plain sailing, but my thesis will probably grow less
plausible to the reader when I pass from percepts to concepts, or from
the case of things presented to that of things remote. I believe,
nevertheless, that here also the same law holds good. If we take
conceptual manifolds, or memories, or fancies, they also are in their
first intention mere bits of pure experience, and, as such, are single
_thats_ which act in one context as objects, and in another context
figure as mental states. By taking them in their first intention, I mean
ignoring their relation to possible perceptual experiences with which
they may be connected, which they may lead to and terminate in, and
which then they may be supposed to 'represent.' Taking them in this way
first, we confine the problem to a world merely 'thought-of' and not
directly felt or seen.[10] This world, just like the world of percepts,
comes to us at first as a chaos of experiences, but lines of order soon
get traced. We find that any bit of it which we may cut out as an
example is connected with distinct groups of associates, just as our
perceptual experiences are, that these associates link themselves with
it by different relations,[11] and that one forms the inner history of a
person, while the other acts as an impersonal 'objective' world, either
spatial and temporal, or else merely logical or mathematical, or
otherwise 'ideal.'

The first obstacle on the part of the reader to seeing that these
non-perceptual experiences have objectivity as well as subjectivity will
probably be due to the intrusion into his mind of _percepts_, that third
group of associates with which the non-perceptual experiences have
relations, and which, as a whole, they 'represent,' standing to them as
thoughts to things. This important function of the non-perceptual
experiences complicates the question and confuses it; for, so used are
we to treat percepts as the sole genuine realities that, unless we keep
them out of the discussion, we tend altogether to overlook the
objectivity that lies in non-perceptual experiences by themselves. We
treat them, 'knowing' percepts as they do, as through and through
subjective, and say that they are wholly constituted of the stuff called
consciousness, using this term now for a kind of entity, after the
fashion which I am seeking to refute.[12]

Abstracting, then, from percepts altogether, what I maintain is, that
any single non-perceptual experience tends to get cou