Kenny Anthony. Ancient Philosophy: A New History of Western Philosophy Volume 1
Подождите немного. Документ загружается.
to have made an extreme statement of scepticism: ‘None of us knows
anything, not even whether we know or do not know, nor even what
knowing and not knowing are’ (DK 70 B1). But this was at the beginning of
a book of atomistic physi cs, so it is hard to know how seriously to take this
manifesto. The sophist Gorgias, on the other hand, oVered an argument to
show that knowledge of reality was impossible. It went like this. If objects of
thought (ta phronoumena) are not real (onta), then what is real is not an object
of thought. But objects of thought are not real; for if any of them are, all of
them are, just as they are thought. But just because someone thinks of a
man Xying or chariots running on the sea, that does not mean that there is
a Xying man or chariots running on the sea. Hence it is not the case that
what is thought of is real; and therefore what is real is not an object of
thought (DK 82 B3).
We do not know whether Gorgias meant this argument seriously or not.
We need not question that if no object of thought is a reality, then no
reality is an object of thought. The weak point in the argument seems to be
the claim that if some object of thought is real, then all objects of thought
are real. The very choice of examples suggests that we can distinguish
between those cases where an object of thought is not real and those cases
where it is real (i.e. when the thought has a reality corresponding to it).
Socrates, Knowledge, and Ignorance
Protagoras and Gorgias were sophists, and it was a regular complaint
against the sophists that they were purveyors of scepticism. Some thought
that Socrates was tarred with the same brush. Socrates certainly went
round puncturing other people’s claims to knowledge, and prided himself
on his awareness of his own ignorance. But he never challenged claims to
knowledge when made by craftsmen and experts in their own particular
Welds. Indeed, over and over again, in Plato’s Socratic dialogues, we are
given a run through half a dozen arts and crafts—shoemaking, shipbuild-
ing, navigation, cookery, medicine—to provide a paradigm of knowledge
against which to test and Wnd wanting the pretensions of those who
claim moral and political knowledge. If Socrates was a sceptic, his scepti-
cism was of a limited and contingent kind. It was only of certain important
things that knowledge was unavailable; and it was not necessarily
148
EPISTEMOLOGY
Socrates seen through Roman eyes in a wall-painting from Ephesus
unavailable to human beings, it was just not to be found in the Athens of
the day.
But in order to evaluate Socrates’ epistemology, and still more in order
to understand the epistemological theses that Plato in his dialogues puts in
Socrates’ mouth, it is necessary to discuss the diVerent Greek words that
correspond more or less to the English word ‘knowledge’. The word
‘epistemology’ itself is derived from the Greek word ‘episteme’ a word
that is often used to indicate knowledge of a rather grand kind, so that one
of its English equivalents is ‘science’. Besides the verb ‘epistamai’ which goes
with this noun, there are humbler words for more everyday knowledge
and acquaintance. Hence, someone who denies the possibility of episteme in a
particular area is not necessarily a sceptic ruling out the possibility of all
knowledge.
The Delphic oracle pronounced that no one was wiser than Socrates.
After interrogating those who had a reputation for wisdom (sophia),
Socrates came to the conclusion that he was wiser than them in that he
did not falsely believe he knew matters that he did not know. Questioning
the politicians and the poets, he concluded that they did not have any real
knowledge of the areas in which they had made their reputation. When he
went to the craftsmen, however, he did Wnd that they had knowledge
(episteme) of many things where he was ignorant, and to that extent they
were wiser than he was. The problem was that on the basis of their
particular expertise, they foolishly thought themselves wise on totally
diVerent, and more important, topics. Socrates decided that he was better
oV than they, lacking both their wisdom and their ignorance (Apol. 22d–e).
