly satisfactory answer.
140
ethics
The unity of the virtues
There seems, according to the Republic, to be a very close connection

between the various virtues. In the Protagoras, Socrates seemed to
suggest that the various virtue words – courage, temperance, and so
on – were in fact names for one state: the knowledge of good and evil,
which ensures virtuous behaviour (see Prt. 329d ff.). In the Republic
this view has certainly been abandoned; because of the distinction
between parts of the soul, there is also a distinction between the
virtues, which relate to the different parts. Yet it seems that the vir-
tues, although different in essence, will be inseparable. I, as a person,
will not be wise or brave unless the parts of my soul are performing
their proper function. My rational part may display cleverness or
my spirited part daring, but these do not become virtues of me as a
whole unless used for their proper purpose. My parts, in turn, will
not be performing their proper functions unless my soul is directed
by reason. Government by reason, although officially the definition
of temperance, is in fact central to all the virtues, including justice.
This view, that virtues are inseparable, is implausible if they are
seen just as natural tendencies to behave in a particular way; there
seems no reason why a person who is naturally courageous will
also be naturally temperate. However, if we see virtues as grounded
in knowledge or true belief, it is reasonable to see them as going
together; it is the same kind of knowledge of what is right or good
that is manifested in various fields, including both running risks and
controlling appetites. Aristotle also commits himself to the unity
of the virtues, because he sees all virtues as grounded in practi-
cal wisdom (Ar. Eth. Nic. VI.13); but he distinguishes between the
fully fledged virtues and natural predispositions to virtue, which
are separable.
Socrates accepts that there are natural tendencies that predispose
people to virtue, and he allows that these do not always go together.
Indeed, he recognizes that two qualities that are desirable in guard-
ians, the spirited and the philosophical, are hard to combine (Resp.
375c–d). In a similar way, at Meno 88b there is a reference to courage
and temperance as tendencies that are not beneficial except when
141
plato
combined with knowledge; these are treated as not being true vir-
tues. Likewise in the later Statesman (306a ff.) the Eleatic Stranger
argues that temperance and courage, again probably seen as natural
tendencies, do not generally arise together, and it is the job of a
statesman to ensure their combination, either in one person or in
a team (311a); although by themselves they may be harmful, they
become beneficial when combined with right belief about what is
fine, just and good (309b ff.).
Virtue and action
Socrates says that justice – and the same should go for the other
virtues as well – is primarily concerned with the inner condition of
the soul; just actions are those that promote and preserve this condi-
tion (Resp. 443c–e). In saying this, he anticipates a central principle
of modern virtue ethics: that virtue, a state of character, is in some
way more fundamental than moral qualities of actions. In fact it
seems likely that we should take him as saying that it is more fun-
damental in two ways: just actions are defined in terms of the state
of soul, and they are valuable because of their connection with the
state of soul. Virtue ethicists often look to Aristotle as a forerunner,
but in fact this view put forward in Plato’s work anticipates modern
virtue ethics more clearly than anything in Aristotle. Whereas for
both philosophers virtue is a central concern of ethics, it is clearer
in Plato that it is foundational, with actions depending on it for their
moral value.
This view can also overcome a common objection to virtue eth-
ics. It is often thought that virtue must itself be defined by reference
to right action: that virtues are something like stable dispositions
that enable us to choose rightly. But if this is so, we cannot go on to
define right action in terms of virtue, without circularity. However,
the Republic, by proposing a definition of the virtues in terms of the
inner relation between parts of the soul, allows them to be defined
without reference to right action, so that they can indeed be seen
as fundamental.
142
ethics
Platonic justice and ordinary justice
However, this conception of justice leaves us with a puzzle. It may

seem to be too far removed from our – or the ancient Greeks’ –
ordinary conception of justice, which is, surely, concerned with our
actions towards other people. Hence, Plato has been accused of sim-
ply changing the subject (see Sachs 1963). He showed Socrates set-
ting out to demonstrate that justice as it is normally understood was
beneficial; but in fact what he demonstrates – even if his arguments
are successful – is that justice in a new, revisionary sense is beneficial.
The first thing to say in response to this is that it is not clear
Plato would have accepted the view that there is a clearly definable
ordinary conception of “justice”; he may have thought that ordinary
people’s use of the term was fundamentally confused. Certainly in
Book 1 of the Republic the characters search for a definition of jus-
tice that reflects people’s normal understanding of the term, and is
concerned with actions towards other people, but they fail to find
one. They consider the definitions “paying to people what we owe
them” and “doing good to friends and harm to enemies”, but find
these unsatisfactory (Resp. 331e ff.). If the ordinary use of a term is
unclear or confused, it is not always unreasonable of philosophers
to give it a new definition that is clearer, and that can then be used
in the search for definite answers to philosophical questions.
But even if Socrates is proposing a new concept of justice that does
not seek to capture precisely the ordinary use of the term, his concept
should have something to do with the ordinary use of the term if
it is to count as a new concept of justice, rather than, say, courage,
or jealousy, or intelligence. A person who is just in his sense may
not always act in ways we would naturally call just, but one would
think that he would often do so, or at least that his behaviour has
something significant in common with what we normally call just
behaviour.
Have we any reason to think this? A just person’s soul, according
to Socrates, is governed by reason. This reason certainly enables him
to pursue his own good; reason is that element in a person that takes
care of the whole (441e). Does it also motivate him to respect the
143
plato
rights or pursue the good of others? It will motivate him to do this,

if it is rational to do so: and it is natural to believe that it is. But does
Socrates say anything that can explain or justify this view?
There is no doubt that Socrates is shown as holding that the just
person in his sense will typically behave in a just way towards others.
This first becomes apparent in Book 4 when he says that a person
in whom each element performs its own function will himself be
just and perform his own function; this means that he will do what
he is best able to do for the good of the community (441d–e). Later,
Socrates goes on to say that a person who is just in his sense will
not be likely to perform any of the acts typically seen as unjust; he
will not steal, break oaths, betray his friends or his country, commit
adultery and so on (442e ff.). But at this point it remains unclear
why this should be so.
A partial answer appears in Book 6, where Socrates is arguing for
the suitability of philosophers as rulers. As we have seen, a philoso-
pher is for Socrates the archetypal just person; and here he argues
that a philosopher will not have many of the motives that lead people
to act unjustly (484b ff.): he will not have strong bodily desires or love
of money; and he will not think human life important, and so will
not be afraid of death. In addition, he will love truth, since truth is
akin to wisdom, and therefore will not normally be motivated to lie.
But while these considerations may show that the philosopher
will, in general, lack motives positively to wrong others, they do
not show that he will actually seek to benefit others. Moreover, they
leave room for his being motivated to wrong others, when the love
of wisdom itself is the motivating factor. This becomes important
because of a special problem that arises later in the Republic.
The philosophers’ choice
Socrates argues that philosophers should rule in the ideal state, but
he also recognizes that they will not want to do so; they would pre-
fer to devote their lives to philosophy. In the image of the cave, he
describes a person being liberated from a dark cave in which he has
144
ethics
been imprisoned, where he saw only shadows, and allowed to escape

to the outside world, where he can see real things, finally looking
at the sun (Resp. 514a ff.). This represents the philosopher attaining
knowledge of the Forms, and in particular the Form of the good. But
then, we are told, he must return to the cave – that is to the realm
of human affairs – in order to take part in government (519d ff.).
Obviously he will be reluctant to do so.
Socrates recognizes this reluctance, and welcomes it. He thinks
a state where rulers do not enjoy ruling is a desirable one; if rulers
actually love power this will lead to conflict and it is better that
they approach ruling as an unavoidable duty (520d). So he speaks
of philosophers being compelled to rule (520a). But clearly this does
not mean physical compulsion; the rulers are told what they must
do, and agree to it. Why do they agree? Glaucon says – and Socrates
does not dispute it – that they agree because “we are making a just
demand of just people” (520e). It is to the state that they owe the
education that enables them to do philosophy; being just, they will
be ready to repay that debt by serving the state.
Plato’s language here, with its explicit reference to something that
is owed (520b), seems to introduce a conventional conception of jus-
tice, one in which it consists of paying what we owe to others. Why
should someone who is just in the distinctive Platonic sense – one
in whom each part of the soul performs its function effectively – be
motivated to pay debts in this way? Since such a soul is governed by
reason, it will be motivated to pay debts if it is rational to do so. But
why should it be seen as rational?
In fact, there are in this particular case some fairly clear reasons
why the rulers should be motivated to take part in government; but
they are not directly linked with the thought that this is owed to
someone. In Book 4 it was argued that guardians (rulers and sol-
diers) should be chosen from among people who identify their inter-
est with that of the community (412d–e). One thing this might mean
is that their self-interest, in a narrow sense, is bound up with the
interest of the community; and in Socrates’ ideal state this seems to
be true, because the division of labour that is a mark of that state
makes everyone dependent on one another, enabling everyone to
145
plato
engage in the activities for which they have the greatest aptitude.
Philosophers may be motivated to be outwardly just, in the sense of
making their proper contribution to society, because if they do not
make this contribution the society will be endangered, and in this
case they may lose their opportunity to do philosophy.
However, the passage mentioned above can also be read in
another way: not that the rulers’ self-interest, narrowly interpreted,
coincides with that of the city, but that they see the interest of the
city as part of their own. This is supported by some of the propos-
als for unity in the state found later in the work, and in particular
the “community of women and children”. Because the guardians
are related they care about one another as we do about family. We
care about our children, not because our narrowly defined interest
coincides with theirs – it may well not – but because we see their
interest as part of ours. This is especially emphasized at 462c–e,
where Socrates claims that the guardians will actually feel injury to
another member of the community as injury to them. In this case,
the philosopher will have a direct motive to do his part in promot-
ing the good of the state as a whole.
However, the claim that the philosophers should take part in
government because it is just seems to go beyond this. It seems that
they are being asked to do this even if it is not, at least directly,
in their own interest, since otherwise there would be no reason to
speak of compulsion; but also that acting in this way is rational,
since otherwise it would not be called just. It appears, then, that
we should accept that we have reason to pursue the good of others:
not, perhaps, of all others, but of others with whom we have a spe-
cial connection, which may well include those who have benefited
us. (Socrates says that justice requires philosophers to take part in
government only in the special circumstances of the ideal state, not
in ordinary states where they arise by chance [520b].) While there
seems to be nothing in Plato’s conception of reason to rule this
out, it does not seem that he ever gives a full explanation of why it
should be so.
146
ethics
Virtue and the good of others
The reasons given here for the philosophers to take part in govern-
ment seem to apply only in the ideal state. Do rational people, in
Plato’s view, have any more general reason for benefiting others?
After all Socrates himself, who certainly did not live in an ideal state,
did not simply devote his life to contemplation, but tried to stimulate
thought among the Athenians – although in the Apology he says his
reason for doing this is that it was in response to a divine command
(Ap. 23c).
One reason why we might want to benefit others is suggested by
Diotima’s speech in the Symposium (206b ff.).4 This puts forward the
idea that we are all seeking to leave behind something through which
we can, in a sense, live on after our deaths, thus achieving a kind of
immortality, and so continuing to have the opportunity to possess
good things forever. One way of doing this is by physical reproduc-
tion: having children. Their life can be seen as a continuation of
ours. But another way is by spiritual reproduction: influencing the
souls of others. This can be done at an individual level by educat-
ing our partner in a personal relationship; it can also be done on a
wider scale by, for instance, writing poems, producing laws or doing
notable deeds that will achieve fame and so act as a moral example.
Here, then, we have a reason to benefit others that stretches beyond
the confines of the ideal state, but it is still limited; the people whose
lives we affect must be connected in some way with us, so that their
lives can be seen as continuations of our own.
Is there room in Plato for a more universal altruism? It is possible
that he came to accept such a position later in his career. In the
Timaeus we are told of God’s motive for creating the cosmos; he
was good, and what is good is not grudging, but desires everything
to be good; finding the realm of matter in a confused and disor-
derly state, he set out to give it order (Ti. 29d–30a). This kind of
motivation seems to be one that human beings can share as well;
for the language here used of God echoes a passage in the prologue
to the Timaeus, which includes the earliest version of the Atlantis
legend. Atlantis was a tyrannical power that oppressed the people
147
plato
of the Mediterranean. The Athenians fought against Atlantis and

regained their freedom, but then ungrudgingly liberated the other
Mediterranean peoples as well (25c). Here the Athenians are act-
ing to benefit people who have no special connection with them;
virtue seems to have a wider scope than it has in the Republic. (It is
interesting that in the Republic Socrates says that Greeks should see
Greeks as natural friends [Resp. 470c], and only barbarians – non-
Greeks – as enemies; in the Timaeus non-Greeks are also brought
within the circle of concern.)
It is sometimes suggested that the rational person, for Plato,
should be seen as pursuing an impersonal good; not just aiming to
achieve the good for himself and those in some way linked with him,
but to increase the amount of good in the world (see e.g. Annas 1981:
259ff.). This can be seen as following from the theory of Forms. The
Form of the good is something impersonal, not the good for any one
person or thing; and so the person who contemplates the Form of
the good and wishes to imitate it will aim not just at his own good or
that of his community, but at an impersonal good – the production
of as much good as possible in the world.
However, I suggest that the theory of Forms need not be read in
this way. Certainly, the Form of the good is not the good for any
specific person or thing; it is the nature that all goods have in com-
mon. To know the Form of the good is not to know what is good for
me or for you. This, however, is compatible with saying that each
instance of the good is a good for someone or something. Consider
the Form of father. It is not any individual person’s father, but rather
the nature that all fathers have in common; but each individual father
is the father of someone. In the same way it may be that what is good
is always good for someone or something, although there is a nature
that all good things have in common. Knowledge of this nature will
inspire us and help us to achieve what is good for ourselves and
those dear to us.
Even if what is good is always the good of someone or something,
this does not mean that we have reason to pursue only our own
good; we may have reason to pursue the good of others, and this
may extend more narrowly, as in the Republic, or more widely, as in
148
ethics
the Timaeus. However, while it seems clear that Plato is committed

to the thought that it is rational, and therefore just, to pursue the
good of others, it does not become clear exactly why this should be.
Given that some – such as Callicles in the Gorgias and Thrasymachus
in Book 1 of the Republic – did challenge that view, this is a gap in
Plato’s arguments.
Justice and happiness: the challenge
We now turn to the second major question of the Republic: how

justice is connected with happiness. “Happiness” here does not mean
a feeling, but a life that is worthwhile as a whole. This is how the
term commonly translated “happiness” (eudaimonia) was normally
understood in ancient Greece, and it was a central concept of Greek
ethics.
The relation between virtue – being a good or admirable person
– and happiness – having a good or worthwhile life – was much
debated in antiquity. Most philosophers saw them as going together;
justice, temperance and so on are virtues and also promote happi-
ness. Socrates, Plato, Aristotle, Epicurus and the Stoics all thought
this, although they would have disagreed about just why it was so.
We must be careful, however, about just how we understand this
connection. On one possible reading of the relation between virtue
and happiness, happiness is the more fundamental concept; virtue
may be defined as a state of character that promotes the agent’s hap-
piness, or valued because it does so. This reading of the connection
produces what may well be seen as an egoistic moral theory: perhaps
not narrowly egoistic, if we see the happiness of others as bound
up with our own, but still one that treats our own good as the most
fundamental reason for action.
However, this is not the only way to see the connection between
virtue and happiness. One might see virtue as the more fundamental
concept, with happiness actually consisting in the exercise of vir-
tue; or one might see the two as equally fundamental concepts, but
nevertheless as linked. On such a reading, it would not in general
149
plato
be true that virtue consisted only of pursuing our own good, or that
our most basic reasons for action always relate to our own good; they
might equally well relate to the good of others. But nevertheless act-
ing virtuously might be something that contributes to our own good,
something to value and take pleasure in. On such a reading, the point
of demonstrating that virtue contributes to happiness is not to give us
a reason for acting virtuously when otherwise we would have none.
Rather, it is to show that there is not, as one might think, a conflict
involved in the good life; the requirements of virtue and happiness
are not, as many suppose them to be, opposed to one another.
I have suggested that for Plato virtue does not always consist
of the pursuit of one’s own good; rather, it consists in the effective
functioning of the soul, governed by reason. While reason is often
directed to the agent’s own good, it may also be directed to the good
of others. Hence, when Socrates sets out to investigate the connec-
tion between justice and happiness, he is not asking what makes
justice a virtue. He is not here giving an account of the foundations
of ethics as we understand them, saying what makes an act right or
a state of character virtuous. Hence, he need not be seen as holding
an egoistic moral theory.
In the main argument of the Republic, it seems that Socrates is not
being asked to show that justice is a virtue. It is true that Thrasyma-
chus in Book 1 had challenged this view, but Glaucon and Adeiman-
tus, whose speeches in Book 2 set the agenda for the greater part of
the work, do not. Adeimantus, at least, or the ordinary people for
whom he speaks, seems to accept that justice is a virtue; he speaks
of justice and temperance being admirable, but difficult and burden-
some (Resp. 364a).5 Socrates is later to assume that justice is a virtue
when describing the virtues of the ideal state (327e). What he is being
asked to show here is that justice is beneficial to the just person.
The challenge that Socrates is here asked to answer is a plausible
one. As we have seen, many ancient philosophers saw virtue and
happiness as going together, and would have said that justice is both
a virtue and conducive to happiness; at the other extreme there were
people, represented in Plato by Callicles in the Gorgias and Thra-
symachus in the Republic, who took an egoistic view, according to
150
ethics
which justice is neither a virtue nor conducive to happiness. But

