25. The main question for those who hold to the doctrine of
ideas is how to account for propositions. Only if we are able to give
some account of propositions shall we be able to distinguish between
truth and error and therefore to develop a scientific theory.

Now in a proposition it may be said roughly that we have a
certain character or attribute applied to some particular thing or
group of things. Now the question is What exactly is the nature of
this thing, that is, of what can be the subject of a proposition? As
far as such subjects or things are concerned, Berkeley himself appears
to be in entire agreement with Locke---as is indicated, for example,
in the first paragraph of the “Principles”---in holding that such
things consist of a number of different ideas which may be said to
go together or accompany one another. Locke, of course, maintains that
in thinking of any definite thing we have in mind, not merely the
various ideas, but also a certain substance in which they inhere and

  ― 48. ―
view facsimile

which accounts for their cohering or going together. Now, as Berkeley
points out, in terms of the theory of ideas this substance would
simply be one idea among others, of the same order as the rest and
therefore incapable of explaining how they could go together. Accordingly
the thing, as conceived by Berkeley, must be understood as a
proposition of the form that “The idea A accompanies the idea B”, and
similar as a proposition, although not of precisely the same form, to
those which he calls laws of nature in which idea A is said to precede
idea B. In fact these are simply two special cases of the general
relation of significance, the idea A being a sign of the idea B
whether the latter accompanies or succeeds it; that is, of course,
leaving aside the difficulties connected with this distinction between
accompanying and succeeding, seeing that Berkeley holds that,
instead of our ideas being in time, time consists of nothing more
than a number of our ideas.


As far as Locke's theory goes, then, we find that by means
of this doctrine of substance, or that in which different ideas inhere,
the attempt is made to account for our knowledge of complex things
which can be made the subject of different propositions and the
different characters attributed to them; and Berkeley's criticism
amounts to this that, starting with the ideas, we cannot arrive at the
substance and therefore we cannot have propositions in the form suggested
by Locke. This substance is perfectly abstract; it does not
account for any actual coherence of ideas but says at most that ideas
are capable of cohering. But the criticism of abstraction is more
closely connected with the attempt to account for predicates. When N.B.
we say that a thing is of a certain character we are using that character,
not as a separate and concrete idea, but in subordination to the
thing which has that character. Moreover, as Locke points out, different
things can have the same character, that is, we can have the same
predicate attributed to them; and, in order that this should be possible,
we must, he thinks, regard that character abstractly and independently
of any of its particular embodiments since if we did not do so, we
should have to say that any two things really had different characters
and were incapable of being described in the same way. He suggests,
then, that these abstract ideas are arrived at by considering a number
of things which resemble one another; by then omitting all the
points in which they differ and leaving only what is common to them
all. And that common element is the predicate or universal arrived
at by abstraction, that is, by leaving out differences; for example, the
abstract idea of a triangle, that is, what we mean by the term triangle
used generically, has not any of the peculiar qualities of particular
classes of triangles but has only what is common to all triangles.
And what Berkeley contends is that it is really impossible to formulate
such an idea, that is, to have before our minds as one particular object
that which is common to all triangles and nothing at all which
is peculiar to some triangles. And even Locke seems to have had
doubts about the possibility of such a conception since he says that
the abstract idea of a triangle is of one which is neither acute
angled, right angled nor obtuse angled; neither equilateral, isosceles
nor scalene but, as he puts it, all and none of these at once.

Now this contradiction shows that Locke himself was incapable
of thinking of a triangle which was neither equilateral, isosceles
nor scalene and that he was working out his theory not in accordance
with the ideas which actually confronted him, but in accordance with

  ― 49. ―
view facsimile

what he thought ought to be the case in order to account for the way
in which “thing” is employed. In our ordinary speech we do use predicates
and it does seem impossible to identify the predicate triangle
with any particular sort of triangle because then it would not be
applicable to the other sorts of triangles. But if we identify it
with none of these kinds of triangle, then, whatever we have, we do not
appear to have any particular idea before our minds.


Now Berkeley points out this contradiction and maintains
quite correctly that we never have, as a single object of thought,
something that can be described as an universal or abstraction; but
what Berkeley does not see is that similar contradictions are bound
to arise so long as we do try to reduce our knowledge to certain
ultimate and separable elements, because we are then unable to find
the connection between these elements and thus to account for the
complex situations which we actually recognise in the statements we
make. Now considering simply Locke's theory we find that he had
begun by saying that knowledge was made up from a number of particular
ideas, and it is only by a combination of a number of those ideas
that we get the notion of a particular sort of thing. If we then
proceed to break up that complex notion, then all that we can get are
the simple ideas that we started with and therefore we can never
arrive at anything by abstraction beyond what we knew already; we can
arrive at the simple ideas which are the elements of the complex ideas
but if these simple ideas are abstract, then we have no need for any
special process of abstraction; but if they are not abstract, then the
process of abstraction is impossible. In fact here again, just as in
his distinction between primary and secondary qualities, Locke is
assuming at one and the same time that our knowledge is limited to
ideas and that we have a direct acquaintance with things for it is
note only by comparing direct things in their complexity that we can discover
their resemblances and difference. But there is nothing in such
comparisons to show that we could arrive at elements which could not
be further analysed. The abstract ideas of Locke, then, must be identified
with the simple ideas which he regarded as the original contents
of our knowledge; and, as he himself admits, the only relation which we
can find between two abstract ideas, each being perfectly simple, is
the relation of difference, that is, the fact that one is not the other;
but we could never arrive at those complex relations of which our
ordinary statements appear to give us information. And the same
criticism applies to Berkeley because, although he has seen some of
Locke's difficulties, he himself likewise takes, as the foundation of
knowledge, a plurality of individual ideas which are perfectly simple.