In Plato’s Socratic dialogues there is always someone who claims know-
ledge in a particular area; typically, a character will claim to know the
nature of a particular virtue or craft. Thus, Euthyphro claims to have
knowledge of piety and impiety (Euthphr. 4e–5a), Meno is happy to accept
that he knows what virtue is (Men. 71d–e), and even the modest Charmides
thinks he knows what modesty is. Socrates then questions such a character
in order to get the knowledge expressed in a deWnition. As each deWnition is
produced, he declares it deWcient, e ither producin g counter-examples or
revealing ambiguities in its terms. Counter-examples can take two forms:
they can show either that the deWnition covers more than it should, or that
it covers less than it should. Thus, when Cephalus in Republic 1 says that
justice is telling the truth and returning what is borrowed, Socrates
150
EPISTEMOLOGY
complains that returning a borrowed weapon to a mad friend is not just
(Rep. 3 31c–d). On the other hand, when Laches, in the dialogue called after
him, says that courage is standing at one’s post without running away,
Socrates points out that tactical retreat can be an expression of courage
(191c). Sooner or later, the alleged expert has to admit that his deWnition
breaks down; and the failure to be able to produce a satisfactory deWni-
tion is taken to show that the claim to knowledge was unjustiWed.
The questioning Socrates, in Plato’s dialogues, is never satisWed with
being oVered a list of items falling under a certain concept such as virtue or
knowledge. Meno tells him that there are many diVerent kinds of virtue: one
for males, one for females, one for children; one for slaves and one for
freemen; one for the young and one for the old. Socrates says that this is no
use: it is like telling someone who wants to know what a bee is that there
are many diVerent kinds of bee. Bees of diVerent kinds, Socrates says, do
not diVer from one another in so far as they are bees; and what we want
to Wnd out is that very thing in which they are all the same and do not diVer
from one another (Men. 72c). So too with virtue. Socrates, we might say, is
looking for the essence of virtue.
Knowledge of the essence of something is clearly a very special kind of
knowledge: and ever since Plato’s Socrates it has been for many philoso-
phers a paradigm of knowledge. Other philosophers, in recent times, have
criticized Socratic insistence on knowledge of essences. Wittgenstein
pointed out that among the items that most interest philosophers some
may not have such an essence at all. He denied, for instance, that every-
thing we call language possesses one feature in common which makes us
use the same word for all. Rather, these phenomena are related to one
another in many diVerent ways, just as diVerent members of the same
family will resemble each other in diVerent features such as build, gait,
colour, temperament, and so on.2 Even where X does have an essence,
being able to de W ne that essence or to articul ate an exceptionless criterion
for distinguishing Xs from not Xs is not a necessary condition for being
genuinely able to tell an X when one sees one. Thus, I can know that a
computer is not alive without being able to produce a watertight criterion
to separate life from non-life.3
2 L. Wittgenstein, Philosophical Investigations (Oxford: Blackwell, 1958), 1. 66–7.
3 The denial of this is called by Peter Geach ‘the Socratic Fallacy’ (God and the Soul (London:
Routledge, 1969), 40).
151
EPISTEMOLOGY
We can agree that knowledge, in the everyday sense of the English word,
can be present in the absence of the power to deWne and delimit. It might
well be thought, however, to be the special task of the philosopher to seek
for essences, or, as the case may be, to lay out the family resemblances
between diV erent applications of a concept. The goal of this special task is
to reach a level of knowledge, or at least understanding, that is superior to
that possessed by the ordinary informal employer of a concept. And it was
for this level of insight that Plato, in his mature dialogues, came to reserve
the Greek word ‘episteme’.
Knowledge in the Theaetetus
One of the richest of Plato’s dialogues, the Theaetetus, is devoted to the
question: What is knowledge (episteme)? (145e). This dialogue, though not
an early one, has the structure usual for a Socratic dialogue: the protagon-
ist (in this case a brilliant young mathematician) oVers a series of
deWnitions, all in turn are rejected by Socrates, and the drama ends with
a proclamation of ignorance. The young Theaetetus, at the beginning
of the dialogue, is pregnant with a reply to the question, ‘What is know-
ledge?’ Socrates oVers himself as midwife to bring it to birth (149a–151d);
but the pregnancy turns out to be imaginary, with only a phantom
oVspring.