between these two extremes there were many people – of whom
Polus in the Gorgias and Glaucon and Adeimantus here are repre-
sentatives – for whom justice is fine and admirable, but not good
(for the just person); it is a virtue, but does not promote happiness
(or at least not always, and contingently when it does). This might
be seen as a common-sense view; but it is the view that Socrates here
undertakes to answer.
We must now look at the challenge in more detail. In Book 2,
Glaucon and Adeimantus challenge Socrates to show that justice
is to be valued for its own sake, and not only for its consequences
(358a ff.). Most people think that justice is in itself something bur-
densome, and is only valued because of its consequences; Glaucon
and Adeimantus say that they themselves do not accept this view,
but that it is the common view, and therefore the alternative position
needs defence.
The claim that justice is to be valued both for itself and for its
consequences seems to mean that it contributes to our happiness,
both directly and indirectly.6 “Good in itself ” should not be read in
a Kantian sense as meaning that it is admirable, independently of
its effect on our happiness; it means only that it benefits us directly,
through the effect that it has on our soul, rather than through some-
thing that is only contingently connected with it.7 It can even be
taken to include the case where justice is beneficial because it pro-
duces a further good thing, provided its connection with that thing
is necessary. It appears that the consequences that are here contrasted
with the direct effect of justice are external ones, ones that arise from
other people’s reaction to our just behaviour.
While Socrates holds that justice is to be valued both for itself
and for its consequences, what is important is to show that it is to
be valued for itself, since it is generally agreed that its consequences
are valuable.8
Two rather different aspects of this are emphasized by the broth-
ers. First, Glaucon talks about the origin of justice (358e ff.). He pro-
poses – not as his own view, but as a widely held one that needs an
answer – an early form of the social contract theory: people came to
151
plato
an agreement that they would act justly towards one another, because
they found this preferable to a state in which everyone constantly has
to suffer the unjust behaviour of others. What would be best for the
individual is if he could act unjustly, but not expect unjust action in
return; but this, in general (as opposed to some special cases) is not
achievable. What is really beneficial on this view is not justice itself,
but the expectation that others will act justly towards one; but acting
justly is, normally, the price of this.
Secondly, Glaucon goes on to talk about the fear of punishment
as a motive for acting justly (359b ff.); and Adeimantus likewise talks
about the rewards that are given to just people, both by other people
during life, and by the gods after death (362e ff.). But these conse-
quences, as they point out, arise from other people, including the
gods, recognizing us as just, not from justice itself; and as Glaucon
says, a person who was unjust but had a reputation for justice would
be rewarded, and a person who was just but had a reputation for
injustice punished (361a–d). In such circumstances, is a just life still
worth living? Socrates has to show that it is.
Justice and happiness: the response
At the end of Book 4, after defining justice, Socrates draws an ana-

logy between justice and health: justice is the natural ordering of
the components of the soul, as health is the natural ordering of the
components of the body (Resp. 444c ff.). His official purpose in
drawing this analogy is to illuminate the nature of injustice, and of
just and unjust behaviour. Injustice, like disease, is a disruption of
the natural order; just behaviour, like healthy behaviour, builds up
the natural order, while unjust behaviour, like unhealthy behaviour,
disrupts it. However, Glaucon points out that this analogy makes it
plausible that justice is a desirable state, to be preferred to injustice,
and Socrates does not dispute this (445a–b). This seems right; if the
various parts in us are functioning in such a way as to preserve our
natural condition, each functioning in the way it is best equipped to
do, this seems a desirable state to be in.
152
ethics
Nevertheless, Socrates and Glaucon do not immediately draw

the conclusion that justice is better than injustice, since a formal
defence of this view cannot be offered until injustice itself has been
examined in more detail. Because of the long digression that begins
at this point, the discussion of injustice is delayed until Book 8, and
it is in the latter part of Book 9 that the official arguments for the
superiority of justice begin.
The first argument turns on the analogy of an unjust soul and an
unjust state. As the most unjust state is a tyranny, the most unjust
soul is a tyrannical one; this does not mean the soul of a tyrant,
but one that resembles a tyrannized state (although a person with a
tyrannical soul may also be a tyrant himself).9 As a tyrannical state
is dominated by one person, so a tyrannical soul is dominated by
one appetite; Socrates uses the example of sexual appetite, although
others are presumably possible.
A tyrannical soul, Socrates argues, is not free, because it is domi-
nated by one desire and so cannot do what it wishes (577e ff.). It
is also impoverished, perhaps because the domination of this one
desire prevents it getting what it needs or wants, and fearful, presum-
ably of what the dominating desire might force it to do. Hence, the
unjust life is undesirable, by comparison with the just life, which is
governed by reason. The argument here seems to depend implicitly
on the idea that our rational nature is our true self; hence, when
reason is in control we are free, making decisions for ourselves, while
when appetite is in control we are enslaved, being forced to act in
ways we do not choose.
The two remaining arguments focus on the claim that the just life
is pleasanter than the unjust. While Plato does not approve of a life
of pleasure pursued for its own sake, he never denies that a good life
will be pleasant, and indeed the most pleasant; we should indeed
gain pleasure from attaining what is truly good.10 Both of these argu-
ments turn on identifying the just life with the philosophical life,
which is here contrasted (in accordance with a tradition said to go
back to Pythagoras) with the lives devoted to the love of honour or
gain. According to the first of these arguments (581c ff.), in decid-
ing which life is best we should rely on the person who has wisdom
153
plato
and experience. The philosophical person obviously has wisdom;

he also has experience, since lovers of honour and gain do not pos-
sess wisdom, while the philosopher will typically experience some
of the pleasures of honour and gain during his life. And clearly, the
philosopher chooses the philosophical life.
This is an odd argument. It does not seem enough to say that a life
is the best if the person with wisdom and experience chooses it. This
would be evidence that it was the best only if he chose it because of
his wisdom and experience. But it does not seem that the philoso-
pher did that; he must have chosen the philosophical life before he
acquired the relevant wisdom and experience, presumably because
he had a natural attraction to it. At any rate this argument does not
tell us why the philosophical life is superior.
The last argument does so (583b ff.). It is a complex argument, but
the central point is that the things that give pleasure to the mind –
truth, knowledge and understanding – have a greater reality than
the things that give pleasure to the body – food, drink and so on.
This is because they are seen as belonging to the unchanging realm
that, earlier in the Republic, was seen as having greater reality. Hence
the pleasures derived from those things – which are the pleasures of
the philosopher – are also more genuine. However, it is also argued
that the lovers of honour and of gain will also achieve the pleasures
most appropriate to them when guided by wisdom (586d ff.); wis-
dom guides us towards achieving these pleasures in the most effec-
tive way, by moderating our desires and preventing us reaching at
things that are beyond our powers.
In a rhetorical conclusion, Socrates introduces the famous image
of the soul as composed of a human being, a lion and a many-headed
monster (588b ff.). The human being, of course, represents reason,
and Socrates and Glaucon agree that it is best that it should be in
control; this reaffirms the idea that our rational nature is our true self.
Once again, to allow the other elements in the soul to take control is
to allow our reason to be enslaved.
These arguments involve both aspects of reason identified ear-
lier: as the faculty that plans for the good of the whole, and as an
element in the soul having aims and desires of its own, for truth
154
ethics
and knowledge. Some of the arguments focus on the latter, drawing

attention to the goodness of knowledge itself, and so the happiness
of the life of a lover of knowledge. But the way in which government
by reason enables other aspects of the soul to achieve satisfaction is
also emphasized. Finally, some arguments seem to point to the just
condition of the soul as intrinsically worthwhile, independent of the
achievement of any particular aim. This is true of the arguments in
Book 9, which focus on the freedom of the just person in the way he
controls his own life, as opposed to the enslavement of the unjust.
It is also true of the considerations put forward at the end of Book
4; although not a formal argument, these suggest that the just state
of the soul is desirable simply because it is a natural state, in which
each part functions well.
It is important that not all the arguments put forward here turn on
the goodness of wisdom itself, for not everyone can be a philosopher.
Socrates wants us to think that even people without philosophical
ability can be just to an extent, if they allow themselves to be guided
by wisdom by submitting to philosophical government (590c–d);
then they, too, will be happy. This cannot be because they enjoy the
pleasures of wisdom itself, but because wisdom guides their lives in
such a way as to attain the overall good.
One concern that is sometimes raised is whether the view that
Socrates is defending in the Republic as a whole – that justice is
always better than injustice for the person who practises it – is com-
patible with the passage about the return of the philosophers to the
cave. There, the philosophers seem to be giving up what is best for
them – philosophical contemplation – in response to a demand of
justice; indeed, it is emphasized that it is not their happiness but the
happiness of the state as a whole that matters (519e). The answer, I
suggest, is that justice, as a state of soul, may always be better than
its opposite, but it does not follow that every just act, considered by
itself, is better than its opposite. If we take the acts in isolation, con-
templation is better than taking part in government; if someone can
devote his life to contemplation without injustice (as perhaps the rare
philosopher outside the ideal state can) their life will be happier than
that of someone who has to govern. But if a philosopher in the ideal
155
plato
state were to choose to devote his life to contemplation, he would be

unjust; that means that his soul would be in a worse state than that
of the just person, and so he would not be happy.
Happiness and the Form of the good
In the last sections of this chapter, I shall mention some questions

that relate Plato’s ethics to his larger philosophical concerns. At
Republic 505a Socrates says that justice and everything else derives
its value from the Form of the good. It seems, therefore, that this
Form should help to explain what happiness is and why justice con-
tributes to our happiness. However, within the Republic Socrates
says relatively little about this Form, discussing it only in the central
books, and there using images to describe it. Can we say how the
Form of the good is relevant to the goodness of justice and the other
virtues?
If we focus on justice as philosophy the answer is clear: justice
enables us to contemplate the Forms, and among them the good,
which is central to the system of Forms. It is the position of the
good in the realm of Forms that gives value to them, and so to the
knowledge and contemplation of them.
However, we have seen that justice also has a wider aspect, con-
cerned not only with philosophy but with how we live our life as a
whole; the condition of soul that justice produces is in itself a good
one even if it does not lead to contemplation of the Forms. How is
this related to the Form of the good? I think we must say that a well-
ordered soul – one that is harmonious, in control of itself, and in its
natural condition – is itself an instance of the Form of the good. This
condition of the soul is therefore a good or worthwhile one to be in,
and the individual virtues are good in so far as they promote this
condition. Other things, too, can be seen as instantiating the Form
of the good if they are well ordered, with their parts fitting together
harmoniously to promote the good condition of the whole, although
obviously what exactly this good condition amounts to will differ
from case to case. This includes just states, as well as healthy bodies,
156
ethics
and perhaps other things as well: artefacts that are well adapted for
their purpose, and even the universe as a whole.
It may be that there is a third way in which justice is related to
the Form of the good: it can enable us to be a part of something that
instantiates the Form. This will be true in the ideal state, where the
just agent contributes to the good of the whole, but also identifies
the good of the whole with his own. However, because this way of
being related to the good is only possible in the ideal state, it is less
relevant than the other two to the project of the Republic of showing
that justice is always more beneficial than injustice.11
We saw in Chapter 3 that the Form of the good is not only what
all good things have in common, but also is itself a good thing: the
“best of realities” (Resp. 532c). One may wonder how this can be.
But if a good thing is something worth possessing, it seems right to
say that the Form is a good thing, for contemplating it, instantiating
it and being part of something that instantiates it are good states to
be in, and these can be seen as ways of possessing it. Hence, while it
is indeed puzzling that the nature shared by all large things should
be seen as large, or the nature shared by all animals as an animal, it
does make sense to see the nature shared by all good things as good,
simply because possessing that nature is a good thing, worth aiming
at, and worth taking pleasure in when achieved.
The nature of the good
This, however, may leave us still wondering about one extremely

puzzling question; what is the good? What do all well-ordered things
have in common? Can we say any more about it than that it is good?
Socrates in the Republic is very cryptic about this; he points in the
direction of knowledge of the good, as something that philosophers
in the ideal state ought to know (Resp. 506d ff.), but he does not tell
his listeners what it is, approaching it through images (although he
does insist that it is not pleasure or knowledge [505b–d], two wide-
spread views, and ones with which Socrates in earlier dialogues had
shown some sympathy12). Does this mean that Plato himself did
157
plato
not claim to know the nature of the good? Or, is it, perhaps, that he
thought it unsafe to commit it to writing?
We do have some evidence from outside the dialogues for Plato’s
view of the good, but it is hard to interpret. There is a story that
he once gave a public lecture on the good; unfortunately our only
report of this comes from a hostile witness, Aristoxenus (Aristox.
Elem. Harm. II.30–31). According to him, the lecture was largely
concerned with mathematics, and ended with the claim that “good
is one”. This could mean simply that there is one good; but it could
also mean that the good is the one, that goodness is unity. If so, it fits
in with an equally mysterious claim of Aristotle, who says that Plato
takes unity and duality as his two basic metaphysical principles, and
that he identifies one of those principle (presumably unity) as the
source of good, the other of evil (Arist. Metaph. A.6).
Puzzling though these indications are, they do seem to fit well
with a theme that is manifested in several places in Plato’s work: link-
ing the good with unity and harmony. In the Republic this is applied
to the virtuous soul in the passage that compares it with a musical
harmony: “binding all these parts together and becoming altogether
one out of many, temperate and harmonious” (Resp. 443d–e).
Also in the Republic, this same idea is applied to the state in a
number of places. At 422e, having urged that extremes of wealth
and poverty should be excluded from the state because they tend to
produce division, Socrates argues that the state he is describing is
truly a state because it is unified; an ordinary community, because
there are divisions between rich and poor, is really a multiplicity of
states. At 462b he praises the “community of women and children”
because it ensures that people have feelings in common, and this
binds them together and makes them one.
In the Timaeus the same principle is applied to the world as a
whole; the elements that make up the world are combined in a har-
monious proportion, so that they may achieve “agreement” and
“friendship”, which makes the whole composed of them indissolu-
ble (Ti. 32c). This echoes a passage in the Gorgias, where we are told
that community and friendship hold the universe together (Grg.
508a).
158
ethics
Clearly, saying that the good is unity is not a full explanation of

its nature. We would have to understand more fully the nature of the
unity in question. Plato cannot have believed that anything was good
if it can be described as one object, or as a whole of parts. A heap of
stones is one heap, and the stones are its parts, but it is unlikely that
Plato would have offered this as an example of the good. The kind of
unity he has in mind is one where parts are combined harmoniously
to form something that operates effectively as a whole; to gain a full
understanding of this one would have to know the principles accord-
ing to which they combine. But the link drawn here between good-
ness and unity does help us to gain some understanding of how Plato
sees the good, in the state, in the soul and in the universe at large.
Philosophy and the good life
Plato clearly sees a strong connection between virtue and philosophy.

This connection has two aspects. On the one hand, the pursuit of
wisdom implies a disposition that is already in many ways virtuous;
as we have seen, the philosopher is free from strong bodily desires
and love of riches, does not fear death, and loves the truth (Resp.
485a ff.). On the other hand, when wisdom is attained and we have
knowledge of Forms, this will inspire us to order our own lives in
their image, and will give us the knowledge we need to order both
our own lives and our community (500b ff.).
Because of the central place of the good in Plato’s philosophy,
he sees philosophical knowledge as intensely relevant to practical
affairs. In this he contrasts to some degree with his pupil, Aristotle.
For him there are two kinds of wisdom, philosophical and practical,
and two kinds of life – both admirable, but distinct – linked with
them: the political life, to which moral virtue is central, and the life
of contemplation. For Plato, by contrast, it is philosophical wisdom
that we need to guide our life. What is more, this wisdom does not
come from a distinctively practical philosophy, separate from the-
oretical philosophy; the good is central to the whole of philosophy,
including areas that we might see as wholly theoretical.
159
plato
But while there are not two kinds of philosophy, there are two
aims associated with it; the contemplation of Forms itself, and the
use of the knowledge thus gained in practical life. While philosophy
is worth studying for its practical value, it also gives us an aim of
its own, an aim of overwhelming attraction, as it is presented both
in the Republic and in the account of love in the Symposium. It is
often pointed out that for ancient thinkers philosophy was a way of
life. But this need not imply that it was a wholly practical kind of
enquiry; while on the one hand philosophers were concerned with
the question how we should live our life, they were on the other hand
dedicated, as part of their way of life, to the pursuit of knowledge
or truth for its own sake. In one way, for Plato, philosophy exists for
the sake of practice; it aims to help us make good practical decisions,
both in our own lives and in the state. But in another way, practice
exists for the sake of philosophy; at least one of the aims for practi-
cal decisions is to make knowledge and contemplation of Forms
possible. There remains a tension at the heart of Plato’s philosophy
regarding how we can reconcile these two aims.
160
eight
God and nature
Plato’s God
In many places in his work, Plato writes about a God who brings
the world into being. He is often envisaged as the craftsman of the
universe, and is therefore sometimes referred to by modern scholars
as the Demiurge (from the Greek for “craftsman”). However, Plato
often refers to him simply as “the God” (ho theos), and while, as we
shall see, he is not exactly like the God of traditional theism, they
are similar enough to make it reasonable to call them by the same
name. Plato may well have played a major part in making belief
in a single supreme God more widespread; philosophers after him,
including Aristotle, the Stoics and later Platonists, took up the idea
and developed it in various ways.
The idea of a creator God was not a widespread one in Greek
thought in Plato’s time. Nevertheless, there is some evidence that
Socrates believed in such a God (see Xen. Mem. I.4, IV.3), and it
is likely, therefore, that Plato had such a belief from the start of his
career. Certainly the creator seems to be present in the Republic,
although he makes only two brief appearances there, as “the crafts-
man of the senses” (Resp. 507c) and as “the craftsman of the heav-
ens” (530a). However, it is in Plato’s later works that God becomes a
central figure. In the Timaeus, which deals with the creation of the
161
plato
world, he dominates the narrative; he is also present in the Sophist

