114 J.J. Haldane
To bring this out consider a further ambiguity. What is meant by
talking about ‘many universes’? In futuristic fantasies, space travellers often
journey to ‘other worlds’. This way of speaking of far away and hitherto
unknown places pre-dates science fiction. The European explorers of the
fifteenth and sixteenth centuries sailed from the ‘old’ world and discovered
the ‘new’, but in saying so no one intends that they left the planet. Similarly
one might speak of ‘other universes’ meaning far distant and currently
unobservable regions of the Universe – the one spatio-temporal-causal
continuum. Alternatively one might mean, though this is much harder to
make sense of, entirely distinct cosmic set-ups, wholly discontinuous with
the Universe we inhabit.
If the hypothesis of plural universes invokes the former idea then it is clear
enough what is being said, but it should also be evident that it fails to serve
the purpose intended. Any evidence we could have for the existence of
spatially or temporally distant regions and systems would necessarily be evid-
ence for situations generally like those obtaining in our sector – that is to say
situations exhibiting the same finely tuned features whose existence seemed
to call for explanation. This is so because all that could lead us to postulate
and predict the character of distant universes would be the application
of observational-cum-inferential methods to empirical-cum-theoretical data
available to us here. So if ‘many universes’ means ‘many local set-ups’ within
the Universe the hypothesis fails to defuse the power of the new design
argument. If on the other hand it is being claimed that there could be many
Universes – entirely distinct realities, wholly discontinuous and sharing no
common elements – then, while it is uncertain how to interpret this, it is clear
that there could be no empirical evidence in support of the hypothesis, and
nor could it be derived as a necessary condition of the possible existence
and character of the only universe of which we have or could have scientific
knowledge. In short the hypothesis appears as entirely ad hoc, introduced only
to avoid what for the naturalist is an unpalatable conclusion, viz., that the
general regularities and particular fine tuning are due to the agency of a
designer –et hoc dicimus Deum.
Some readers will be struck by the parallels between the many universes
hypothesis and another theoretical construction, namely the so-called ‘Many-
Worlds Interpretation’ of quantum mechanics. This is a response to a deeply
puzzling feature of a major part of fundamental physics. In a quantum-
mechanical situation it seems that there are indeterministic transitions
between states. The theory tells us that a system will go from A to either B or
C, but in principle it cannot tell us which one it will go to – the outcome is
indeterminate. Among those who find this situation unacceptable some main-
tain that the uncertainty is only epistemological. There is a fact of the matter
involving ‘hidden variables’ but for one reason or another we do not, or