Theaetetus’ Wrst proposal is that knowledge consists of things like
geometry and astronomy, on the one hand, and shoemaking and carpen-
try on the other (146d). This will not do: Socrates is never happy with a list,
and he says that if we tried to deWne geometry and carpentry the word
‘knowledge’ itself would turn up in the deWnition. Theaetetus next sug-
gests that knowledge is perception: to know something is to perceive it
with the senses (151e). Socrates observes that since only what is true can be
known, knowledge can be sense-perception only if such perception is
always correct. But this can only be the case if we accept the thesis of
Protagoras that whatever seems to a particular person is true for him.
With regard to momentary sensations, Protagoras’ thesis may be given
plausibility by the thesis of Heraclitus that the world is in constant Xux. The
colours we see are not stable objects: when my eye encounters a piece of
marble, the whiteness of the marble and my vision of that whiteness are two
152
EPISTEMOLOGY
momentary items, twins begotten together by the encounter of the parent
eye and the parent marble (156c–d). If, then, on a particular occasion I say
‘This is white’, I cannot be wrong: no one else is in a position to contradict
me. The same is true of other kinds of sense-perception (157a).
Let us suppose we concede to Protagoras that, in such a case, what the
perceiver says goes. Still, Socrates insists, there are many other kinds of case
where it would be absurd to make such a claim. We have dreams in which
we think we are Xying; a man may go mad and think he is a god. Surely
these are cases where what seems to a person is not true? And even the
ordinary cases, where the perception is not erroneous, cannot be cases of
real knowledge. For how can we be sure that we are not dreaming? Half
our life is spent abed, and it is a commonplace that it is impossible to prove
that one is awake and not asleep (158c–e).
At this point Socrates oVers Theaetetus (and Protagoras) a response—
rather a feeble response, since it deals with the case not of dreamers or
madmen, but of sick people whose senses are aVected by their disease.
Suppose Socrates falls ill, and sweet wine begins to taste sour to him. On
the Heraclitean account, the taste of the wine is the oVspring of the wine
and the taster. Socrates sick is a diVerent taster from Socrates healthy, and
with a diVerent parent naturally the oVspring diVers. It may not be true
that the wine is sour, but it is, in his sickness, sour for Socrates. So we do
not have here a case of erroneous perception, and the equivalence of
knowledge and perception is not yet defeated.
Socrates in the dialogue moves on to diVerent terrain. There are cases of
perception without knowledge: we can hear a foreign language spoken,
and yet not know the language (163b). There are cases of knowledge
without perception: when we shut our eyes and recall something we
have seen, we know what it looks like and yet are no longer seeing it
(164a). But if knowing ¼ perception, then both these must be cases of
simultaneously knowing and not knowing, and surely that is an absurdity?
But even now, Socrates is willing to allow Protagoras a way out. It is easy to
have cases of simultaneous perception and non-perception: if you wear an
eyepatch you see something with one eye but not with the other. So if
perception ¼ knowledge, it is no surprise that you can both know and
not know at the same time (165c).
In discussing Theaetetus’ identiWcation of knowledge and perception,
Plato’s Socrates gives Protagoras a surprising amount of rope. But he is, in
153
EPISTEMOLOGY
the end, conWdent that Protagoras hangs himself on Democritus’ hook. It
seems to all men that some men know better than others: if so, that
must—according to Protagoras—be true for all men. It seems to most
people that Protagoras’ thesis is false; if so, his thesis must be, on his own
account, more false than true, since the unbelievers outnumber the
believers (170b–171d). But his thesis can be attacked more directly. How-
ever plausible it may be if applied to sense-perception, it cannot apply to
medical diagnosis or political prediction. Even if each man is an authority
on what he senses now, he is not the measure of what he will feel or perceive:
a physician knows better than a patient whether the patient will later feel
hot or cold, and a vintner will know better than a drinker whether a wine,
come next year, will taste sweet or dry (178c).