(265b ff.), Statesman (269d ff.), Philebus (28d ff.) and Laws (903b ff.).
This seems to be related to a shift of perspective in Plato’s thought
– which need not imply an actual change of doctrine – in which he
begins to take this world, and therefore its creator, more seriously,
emphasizing the goodness of the world more than its imperfection
and the way it contrasts with Forms.
In what follows we shall focus mainly on the Timaeus, and here a
warning is necessary; in that work the chief speaker, Timaeus, says
that his account is conjectural (Ti. 29b–d). He claims that we cannot
have true knowledge of things that change, so, since his account is
concerned with the changing world rather than the eternal Forms,
it can be only a “likely story”. However, this need not be taken to
mean that it does not aim at truth; rather, it tries to come as close to
the truth as possible, although it does not claim perfect success. It is
therefore reasonable to use it, with caution, as a guide to Plato’s views.
Why does Plato believe in God?
At the beginning of the Timaeus there is a simple argument for the

existence of a creator. Everything visible and tangible has come into
being, and everything that comes into being must have a cause. The
universe is visible and tangible; therefore it has come into being, and
accordingly has a cause (Ti. 27d–8c).
Of the two assumptions that ground this argument, the second is
a widespread, although not universal, philosophical view. The first
reflects Plato’s own way of thinking, in which eternal and unchang-
ing Forms contrast with transient sensible things. It is sometimes
objected that even if each particular sensible thing comes into being,
it does not follow that the totality of sensible things must do so. But
for Plato the universe – that is the kosmos, whose name in Greek
means “order” – is not just the totality of sensible things. It is a spe-
cific ordered state, linked with the regular motions of the heavens;
and this specific state must have had a beginning. Indeed, Plato does
not think that the totality of sensible things had a beginning. He
162
god and nature
believes, as we shall see, that there were sensible things, in a state of

chaos, before the beginning of the ordered world.
This argument is a version of what was later to become known
as the cosmological argument. By itself, however, it can prove only
the existence of a cause of the world, not of an intelligent cause that
deserves the name of “maker”. In fact, as soon as this argument had
been given, Timaeus begins referring to a “maker and father” of the
world. But only later does he say anything that might justify this
way of speaking.
This argument comes from a passage we have already looked at,
Timaeus 46d–e, where a contrast is drawn between the kinds of cau-
sation typical of body and of soul. Bodies, because they are moved
from outside, are governed by necessity, and this means that they
cannot act for a purpose; soul, by contrast, because it moves itself,
can be responsive to reasons and so can act for a purpose.1 This
claim is of quite general relevance, and can relate to the activities of
human and animal souls, but in the context where Timaeus makes
it it does in fact relate to divine activity in creation. If we see things
in the world as existing for a purpose, as Plato does, then a soul,
something living, must have ordered them so as to enable them to
serve that purpose. If the universe as a whole has a purpose, then
it must have a cause that is intelligent and hence living, and this is
Plato’s conception of God.
Much of the Timaeus is taken up with a discussion of the purposes
that things in the world serve, and the way they contribute to the
good. Like many thinkers who have argued for the existence of God,
Plato is impressed with the way in which parts of human and animal
bodies, and their environment, are adapted to promote survival and
the continuance of the species, and also, in his view, to make intel-
lectual activity possible. However, for him the strongest proof that
the universe is well ordered – emphasized in both the Timaeus and
the Laws – is the movements of the heavenly bodies; they are both
regular in themselves, and form a harmonious pattern (Ti. 37c ff.,
47a–b; Leg. 898c–9b). Such regular motions, Plato thinks, could not
be produced by material nature on its own, but require a living and
intelligent cause.
163
plato
Plato does not infer the existence of a creator directly from the
orderly movements of the heavens; rather, he holds that the universe
and the heavenly bodies are themselves alive and rational (Ti. 34b ff.,
39e ff.). This is another application of the claim that only soul can
act for a purpose; the heavenly movements are seen as good, and so
as requiring a living cause. In the first instance this is a reason for
believing in souls internal to the world. But the presence of such souls
is itself a good state of affairs; the universe is a living and intelligent
being, which Plato thinks is the best state for anything to be in (30b).
Since this state of affairs had a beginning, it must have had a cause,
and since it is a good state of affairs, this must be a purposive cause.
Plato’s God, therefore, is not related to the universe in quite the same
way as the God of eighteenth-century theism, who is often compared
to a clockmaker; what he makes is alive, not purely mechanical.
Plato’s argument here, from the purposive arrangement of the
universe to the existence of a creator, is a version of what has come to
be known as the teleological argument or the argument from design.
However, it is not the same as the version of that argument most
often discussed nowadays, made famous by David Hume in his Dia-
logues Concerning Natural Religion. This argument turns on an ana-
logy: the universe and living things resemble works of human design,
and there is therefore reason to believe that their cause resembles a
designer. This argument is questionable, because there seem to be,
in nature, causes of purposive arrangement other than design; for
instance living things produce purposive arrangements by reproduc-
tion. While there may, for all we know, be a designing mind behind
every natural phenomenon, acting either directly or indirectly, we
do not observe this, and so cannot claim empirical justification for
the claim that purposive arrangement must proceed from a mind.
Plato’s argument, by contrast, is conceptual. If something exists for a
purpose this must have been given it by something living and intel-
ligent, since material things cannot, on their own, act for a reason.
Of course, for this argument to have any force we must accept that
the universe and things in it have purposes. This does not only mean
that they serve purposes – that they are so arranged as to promote
good ends – something that Plato believes, but that others might
164
god and nature
question. It also means that this is not an accident. In some cases

something might produce a good although it does not exist for the
sake of that good, since its cause would have produced it anyway
whether it was good or not; in such a case that good is accidental.
Even if things in the universe seem to have a purpose, and so call
for a rational explanation, we need not accept the idea that there
is a creator if we have strong enough other reasons for rejecting it.
Although this is disputed, this argument seems to show that the
creator is a soul. It also shows in what respect he is like the human
soul: in being a self-mover, and therefore able to act for a reason. It
allows him to be very unlike us in other ways. He is a purely rational
being, without spirited or appetitive elements, and it is reasonable
to think that his response to reason is immediate, with no need to
make up his mind. Conceptions of God often have to steer a course
between the twin dangers of making him too anthropomorphic, and
making him so unlike us that we cannot conceive his nature at all.
Plato’s conception can overcome this problem.
Plato’s God and the God of theism
Plato’s God, a creator external to and prior to the world, is clearly in

many respects like the God of the traditional monotheistic religions,
and certainly more so than some other ancient conceptions of God:
Aristotle’s detached “unmoved mover” or the immanent spirit of the
Stoics. However, there are some significant differences.
Perhaps the most immediately striking is that Plato’s God is not
alone: there are other gods, his “children”, who help him. The uni-
verse is a god, since it is an immortal living being (Ti. 34a, 92c),
the heavenly bodies are gods (40a ff.), and at least the possibility of
other gods is left open (40d ff.). These gods take part in the work of
creation, making human and animal bodies and the lower elements
in the soul (42e ff., 69c ff.).
One may ask whether Plato is a monotheist or a polytheist, but
in fact it seems quite possible to be both, since monotheists and
polytheists will typically use the word “god” in different ways. If
165
plato
“god” means a spiritual being of superhuman power and virtue, Plato

believes in many of them; if it means a perfect creator and governor
of the universe, he believes in only one.
Plato’s creator is detached from the day-to-day workings of the
universe, so if there are divine powers within it – exercising provi-
dence, answering prayers, providing inspiration to prophets and
poets and so on – they must be distinct from him (see Symp. 202e–
3a). But when Plato thinks of the whole universe as embodying a
design, he naturally thinks of a single designing mind.
This detachment from the universe is another difference between
Plato’s God and the God of traditional theism. After the initial act
of creation he withdraws, leaving the ongoing work of running the
world to the lesser gods (Ti. 42e), probably because if he remained
involved in the world this would involve him in its processes of
change and so detract from his perfection.
Yet this detachment should not be exaggerated. At Timaeus 41b
we are told that it is only through his will that anything generated
can remain in being forever. In the immediate context this relates to
the heavenly gods, but it must also apply to the universe as a whole
and to the rational element in the human soul, which are likewise
generated but immortal; it is the creator’s will that sustains them in
being. God does not intervene – his relation with the universe is
unchanging – but he is still in a sense present.
Another striking difference is that Plato’s God is not the source
of all being. First, he does not create the Forms, which are uncre-
ated and eternal. Rather, he takes the Forms as his model in making
the world, trying to reproduce in the world the goodness that he
finds in them (29a, 37c). The Forms, because they are timeless and
non-spatial, cannot by themselves be the cause of events in time
and space. God bridges the gap between Forms and the material
world, producing events that happen in time and space, but using
the Forms as a guide.
Secondly, God creates the world not out of nothing, but out of
a pre-existing material (30a, 52d ff.). This is important, because it
provides the motive for creation. God wanted everything to be good;
finding the sensible realm in a state of disorder, which is not good, he
166
god and nature
set out to improve it. If all that had existed in the first place had been
perfect – God himself and the Forms – it is not clear that it would
have been good to create a universe, which is necessarily imperfect.
But if things were originally in a bad condition, God is to be praised
for making them better; and the magnitude of his achievement, even
if it is imperfect, can be appreciated by comparing it with the chaos
that was there before.
This brings us to our final point: it does not seem right to describe
Plato’s God as omnipotent. While this term is hard to define, it seems
that we should withhold it in this case, for there is something exter-
nal that sets a limit to what he can do. Although he controls the
world and makes everything serve his purpose, there remains a
disorder that he cannot overcome, an imperfection that he cannot
eliminate. Thus at Timaeus 48a we are told that (divine) intelligence
persuaded (material) necessity to guide most of the things that come
to be towards the best, and at many places that good results were
achieved as far as possible.
The constraint on God’s power arises from the nature of the
materials with which he works. An example is found at 75a ff., in
the description of the creation of the human head. Here we are told
that it was not possible to combine thickness (which would have
given more effective protection) and sensitivity; the creator had to
choose, and chose sensitivity. It does not seem that the combination
would have been logically impossible; rather, the necessity inherent
in the materials prevents it.
It is important not to overstate these limits on God’s power. We
must not suppose that his purpose is frustrated; his goal is to make
the world as good as possible, and this is achieved. Yet there are per-
fections that the world cannot achieve, so that it does not perfectly
represent its model. It falls short, not of the goal, but of the ideal.
Nor should we think that there is an arbitrary limit to what God
can do; the limit is inherent in the project he is involved in. Any
image of Forms will fall short of Forms in some respect. It will be
changeable, and will be embodied in some material (see Ti. 52c),
which will set limits to what can be done, although the precise nature
of those limits may depend on what the material is.
167
plato
The phrase “the best of all possible worlds” belongs to an age much
later than Plato’s, but it expresses well the spirit of Plato’s thought
about the universe. But the phrase is ambivalent: on the one hand it
celebrates the goodness of the world; on the other it laments its lim-
its. It means that the evil that is present in the world is inescapable;
nothing better is possible. Both aspects of the concept are present
in Plato. He sees the goodness of the world as cause for celebration,
but he also sees evil as inevitable.
Plato’s conception of God is an interesting one, and deserves to be
taken seriously. It may overcome some problems that the traditional
conception gives rise to, in particular the problem of evil, which
arises in a particularly acute way if God is seen as omnipotent and
as the source of all being. It is often thought that we have a choice
between the traditional conception of God and no God at all. But
if we accepted a Platonic God – a powerful, wise, benevolent, but
limited, designer and maker of the universe – we could not dismiss
the idea as unimportant; such a being would be an appropriate object
of religious feeling.
Necessity
We should now look more closely at necessity, the constraint under

which God works.2 We have seen that, while only soul is capable of
intelligence, material things are governed by necessity. This gives
rise to two different kinds of explanation: things brought about by
an intelligent cause are explained in terms of the purpose they serve;
things brought about by material causes in terms of necessity, that
is by showing how, given a certain event or arrangement of things,
a particular result is necessary (Ti. 46d–e). When intelligence and
necessity work together – when an intelligent being makes use of
material things and exploits their powers – both kinds of explana-
tion may be relevant. For instance, vision is explained in both these
ways. At 45b–6c we have an account of the mechanism of vision. A
beam of light, flowing from the eyes, picks up the motions of what-
ever it comes into contact with, and transmits them back to the eyes.
168
god and nature
At 47b–e the purpose of vision is explained. Its primary purpose,

Timaeus claims, is to enable us to see the movements of the heav-
ens, study of which will move us to philosophy. Many such double
explanations are found in the later part of the dialogue.
This distinction determines the shape of the Timaeus. The first
part, which deals with the creation of the world as a whole, the heav-
enly bodies and the rational human soul, focuses on intelligence. The
second part is concerned with materials used by God in creation and
focuses on necessity. The last part tells us how these materials were
used in the making of the human body, and here there are explana-
tions in terms of both intelligence and necessity.3
While necessity is a constraint on the power of intelligence, it
can also have a positive significance. Timaeus says that “intelligence
prevailed over necessity by persuading it to guide most of the things
which come to be towards the best, so that this cosmos was produced
by necessity subordinated to wise persuasion” (48a, emphasis added).
This, I suggest, should be taken seriously; it is necessity that guides
things towards the best, and plays a part in producing the cosmos;
the natural powers of things, governed by necessity, are put to posi-
tive use. Yet the references to intelligence prevailing and persuading
show that necessity originally existed independently of it, and was
not then a force for order.
There has been much controversy about just how we should
understand Plato’s use of the term “necessity” here, but I want to
defend the view that it refers to what we would now call natural or
causal necessity. Material things, for Plato, have natures that deter-
mine how they behave in specific circumstances.4
There are a number of reasons for thinking this. First, it makes
sense of the original contrast between intelligence and necessity.
Material things are seen as incapable of intelligence because, given
the circumstances, they must behave in a certain way, and have no
control over their movements. Next, the part of the Timaeus devoted
to necessity bears this out. It shows that material things have definite
natures (based on geometrical forms), and that these determine the
ways they move and interact, the ways in which different bodies can
be transformed into one another and the effect they have on us in
169
plato
perception (Ti. 53d–68c). This way of reading necessity shows how

the creator can make use of material things as instruments in pro-
ducing the cosmos – a tool, to be useful, must be relied on to behave
in a particular way – but also how they can set limits to his powers.
Finally, this reading shows how necessity can provide an explanation
of a state of affairs, since we have explained an outcome if we show
how it arises necessarily from what goes before.
However, the concept of necessity, so understood, gives rise to two
major problems. First, Plato associates necessity with disorder. He
describes it as “the wandering cause” (48a); he says that when sepa-
rated from intelligence it produces “chance and disorderly effects”
(46e); and he links it with the state of things before the universe was
made, which was one of “discordant and unordered motion” (30a).5
When controlled by intelligence it can be a force for order, but left
to itself it is not. We naturally associate necessity with order. Can we
make sense of Plato’s linking it with disorder?
Secondly, we are told that intelligence persuaded necessity to guide
most of the things that come to be towards the best; that is, the
creator induced necessity to do something it would not have done
of its own accord. But is not necessity unalterable? How can it be
persuaded to serve a purpose?
The problem of disorder
We tend to associate natural necessity with order and regularity;

and there certainly is a link. If anything is governed by natural
necessity, its behaviour is regular in the sense that there is a rule
to it; given specific circumstances, it can be relied on to behave in
a specific way. This does not, however, mean that it is regular in the
way that the heavenly bodies are (or seem to be) regular. It does
not have to be periodic, returning to its starting-point at definite
intervals; nor need it be easily predictable. It is plausible that when
Plato refers to necessity in the absence of intelligence as producing
disorder, this is what he has in mind: periodicity is a kind of stabil-
ity, which is something he values, and predictability helps to make
170
god and nature
the world intelligible; their absence would indeed make the world
a worse place.
We are also told that necessity in the absence of intelligence pro-
duces chance events; but this need not mean that they are wholly
undetermined; it may mean only that they have no purpose and are
unpredictable. That “chance” can be understood in this way is shown
by the existence of games of “chance”. The fall of a die is governed by
deterministic laws, so that if I know the exact speed and direction
with which it is thrown I can predict how it will land; but in practice
it is unpredictable and so can be called a chance event.
In fact, while we generally think of the material world as, at least
to a large extent, governed by deterministic laws, much of it is in
practice unpredictable; the weather, the incidence of natural events
such as earthquakes, or indeed health and disease in the human
body, cannot be predicted in detail. For a long time, science, while
not of course denying this, tended to neglect it, focusing on systems
that were predictable, but recently the situation has changed, and
scientists have begun to take fuller account of the disorderliness and
unpredictability of the natural world; this may make it easier for us
to appreciate Plato’s point of view.
One aspect of this change has been the development of chaos
theory, according to which some systems have sensitive depend-
ence on initial conditions, which means that if their starting-point is
changed by however small an amount, they will develop completely
differently. Such systems are deterministic – there is a rule that gov-
erns their behaviour – but not periodic, and not predictable unless
the initial conditions are known with perfect precision, which is not
possible for a finite being.
There are also some systems that are not chaotic in the technical
sense, so that they are in principle predictable, but that cannot be
predicted in practice because of the complexity of the phenomena
involved. Here again scientists have recently begun to focus on this
area, treating complex systems as interesting in their own right rather
than neglecting them in favour of simple ones.
It is clear that, much of the time, natural necessity does indeed
produce systems that are in some way disorderly. Why, then, does
171
plato
the idea of a connection between necessity and disorder often seem

alien to us?
One reason may be the fact that we are familiar with large-scale
machines, and often take them as a paradigm of the behaviour of
things governed by necessity; it is very natural to refer to “mechani-
cal” necessity. Machines do indeed behave in a visibly regular and
orderly way. But in fact they are not typical of things governed by
necessity; they are cases where necessity has been organized by intel-
ligence to serve a purpose.
However, perhaps the major reason why Plato’s way of thinking
looks alien to us is that, since Newton, we have had an explanation,
in terms of necessity, of the movement of the heavenly bodies: the
very phenomenon that Plato saw as evidence of intelligence (not just
in the creator, but in the world itself), because they are so orderly,
and necessity for him was linked with disorder. Newton showed
that necessity can produce this kind of order. But after Newton it
became easy to see the heavens as paradigms of necessity, giving rise
to the sense that whatever is governed by necessity should have the
same regularity and predictability. This, I suggest, was a mistake; the
heavenly bodies are in fact unusual among natural phenomena, and
while Plato was wrong to see them as rational beings, his view was
not absurd given how different they are from most things governed
by necessity. In many cases Plato was right to see necessity, left to
itself, producing disorder, and intelligence introducing order.
The problem of persuasion
The other puzzle that Plato’s use of the concept of necessity poses
is that he refers to intelligence persuading necessity to guide things
towards the best. Surely necessity is unalterable, in which case it
makes no sense to speak of persuading it.
Glenn Morrow (1950) has proposed an answer to this problem
that I think we should in essence accept. Persuasion contrasts with
compulsion; whereas compelling something means making it act
in a way contrary to its nature, persuading it involves exploiting its
172
god and nature
nature to bring about a desired result. In the Phaedrus, when dis-

