Reply to Haldane 155the molecules might be rather large and complex and a complete quantum
mechanical account of the process could be beyond the fastest computers. We
could still feel, as Weinberg suggested, that there is some sort of explanation
of the chemical process. We could say that the chemical process is similar
to that in the simple case, and that prediction is here defeated only by sheer
complexity. Similar situations of course exist with respect to deterministic
mechanisms in classical mechanics, as it is demonstrated in chaos theory.
Chaotic systems can indeed be deterministic but unpredictable.
On p. 93 Haldane says that if a materialist explanation of life seems
incomplete then ‘only a non-scientific insistence on reductionism motiv-
ates the thesis that [living systems] must be no more than mechanism even
where there could be no deductive explanation of how it is so’. I would reply
that if there are plausible ideas about how something could be so, in accord-
ance with naturalistic principles, even though this cannot be deduced in
detail, and if there are no plausible alternative naturalistic explanations, then
it is reasonable to suppose that things did come about in the hypothesized
way. I do not think that scientists regard this sort of reasoning as ‘unscien-
tific’, even though (often per impossibile) detailed predictions or retrodictions
would be regarded by them as better. Rejection of appeal to non-natural
causes (in any of the senses of ‘cause’ distinguished by Haldane) is really
only an application of Ockham’s razor, the principle that entities should
not be multiplied beyond necessity. No doubt Haldane holds that the non-
naturalistic explanation is simpler, but any appearance of simplicity could
be deceptive, if the appeal is simply to a God whose ways are beyond our ken.
I agree that the argument is not over. Readers of this book must make up
their own minds.
2 Representation and Intentionality
On p. 91 Haldane sees difficulties for naturalism in the notions of repre-
sentation and intentionality. Now undoubtedly there are such things as re-
presentations. A portrait is a representation of a person, and an irregular blue
line on a map is a representation of the twists and turns of a river. Whether
there are representations in the brain or mind is a further matter, and quite
controversial. For example, if the brain is entirely a connectionist device then
there is no place for representations (pictures?) in any obvious sense. Perhaps
‘information’ is a more useful word than is ‘representation’, in a rather
abstract, information-theoretic use of the term ‘information’. It is a familiar
thought that DNA codes genetic information, much as instructions in a
computer are programmed in. There seems to be nothing very difficult
for naturalism in supposing that such capacities for acquiring and storing