370 | 33 PIONEER Of ARTIfICIAl lIfE
Now, half a century later, such machines exist—and they are invaluable in the study of spe-
cific complex systems and of the mathematics of complexity in general. For instance, Turk was
able to use powerful computers to play around with the numbers in Turing’s own equations.
Without them he could not have done the calculations and explored the huge set of possibili-
ties, only some of which result in life-like forms. Similarly, he could not have produced even a
single leopard spot without the aid of computer graphics: indeed, computer graphics have been
invaluable in many areas of mathematical modelling.
But plus ça change, plus c’est la même chose: Turing’s 21st-century successors face updated
versions of the methodological problems that faced him. We have seen, for example, that
Goodwin’s Acetabularia model couldn’t illustrate some of the implications of his mathematics
because of lack of computer power. So it is still the case that Turing-rooted biological ideas
cannot always be applied in practice, even if their relevance is evident in theory.
A third obstacle to the take-up of Turing’s work was that even if his mathematics was impec-
cable as mathematics, he didn’t (and couldn’t) prove that real biological structures actually do
emerge in this way. Only experimental developmental biology could show that. (Likewise,
only psychological experiments can confirm the psychological reality of a computer model
of mind.)
This question couldn’t be experimentally addressed in Turing’s lifetime. Strictly, that’s not
quite true: one of his six references cited then-recent research showing that the open (‘head’)
end of a growing Hydra tube develops patches of chemicals that show up with a particular stain,
and that the tentacles subsequently arise at those points.^35 But the nature of the chemicals, and
just how they could form the whorl of tentacles, were mysteries. The genetic and biochemical
techniques needed to solve such mysteries were not yet available. Today, the empirical ques-
tion raised by his theory—or rather, the host of different questions—are at last being debated.
Goodwin’s study of Acetabularia is just one example.
The last reason for the delay in following up the work of Turing (and of D’Arcy Thompson)
was that it approached biology in a deeply unfashionable way. D’Arcy Thompson had been
clear on that point: natural selection, he said, was a secondary factor in morphogenesis. If he
had lived to see the rise of molecular biology after Crick and Watson’s discoveries, he would
have had reservations about that too, for its reductionist approach discourages biologists from
asking morphological questions. Unlike D’Arcy Thompson, Turing didn’t mount an explicit
attack on neo-Darwinism. Nevertheless, his ideas, too, were out of line with the biological
orthodoxy.
Today Turing’s conceptual chickens have come home to roost and are sitting pretty on their
perches: indeed, they are becoming increasingly fashionable. The mathematical verdict on ‘The
chemical basis of morphogenesis’ is highly positive. The biological verdict is generally positive
also. Even orthodox neo-Darwinists allow that the questions asked by the structuralist biolo-
gists are interesting—and that many of their answers are valuable, too.
The editor of a collection of Turing’s writing on morphogenesis (one of Goodwin’s collabora-
tors) has even said that his 1952 paper has been cited more often ‘than the rest of his works
taken together’.^36 I doubt this, because the Turing test paper has been cited countless times
by philosophers, computer scientists, and members of the general public,^37 but undoubtedly
the reputation of this long-neglected work has rocketed. Given the surging public and profes-
sional interest in various areas of biology, that rocket will still be airborne—and visible to ‘every
schoolboy’—one hundred years from now.