cussing literal persuasion in the context of rhetoric, Socrates claims
that we tailor our methods to the nature of the soul of our audience,
exploiting the qualities of that soul (Phdr. 271b ff.).
Material things have natural powers that determine how they will
act when placed in a particular arrangement. Intelligence – human or
divine – can then exploit this by creating new arrangements in which
the powers can interact so as to produce a desired result. This result
will be explained through necessity (since, given the natural pow-
ers of the materials, it followed necessarily from the way they were
arranged), but also through intelligence (since intelligence brought
about the arrangement).
However, this account still leaves us with a puzzle. Before intel-
ligence acts there will already be an arrangement of materials that,
left to itself, should determine what events will follow. Hence, intel-
ligence does not just produce new arrangements, but alters the
arrangement things are already in, and in doing so it seems that it
overrules the necessity that governs their movements.
This means, I think, that we have to accept that the necessity that
governs material things is not absolute; it determines how things will
behave if nothing else intervenes. From this it follows that if the mater-
ial world were isolated from outside influences, its behaviour would
be wholly determined; it would be as it was imagined by Plato’s older
contemporary Democritus, a materialist, who held that everything
happens by necessity (DK Democritus 68 A1, A66), and as we have
seen this would mean that nothing happened for a purpose. But for
Plato the world is not so isolated; it is constantly influenced by souls,
not only by God in creation but by other souls, human and divine,
throughout its history
Souls, because they move themselves, are independent of material
forces, and they in turn can act on the material world. They do not
eliminate necessity, since material things go on behaving according
to their determinate nature, but they modify its effects, creating new
arrangements in which it has desirable results. If there were not souls
the universe would be deterministic and so bereft of purpose; but
because there are souls it is not.
173
plato
The origin of evil
Although Plato believes that this world is good, he recognizes that

there are many imperfections in it. In the first place, simply because
it is a material thing and not a Form, it will fall short of its model in a
number of ways. One of these is mutability; it cannot be unchanging
as the model is. This is revealed, for instance, in the way in which
it cannot possess eternity in the fullest sense, but only time, which
is an image of eternity (Ti. 37c ff.). Another is the way in which it is
necessarily embodied in some material, which will impose limits on
what can be achieved (52b–c).
However, these factors mean only that some goods cannot be
achieved. Plato also believes that there are some positive evils in the
world: disorder in the natural world; irrationality in the soul. How
do these arise?
We have seen that before God took control of the material world
it was in a chaotic state of “discordant and unordered motion” (30a).
While God succeeded in making the world for the most part orderly,
he did not completely eliminate this disorderly motion; we are told
that it continues within the ordered cosmos (57d ff.). It seems that
even within the cosmos the material factor cannot be brought com-
pletely under divine control, and therefore continues to produce
damaging results.6
Timaeus does not say much about the effect of material disorder
on the world at large. (Elsewhere Plato shows some interest in catas-
trophes, such as the legendary flood, and their effect on history. These
are discussed at the beginning of the Timaeus, in the Atlantis story
[22c ff.], and in the Statesman [270c–d, 273a] and Laws [676a ff.].)
The emphasis here is rather on the effect of material disorder on
individuals. Various kinds of evil that can affect human beings are
traced to physical causes. Old age and death are caused by the decay
of the particles that make up our bodies (Ti. 81c ff.); disease is caused
by imbalance among the body’s material constituents (82a ff.). But
material causes are also found for the evils that affect the soul.
As we have seen, there are for Plato two primary kinds of evil, or
sources of wrongdoing, in the soul: ignorance, affecting the rational
174
god and nature
part of the soul, and the rebellious tendencies of the lower parts
of soul, spirit and appetite. Both of these in the Timaeus seem to
have bodily causes. At 43a ff. we are told how the rational soul, in
infancy, is plunged into the flood of matter, both internal to the body
(the bloodstream) and external to it (in sensation); this disrupts it,
halting the operation of pure reason, disordering that of perceptual
judgement. Later the soul recovers and we achieve intelligence, but
we need proper nurturing to recover fully.
The lower parts of the soul are not themselves of material ori-
gin, but are created by the gods. They are described as dangerous,
but necessary; presumably we need them to survive in the material
world (69c ff.). However, the disorder of these parts, which leads
them to behave in a rebellious manner, is later shown to have a bod-
ily cause (86b ff.). Diseases of the soul are divided into ignorance
and madness. Madness is said to be produced by excessive pleasures
and pains, pleasure producing indulgence, while pain produces bad
temper and melancholy, cowardice and recklessness. These pleasures
and pains are themselves traced to bodily causes, and the fact that
disorder in the soul has a bodily origin is used in support of the claim
that “no one is willingly bad” (Ti. 86d–e).
It seems, then, that matter is both disorderly in itself, and a source
of disorder in the soul. This harmonizes well with the Phaedo and
Republic, in both of which body is seen as a cause of evil. Yet this
is in some ways puzzling. For one thing, if matter has a disorderly
motion of its own, going back to the primeval chaos, before soul
was introduced into the world, what is the source of that motion? If
self-motion is seen as a distinctive characteristic of soul – and the
language of self-motion is used repeatedly in the Timaeus – then it
might seem to follow that soul is the source of all motion. I suggested
in Chapter 5 that this problem can be overcome if we are prepared
to tolerate an infinite regress. If we trace a series of motions to a
source, this source must be a self-mover; but it is possible that there
are series of motions that have no source, but go back infinitely into
the past; and the disorderly motion of matter in the Timaeus, which
goes back to the primeval chaos, may be of this kind. Nevertheless,
some readers will find this solution an uncomfortable one.
175
plato
In any case, there is another problem that does not turn on a

particular theory of soul; Plato’s own conception of matter makes it
hard to see how it could give rise to disorder. For him the material
world consists fundamentally of the receptacle, space, characterized
by qualities that are images of the Forms. The receptacle itself is said
to be neutral and not to distort the images that it receives (50d–51b).
But in this case, while it may indeed set limits to what the creator
can do, it is hard to see how it could give rise to positive disorder.
There seems to be a real tension in Plato’s thought here. On the
one hand, as part of his general project of contrasting the material
world and the Forms, he seeks to give as reductive an account of
matter as possible. On the other, he wants to explain disorder in the
world, without attributing it to God, and this requires a positive,
active independent element. This is also required by Plato’s view of
the motive of creation, according to which there was once much
more disorder in the material world, before God took it over, and
God’s aim was to eliminate that disorder.
Cosmology in the Laws
In Book 10 of the Laws, in the course of a defence of religion, Plato

returns to the topic of cosmology. In some ways the position taken
here is very similar to that of the Timaeus. There is a creator God,7
although he remains in the background; the gods we are urged to
worship are those of the heavens. The world is good and ordered for
a purpose, and the regular movements of the heavenly bodies are
used as evidence for this. But in some other ways, the Laws seems
to reveal an outlook very different from that of the Timaeus. Since
this is probably Plato’s last treatment of the subject, we should briefly
focus on it.
The chief speaker of the Laws, the Athenian Stranger, advances
an argument against materialism that rests on the concept of self-
motion (Leg. 894d ff.). Everything that moves, he argues, must
be moved either by itself or by something else; and if we trace a
sequence of motions back to its source, we will find that the first in
176
god and nature
every sequence must be a motion that moves itself. For this reason
it cannot be true, as materialists think, that the soul is just an epi-
phenomenon of material changes; it is something fundamental to
the universe.
The Athenian claims that if soul is the source of all motion, it is
the source of all good and evil; and there must be more than one soul
at work in the world, since there is at least one soul producing good
and one producing evil (896d–e). He goes on to argue, however, on
the basis of the heavenly movements, that the souls that are most
powerful in the universe are good ones, and can be identified with
the gods (898c ff.).
It will be seen that the Athenian’s argument turns on an assump-
tion that might be questioned: that there cannot be an infinite regress
of motions. Timaeus (and the Stranger in the Statesman) seemed
prepared to accept an infinite regress, tracing material motion back
to the primeval chaos (Ti. 30a, 52d ff.; Plt. 273b–c). In the Laws, by
contrast, the Athenian starts from an assumption, which he says
many materialists share, that the world was once at rest; in this case,
he argues, only soul could have set it in motion (Leg. 895a).
This at once gives us a view very different from that of the Timaeus;
it makes soul prior to all bodily motions. In fact, however, it seems
that an even stronger claim is being defended: that soul is prior to
body itself (892a, 896a ff.). This can be seen as justified if we see
generation as a motion; if soul is prior to all bodily processes, this
must include the process that brings bodies into being. (Plato may
well be assuming, as he does in the Timaeus, that all bodies are gen-
erated in some way.) It seems, therefore, that in the world as it is
envisaged here, God does not create order out of disorder. It may be
that, like the God of traditional theism, he is to be seen as creating
the world out of nothing; soul is referred to at a number of points
as generated, presumably by God (892c, 896a, 967d), and body in
turn is generated by soul.
The Athenian draws two further consequences from the claim
that soul is the source of all motion: first, as we have seen, that it is
the source of good and evil, and secondly that all the attributes of
soul are prior to matter – these turn out to include not only rational
177
plato
attributes, but such things as opinion true and false, joy and grief,
hope and fear, love and hate (896d–7a). This seems to go against
the position expressed in the Timaeus, and also to some extent in
the Phaedo and Republic, that body is a source of evil, and that the
disorderly elements in soul arise from the body.
If the independent bodily factor in the world is removed, one
might well think that the element of necessity, which limits the cre-
ator’s power, will also disappear. The Laws is not very explicit on this;
we are told that the gods “can do anything which is within the power
of mortals and immortals” (901d) – that is, if anyone can do it, they
can – but this need not mean that they are omnipotent. However,
earlier in the Laws the Athenian said that gods are subject to “divine”
mathematical necessities but not to the necessities of ordinary life
(818a–b); this could be read as meaning that the only limits on their
power are logical and mathematical ones.
Given this new perspective, which seems to eliminate the inde-
pendent, disorderly material factor, we need to reconsider the origin
of evil. Book 10 of the Laws does include a consideration of the prob-
lem of evil, but in a particular form: the problem of injustice, why
the wicked prosper and the innocent suffer (899d ff.). The Athenian
claims that God has so ordered the universe that this is rectified in
future lives, with good souls ascending to a better situation, bad
souls descending to a worse one. This, however, does not explain
how wickedness arises in the first place. The Athenian insists, in
line with the interpretation of self-motion found in this work, that
the cause of good or evil character lies within the soul, although
we may sometimes be influenced by other souls (904b–d). But this
remains puzzling; if there is no material cause of evil, why should it
arise within the soul? Certainly not by a deliberate act of choice; it
remains the case in the Laws that “no one acts unjustly except against
his will” (860d). The causes of wrongdoing are ignorance, anger and
pleasure; these are internal to the soul, but it is not clear how they
arise there if not by a bodily cause.
Hence, it seems that the position taken in the Laws overcomes
some of the problems found in the Timaeus, in particular that of
how the bodily element, as Plato conceives it, can be a source of
178
god and nature
disorder, but at the same time gives rise to problems of its own. It is
interesting, however, as it gives evidence that Plato was rethinking
these issues up to the very end of his life.
179
nine
Aesthetics
One of the most striking features of Plato’s thought is his attitude

to the arts, especially poetry, revealed in the severe restrictions he
places on them in his account of the ideal state. It is sometimes
thought that he banished poetry completely from that state, and
indeed some of his own language suggests that,1 although in fact
he allows a place for hymns to the gods and odes in praise of good
people (Resp. 607a). But it does seem that he condemns almost
all of the poetry that existed in his time, and in particular that of
Homer and the tragedians,2 often seen as the supreme achieve-
ments of ancient Greek literature. Moreover, while the main focus
of his criticism is poetry, he does also propose restrictions on music
(398c ff.), and on visual art, including the design of buildings, fur-
niture, clothing and so on (401a ff.). In this chapter, we shall look at
Plato’s thought about the arts so as to see his reasons for restricting
them in this way.
The term “aesthetics” is in some ways anachronistic in the dis-
cussion of Plato. For us, it includes both the philosophical consid-
eration of the arts, and that of beauty and related properties and
our experience of them. While, as we shall see, Plato did think that
beauty can be manifested in the products of art, he would not have
seen the topics as particularly closely connected. We shall be focus-
ing here on his account of the arts.
181
plato
However, even the term “art” is anachronistic in discussing Plato;

he has no term that translates as “art”, except technē, which corre-
sponds to the wider sense of “art” in which it is equivalent to “craft”
or “skill”. In fact, the idea that painting, poetry, music and so on
form a distinctive group, which can be identified as “art” in a more
restricted sense, did not develop in its modern form until the eight-
eenth century. However, Plato does have definite thoughts about
many of the activities and products that we would describe as art.
He is mostly concerned with poetry, but poetry for him is closely
connected with music – the Greek word mousikē covers both – as
it was often written to be sung or recited publicly. He also draws
analogies between painting and poetry.
Poetry and censorship
Discussions of poetry are found in two places in the Republic: first in

Books 2 and 3, and then in Book 10. The first discussion forms part
of Socrates’ account of the education of the guardians, the military
class, from whom the future rulers of the state will be drawn; the
question is what kind of poetic works are suitable for use in their
education. In the later discussion, however, Socrates seems to make
a more general claim: that poetry should be severely restricted in the
ideal state as a whole, not just in education.3 Moreover, the later dis-
cussion seems to propose more radical restrictions on both the form
and the content of poetry. Although the earlier discussion proposed
fairly severe censorship, it was made clear that some existing poetry,
including passages from Homer, would survive (see e.g. Resp. 389e).
In the later discussion this no longer seems to be the case.
Socrates proposes restrictions on both the content and the form of
poetry. His restrictions on form will be discussed in later sections. As
far as content goes, he seeks to forbid poetry that promotes views that
he finds damaging: for instance, that the gods produce evil as well as
good (397e ff.); that they are changeable, and that they can be deceit-
ful (380d ff.); that death is an evil and should be feared (386b ff.); and,
by implication, that a just life is an unhappy one (392b–c).4
182
aesthetics
Plato’s apparent support of such radical censorship is in itself dis-

turbing. In the context of education, the idea that books should be
selected in such a way as to encourage the development of enlight-
ened views is not wholly unreasonable, but even there we might
well think that students should be presented with different views
and allowed to discuss them and make up their own minds, rather
than simply being fed a diet of material that presents views that
their teachers approve of. The way in which Socrates restricts the
educational diet of his young guardians is particularly striking when
we consider that he is here concerned with the education of the
most intelligent part of the population, some of whom will go on to
become philosophers.
In any case, in Book 10 Socrates goes on to propose restric-
tions on poetry in the state as a whole, not just in formal educa-
tion (595a, 607a ff.). However, it is clear there that he is still looking
at poetry from an educational point of view; he is concerned that
people will look to it for guidance and understanding, and are liable
to be deceived by it if it is not grounded on knowledge. This does
indeed seem to be how poetry was often seen at the time. Homer
was regarded as “the teacher of Greece” (606e), and people looked
to him as a source of wisdom. In the light of this Socrates’ readiness
to censor his work is more understandable.
Nevertheless, it is still disturbing if Plato envisaged the people
of his ideal state being taught only the truth as he saw it, and not
being presented with alternative positions. Perhaps the explanation
of this is, once again, a strong conviction of the power of reason to
find the right answer. If we think the truth is uncertain, we will want
to consider different possible views and make up our own minds;
but if the truth is known, why should we concern ourselves with
deceptive views? This does not mean that Plato expects ethical and
philosophical truths to be accepted simply on the basis of authority;
he believes that philosophers should come to grasp these truths for
themselves, by the use of their own reason. But perhaps he sees ethi-
cal and philosophical truth as rather like mathematical truth. There,
although we should indeed follow proofs for ourselves, we can be
confident that if we reason properly we will reach the right answer
183
plato
and agree on it, and therefore it is pointless to consider rival views.

For Plato, the same will apply to moral and philosophical truth.
The concept of imitation
In Book 3 of the Republic, having set out his proposals for the content
of poetry, Socrates turns to its form. His discussion here turns on the
concept of mimesis, traditionally translated “imitation”; this concept
will also be central to his discussion of poetry in Book 10, and has
had an important impact on later aesthetic theory, so it deserves to
be looked at in some depth. The basic idea expressed by the term
is that of copying someone or producing a copy of something. It
can be used in contexts as various as mimicking someone’s voice,
and imitating someone’s behaviour when one takes them as a moral
example. There is always an implication of likeness; the imitator
becomes like the thing imitated or produces something like it. For
this reason, “imitation” is a translation that does capture the basic
force of the Greek term, although it will not be idiomatic in all the
contexts where Socrates uses it. For instance, we do not normally
speak of an actor imitating the character he plays, but one can see
what would be meant by saying this; he acts, in some respects at least,
as that character would act.
In Book 3 the term “imitation” is used to denote a specific kind
of poetry (Resp. 392d ff.): dramatic poetry, where an actor takes the
part of a character, and some parts of epic poetry, the speeches, in
which the reciter likewise takes the part of a character and says what
that character would say. Both the actor or reciter and the author
of such poetry can be spoken of as “imitating”. This contrasts with
other kinds of poetry, including other parts of epic, which proceed
in a straightforwardly narrative way; the reciter or singer simply
recounts events and does not speak in the voice of a character. In
imitative poetry, the speaker is himself becoming like the character
whose part he plays. Socrates seeks to restrict, although not wholly
eliminate, poetry that is imitative in this limited sense. In particu-
lar, he forbids the imitation of morally bad people, mad people and
184
aesthetics
people of lower classes; he does not here forbid the description of

such characters, and it is not clear what stories it would be possible
to tell without mentioning them.
In Book 10, by contrast, “imitation” is given a wider sense. It is
applied to painting (597e), and also to narrative poetry. Socrates
speaks of Homer “imitating medical language” (599c), although in
the passages he seems to have in mind, Homer is not actually taking
the part of a doctor but simply describing what a doctor might do.5
In these passages, therefore, the point is clearly not that the artist
becomes like what he depicts, but only that he produces something
like it. In the case of visual art it is clear enough how he does this; a
painting does indeed look, in some ways, like the thing it represents.
In poetry the relation is more complex; a poem about a battle, for
instance, does not actually look or sound like a battle. But it may
bring a battle before the mind, in such a way that the listener will
say “a battle is like that”.
However, in Book 10 Socrates gives a further twist to the concept
of imitation: he uses it specifically for something that reproduces
the appearance of a thing rather than the reality – that looks like the
thing it represents without being really like it (597e–8c). This is the
sense of “imitation” found in expressions such as “imitation marble”,
or in the warning to beware of cheap imitations. In the case of paint-
ings this applies in a fairly simple way; a painting of a bed looks like
a bed, but is not really like a bed – one cannot sleep on it. In the case
of poetry, even the similarity of appearance is more abstract, but once
again the poem is not really like the thing it represents. This new
conception of imitation is particularly emphasized in the first argu-
ment of Book 10; there poets are presented as imitating their subject
matter in the same sense in which painters do.6 In the later part of
Book 10 Socrates goes back to speaking of poets imitating people,7
and once again largely has dramatic poetry in mind; the arguments
here are more closely linked with those of Book 3.
These two uses of “imitation” fit together rather uncomfortably.
It is quite possible that the same work might be imitative in both
senses. But they ground different criticisms of poetry that are hard
to combine. In Book 3, and the later part of Book 10, the objection to
185
plato
imitation turns largely on the thought that we might actually become

