26. Berkeley, in criticising abstraction as it is conceived by
Locke, points out that there is quite a sound method of abstraction;
that there is an actual process which we can perform and which could
reasonably be described by means of that term, namely, that of drawing
a distinction between things which really do exist apart from one
another. What he objects to; what he regards as “vicious abstraction”,
is drawing a distinction between things which are not really separable
from one another. But, as we have already seen, the difficulty
in this case is to show how this vicious abstraction is really possible.
If two things or what we suppose to be two things are really
one thing, if there is no distinction whatever between them, then it
would not appear to be possible even to say wrongly that the one was
distinct from the other; we should be formulating a proposition of the
form: “A is not A” and no one could ever possibly have believed a

  ― 50. ―
view facsimile

proposition of that type. Thus until he has shown how the kind of
abstraction he objects to is possible, Berkeley's criticism is not
complete. He appears to consider that it arises from the wrong use
of words but he is unable to show how words can be thus wrongly
employed. For example, he criticises Locke's statement that words
become general by being made to stand for an abstract general idea,
and he says, in opposition to this view, that words become general by
note being made to stand for several particular ideas; but this does not
noteexplain what “becoming general” means, What we really have in mind
when we employ a word in that way; that is, whether we use a word particularly
or generally, we must have something definite in mind in
either case since otherwise we could not distinguish between the two
uses. We may say, then, that Berkeley's criticism of Locke may be valid
in so far as he demonstrates that we never have before our minds, as a
single object of thought, something which we could call a general idea.
Obviously, if we had any one such idea before our minds, it would be
a particular idea; and Locke appears to feel that difficulty himself
when he says that the general idea of a triangle is neither equilateral,
isosceles nor scalene and so on but is all and none of these
at once. But if we do not arrive at general ideas by taking a number
of distinct things, neglecting the points in which they differ and
thus thinking solely about what they have in common, the question still
remains: What are we to make of generality, what are we to make of the
fact that, according to our statements, many different things may have
the same character?

Now Berkeley's explanation of generality is that it is a
relation among a number of different things such that any one can be
taken to represent all the others. Thus in geometry we work with a
particular triangle, but we do so on the understanding that this
triangle is to be taken as representing all other triangles, so that
whatever we discover about this triangle can also be attributed to other
triangles. Now a theory of this kind does not really explain
the use of general terms or even the procedure adopted in geometry.
If we are dealing with a number of pure particulars, then it would be
impossible to say exactly what others a certain one could be taken to
stand for. On the other hand, if we have determined what particular
objects or ideas a certain one was to represent, then it would have to
represent just these and no others; that is to say, we should have to
say quite arbitrarily that the idea A has to stand for ideas B,C, E, etc.
But if this were so, then it would be impossible to make those discoveries
which we actually do make in ordinary experience. If we say
that the idea A is the idea “man” and that it stands for the ideas B,
C, D, E, etc., then we should be entitled to applyemploy the same term “man”
to any one of these represented particulars. But if we come upon a
new particular, one which has not been already labelled in accordance
with this finding, then we should never think of saying that this new
particular was a man; that is, we should have an absolutely unexplained
relation of representation among a group of ideas and we should be
incapable of making any new discovery. In fact, again taking the case
of men, we should require, in terms of this theory, to say that any new
particular that was presented to us was not a man. The point is, then,
that in recognising a certain individual as belonging to a certain
class we must proceed, not on the basis of any relation of representation,
but on the basis of something that we actually find in this new
particular; that is, we must find that the individual possesses

  ― 51. ―
view facsimile

some characteristic in virtue of which we apply the term “man” to him.
Starting with isolated ideas which could only represent one another,
we could never recognise the qualities of things, that is, we could
never have propositions in which a certain subject is said to have a
certain predicate as we actually do. It is necessary, therefore, that
we should be able to distinguish the various characters of any particular
thing although this does not mean, as Locke supposed, that we
can separate these characters and take any one of them as an isolated
object of thought because in that case there would be just as great a
difficulty, as there is in Berkeley's theory, in determining how that
particular idea could stand for a number of others and what others it
actually did stand for.

If we take the case of this triangle, then if we speak about
it as a triangle that means that we are considering that particular
character which it has. But it would be equally possible to speak of
it as a scalene triangle and to consider what was implied by its having
that peculiar character; and unless we can distinguish within it
those two characters, being scalene and being a triangle, we could not
decide how, in a particular case, it might be taken as standing for all
triangles and how, in another case, it might be taken as typical or
representative of that particular species of triangles which are
scalene. In fact if we were dealing with pure particulars, things in
which no special characteristics could be distinguished, then any one
might stand for any other and it would be quite impossible to justify
the various terms we employ and the distinctions which we make. We
may say, then, that Berkeley is right as against Locke in considering
that the characters of a thing are not parts of it, and that we dod not
arrive at a knowledge of the thing by adding its characters together.
But so long as he accepts the theory of simple ideas, so long, that is,
as he considers that any really particular thing can have only one
character, he cannot really refute Locke's theory of abstraction since
these simple ideas are as abstract as anything can be; and he cannot
give any account of predication, that is, of the attribution of characters
to things. In saying, for example, that he has no abstract idea
of a man, he says the idea of man that “I frame to myself, must be either
of a white, or a black, or a tawny, a straight, or a crooked, a tall, or a
low, or a middle-sized man”. (Introduction: Paragraph X). In that way
Berkeley considers we have an idea which is really particular but which is
capable of standing for other particulars. But what he does
not observe is that all those terms by means of which he could particularise
the kind of man he had in mind are just as general as the
term “man” itself. If it were a white man that was in question, then,
according to the same argument, we should have to state the precise
note shade of whiteness in order to make the idea of whiteness particular
enough for Berkeley's requirements. note And in the same way we should have
to state exactly the height of the man and so on with any description
which might be ascribed to him, so that by the time we should arrive
at a pure particular we should have something to which no particular
idea could be applied; the most that we could do would be to point and
say “this” or “that”. But once we had an indescribably particular of
this kind, we should have reason for taking it to stand for any
note other particular; no ground of distinction between those other particulars
which it did represent and those which it did not represent.
The results of this argument, then, are that:
I. we can use no descriptive epithets; we simply have a pure individual
which cannot intelligibly be called black or white or said to be a
man or anything else.

  ― 52. ―
view facsimile

II. if we did use any term to stand for that particular individual,
then that would mean the individual as a whole and not anything
particular about the individual, and consequently, if two different
terms were applied to the same individual, these two different terms
would have exactly the same meaning; and this is the basis of the
argument that the being of sensible things is their being perceived;
that is, sensible things being those things to which we can attribute
”being” and also “being perceived”. And those two terms each meaning
the thing as a whole and not any particular character of it, the result
is that the terms have exactly the same significance.