Back to September, 2004 : Forward to November, 2004
THE COMING DARK AGE
Newsletter
October, 2004
1. INTRODUCTION
This month's newsletter discusses the way that setbacks occur on all temporal and spatial scales in complex evolving
systems, not just social systems but also biological and physical systems.
Past editions of the newsletter are at the following address: http://www.darkage.fsnet.co.uk/Newsletter.htm I welcome
all comments, suggestions and contributions, especially the latter. Please forward this newsletter to anyone you
think might be interested. Marc Widdowson
2. ROME AND THE DINOSAURS
Recent articles in the scientific press have concerned new theories about that hoary old chestnut, the demise of
the dinosaurs. In the early 1990s, the suggestion that the dinosaurs had been wiped out by a meteorite impact began
to gain acceptance. The evidence consisted of a high concentration of iodine at the geological Cretaceous-Tertiary
boundary, which supposedly pointed to some kind of impact event, plus the existence of a meteorite crater of the
right size and age in the Gulf of Mexico. However, some scientists disagreed. Rival theories included the idea
that the extinction was linked to the prolonged volcanic activity that laid down India's so-called Deccan traps,
the idea that dinosaurs had been wiped out by disease, the idea that it was the result of climate change, and the
idea that the dinosaurs simply lost out to more efficient creatures i.e. the mammals. Most recently, University
of Colorado physicist, Brian Toon, has added a further twist to the meteorite impact theory, arguing that it would
have created such a high level of lethal radiation that the dinosaurs would simply have been burnt to a cinder.
Meanwhile, he has been opposed by other scientists who point out that the dates calculated for the meteorite and
the dates for the disappearance of the dinosaurs are in disagreement by something like half a million years (the
death of the dinosaurs coming later). Paleontologists have also shown that the demise of the dinosaurs was no sudden
event, and that their branch of the animal kingdom had actually been in decline for millions of years before the
Cretaceous-Tertiary cut-off. This proliferation of competing theories for the dinosaur extinction has parallels
with the proliferation of theories for events in human history, such as the disappearance of the Roman empire (explained
by everything from bad plumbing to barbarian invasion). In both cases there is a similar conclusion to be drawn
- all the theories are partly correct, and none of them has a monopoly on the truth. The fall of Rome was not due
to one single cause. Rome had many problems and it was their interaction and the way that they compounded each
other which ensured that the empire was doomed to failure. Similarly, with the dinosaurs, meteorite impacts, volcanic
eruptions, climate change and competition from other lifeforms may have all been part of the mix. The dinosaurs
could probably have survived any one of these things in isolation. It was their combined effect that proved so
devastating. There is a further point, which is that problems and disasters are always happening to complex systems,
whether it be biological systems or socio-historical systems. These can have effects that range continuously from
the mild up to the truly catastrophic. The catastrophic incidents do not need any special explanation - they are
simply part of this overall continuum. What one observes is that the larger the event, the more rarely it occurs.
Nevertheless, the large events are due to the same general background of knocks and insults that cause the much
more frequent minor events. This largely connects with the theory of 'self-organised criticality', developed by
the physicist Per Bak (for a popular exposition see his book "How Nature Works" and/or the book "Ubiquity"
by science journalist Mark Buchanan). According to this theory, the large disasters are not necessarily caused
by large knocks. Whether a particular knock has a large or a small effect depends on the configuration of the system
at the given time. It is a bit like the childhood game of 'boxes' - where you draw a grid of dots on a piece of
paper and you and your opponent take turns joining the dots, with the aim being to enclose the largest number of
boxes. Sometimes your opponent's move allows you to complete just one or two boxes, but sometimes it unleashes
a chain reaction that allows you to fill in a large area of boxes. The input is always the same - drawing one line
at a time - but the output can vary enormously depending on the configuration of the system. The theory of self-organised
criticality goes beyond this simple observation to assert that a complex system tends naturally to evolve to this
critical state where events of any magnitude can occur, and that there is a particular mathematical relation between
the size of the event and its frequency of occurrence. When you hear anyone offering a simple explanation of some
natural or historical phenomenon, therefore, be very suspicious. Such explanations can seem satisfying, but there
are often problems with them, as indicated by the existence of rival theories. For me, the rejection of these special
theories in favour of more general, abstract theories such as that of self-organised criticality is actually more
satisfying in the long run. It offers the possibility that we can understand how history works, rather than being
forced to regard it as a series of random accidents. Similarly, in the natural world, the phenomenon of convergent
evolution - e.g. that the eye evolved independently but along essentially the same lines in both vertebrates and
molluscs - reinforces the idea that there may be a deeper logic to the evolution of the biosphere, and it is not
simply a case of chance meteorite impacts. Chance plays its role, but there are more fundamental patterns that
guide the process - just as chance molecular collisions determine how individual smoke particles rise from a bonfire,
and yet at a suitable level of abstraction all columns of smoke look the same.
Back to September, 2004 : Forward to November, 2004