like what we imitate. In the first argument of Book 10, however, the
central point is that an imitation is not genuinely like what it imitates;
it reproduces only the appearance and not the reality. There is no
direct contradiction, because the objects of imitation discussed in
the two passages are different; poetry might produce a false appear-
ance of some craft – war, statecraft or education – while actually
producing harmful emotions, bringing about a genuine likeness to
the object of imitation in that sense. Nevertheless the double use of
the term “imitation” is confusing.
Before going on to look at Socrates’ arguments in more detail,
one other point should be noted. Plato has inspired the view, an
influential one when the modern concept of art was being developed
in the eighteenth century, that imitation is the distinctive feature of
art: what all the arts have in common, and perhaps what gives art its
purpose. However, Plato’s own works contain no such view; indeed,
they could not, since Plato has no general term for art. Nor is there
any reason to think that he would have seen it as essential to all the
activities that we describe as art. Although he uses the term in con-
nection with poetry, visual art and music (see Resp. 399a–c), he need
not have thought that, for instance, architecture was an example of
imitation. Indeed, it does not even apply to all poetry; in the sense of
“imitation” used in Book 3 it is clear that much poetry is not imita-
tive, whereas in the wider sense used in Book 10 it is less clear, but
he does seem to leave room for some that is not.8 Plato’s Socrates
is simply identifying imitation as a central feature of some of the
activities we call art, and criticizing them on that basis.
Imitation and character
In Book 3 of the Republic Socrates raises the question whether the

poetry used in the education of the guardians should be imitative,
in the sense where this involves the speaker taking a part (Resp.
394e). Since the students would themselves have recited the poetry
they were learning, not just listened or read silently, this question
186
aesthetics
can also be cast as whether the young guardians themselves should

be imitative. One might think that Socrates was going to say the
young guardians should not imitate at all, but in fact he allows some
imitation of good people: those who are suitable role models for
the guardians (385c). However, he rules out imitation of the bad,
the mad, those of lower classes and those in damaging emotional
states (385d ff.), and he thinks that as a result of this the amount of
imitation in the poetry taught to the guardians will be small (386e).
Alexander Nehamas (1988) and others have suggested that what
Socrates is really opposing here is poetry which pursues imitation
as an end in itself; such poetry would try to imitate as many kinds of
person and situations as possible (see also Janaway 1995: 100). Aris-
totle notes that we take pleasure in imitation, even of things that are
in themselves ugly (Ar. Poet. 1448b5–12). Plato holds that imitation
of this kind is undesirable. The guardians should imitate only the
kind of person they aim to become. Imitation is not an end in itself,
but something pursued for the sake of development of character.
Socrates’ official reason for rejecting indiscriminate imitation is
that it conflicts with the principle, central to his ideal state, that each
person should perform just one function (Resp. 394e). Just as people
in actual life should do just one kind of work, the kind they are best
at, so in poetic performances they should imitate just one kind of
person. One may wonder why this principle applies to imitation as
well as to actual life. To start with, Socrates seems simply to be draw-
ing an analogy: as a person cannot be good at doing many kinds of
work, so he cannot be good at many kinds of imitation (395a–b).
However, he then goes on to make a further point: that imitation, if
engaged in repeatedly, actually has an effect on one’s character, and
one becomes like what one imitates (395d). Here, there is not just an
analogy between doing and imitating. What we imitate affects what
we do, and if the guardians imitate bad people this will interfere with
their ability to do their distinctive function of protecting the state.
It certainly seems possible that imitation should have this effect;
taking someone’s part can lead us to sympathize with him, and so
may lead us to become more like him. It is not clear, though, why
this should always be the case. Might we not imitate someone while
187
plato
remaining clear that we are not him, and so avoid corruption? In

fact Plato might have accepted this; Socrates says at one point that
the guardians may occasionally imitate a bad person “for fun” or
“in play” (396e). But the danger of corruption seems always to be
present.
Plato returns to this theme in Book 10 of the Republic. The first
argument against poetry of that book turns on the false appearance
of knowledge it produces, and will be considered in the next sec-
tion. However, there is also a group of arguments that relate to the
themes of Books 2 and 3, but develop them further: Socrates argues
that poetry appeals to damaging emotions, and can corrupt a good
character (605c ff.).
Socrates claims that when we see, presented on the stage, some-
thing we would be ashamed of in real life, we nevertheless enjoy it
and sympathetically share the feelings of the people we see presented.
A part of the soul that we normally restrain – the part that contains
these shameful feelings – is strengthened and encouraged by this
process, so that we come to have these feelings in our own life. This
is developed at greatest length with grief. Great grief, according to
Socrates, is not an appropriate reaction to the death of loved ones,
but in tragedy we often see people indulging in grief, admire the
performance and feel sympathy with them. Something similar hap-
pens in comedy; laughable acts that we would normally be ashamed
of are accepted and treated with sympathy because they are seen as
just a joke. The same effect may result from depictions of other feel-
ings such as anger and sexual passion. In this way, we become like
the people we see represented on stage.
This resembles the claim of Book 3 that we become like what we
imitate, but goes beyond it in two ways. First, it extends the claim to
the audience; we do not have to be acting the parts ourselves to have
sympathy with the characters. Secondly, we are told something about
how it works; we feel able to sympathize with the characters because
we think of them as someone else, not ourselves, and so do not see
the emotions we are feeling as damaging to ourselves (606a–b). The
point is not that we directly identify with the character, but that,
precisely because we are aware of the difference, we allow ourselves
188
aesthetics
to feel what we would not otherwise feel. But, Socrates suggests, this
can be damaging; it is dangerous to allow these feelings even in a
purely fictional context.
This passage does not simply ignore the difference between enjoy-
ing a performance and taking pleasure in the act depicted. Rather, it
argues that, despite the difference, one can lead to the other. That it
can do so seems true; often a powerful performance will lead us to
sympathize with the character, even if he is not a person we would
normally have sympathy with. We may question whether this hap-
pens as widely as Plato seems to think.
Imitation and knowledge
Socrates’ principal attack on poetry in Book 10 of the Republic turns

on the idea that it makes deceptive and dangerous claims to know-
ledge. Socrates begins by considering visual art. This is done in order
to show that there is an activity that reproduces the appearance of
a thing, not the reality, and that this activity need not involve any
knowledge of the nature of the thing; this activity is then named
imitation. Having established this, he goes on to argue that poetry
is also an example of imitation in this sense.
Socrates approaches the subject through the theory of Forms
(596a ff.). A craftsman, producing, for instance, a bed, works “look-
ing to the Form” (596b). This need not mean, as becomes clear later
(601c ff.), that he has knowledge of the Form of bed – his grasp of it
may be indirect, and amount only to true belief – but he has some
grasp of what a bed is, of its purpose and how it fits together to serve
that purpose. A painter, by contrast, painting a bed, need not have
any awareness of the Form; he takes the visible bed as his model, and
paints on the basis of familiarity with that.
The next step is to claim that if something is copied from visible
examples of a kind rather than from the Form, it only reproduces the
appearance of the thing, not the reality (598a ff.). A painting of a bed,
for instance, reproduces the way it looks from a certain perspective,
not how it is in itself. One might try to reproduce something as it is,
189
plato
that is to make a new object of the same kind as it, as, for instance,
a maker of reproduction furniture does, but one could not do this
simply on the basis of looking at the original object; one would need
some grasp of the Form, that is of the purpose of the object and how
it worked. However, one can, just by looking at an object, reproduce
its appearance.
One should be careful not to over-interpret these claims. They
need not be taken to mean that the painter must be copying specific
individual beds that he can see; nor do they mean that his aim must
be to reproduce the exact appearance of the things he is taking as his
model. It means only that his work is based on his familiarity with
visible things, and aims to look like them. It need not, therefore, be
seen as a totally mindless activity. It can involve skill, and this will
become important when we come to apply the analogy to poetry.
Socrates is not denying that poets are skilful; if they were not, their
work would not be so dangerous.
While Socrates is here shown as having a low opinion of visual
art, seeing it as trivial, he does not seek to ban it from the ideal state.
The discussion of painting is intended to show that it is possible to
reproduce the appearance of something without any grasp of the
reality. Socrates then goes on to ask if poetry does the same (598d ff.).
When painting does this, it is not normally deceptive; most people
are able to distinguish between a painting of a bed and a real bed.
Socrates does point out that children and stupid people – who have
no experience on the basis of which to recognize a work of art – may
in fact confuse a painting with what it represents (598c). But such
confusion, one may think, is easy to overcome, and in any case will
not normally have disastrous consequences. However, if poetry is
imitative in the same way, it is much more likely to be deceptive, and
this deception will be dangerous.
Clearly, Socrates does not mean that in listening to a poem about
a battle or a feast, we may be led to think that a battle or a feast is
actually taking place. Rather, his point is that poetry may produce a
false impression of knowledge. The poetry of Homer deals with such
matters as medicine, war and statecraft. It reproduces the appearance
of such things, in that it describes them in ways that remind us of
190
aesthetics
our experience of them, and so it may give us the impression that

the poet has knowledge of such matters. This, then, may lead us to
see the poet as a source of guidance; and indeed Homer was seen in
this way in ancient Greece (598d–e).
However, Socrates argues, Homer does not really have such
knowledge. Socrates does not simply rely on the analogy of painting
in showing this;9 rather, he focuses on what was known, or thought
to be true, of Homer’s life (599b ff.). He argues that we can know
that Homer did not have knowledge of central areas such as war,
politics and education, since he had no achievements in those fields:
he was not a general, a legislator, the founder of a city, an inventor,
a practical educator or the founder of a school. In this, Socrates
proposes, he is typical of poets more generally.10 If poets really had
knowledge of the matters they write about, he suggests, they would
not choose to devote themselves to poetry, rather than to genuinely
useful achievements (599a–b); nor would others allow them to do
so (600d–e).
Many things can be questioned about this argument. For one
thing, it seems to assume that the actual achievements of a statesman,
a general and so on are more valuable than those of a poet. Moreover,
even if we grant that they are, it does not follow that everyone thinks
the same; if Homer, or his contemporaries, believed that poetry was
more valuable, even if they were wrong, this would explain why
he devoted himself to poetry. Some people are frustrated by public
opinion in carrying out their true vocations; Socrates himself may
be an example. Moreover, there were some poets who had other
notable achievements; for instance the Athenian statesman Solon,
who is mentioned in the Timaeus, was both a poet and an important
legislator who played a significant part in the development of the
Athenian constitution.11
Nevertheless there seems to be something right in what Socrates
says; granted that it is possible to create a plausible appearance of
something without an understanding of the reality, it is reasonable to
think that poets are often doing so, and that they have no expertise in
the matters they write about, and in this case it is indeed dangerous
to look to them for guidance.
191
plato
However, we should notice an ambiguity in the concept of imita-

tion that Socrates is using here. There is an activity that reproduces
the appearance of things, not the reality, and this activity need not
involve any knowledge of, or even true belief about, the reality. It
can be based, not on a grasp of what the kind of thing in question
is, but just on a familiarity with examples and their appearance. But
it does not follow that no one who engages in this activity has any
grasp of the reality, nor that a grasp of the reality cannot help them
in performing it (as when artists study anatomy in order to help
them in drawing human figures). In saying that poetry is imitative,
does Socrates mean only that it reproduces appearances of things,
not the things themselves, or that it is not based on knowledge of
the things themselves?
In his actual discussion of Homer he certainly seems to mean the
latter; he seeks to show that Homer lacked real expertise, and con-
cludes on that basis that he was an imitator (Resp. 600e). However,
if this is right, it seems to leave room for a kind of poetry that is
grounded on knowledge. Such poetry would still be imitative in the
sense that it reproduced appearances rather than realities; a poem
about a city, for instance, is not a city, and it is at least plausible to
think that producing such a poem is less valuable than producing
an actual city. But it would not be deceptive as the poetry of Homer
and others are; it would be guided by a real knowledge of statecraft.
While Socrates seems to condemn all existing poets as imitators in
a pejorative sense, he may leave open some room for a possible kind
of poetry that does not deserve this kind of condemnation. On the
other hand one may still ask what the value of this kind of poetry
would be. If one has real knowledge of important subjects, would
it not be better to make use of it in practical activity or in direct
teaching, rather than in the creation of poems? I shall return to this
subject later, after discussing another aspect of Plato’s view of poetry.
192
aesthetics
Poetry and inspiration
When we turn from the Republic to some other works of Plato, we

may be surprised to find a view of poetry that is at first sight very
different from that of the Republic. According to this view, poetry
is the result of divine inspiration, and so can be a vehicle of truth,
although not of knowledge. The gods speak through the poets, as
they do through prophets, although the poets do not understand
what they are saying. This idea appears in some Socratic dialogues,
the Apology and Ion, and is briefly alluded to in the Meno (99d),
which is generally thought to have both Socratic and Platonic ele-
ments, and also appears in the Phaedrus, a work normally seen as
reflecting Plato’s own mature thought.
The idea of divine inspiration of poets was not, of course, new
– ancient poets themselves, such as Homer and Hesiod, invoke the
muses to inspire them – but Plato emphasizes the irrational nature
of the process; rather than speaking to the poets the god possesses
them, as he does prophets and initiates in some mystery cults, and
speaks through them. This idea can be seen as having both positive
and negative implications. On the one hand, if we emphasize irra-
tionality and the fact that poets do not have knowledge, it seems a
negative view of them, opposing their pretensions to wisdom; if we
emphasize the connection with the gods, and the way in which this
enables the poets to reveal truth, it becomes more positive.
In the Socratic dialogues it is the negative aspect of this view of
poetry that is emphasized. In the Apology (22a–c), it appears as part
of Socrates’ unsuccessful search for someone who is really wise; poets
claim wisdom, but their claim turns out to be false, since their poetic
ability comes from inspiration, not knowledge. The Ion is a satirical
work, in which Socrates makes fun of Ion, a reciter of epic poetry
who claims great wisdom. Socrates tries to convince him that he is
not in fact wise, and that his success is due not to expertise, but to
inspiration; the god’s inspiration of the poet passes itself on to the
reciter and through him to the audience (see esp. Ion 533c ff.).
However, in the Phaedrus the same general account of poetry is
used to show it in a more positive light. Socrates is there defending
193
plato
(erotic) love. He has previously condemned it as a kind of madness,

but now he argues that, although it is indeed a kind of madness, it is
a sort of divinely inspired madness and can therefore be beneficial.
To support this, Socrates refers to other examples of divine madness,
which are found in prophecy and mystic initiation, but also in poetry,
which is here said to recount the acts of heroes for the instruction of
future generations (Phdr. 245a). It seems that this should be taken
seriously, not seen as simply ironic; the praise of love is presumably
meant seriously, so the praise of poetry should be likewise.
Of course, Plato certainly thinks that knowledge, gained by rea-
son, is superior to any kind of truth gained by an irrational process;
poetry, therefore, continues to be ranked below philosophy. But this
does not mean that these irrational ways of gaining access to truth
are insignificant. Indeed, the description of love in the Phaedrus
suggests that it is often through this divinely inspired form of irra-
tionality that we are first moved to do philosophy (252e–3a); the love
of wisdom must come before wisdom itself, and this is the result of
inspiration.
One might ask how poetry can be a vehicle of truth if poets do
not agree with one another. An answer to this is suggested by the
Timaeus, in its account of prophecy: although prophets are inspired,
their words need interpretation, which must be done by someone
in his right mind (Ti. 71d–2b). This may refer to the ecstatic utter-
ances of some prophets, but can also apply to the riddling ways in
which prophets often spoke, as the Delphic oracle famously did;
their utterances, although they contain truth, are not clear and need
interpretation. The same may be true of poetry.
Can this view be reconciled with that found in the Republic? In
one way they are certainly similar. They both involve the claim that
poets do not have knowledge. Indeed, some arguments found in the
Ion for the view of poetry presented there are very like arguments
from the Republic: that poets (or, in the Ion, reciters) do not have
expertise in the subjects they talk about, or people would not allow
them to devote themselves to poetry (Ion 541b–c); that they produce
irrational emotions in their audience (535d–e). But in another way
they seem strongly opposed; if poetry is divinely inspired, it should
194
aesthetics
be a source of truth, even if not of knowledge, whereas in the Repub-

lic it is deceptive.
One might try to reconcile the different approaches by pointing
out that the Republic deals with an ideal state, in which there are
philosopher rulers who can themselves impart truth to the people,
and train other philosophers to come after them. In such a state,
one may think, poetry would not have even the limited value that
Plato sometimes allows it in the actual world. Even if it contains
some truth, it is inferior to philosophical teaching, and it is danger-
ous, because people may look to it as a source of knowledge; only
philosophers may be able to interpret it rightly and so avoid damage.
Indeed, even in the Republic Socrates does allow poetry some value;
he permits hymns to the gods and praises of good people to remain,
the latter being exactly the function he allows to inspired poets in the
Phaedrus. (It is often said that works of this kind cannot be inspired,
and are an inferior form of poetry; but this is not clearly true. Hymns,
such as the works of Charles Wesley, and eulogies, such as Ben Jon-
son’s dedication to Shakespeare, can surely be of high poetic quality.)
Some aspects of the Republic, however, are hard to reconcile with
the idea of poetry as divinely inspired. One is the claim that imita-
tion, which includes poetry, is merely a game (Resp. 602b), although
this might perhaps mean only a game in comparison with philoso-
phy. Another is the claim that imitators lack true belief, not just
knowledge (602a), although here one might say that they sometimes
speak the truth, through inspiration, even without believing it. But
the biggest problem may not be in reconciling specific claims made
in the different works, but simply in their tone. Would Plato be as
hostile to poetry in general as he seems to be in the Republic if he
thought of it as a divine gift? We should perhaps simply take these
as different views of poetry to which Plato was attracted, either at
different moments in his career, or even at the same time, being
unable to choose decisively between them.
Finally, we could ask whether this idea of inspired speech might
help us to understand some parts of Plato’s own works. Socrates’
speech in the Phaedrus may itself be seen as an example of inspired
speech. It is said to be the result of inspiration by the muses (Phdr.
195
plato
262d), and to have been largely spoken in play (266c), yet it contains
philosophical content that seems to be meant seriously and to resem-
ble, to some extent, positions that Plato favours elsewhere. Might it
be seen as a poetic utterance, pointing to truth, but not expounding
it with philosophical rigour? Likewise the Symposium contains two
speeches in praise of love by poets, the comedian Aristophanes and
the tragedian Agathon; clearly, neither of these speeches represents
the philosophical truth, and both are in fact criticized later by Soc-
rates,12 yet they seem to contain insights that Plato finds it hard to
incorporate within his own thought. Aristophanes presents us with
the idea of love for an individual because of the individual he is,
not for the sake of the beauty he shares with others (Symp. 191d ff.),
Agathon with the idea of love as something outgoing, producing
peace and harmony, which seeks to benefit others rather than simply
to achieve the good for oneself (197d). Both these speeches may be
seen as examples of poetic inspiration, pointing indirectly to a truth
that is not fully understood.
Art and beauty
As we have seen, Plato has no term corresponding exactly to our “art”.

