no adaptive action of quite a novel kind is ever
performed from the first without consciousness of its performance, and
therefore, although it is true that by repetition its performance may
become mechanical or unconscious, this does not prove that consciousness
was without use in producing the adaptive action. It only proves that
after a nervous mechanism has been elaborated by the help of
consciousness, consciousness may be withdrawn and leave the finished
mechanism to work alone; the structure having been completed, the
scaffolding necessary to its completion may be removed.

But passing over this difficulty which the theory of conscious
automatism seems bound to encounter in its collision with the theory of
natural selection, the most insuperable of all its difficulties arises
from the bare fact, which it cannot explain, that conscious intelligence
exists, and exists in the most intimate relation with one peculiar kind
of material structure. For automatists must concede that the evidence
of causation in the region of mind is at least as cogent as it is in the
region of matter, seeing that the whole science of psychology is only
rendered possible as a science by the fundamental fact of observation
that mental antecedents determine mental consequents. Therefore, if we
call a physical sequence _A, B, C_, and a mental sequence _a, b, c_
automatists have to explain, not merely why there should be such a thing
as a mental sequence at all, but also why the sequence _a, b, c_ should
always proceed, link for link, with the sequence _A, B, C_. It clearly
is no answer to say that the sequence _A, B, C_ implies the successive
activity of certain definite nerve-centres _A', B', C'_ which have for
their subjective effects the sequence _a, b, c_ so that whenever the
sequence _A, B, C_ occurs the sequence _a, b, c_ must likewise occur.
This is no answer, because it merely restates the hypothesis of
automatism, and begs the whole question to be discussed. What methodical
reason demands as an answer is simply why the sequence _A, B, C_ even
though we freely grant it due to the successive activity of certain
definite nerve-centres, should be attended by the sequence _a, b, c_.
Reason perceives clearly enough that the sequence _a, b, c_ belongs to a
wholly different category from the sequence _A, B, C_ the one being
immediately known as a process taking place in a something which is
without extension or physical properties of any kind,