The Wnal argument by which Socrates leads Theaetetus to abandon the
proposal that knowledge is perception is this. The objects of the senses
are delivered to us through diVerent channels: we see with our eyes and
hear with our ears. Colours are not the same as sounds; we cannot hear
colours and we cannot see sounds. But what of the judgement ‘Colours are
not the same as sounds’? Where does that piece of knowledge come from? It
cannot come from the eyes, since they cannot see sounds; it cannot come
from the ears, since they cannot hear colours. Moreover, there are no
organs for detecting sameness, in the way that there are organs for seeing
and hearing . The soul itself contemplates the common terms that apply to
the deliverances of all of the senses (184b–185d).
Theaetetus, in response to this argument, moves to a second proposed
deWnition of knowledge. Knowledge is not perception (aesthesis); it is
thought (doxa), and thought is an activity of the soul by itself. When the
mind is thinking, it is as if it were talking to itself, asking and answering
questions, and silently forming opinions. Knowledge cannot be identiWed
outright with thought, because there are false thoughts; but perhaps we
can say that knowledge is true thought (187a5).
Socrates, after an interesting diversion in which he points out that the
notion of ‘false thought’ is not without its problems, oVers an objection to
this de W nition. There are cases where people have true thoughts, and form
true opinions, without having actual knowledge. If a jury is persuaded by a
clever attorney to bring in a certain verdict, then if the verdict accords with
the facts, the jurors will have formed a true opinion. But do their true
thoughts amount to knowledge? Not really, says Socrates: only an eyewit-
154
EPISTEMOLOGY
ness is in a position really to know what happened in a case of alleged
assault or robbery. So knowledge cannot be deWned as true thought.
Socrates showed earlier that knowledge is not perception by giving an
example of a piece of knowledge for which perception was insuYcient. He
has now shown that knowledge is not true opinion by giving an example of
a piece of knowledge for which perception was necessary. It might be
expected that Theaetetus might have responded by oVering an account
of knowledge embracing both perception and thought in some relation to
each other. Instead he oVers an elaboration of his second deWnition.
Knowledge, he now suggests, is true thought plus a logos; and he proposes
three forms that the logos may take (206c).
‘Logos’, as has been remarked, is a diYcult word to translate because it
corresponds to many diV erent English words: ‘word’ itself, ‘sent ence’,
‘discourse’, ‘reason’. In the present context it is clear that for Theaetetus
a true thought with a logos is a thought that is somehow articulated in a
way that a thought without a logos is not: but I shall leave the word
untranslated while explaining the diVerent kinds of articulation he has in
mind.
One way in which one can give a logos of a thought is by expressing it in
words. But being able to articulate a thought in this sense cannot be what
makes the diVerence between true thought and knowledge, since anyone
who is not dumb is capable of doing so (206d–e).
More plausibly, a logos may be a kind of analysis. To know what X is is to
be able to analyse it into its elements. Thus one can exhibit knowledge of a
word by spelling it out in letters. If that is what knowledge is, then
knowledge of reality must be exhibited in analysing it into the ultimate
elements of which it is composed. But the analogy with spelling places us
in a diYculty. The word ‘Socrates’ can be analysed into its elements, such
as the letter S. But the letter S cannot be further analysed; unlike the word
‘Socrates’, the letter S has no spelling. So if knowledge involves analysis, the
ultimate, unanalysable elements of the universe cannot be known. And if
the elements of a complex cannot be known, how can the complex itself be
known? Moreover, a mere listing of the elements of a complex is insuY-
cient for knowledge unless the elements are put together in the correct
way (207b).
Theaetetus’ Wnal account of giving a logos of an object is giving a descrip-
tion that is uniquely true of it: thus one might give a logos of the sun by
155
EPISTEMOLOGY
saying that it is the brightest of the heavenly bodies. But does this amount
to real knowledge of the sun? Surely, being able to oVer some deWnite
description of X is a necessary condition of having any thought at all about
X; it is not su Ycient to turn a true thought about X into a piece of genuine
knowledge.