Nevertheless, we can ask about the value he ascribes to the kind of
activities and products that we would call art: to painting and sculp-
ture, poetry, music and so on. We know that he banishes most poetry
from the ideal state, and also severely censors music and design.
However, this is not to say that he denies all value in the products of
the arts. I would suggest that there are two kinds of value in particular
that, within a Platonic way of thinking, can be ascribed to them.
First, works of art may have value simply by being instances of
Forms such as beauty, harmony and so on. As such, they can inspire
us to philosophy. In the Phaedrus (250b ff.), Socrates draws atten-
tion to beauty as the only one of the most significant Forms that has
visible images; justice, temperance and so on are not visible in the
same way. So beauty plays a special role in prompting the soul to
recollection of Forms. Not all beauty is visible, for there are beauti-
196
aesthetics
ful souls and so on. But visible beauty is valuable in itself, as well as
prompting people to seek for higher kinds of beauty. Here Socrates
is talking about natural beauty, that of people; he is describing the
experience of falling in love. But could not what he says be relevant
to the arts as well? Could they not produce instances of beauty that
are valuable as such? This would not be representational art, or at
least it is not as representational that it has this value. A picture of
a man – even an idealized man – is an image not of a Form but of
a (real or imagined) particular. But a work of art, simply because it
is beautiful, may be an image – that is an instance – of beauty itself.
In several places Plato seems to allow works of art this kind of
value. In Republic Book 3 (399d ff.) Socrates says that young people
should be encouraged to develop such qualities as harmony, grace
and simplicity in their characters, and should therefore be sur-
rounded by things that have those qualities. Hence craftsmen should
produce things that have those qualities. This relates to painting,
but also to weaving, embroidery, architecture and furniture design.
These arts, Socrates claims, can produce qualities that resemble those
desirable in the soul, although clearly works of art possess them a
very different way from the soul.
Music is seen by Plato as especially valuable because of the har-
monious relationships between musical sounds, which is in some
way parallel to the kind of harmony that should exist in the soul.
This aspect of music is first emphasized in the Republic (531c), but
further developed in the Timaeus, which includes the passages that
give rise to the famous concept of the harmony of the spheres (Ti.
35b–d, 38c–d). For Plato this does not mean that the heavenly bod-
ies literally make musical sounds, but that their movements stand in
mathematical relations that embody harmonious proportions. Later
in the same work he says that musical sounds give pleasure, not only
to foolish people but to the wise, “because they imitate the divine
harmony in mortal movements” (80b). So music is valuable as an
instance of harmonious relationships, which are also manifested in
the heavenly movements and, ideally, in the soul.
In the Philebus, there is a curious passage in which Socrates dis-
cusses beautiful things which produce pure pleasures (Phlb. 51c–
197
plato
d). These include simple, clear, musical notes, but also pure shapes
and colours. He emphasizes that by a beautiful shape he does not
mean a living thing or a picture (a piece of mimetic art), but just a
simple geometrical shape. This would seem to leave room for some
abstract visual art, although only of a very simple kind. It appears,
then, that the arts can produce some things that are valuable in
their own right.
Poetry and philosophy
But can Plato find any distinctive value for representational art,
including poetry? We saw earlier that the idea of imitation, as it
is explained in Book 10 of the Republic, is ambiguous; it can mean
either simply work that reproduces the appearance of something,
not the reality, or work that is not based on knowledge of the real-
ity. While poetry cannot fail to be imitative in the first sense, it may
not always be imitative in the second; there could be a poetry that is
based on knowledge. Might poetry of this kind serve the purposes
of philosophy?
One may think that we find something of the sort in Plato’s own
works. Although they are not in verse, they achieve in prose effects
similar to those that poets can achieve; indeed, Plato may be thought
of as the first writer to do so. The Republic can be seen as, among
other things, an imitation of an ideal state, intended to create in
people’s minds an image of what such a state would be like. The dia-
logues more generally can be seen as imitations of a philosophical
life, showing philosophers – primarily Socrates, but others as well
– engaged in philosophical enquiry and discussion.
Such imitations still fall short of what they represent. Just as a
description of a city is not a city, a description or dramatic recon-
struction of a philosophical debate is not a philosophical debate. A
real philosophical debate requires live participants, able to think
up answers for themselves, and to explain, when questioned, what
they mean. No one can actually become wise just by reading Plato’s
works. But such imitations, if guided by knowledge, can still serve a
198
aesthetics
valuable purpose; they can set before us the ideal of a philosophical

life; and they can be a stimulus to philosophical thought.
It is not clear that all of Plato’s use of poetic language and methods
can be explained in this way. Perhaps he does sometimes speak in a
poetic way in an attempt – like the inspired poets he describes – to
express something he does not fully understand, and of which he
cannot give a philosophical account. This may be the explanation of
the various myths that occur in his work, particularly those regard-
ing life after death that appear in the Gorgias (524d ff.), the Phaedo
(108e ff.), the Phaedrus (246a ff.) and the Republic (614b ff.). This is
not to say that no philosophical account of these matters is possible
– it is just that it has not yet been achieved – and Plato would have
said that a philosophical account was preferable. But when it has not
been achieved, a poetic one may have some value. Sometimes, too,
Plato may simply be trying to make his work attractive to outsiders;
and sometimes his pen may run away with him and lead him to do
things that he would find it hard to justify philosophically.
The paradox of Plato
Plato’s works have a puzzling feature: they are extremely imitative,

not only in the sense found in Book 10 of the Republic – that they
present the appearance of things – but in that found in Books 2 and
3 – that they are dramatic, with the speaker taking on a part. In
fact, both kinds of imitation described in Book 3 are found in the
dialogues. Some, such as the Gorgias and Phaedrus, are actually dra-
matic in form; others, like the Symposium, Phaedo and Republic, have
narrators, but contain many speeches where the narrator repeats
someone else’s words. Many of the dialogues represent conflict, and
therefore involve the presentation of characters whose state of soul is
less than desirable. There is imitation of bad characters, like Critias
in the Charmides, Callicles in the Gorgias or Thrasymachus in the
Republic. There is imitation of figures of fun, like Hippias or Ion in
the works called after them. There is imitation of unphilosophical
poets, like Agathon and Aristophanes in the Symposium. The Ion and
199
plato
Hippias Major can indeed be seen as comedies. And the Phaedo is,
in a way, a tragedy; it does not actually recommend a tragic view of
life, rather telling us not to fear death or mourn for our loved ones,
yet it does in fact convey the sorrow that Socrates’ companions felt
at his death.
This may lead us to ask whether Plato’s works would be allowed
in the ideal state. One may well think that they would not. They are
not hymns to the gods or (only) praises of good people (although
of course they do contain the praise of Socrates); they are dramatic
and highly imitative works. It is true that in his late work, the Laws,
he does recommend his own work for use in the state he is there
proposing (Leg. 811c–12a), but the Laws lacks much of the dramatic
quality of earlier works.
Plato himself might not have seen this as a problem. He might say
that in the ideal state philosophers would be present, could them-
selves give philosophical guidance and reveal in their own lives what
a philosophical life was like. Hence, a literary representation of phil-
osophy would not be needed. But we may well feel that there is in
fact something valuable in the works of Plato as we have them – the
interplay of characters, the interaction of philosophical and poetic
elements – that would be missing in the ideal state. In this case there
is something of value in Plato’s works that his own principles would
exclude, and this is a paradox he may not have been able to over-
come.
200
Notes
1. Introduction
1. The development of Plato’s ideas is discussed in Chapter 2.

2. Certainly philosophers can arise by chance outside the ideal state (Resp. 520b),
but they are not as well educated as those who are brought up within that
state.
2. Plato’s development and Plato’s Socrates
1. A historic defence of the unity of Plato’s thought is P. Shorey, The Unity of Plato’s
Thought (Chicago, IL: University of Chicago Press, 1960). The widely accepted
developmental view can be found at, for instance, G. Vlastos, Socrates: Ironist
and Moral Philosopher (Cambridge: Cambridge University Press, 1991), 46–7.
2. The abandonment of the theory is argued for by G. E. L. Owen, “The Place of the
Timaeus in Plato’s Dialogues”, Classical Quarterly 3 (1953), 79–95. For the view
that Plato came to attach less importance to it see for instance I. M. Crombie,
An Examination of Plato’s Doctrines, 2 vols (London: Routledge & Kegan Paul,
1962–3), vol. 2, 356 ff.
3. “No harm comes to a good man either in life or in death” (Ap. 41c–d).
4. Plato is not mentioned by name in the second passage, but the parallel with the
first shows clearly that he is being discussed.
5. The first position is argued at length in Vlastos, Socrates, chs 2–4. For the second
see ibid., 117 n.50.
6. See, for instance: “the living thing of which all other living things are parts” (Ti.
30c).
201
plato
3. Plato’s metaphysics: the “theory of Forms”
1. I capitalize “Form” in order to make it clear when I am discussing the Forms

of Plato’s theory; but this should not be taken to imply that “Form” in Plato is a
technical term; his use of it is continuous with its use in fairly natural ways of
speaking, as, for instance, when we speak of different forms of animal.
2. This is argued by G. Vlastos, “Reasons and Causes in the Phaedo”, Philosophical
Review 78 (1969), 291–325, reprinted in his Plato: A Collection of Critical Essays,
vol. 1: Metaphysics and Epistemology, 132–66 (Garden City, NY: Doubleday,
1970).
3. See Laws 822a, which seems to affirm that the movements of the heavenly bod-
ies are perfectly regular.
4. This is argued by D. Keyt, “Aristotle on Plato’s Receptacle”, American Journal of
Philology 72 (1961), 291–300.
5. See for instance C. Meinwald, “Good-bye to the Third Man”, in The Cambridge
Companion to Plato, R. Kraut (ed.), 365–96 (Cambridge: Cambridge University
Press, 1992). The basic distinction was first proposed by Michael Frede, Prädika-
tion und Existenzaussage (Göttingen: Vandenhoeck & Ruprecht, 1967).
6. See for instance Timaeus 50d, where the Forms are compared to a father and
the receptacle to a mother.
7. This analogy seems to have been introduced into the debate by Peter Geach,
who says he was inspired to use it by Wittgenstein. See P. Geach, “The Third
Man Again”, Philosophical Review 65 (1956), 72–82, reprinted in R. E. Allen
(ed.), Studies in Plato’s Metaphysics, 265–77 (London: Routledge & Kegan Paul,
1965).
4. Knowledge
1. See Sophist 253c, where the Stranger suggests there are some Forms that are
responsible for the combination and division of other Forms.
2. Such a view is expressed by Cephalus at Republic 331b.
5. The soul
1. Phaedrus 248a ff. describes the lower elements being part of the soul even before
incarnation. See also Laws 896c–d, which seems to imply that all the activities
of soul are prior to body.
2. Some of the arguments in this and the following section have previously been
published in my “Plato on the Self-moving Soul”, Philosophical Inquiry 20
(1998), 18–28.
3. The argument starts at Phaedo 100c; the crucial passage is 105c ff.
4. The “good horse” in this passage is an image of the spirited part.
202
notes
5. See for instance Timaeus 36c–d, 43a ff., where the rational soul has a circular
motion; and 41d–e, where human souls, before incarnation, are set in the stars.
6. It is so presented at Republic 442c.
7. See Republic 442c–d (the definition of temperance), and 443d–e.
6. Politics
1. The name “Republic” represents Greek Politeia, whose basic meaning is simply
“constitution” or “system of government”.
2. The translation “lie” is sometimes condemned as deceptive; the Greek word
in question, pseudos, can just mean “falsehood”, so as to include fiction. But
Socrates does say that he hopes the people can be induced to believe the “lie”.
3. On this see, for instance, J. Annas, An Introduction to Plato’s Republic (Oxford:
Clarendon Press, 1981), 179–81.
4. For instance at 374b, “We prevented a shoemaker trying to be a farmer …”.
5. Some of these points are made by R. Bambrough, “Plato’s Political Analogies”,
in Philosophy, Politics and Society, P. Laslett (ed.), 98–115 (Oxford: Blackwell,
1956), reprinted in Plato: A Collection of Critical Essays, vol. 2: Ethics, Politics
and Philosophy of Art and Religion, G. Vlastos (ed.), 187–205 (Garden City, NY:
Doubleday, 1971).
7. Ethics
1. A principle on these lines is stated at Republic 435e–6a, although it is not clear

how wide its scope is supposed to be; it is not explicitly applied to justice.
2. See Republic 441d–e: “when each element in someone performs its own func-
tion, he will be just and perform his own function”.
3. See Laws 630d ff., 705d ff. on virtue as the aim of law.
4. This passage is used as a basis for an interpretation of Plato’s ethics by Terence
Irwin in Plato’s Ethics (Oxford: Oxford University Press, 1995), ch. 18.
5. And at Republic 364a–b, “they dishonour and look down on those who are weak
and poor, even while agreeing that they are better than others” (emphasis added).
6. See Republic 358a: “which the person who intends to be happy should welcome
both for itself and for its consequences”.
7. The things listed as examples of things good in themselves at Republic 357b–c
support this: enjoyment and harmless pleasures are things we value only for
their own sake; intelligence, sight and health both for themselves and their
consequences. These are beneficial things, things that contribute to our happi-
ness, rather than admirable things.
8. See for instance Republic 367d: “praise justice for the way in which it by itself ben-
efits us, and injustice harms us, and leave others to praise rewards and reputation”.
9. The description of the tyrannical soul begins at Republic 572d.
203
plato
10. See Philebus 21d–e, where it is agreed that no one would accept a life without
pleasure.
11. This reading of Plato is in some ways similar to that put forward by R. Kraut,
“The Defence of Justice in Plato’s Republic”, in Kraut, The Cambridge Companion
to Plato, 311–37.
12. On pleasure see Protagoras 351c ff.; on knowledge Meno 87e ff.
8. God and nature
1. This position can, of course, be challenged. As we saw in Chapter 5, it was chal-

lenged by Aristotle, for whom unthinking nature, as well as intelligence, can
(of its own accord) act for a purpose.
2. Some of the arguments of this and the following section are taken from my
“Plato on Necessity and Chaos”, Philosophical Studies 127 (2006), 283–98.
3. This structure is explained at Timaeus 47e, 68e–9b.
4. This view was proposed by G. Morrow, “Necessity and Persuasion in the
Timaeus”, Philosophical Review 59 (1950), 147–64, reprinted in Allen (ed.),
Studies in Plato’s Metaphysics, 421–37.
5. The precosmic state is linked with necessity at Timaeus 48b.
6. A rather more explicit statement of this, although in a mythic context, can be
found in the Statesman 273b ff.; the bodily element of the universe was in great
disorder before it came into its present order, and this is the source of the evil
that is still present in the world.
7. Called “the overseer” at Laws 903b and “our king” at 904a, he is compared to a
wise craftsman at 903c.
9. Aesthetics
1. Most explicitly, perhaps, Republic 607b: “we were right to banish it [poetry]
from the state”.
2. See for instance Republic 600d: “all poets, starting with Homer, are imitators
of images of virtue … and are not in contact with truth”. (“Imitator” is, in the
context of Republic 10, a term of condemnation.)
3. See Republic 595a: “Poetry, so far as it is imitative, should not be accepted” [in
the state]. And 607a: “hymns to the gods and praises of good people are the
only poetry to be accepted into the state”.
4. Officially Socrates does not commit himself on this point, since the question
whether the just life is advantageous is the topic of the Republic as a whole,
and has yet not been settled. But it is clear what Socrates thinks, and he claims
that if a just life is advantageous, poets are wrong to represent it differently. A
more explicit statement of this point is found in the Laws 660b ff., where similar
proposals for censorship are found.
204
notes
5. No passages on medical themes are actually quoted in the Republic, but see Ion
538b–c for the kind of passage Plato probably has in mind.
6. See for instance Republic 599b: “If he [Homer] really had understanding about
those things which he imitates”.
7. For example at Republic 603c: “Imitative poetry imitates people acting under
compulsion or willingly”.
8. This must be so if he bans all imitative poetry, and yet accepts hymns to the
gods and praises of good people.
9. Republic 598d – “after this we must examine tragedy” – makes it clear that only
after this is the status of poetry decided. See also 599d. I take Socrates at 597e
to be saying only that tragedians are two steps removed from truth if they are
imitators (as will be decided later), not, as some read it, since they are.
10. At 598d Socrates refers to “tragedy and its leader Homer”, and at 600e to “all
poets beginning from Homer”. Although Homer was an epic, not a dramatic
poet, he seems to have been widely associated with tragedy.
11. Solon’s poems are mentioned at Timaeus 21b–d.
12. Agathon is criticized at Symposium 199c ff., and Aristophanes at 205d–e.
205
Plato’s works
This list includes all works that are generally agreed to be by Plato, and also some
that, while disputed, are widely thought to be his. Dialogues whose authenticity is
disputed are marked with an asterisk.
In line with the arguments of Chapter 2, Plato’s works are here divided into
three groups: Socratic dialogues, reflecting the thought of the historical Socrates;
Platonic dialogues, reflecting Plato’s thought during the period when he developed
his most famous ideas; and later Platonic dialogues, reflecting Plato’s thought in
the last period of his life. The Socratic dialogues are often believed to have been
written early in Plato’s career, and to precede the Platonic dialogues, but this cannot
be known with certainty.
Socratic dialogues
(In alphabetical order)
Alcibiades*
Apology
Charmides
Crito
Euthydemus
Euthyphro
Gorgias
Hippias Major*
Hippias Minor
Ion
Laches
Lysis
207
plato
Meno
Protagoras
The Gorgias and Meno are often thought to have both Socratic and Platonic ele-
ments.
The Menexenus is also often grouped with the Socratic dialogues, though, being
a satirical work, it does not directly reflect the thought of either Socrates or Plato.
Platonic dialogues
(In conjectural order of composition)
Cratylus
Phaedo
Symposium
Republic
Phaedrus
Parmenides
Theaetetus
The Parmenides and Theaetetus are widely thought to be the last of this group, and
to have some features in common with the late group.
Later Platonic dialogues

