us
appears so simple: 'How do we make mistakes?' The failure of the enquiry
seems to show that we should return to knowledge, and begin with that;
and we may afterwards proceed, with a better hope of success, to the
examination of opinion.

But is true opinion really distinct from knowledge? The difference
between these he seeks to establish by an argument, which to us appears
singular and unsatisfactory. The existence of true opinion is proved
by the rhetoric of the law courts, which cannot give knowledge, but
may give true opinion. The rhetorician cannot put the judge or juror in
possession of all the facts which prove an act of violence, but he may
truly persuade them of the commission of such an act. Here the idea of
true opinion seems to be a right conclusion from imperfect knowledge.
But the correctness of such an opinion will be purely accidental; and is
really the effect of one man, who has the means of knowing, persuading
another who has not. Plato would have done better if he had said that
true opinion was a contradiction in terms.

Assuming the distinction between knowledge and opinion, Theaetetus, in
answer to Socrates, proceeds to define knowledge as true opinion, with
definite or rational explanation. This Socrates identifies with another
and different theory, of those who assert that knowledge first begins
with a proposition.

The elements may be perceived by sense, but they are names, and cannot
be defined. When we assign to them some predicate, they first begin to
have a meaning (onomaton sumploke logou ousia). This seems equivalent
to saying, that the individuals of sense become the subject of knowledge
when they are regarded as they are in nature in relation to other
individuals.

Yet we feel a difficulty in following this new hypothesis. For must not
opinion be equally expressed in a proposition? The difference between
true and false opinion is not the difference between the particular and
the universal, but between the true universal and the false. Thought may
be as much at fault as sight. When we place individuals under a
class, or assign to them attributes, this is not knowledge, but a very
rudimentary process of thought; the first generalization of all, without
which language would be impossible. And has Plato kept altogether clear
of a confusion, which the analogous word logos tends to create, of a
proposition and a definition? And is not the confusion increased by the
use of the analogous