relation to what, if there be any Design in Nature at all, must be
regarded as the higher purpose of such Design. Therefore, cases of
individual or otherwise relative failure cannot be quoted as evidence
against the hypothesis of there being such Design. The fact that the
general system of natural causation has for its eventual result "a fair
order of Nature," cannot of itself be a fact inimical to the hypothesis
of Design in Nature, even though it be true that such causation entails
the continual elimination of the less efficient types.

'To the best of my judgement, then, this argument from failure, random
trial, blind blundering, or in whatever other terminology the argument
may be presented, is only valid as against the theory of what Mr.
Alexander alludes to as a "Carpenter-God," i.e. that if there be Design
in Nature at all, it must everywhere be _special_ Design; so that the
evidence of it may as well be tested by any given minute fragment of
Nature--such as one individual organism or class of organisms--as by
having regard to the whole Cosmos. The evidence of Design in this sense
I fully allow has been totally destroyed by the proof of natural
selection. But such destruction has only brought into clearer relief the
much larger question that rises behind, viz. as before phrased, Is there
anything about the method of natural causation, considered as a whole,
that is inimical to the theory of Design in Nature, considered as a
whole?'


It is true that this argument does not bear directly upon the
_character_ of the God whose 'design' Nature exhibits: but indirectly it
does[33]. For instance, such an argument as that found above (on p. 79:
'we see a rabbit, &c.') seems to be only valid on the postulate here
described as that of the 'Carpenter-God.'