as that of projecting into external
Nature the agency of volition, which was known to each individual as the
apparent fountain-head of causal activity so far as he and his
neighbours were concerned. To reach this most obvious explanation of
causality in Nature, it did not require that primitive man should know,
as we know, that the very conception of causality arises out of our
sense of effort in voluntary action; it only required that this should
be the fact, and then it must needs follow that when any natural
phenomenon was thought about at all with reference to its causality, the
cause inferred should be one of a psychical kind. I need not wait to
trace the gradual integration of this anthropopsychic hypothesis from
its earliest and most diffused form of what we may term polypsychism
(wherein the causes inferred were almost as personally numerous as the
effects contemplated), through polytheism (wherein many effects of a
like kind were referred to one deity, who, as it were, took special
charge over that class), up to monotheism (wherein all causation is
gathered up into the monopsychism of a single personality): it is enough
thus briefly to show that from first to last the hypothesis of
anthropopsychism is a necessary phase of mental evolution under existing
conditions, and this whether or not the hypothesis is true.

Thus viewed, I do not think that 'the general consent of mankind' is a
fact of any argumentative weight in favour of the anthropopsychic
theory--so far, I mean, as the matter of causation is concerned--whether
this be in fetishism or in the teleology of our own day: the general
consent of mankind in the larger question of theism (where sundry other
matters besides causation fail to be considered) does not here concern
us. Indeed, it appears to me that if we are to go back to the savages
for any guarantee of our anthropopsychic theory, the pledge which we
receive is of worse than no value. As well might we conclude that a
match is a living organism, because this is to the mind of a savage the
most obvious explanation of its movements, as conclude on precisely
similar grounds that our belief in teleology derives any real support
from any of the more primitive phases of anthropopsychism.

It seems to me, therefore, that in seeking to estimate the evidence of
design in Nature, we must as it were start _de novo_, without reference
to anterior beliefs upon the subject. The question is essentially one to
be