(In conjectural order of composition)
Timaeus
Critias (incomplete)
Sophist
Statesman (also called Politicus)
Philebus
Laws
Epinomis*
The Seventh Letter, if genuine, also belongs to the last period of Plato’s life.
208
Further reading
Translations
A comprehensive translation of Plato’s works (including doubtful and spurious

works) by various translators, is J. M. Cooper, with D. S. Hutchinson (eds), Plato:
Complete Works (Indianapolis, IN: Hackett, 1997). Many of the individual works
of Plato from this edition have also been published separately. For details see below
under the various dialogues.
An older comprehensive translation is E. Hamilton and H. Cairns (eds), The
Collected Dialogues of Plato (New York: Pantheon, 1961), reprinted (Princeton, NJ:
Princeton University Press, 1971).
There are many translations of individual works, with introductions and notes.
Some important ones are as follows.
Meno
Allen, R. E. (trans.) 1989. The Dialogues of Plato, vol. 1: Euthyphro, Apology, Crito,
Gorgias, Menexenus. New Haven, CT: Yale University Press.
Beresford, A. (trans.), with intro. by L. Brown. 2005. Protagoras and Meno. London:
Penguin.
Day, J. M (ed. and trans.) 1994. Plato’s Meno in Focus. London: Routledge. Includes
an introduction and collected essays by various writers.
Grube, G. M. A. (trans.), J. M. Cooper (rev.) 2002. Five Dialogues. Indianapolis, IN:
Hackett. Also includes Euthyphro, Apology, Crito and Phaedo.
Waterfield, R. (trans.), with intro. by A. Gregory. 2009. Meno and Other Dialogues.
Oxford: Oxford University Press. Also includes Charmides, Laches and Lysis.
209
plato
Phaedo
Gallop, D. (trans.) 1977. Phaedo (with philosophical commentary). Oxford: Claren-
don Press.
Gallop, D. (trans.) 2009. Phaedo. Oxford: Oxford University Press.
Grube, G. M. A. (trans.), J. M. Cooper (rev.) 2002. Five Dialogues. Indianapolis, IN:
Hackett. Also includes Euthyphro, Apology, Crito and Meno.
Tredennick, H. (trans.), H. Tarrant (rev.) 2003. The Last Days of Socrates. London:
Penguin. Also includes Euthyphro, Apology and Crito.
Symposium
Allen, R. E. (trans.) 1993. The Dialogues of Plato, vol. 2: The Symposium. New Haven,
CT: Yale University Press.
Nehamas, A. & P. Woodruff (trans.) 1989. Symposium. Indianapolis, IN: Hackett.
Waterfield, R. (trans.) 2008. Symposium. Oxford: Oxford University Press.
Republic
Allen, R. E. (trans.) 2006. Republic. New Haven, CT: Yale University Press.
Griffith, T. (trans.), with intro. by G. Ferrari (ed.) 2000. Republic. Cambridge: Cam-
bridge University Press.
Grube, G. M. A. (trans.), C. D. C. Reeve (rev.) 1992. Republic. Indianapolis, IN:
Hackett.
Lee, H. D. P. (trans.), with intro. by M. Lane. 1997. Republic. Harmondsworth:
Penguin.
Reeve, C. D. C. (trans.) 2004. Republic (translated from the new standard Greek
text). Indianapolis, IN: Hackett.
Waterfield, R. (trans.) 1998. Republic. Oxford: Oxford University Press.
Phaedrus
Nehamas, A. & P. Woodruff (trans.) 1995. Phaedrus. Indianapolis, IN: Hackett.
Rowe, C. J. (trans.) 2005. Phaedrus. London: Penguin.
Waterfield, R. (trans.) 2009. Phaedrus. Oxford: Oxford University Press.
Parmenides
Allen, R. E. (trans.) 1997. The Dialogues of Plato, vol. 4: The Parmenides. New Haven,
CT: Yale University Press.
Gill, M. L. & P. Ryan 1996. Parmenides. Indianapolis, IN: Hackett.
Theaetetus
Levett, M. (trans.), M. Burnyeat (rev.), with intro. by B. A. O. Williams (ed.) 1992.
Theaetetus. Indianapolis, IN: Hackett. The same translation is also published
with an extended introduction by M. Burnyeat, in The Theaetetus of Plato (Indi-
anapolis, IN: Hackett, 1990).
210
further reading
McDowell, J. (trans.) 1977. Theaetetus (with philosophical commentary). Oxford:

Clarendon Press.
Waterfield, R. (trans.) 1987. Theaetetus. Harmondsworth: Penguin.
Timaeus
Lee, H. D. P. (trans.), T. Johansen (rev.) 2008. Timaeus and Critias. London: Penguin.
Waterfield, R. (trans.), with intro by A. Gregory. 2008. Timaeus and Critias. Oxford:
Oxford University Press.
Zeyl, D. (trans.) 2000. Timaeus. Indianapolis, IN: Hackett.
Sophist
White, N. (trans.) 1993. Sophist. Indianapolis, IN: Hackett.
Statesman
Rowe, C. J. (trans.) 1999. Statesman. Indianapolis, IN: Hackett.
Waterfield, R. (trans.), with intro. by J. Annas (ed.) 1995. Statesman. Cambridge:
Cambridge University Press.
Philebus
Frede, D. (trans.) 1993. Philebus. Indianapolis, IN: Hackett.
Laws
Mayhew, R. 2008. Laws 10 (with philosophical commentary). Oxford: Clarendon
Press. This book contains most of the religious and cosmological material in
the Laws.
Saunders, T. J. (trans.) 1970. Laws. Harmondsworth: Penguin.
Compilations
Chappell, T. (trans.) 1996. The Plato Reader. Edinburgh: Edinburgh University Press.
Murray, P. & T. S. Dorsch (trans.) 2000. Classical Literary Criticism. Harmonds-
worth: Penguin. Includes the Ion and extracts from the Republic, along with
material by Aristotle, Horace and Longinus.
Partenie, C. (trans.) 2009. Selected Myths. Oxford: Oxford University Press. Includes
extracts from the Protagoras, Gorgias, Symposium, Phaedo, Republic, Phaedrus,
Statesman, Timaeus and Critias.
Reeve, C. D. C. (trans.) 2006 Plato on Love. Indianapolis, IN: Hackett. Includes
the Lysis, Symposium, Phaedrus and Alcibiades, with extracts from the Republic
and Laws.)
211
plato
Books on Plato’s work as a whole
There are numerous books on Plato’s work as a whole. Short introductory works
include:
Annas, J. 2003. Plato: A Very Short Introduction. Oxford: Oxford University Press.
Hare, R. M. 1982. Plato. Oxford: Oxford University Press.
Kraut, R. 2008. How to Read Plato. London: Granta.
Williams, B. 1998. Plato. London: Routledge.
Longer works include:

Crombie, I. M. 1964. Plato: The Midwife’s Apprentice. London: Routledge & Kegan
Paul.
Grube, G. M. A. 1935 Plato’s Thought. London: Methuen.
Guthrie, W. K. C. 1975. A History of Greek Philosophy, vol. 4: Plato: The Man and his
Dialogues: Earlier Period. Cambridge: Cambridge University Press.
Guthrie, W. K. C. 1978. A History of Greek Philosophy, vol. 5: The Later Plato and
the Academy. Cambridge: Cambridge University Press.
Melling, D. J. 1987. Understanding Plato. Oxford: Oxford University Press.
Rowe, C. J. 1984. Plato. Brighton: Harvester.
More advanced and complex works are:

Crombie, I. M. 1962. An Examination of Plato’s Doctrines, vol. 1: Plato on Man and
Society. London: Routledge & Kegan Paul.
Crombie, I. M. 1963. An Examination of Plato’s Doctrines, vol. 2: Plato on Knowledge
and Reality. London: Routledge & Kegan Paul.
Gosling, J. 1973 Plato. London: Routledge & Kegan Paul.
Guides and companions
These books consist of new articles, written for the guide in question, aiming to
give a survey of major areas of Plato’s thought:
Benson, H. (ed.) 2006. A Companion to Plato. Oxford: Blackwell.
Fine, G. (ed.) 2008. The Oxford Handbook of Plato. Oxford: Oxford University Press.
Kraut, R. (ed.) 1992. The Cambridge Companion to Plato. Cambridge: Cambridge
University Press.
Collections
These are collections made up of (mostly) existing articles on Plato, aiming to bring
together major works in the field.
212
further reading
Allen, R. E. (ed.) 1965. Studies in Plato’s Metaphysics. London: Routledge & Kegan
Paul.
Fine, G. (ed.) 2000. Plato. Oxford: Oxford University Press. The paperback edition
is published in two volumes: one on metaphysics and epistemology, and one on
ethics, politics, religion and the soul.
Smith, N. (ed.) 1998. Plato: Critical Assessments, 4 vols. London: Routledge.
Vlastos, G. (ed.) 1970. Plato: A Collection of Critical Essays, vol. 1: Metaphysics and
Epistemology. Garden City, NY: Doubleday.
Vlastos, G. (ed.) 1971. Plato: A Collection of Critical Essays, vol. 2: Ethics, Politics
and Philosophy of Art and Religion. Garden City, NY: Doubleday.
Specific areas of Plato’s thought
Metaphysics and theory of knowledge

These topics are closely entwined in Plato’s thought, and many works deal with
both. On metaphysics especially:
Dancy, R. M. 2004. Plato’s Introduction of Forms. Cambridge: Cambridge University
Press.
Fine, G. 1993. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms. Oxford:
Clarendon Press.
Irwin, T. 1977. “Plato’s Heracleitianism. Philosophical Quarterly 27: 1–13. Reprinted
in Smith (ed.), Plato: Critical Assessments, vol. 2, 102–15.
Irwin, T. 2000. “The Theory of Forms”. In Fine (ed.), Plato, 145–72. Adapted from
his Plato’s Ethics (New York: Oxford University Press, 1995), ch. 10.
Malcolm, J. 1991. Plato on the Self-Predication of Forms: Early and Middle Dialogues.
Oxford: Clarendon Press.
McCabe, M. M. 1994. Plato’s Individuals. Princeton, NJ: Princeton University Press.
Nehamas, A. 1975. “Plato on the Imperfection of the Sensible World”. American
Philosophical Quarterly 12: 105–17. Reprinted in Smith (ed.), Plato: Critical
Assessments, vol. 2, 72–92, and in Fine (ed.), Plato, 173–93.
Patterson, R. 1985. Image and Reality in Plato’s Metaphysics. Indianapolis, IN:
Hackett.
Silverman, A. 2002. The Dialectic of Essence: A Study of Plato’s Metaphysics. Princ-
eton, NJ: Princeton University Press.
Vlastos, G. 1965. “Degrees of Reality in Plato”. In New Essays on Plato and Aristotle,
R. Bambrough (ed.), 10–19. London: Routledge & Kegan Paul. Reprinted in his
Platonic Studies, 58–75 (Princeton, NJ: Princeton University Press, 1973) and
in Smith (ed.), Plato: Critical Assessments, vol. 2, 219–34.
On knowledge especially:
Gulley. N. 1962. Plato’s Theory of Knowledge. London: Methuen.
Robinson, R. 1941. Plato’s Earlier Dialectic. Ithaca, NY: Cornell University Press.
213
plato
Scott, D. 2003. Recollection and Experience: Plato’s Theory of Learning and its Suc-
cessors. Cambridge: Cambridge University Press.
On both equally:
Fine, G. 2003. Plato on Knowledge and Forms: Selected Essays. Oxford: Clarendon
Press.
White, N. 1976. Plato on Knowledge and Reality. Indianapolis, IN: Hackett.
Politics
Bobonich, C. 2002. Plato’s Utopia Recast: his Later Ethics and Politics. Oxford: Oxford
University Press.
Schofield, M. 2006. Plato. Oxford: Oxford University Press.
Taylor, C. C. W. 1986. “Plato’s Totalitarianism”. Polis 5: 4–29. Reprinted in Fine
(ed.), Plato, 762–78.
The most famous attack on Plato’s political views is in K. Popper, The Open Society
and its Enemies, vol. 1: The Spell of Plato (London: Routledge & Kegan Paul, 1945).
Issues raised in this book are discussed in R. Bambrough (ed.), Plato, Popper and
Politics: Some Contributions to a Modern Controversy (Cambridge: Heffers, 1967).
Ethics
Annas, J. 1978. “Plato and Common Morality”. Classical Quarterly 28: 437–51.
Reprinted in Smith (ed.), Plato: Critical Assessments, vol. 3, 206–19.
Cooper, J. 1984. “Plato’s Theory of Human Motivation”. History of Philosophy Quar-
terly 1: 3–21. Reprinted in Fine (ed.), Plato, 668–88.
Irwin, T. 1995. Plato’s Ethics. New York: Oxford University Press. This is a recasting
of his earlier work, Plato’s Moral Theory (Oxford: Oxford University Press, 1977).
Sedley, D. 2000. “The Ideal of Godlikeness”. In Fine (ed.), Plato, 791–810.
Cosmology
Sedley, D. 2007. Creationism and its Critics in Antiquity. Berkeley, CA: University of
California Press. Esp. chapter 4. (See also further reading on the Timaeus, below.)
Aesthetics
Janaway, C. 1995. Images Of Excellence: Plato’s Critique of the Arts. Oxford: Claren-
don Press.
Moravcsik, J. & P. Temko (eds) 1982. Plato on Beauty, Wisdom and the Arts. Totowa,
NJ; Rowman & Allanheld.
Murdoch, I. 1977. The Fire and the Sun: Why Plato Banished the Artists. Oxford:
Clarendon Press.
Nehamas, A. 1982. “Plato on Imitation and Poetry in Republic 10”. In Plato on
Beauty, Wisdom and the Arts, J. Moravcsik & P. Temko (eds), 47–78 (Totowa,
214
further reading
NJ; Rowman & Allanheld, 1982). Reprinted in Smith (ed.), Plato: Critical Assess-
ments, vol. 3, 296–323.
Love
Nussbaum, M. C. 1986. The Fragility of Goodness: Luck and Ethics in Greek Tragedy
and Philosophy. Cambridge: Cambridge University Press. Esp. chs 6–7.
Price, A. W. 1990. Love and Friendship in Plato and Aristotle. Oxford: Clarendon
Press. Esp. chs 1–3.
Vlastos, G. 1973. “The Individual as Object of Love in Plato”. In his Platonic Stud-
ies, 1–42 (Princeton, NJ: Princeton University Press, 1973). Reprinted in Fine
(ed.), Plato, 619–45.
Plato’s use of myths

Brisson, L. 1998. Plato the Myth Maker, G. Nadaff (trans.). Chicago, IL: University
of Chicago Press.
Partenie, C. (ed.) 2009. Plato’s Myths. Cambridge: Cambridge University Press.
Works on Plato as a writer
Blondell, R. 2002. The Play of Character in Plato’s Dialogues. Cambridge: Cambridge

University Press.
and Philosophy. Cambridge: Cambridge University Press. Esp. Interlude 1.
Rowe, C. J. 2007. Plato and the Art of Philosophical Writing. Cambridge: Cambridge
University Press.
Szlesak, T. A. 1999. Reading Plato, G. Zanker (trans.). London: Routledge.
Works on particular dialogues
Meno
Scott, D. 2006 Plato’s Meno. Cambridge: Cambridge University Press.
Phaedo
Bostock, D. 1986. Plato’s Phaedo. Oxford: Clarendon Press.
Bostock, D. 2000. “The Soul and Immortality in Plato’s Phaedo”. In Fine (ed.), Plato,
886–906. Adapted from his Plato’s Phaedo (Oxford: Clarendon Press, 1986),
21–41.
Vlastos, G. 1969. “Reasons and Causes in the Phaedo”. Philosophical Review 78:
291–325. Reprinted in Vlastos, Platonic Studies, (Princeton, NJ: Princeton Uni-
versity Press, 1973), 76–110, and in Smith (ed.), Plato: Critical Assessments, vol.
2, 16–44.
215
plato
Symposium
Sheffield, F. 2006. Plato’s Symposium: the Ethics of Desire. Oxford: Oxford University
Press.
Republic
Annas, J. 1981. An Introduction to Plato’s Republic. Oxford: Clarendon Press.
Dahl. N. 1991. “Plato’s Defence of Justice”. Philosophy and Phenomenological
Research 51: 809–34. Reprinted in Fine (ed.), Plato, 689–716.
Fine, G. 1990. “Knowledge and Belief in Republic 5–7”. In Companions to Ancient
Thought 1: Epistemology, S. Everson (ed.), 85–115. Cambridge: Cambridge Uni-
versity Press. Reprinted in Smith (ed.), Plato: Critical Assessments, vol. 2, 234–65.
Kraut, R. 2000. “Return to the Cave: Republic 519–521”. In Fine (ed.), Plato, 717–36.
Mitchell, B. & J. Lucas 2003. An Engagement with Plato’s Republic. Aldershot: Ash-
gate.
and Philosophy. Cambridge: Cambridge University Press. Esp. ch. 4.
Pappas, N. 1995. Plato and the Republic. London: Routledge.
Sachs, D. 1963. “A Fallacy in Plato’s Republic”. Philosophical Review 72: 141–58.
Reprinted in G. Vlastos (ed.), Plato: A Collection of Critical Essays, vol. 2: Eth-
ics, Politics and Philosophy of Art and Religion, 35–51, and in Smith (ed.), Plato:
Critical Assessments, vol. 3, 206–19.
Santas, G. (ed.) 2006. Blackwell Guide to Plato’s Republic.
Smith, N. 1996. “Plato’s Divided Line”. Ancient Philosophy 16: 25–46. Reprinted in
Smith (ed.), Plato: Critical Assessments, vol. 2, 292–315.
White, N. 1979. A Companion to Plato’s Republic. Oxford: Blackwell.
White, N. 1986. “The Rulers’ Choice”. Archiv für Geschichte der Philosophie 68:
22–46.
Williams, B. A. O. 1973. “The Analogy of City and Soul in Plato’s Republic”. In
Exegesis and Argument, E. Lee, A. Mourelatos & R. Rorty (eds), 196–206. Assen:
van Gorcum. Reprinted in Fine (ed.), Plato, 737–46.
The Republic is also treated at length in Irwin, Plato’s Ethics.
Phaedrus
Bett, R. 1986. “Immortality and the Nature of the Soul in the Phaedrus”. Phronesis
31: 1–26. Reprinted in Fine (ed.), Plato, 907–31.
Griswold, C. L. 1986. Self-Knowledge in Plato’s Phaedrus. New Haven, CT: Yale
University Press.
Ferrari, G. 1987. Listening to the Cicadas: A Study of Plato’s Phaedrus. Cambridge:
Cambridge University Press.
Parmenides
Meinwald, C. 1991. Plato’s Parmenides. New York: Oxford University Press.
216
further reading
Miller, M. H. 1986. Plato’s Parmenides: The Conversion of the Soul. Princeton, NJ:
Princeton University Press.
Vlastos, G. 1954. “The Third Man Argument in the Parmenides”. Philosophical
Review 63: 319–49. Reprinted in Smith (ed.), Plato: Critical Assessments, vol.
4, 3–27.
Theaetetus
Bostock, D. 1988. Plato’s Theaetetus. Oxford: Clarendon Press.
Chappell, T. 2005. Reading Plato’s Theaetetus. Indianapolis, IN: Hackett. Also
includes a translation.
Sedley, D. 2004. The Midwife of Platonism: Text and Subtext in Plato’s Theaetetus.
Oxford: Clarendon Press.
Timaeus and Critias