At this point Theaetetus gives up. The thoughts he has delivered with
the aid of Socrates’ midwifery turn out to be mere wind-eggs. We are far
from having reached a deWnition of knowledge; and hence all the use of
words like ‘know’ and ‘not know’ throughout the dialogue turns out to
have been illegitimate (196e).
Perhaps Theaetetus gave up too soon. If he had oVered a fourth account
of ‘logos’ as meaning something like ‘justiWcation’, ‘reason’, or ‘evidence’,
then his deWnition of knowledge as true belief plus logos would have been
found satisfactory by many a philosopher during the subsequent millennia
of philosophy. But Plato’s Socrates was a hard man to satisfy, and Plato
himself, in the sixth and seventh books of his Republic, has his Socrates
present quite a di Verent epistemology in quite a diVerent style.
Knowledge and Ideas
The presentations in the two dialogues diVer above all because the Republic
appeals, as the Theaetetus does not, to Plato’s Theory of Ideas. Common to
both dialogues is the principle that what is known must be true; knowledge
can only be of what is. The Ideas are relevant in the Republic because Plato is
committed to the thesis that only the Ideas really are: that is to say,
everything other than an Idea is what it is only in a qualiWed sense.
Beautiful things other than the Idea of Beauty, for instance, are beautiful
only at one time and not another, or beautiful only in one part and not in
another. Nothing except the Idea of Beauty is just beautiful, period (Smp.
211a). The Ideas make their Wrst appearance in the Republic in the Wfth book,
where Plato is describing the philosopher. He describes him as the lover of
truth, and distinguishes him from the mere dilettante, the lover of sights
and sounds.
The non-philosopher does not know the diVerence between beautiful
objects and beauty itself: he is living in a dream, mistaking an image for
reality (Rep. 476c–d). For the state of mind (dianoia) of such a person, Plato
156
EPISTEMOLOGY
uses the word ‘doxa’, which in the Theaetetus was used for thought, or belief .
He contrasts it with the knowledge that belongs to the philosopher—here
called gnome. If knowledge must be knowledge of what is, and only an
Idea utterly is, then knowledge must be knowledge of Ideas. If there is
anything at the opposite pole from an Idea, something that utterly is
not, that is totally unknowable. But most things that are F are partly
F and partly not F, F in one respect and not in another. They are set in
between what is utterly F and what is utterly not F. These are the objects
of doxa.
At this point a fundamental diVerence emerges between the Republic and
the Theaetetus. In the Theaetetus we sought to locate the essential characteris-
tic of knowledge as a feature of the state of mind of the knower: is it a
matter of sensation? must it include a logos? But in the Republic the
diVerence between knowledge and belief is a diVerence between objects:
between what is known and what is thought of. This point is made quite
explicitly. Knowledge and thought, Plato says, are powers (dynameis), just as
sight and hearing are powers. Powers do not have colours or shapes by
which we can tell one from another. ‘In the case of a power I look only at
what it is concerned with and what it does to it, and by reference to that
I call each the power it is’ (477d). Sight is a power to discriminate colour,
and hearing a power to discriminate sound: it is the diVerence between the
objects, colour and sound, that distinguishes these two powers from each
other. Similarly, Plato proposes, the diVerence between knowledge and
belief is to be determined by noting the diVerences between the two kinds
of object with which they deal (478b6 V.).
In book 6 Plato takes this line of argument further, and subdivides
gnome and doxa. Doxa, or thought, has the visible world as its realm, but
it comes in two diVerent forms that have diVerent objects. One form
is imagination (eikasia), whose objects are shadows and reXections;
another form is belief (pistis), whose objects are the living creatures
about us and the works of nature or of human hands. The realm of
gnosis, of knowledge, is likewise divided into two. Knowledge par excel-
lence is noesis, or understanding, whose object is the Ideas that are the
province of the philosopher. But there is also another kind of know-
ledge, typical of the mathematician, to which Plato gives the name
dianoia (509c5 V.). The abstract objects of the mathematician share with
the Ideas the characteristic of eternity and unchangeability: they
157
EPISTEMOLOGY