Reply to Haldane 163else for its existence. If God also exists then God could be necessary in the
same sense, but this would not be a good enough sense of ‘necessary’ for
Aquinas or Haldane, since the same question ‘Why does it exist?’ would recur
in relation to God. The universe could also be said in a sense to fill the bill for
the other desideratum put forward by Haldane in the quoted passage. If to be
eternal is to be outside space–time, the whole space–time universe including
space–time itself is at least not itself in space–time and so might also merit
the epithet ‘eternal’. Moreover, in a space–time way of looking at things we
do not speak of change or staying the same, except in the sense of temporal
parts of objects differing or being similar, and motion is just relative inclina-
tion of world lines. Similar remarks could be made about a super-universe
if the total universe of everything there is contains many universes as we
normally conceive them, as in the speculations of Carter and others that
I mentioned in chapter 1. This does not of course constitute a conclusive
answer to Haldane. The reader will have to decide for himself or herself
whether he or she understands the notion of necessity that Haldane requires.
Certainly I yearn for such a notion: it might help us to answer the compelling
but apparently unanswerable question ‘Why is there anything at all?’ But
I can’t see how I can find such a notion that would strike me as intelligible.
In my main essay I suggested that an adequate concept of God for the
theist should be that of an atemporal being, not that of a sempiternal being.
I’m not sure that Haldane is right in laying stress on a realistic notion of
causation (see pp. 123ff ). I am myself suspicious about the use of the notion
of causality in fundamental physics and metaphysics. (It is a very useful word
for plumbers, instrument mechanics and brain surgeons.) A key element in
the notion of causation is that of ‘IfA had not happened then B would not
have happened’. I take this as meaning that the happening of B follows by
logic from the happening of A together with contextually agreed background
assumptions. So the notion of causality is a contextual one. Another element
in the notion of causality is a temporal one, but I think that if we do have a
notion of causality it should not rule out backwards causation. Huw Price has
made use of the notion of backwards causation in dealing with the problem of
non-locality in quantum mechanics.^25 My own view is that Aquinas’s third
way (as Haldane states it on pp. 118ff ) might be improved by replacing
‘cause’ and ‘caused by’ with ‘explains’ and ‘is explained by’. Of course the
notion of explanation is contextual too. The main issue between us is over our
relative happiness or unhappiness at the notion of a necessary being in any
other sense than one (such as that of ‘depending on nothing else’) which can
be sliced away by Ockham’s razor. Thus the atheist could say that the universe
depends on nothing else.
Haldane raises questions to do with Hume’s epistemology. Certainly,
I have an inclination to defend something like a regularity view of laws of