Gregory, A. 2000. Plato’s Philosophy of Science. London: Duckworth.
Johansen, T. K. 2004. Plato’s Natural Philosophy. Cambridge: Cambridge University
Press.
Patterson, R. 1985. “On the Eternality of the Platonic Forms”. Archiv für Geschichte
der Philosophie 67: 27–46. Reprinted in Smith (ed.), Plato: Critical Assessments,
vol. 2, 142–60.
Strange, S. 1985. “The Double Explanation in the Timaeus”. Ancient Philosophy 5:
25–39. Reprinted in Fine (ed.), Plato, 399–417.
Vlastos, G. 1975. Plato’s Universe. Oxford: Clarendon Press.
Sophist
Brown, L. 1986. “Being in the Sophist; a Syntactical Enquiry”. Oxford Studies in
Ancient Philosophy 4: 49–70. Reprinted in Fine (ed.), Plato, 457–80.
Notomi, N. 1999. The Unity of Plato’s Sophist: Between the Sophist and the Philoso-
pher. Cambridge: Cambridge University Press.
Statesman
Lane, M. 1998. Method and Politics in Plato’s Statesman. Cambridge: Cambridge
University Press.
Theaetetus, Sophist, Statesman, Philebus

McCabe, M. M. 2000. Plato and his Predecessors: The Dramatisation of Reason.
Cambridge: Cambridge University Press.
Laws
Stalley, R. F. 1983. An Introduction to Plato’s Laws. Oxford: Blackwell.
217
plato
Online resources
The Stanford Encyclopedia of Philosophy (http://plato.stanford.edu) includes a

number of helpful articles on Plato. On topics covered in this book the following
are especially relevant: R. Kraut, “Plato” (a general account of his thought); N. Pap-
pas, “Aesthetics”; D. Frede, “Ethics”; E. Brown, “Ethics and Politics in the Republic”;
A. Silverman, “Middle Period Metaphysics and Epistemology”; C. Bobonich, “On
Utopia”; S. Rickless, “Parmenides”; D. Zeyl, “Timaeus”.
Project Archelogos (www.archelogos.com) has analyses of the arguments of
some of Plato’s dialogues, with commentaries. At the moment it includes: H. Ben-
son, Charmides; T. Chappell, Theaetetus; F. G. Herrmann & D. Robinson, Lysis; R.
Waterfield, Gorgias; C. J. Rowe, Republic V (forthcoming).
218
Bibliography
Allen, R. E. (ed.) 1965. Studies in Plato’s Metaphysics. London: Routledge & Kegan
Paul.
Annas, J. 1981. An Introduction to Plato’s Republic. Oxford: Clarendon Press.
Bambrough, R. 1956. “Plato’s Political Analogies”. In Philosophy, Politics and Society,
P. Laslett (ed.), 98–115 (Oxford: Blackwell, 1956). Reprinted in Plato: A Collec-
tion of Critical Essays, vol. 2: Ethics, Politics and Philosophy of Art and Religion,
G. Vlastos (ed.), 187–205 (Garden City, NY: Doubleday, 1971).
Bambrough, R. (ed.) 1965. New Essays on Plato and Aristotle. London: Routledge
& Kegan Paul.
Bostock, D. 1986. Plato’s Phaedo. Oxford: Clarendon Press.
Brandwood, L. 1992. “Stylometry and Chronology”. See Kraut (1992a): 90–120.
Crombie, I. M. 1962–3. An Examination of Plato’s Doctrines, 2 vols. London:
Routledge & Kegan Paul.
Fine, G. 1993. On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms. Oxford:
Clarendon Press.
Frede, M. 1967. Prädikation und Existenzaussage. Göttingen: Vandenhoeck & Ruprecht.
Geach, P. 1956. “The Third Man Again”. Philosophical Review 65: 72–82. Reprinted
in Allen (1965): 265–77.
Irwin, T. 1995. Plato’s Ethics. Oxford: Oxford University Press.
Janaway, C. 1995. Images of Excellence: Plato’s Critique of the Arts. Oxford: Claren-
don Press.
Keyt, D. 1961. “Aristotle on Plato’s Receptacle”. American Journal of Philology 72:
291–300.
Kraut, R. (ed.) 1992a. The Cambridge Companion to Plato. Cambridge: Cambridge
University Press.
Kraut, R. 1992b. “The Defence of Justice in Plato’s Republic”. See Kraut (1992a):
311–37.
219
plato
Kripke, S. 1981. Naming and Necessity. Oxford: Blackwell.

Laslett, P. (ed.) 1956. Philosophy, Politics and Society. Oxford: Blackwell.
Mason, A. S. 1998. “Plato on the Self-moving Soul”. Philosophical Inquiry 20: 18–28.
Mason, A. S. 2006. “Plato on Necessity and Chaos”. Philosophical Studies 127:
283–98.
Meinwald, C. 1992. “Good-bye to the Third Man”. See Kraut (1992a): 365–96.
Morrow, G. 1950. “Necessity and Persuasion in the Timaeus”. Philosophical Review
59: 147–64. Reprinted in Allen (1965): 421–37.
Nehamas, A. 1988. “Plato and the Mass Media”. The Monist 71: 214–34.
Owen, G. E. L. 1953. “The Place of the Timaeus in Plato’s Dialogues”. Classical
Quarterly 3: 79–95. Reprinted in Allen (1965): 313–38.
Sachs, D. 1963. “A Fallacy in Plato’s Republic”. Philosophical Review 72: 141–58.
Scott, D. 1987. “Platonic Anamnesis Revisited”. Classical Quarterly 37: 36–66.
Scott, D. 2003. Recollection and Experience: Plato’s Theory of Learning and its Suc-
cessors. Cambridge: Cambridge University Press.
Shorey, P. 1960. The Unity of Plato’s Thought. Chicago, IL: University of Chicago
Press.
Vlastos, G. 1965. “Degrees of Reality in Plato”. In New Essays on Plato and Aristotle,
R. Bambrough (ed.), 1–19 (London: Routledge & Kegan Paul, 1965). Reprinted
in Vlastos (1981): 58–75.
Vlastos, G. 1969. “Reasons and Causes in the Phaedo”. Philosophical Review 78:
291–325. Reprinted in his Plato: A Collection of Critical Essays, vol. 1: Metaphys-
ics and Epistemology, 132–66 (Garden City, NY: Doubleday, 1970).
Vlastos, G. (ed.) 1970. Plato: A Collection of Critical Essays, vol. 1: Metaphysics and
Epistemology. Garden City, NY: Doubleday.
Vlastos, G. (ed.) 1971. Plato: A Collection of Critical Essays, vol. 2: Ethics, Politics
and Philosophy of Art and Religion. Garden City, NY: Doubleday.
Vlastos, G. 1972. “The Unity of the Virtues in the Protagoras”. Review of Metaphysics
25: 415–58. Reprinted in Vlastos (1981): 221–69.
Vlastos, G. 1973. Platonic Studies. Princeton, NJ: Princeton University Press.
Vlastos, G. 1991. Socrates: Ironist and Moral Philosopher. Cambridge: Cambridge
University Press.
Wittgenstein, L. 1953. Philosophical Investigations, G. E. M. Anscombe (trans.).
Oxford: Blackwell.
220
Index
Academy 6, 50 censorship 182–4

Adeimantus 120, 124, 128, 150–52 Charmides 5
Agathon 196, 205 Cicero 21
akrasia 115, 117 collection and division 68–72
Alexander of Aphrodisias 53 combination of Forms 66–8
Annas, J. 75, 113, 148, 203 community of women and children
argument from design 162–5 123–4, 133
Aristophanes 4, 196, 205 cosmological argument 163–4
Aristotle 1, 3, 29, 39, 134, 149, 159, Critias 5
161, 165 Crombie, I. M. 45, 83, 201
Aristotle’s works
Magna Moralia 20 definition 65–6, 71–2
Metaphysics 20, 36, 46, 50, 53, 85, Demiurge see God
158 Democritus 173
Nicomachean Ethics 90, 141 dialectic 10, 66, 68, 70, 85–92
Poetics 7, 187 dialogue form 9–13, 198–200
Politics 128, 133 Dionysius the elder (ruler of Syra-
Posterior Analytics 91 cuse) 5
Aristoxenus 6, 158 Dionysius the younger (ruler of Syra-
Athens 4–6, 147–8 cuse) 6
Atlantis 147–8, 174
Epicrates 72
Bambrough, R. 203 Epictetus 21
Beauty 31–2, 81–2, 196–8 Epicurus 149
Bostock, D. 79 eternity 25
Euclid 91
cave, image of 94–8, 144–5, 154–5 Eudoxus 6
221
plato
evil, origin of 174–6, 178 knowledge 61–98

definition of 62–4
Fine, G. 55 and Forms 64–5, 72–6
flux 35–8 and understanding 75–6
Forms (Platonic) 3, 16, 24–5, 27–59, Kraut, R. 202
162, 166–7, 196–7 Kripke, S. 56
arguments against 53–9
arguments for 49–53 law 130–34
as causes 29 sovereignty of 133–4
changelessness of 30–31, 38, 74 and virtue 133, 140
and the good 91–4 lies (in the ideal state) 130–31
as ideals 31–3 line, image of 85–7
and imitation 189–90 love 31–2, 69–70, 81–2, 193–4, 196,
as paradigms 45–7 197
and philosophical enquiry 85–91
range of 40–43 mathematics 6, 77, 82, 85–91, 93, 96
and reality 47–9, 72–4 Meinwald, C. 55, 202
and recollection 76–84 Morrow, G. 172–3, 204
as universals 29 music 197
Frede, M. 44 myths 199
Geach, P. 202 necessity 168–73, 178

Glaucon 120, 145, 150–52, 153, 154 Nehamas, A. 187
God 3, 24, 161–79
good, Form of the 91–4, 148, 156–9 Owen, G. E. L. 24, 201
Heraclitus 36 paradox of enquiry 76

Homer 185, 190–92 philosopher rulers 125–6
Hume, D. 164 philosophy
hypothesis, method of 87–91 and the good life 159–60
and justice 139–40, 153–5
Ideas (Platonic) see Forms as way of life 10, 160
imitation 184–92, 198–200 Plato
immortality 104–10 development 15–17, 22–6, 66–8,
imperfection (of the sensible world) 104, 109, 132, 161–2
33–8 as imitative artist 198–200
Irwin, T. 203 life 4–6
writings (overview of) 7–9
Janaway, C. 187 Plato’s individual works
justice Apology 8, 17, 18–19, 22, 147, 201
and happiness 149–56 Charmides 199
in the soul 135–41 Cratylus 42
in the state 121, 127 Crito 12–13, 17, 22
Euthyphro 29, 65, 126
Keyt, D. 202 Gorgias 70, 110, 158, 199
222
index
Hippias Major 200 on government 130–32

Ion 193–4, 199, 204 Symposium ix, 30, 34–5, 37, 45,
Laws 100–101, 109, 115, 117, 120, 104–5, 126, 147, 160, 196, 199,
174, 200, 202, 203, 204 205
on cosmology 176–9 Theaetetus ix, 36, 65, 66, 74
on law 132–4 on knowledge 62–4
Meno 17–18, 29, 65, 74, 75, 87–8, Timaeus 25, 30, 41, 45, 49, 58, 74,
140, 141, 204 84, 93, 109, 110, 111–14, 117,
on recollection 76–8 132, 147–8, 158, 177, 197, 201,
Parmenides 24, 25, 30, 41, 53–9, 67 202, 204, 205
Phaedo 29, 30, 36–7, 47, 50, 84, on evil 174–6
87–8, 92, 99, 109–10, 200, 202 on God 161–8
on immortality 105–7 on necessity 168–73
on recollection 79–81 on self-motion 101–4
Phaedrus 10, 69–72, 110, 111, 196, on space 38–40
202 poetry 181–96, 198–9
on immortality 107–8 and inspiration 193–6
on poetry 193–6 and philosophy 198–9
on recollection 81–2, 100
Philebus 71, 104, 132, 135–6, receptacle see space
197–8, 203 recollection 76–84, 196
Politicus see Statesman
Protagoras 17–18, 115, 126, 141, Sachs, D. 143
204 Scott, D. 82
Republic 11, 18–19, 33, 42, 45, self-motion 100–104, 107–8, 176–8
47–8, 50, 51, 65, 67, 68, 84, 110, self-predication 43–6, 54–7
131–2, 158, 159–60, 161, 197, separation of Forms 30–31
199, 202, 203, 204, 205 Shorey, P. 201
on the divided soul 111–17 Socrates (as historical figure) 4–5,
on Forms and knowledge 72–4 16–23, 70, 161
on the good 91–4 Socratic dialogues 8, 16–23, 29, 52–3,
on the ideal state 121–31 65, 83
image of the cave 94–8 Solon 191, 205
on immortality 108–9 soul 26, 99–117
on justice and happiness 149–56 as harmony 99
on the nature of justice 135–44 division of 26, 111–17
on the philosophers’ choice 144–6 origin of 26, 108, 109, 177
on philosophical enquiry 85–91 space 38–40
plan of 120–21 Sparta 5–6, 111, 133–4
on poetry 182–92, 194–6 Speusippus 6
Seventh Letter 5, 109 Stoic school 21, 149, 161, 165
Sophist 71–2, 104, 202 sun, image of 91–4
on combination of Forms 65–6 Syracuse 5–6
Statesman 42, 71–2, 142, 174, 177,
204 Theaetetus 6
223
plato
third man argument 53–7 in the soul 137–9

Thirty, the (rulers of Athens) 5 in the state 126–7
virtue ethics 142
unity Vlastos, G. 22–3, 44, 47, 201, 202
and the good 158–9
in the state 124–5, 133 Wittgenstein, L. 56, 202
of the virtues 141–2 women (in the ideal state) 123
Velveteen Rabbit, the 48, 73 Xenophon 7, 70, 161

virtue
and knowledge 139–40 Zeno (of Elea) 59
224

Plato (Ancient Philosophies)

Uploaded by

Copyright:

Available Formats

Plato (Ancient Philosophies)

Uploaded by

Document Information

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

Plato (Ancient Philosophies)

Uploaded by

Copyright:

Available Formats

What are some of the major philosophical ideas discussed in the text?

What are some of the major philosophical ideas discussed in the text?

What are some of the works by Plato that are mentioned?

What are some of the works by Plato that are mentioned?

Plato

This series provides fresh and engaging new introductions to the

© Andrew S. Mason, 2010

This book is copyright under the Berne Convention.

isbn: 978-1-84465-173-3 (hardcover)

British Library Cataloguing-in-Publication Data

This book aims to present Plato’s thought to readers with no previous

thought, is the Theaetetus, and therefore it is only discussed briefly

The figure of Plato

Plato is one of the most significant figures in the history of philoso-

way strikingly modern, as he sees the physical world as having a

Plato was born in 427 bce to a wealthy upper-class family in Athens.

During much of Plato’s youth Athens was at war with Sparta,

Plato founded a school at Athens; it was in the precincts of a

We have handed down to us from antiquity thirty-five dialogues and

them is a single speech; and it seems that the Critias, an unfinished

discussion of metaphysics and the theory of knowledge, as well as

The dialogue form

the dialogue form in works where he seems clearly to be developing

ers of the later dialogues, seem to be seen as philosophers with real

There is another reason why it might be thought dangerous to

The plan of this book

My aim in this work is to introduce some of Plato’s central ideas.

Two major problems confront anyone who wants to study Plato’s

claims, he was exclusively concerned with ethical issues, and his

development, he would have been very much under the influence

Plato’s picture of Socrates

The dialogues commonly called “Socratic” paint a distinctive picture

We should be careful in just how we draw the contrast between

works. By contrast, in the Republic he is shown taking the view that

Is this the real Socrates?

To answer this question we must look at what other authors have to

character in Plato’s works. However, in two places in the Metaphys-

that everything ascribed to Socrates in the dialogues represents his

Plato’s Socratic dialogues

a particular issue or situation. One work that raises particular puz-

Plato’s development: the later phase

What seems to be recognizably the same theory is present in dia-

Continuity and change in Plato’s thought

One of the best-known aspects of Plato’s thought, and one of the

idea is indeed the ancestor of our modern English “idea”, it should

One of the most notable aspects of his theory is the so-called

All these features of Forms are certainly controversial; but none

world, which leads him to contrast it with Forms. It is this contrast

What is wrong with the world?

and nothing happens to instantiate it; it is because “perfect large-

that they are appropriate objects of knowledge; if we seek knowledge

Forms, bodies and space

In the Timaeus, Plato gives a fuller account of the nature of bod-

way underlying and composing bodies, so that bodies come into

Plato, by identifying the underlying thing with space, overcomes

How many Forms?

directly at one point, at the beginning of the Parmenides (130b ff.).

as unity, largeness and equality often feature in arguments intended

the world, and distinguishes between natural and arbitrary ways of

One of the most puzzling aspects of Plato’s discussions of Forms is