At Home in the Universe: The Search for the Laws of Self-Organization and Complexity , by Stuart Kauffman, New York: Oxford University Press, 321 pages, $25.00
Back in the ninth grade, I was subjected to that bogeyman of all liberal intellectuals, a creationist biology teacher. For the most part, he followed the standard curriculum (which even in the mid-1970s was heavy on ecology and "environmental" science), but when it came time to discuss evolution, his heterodoxy appeared. It was mostly standard stuff--nothing likely to faze a free-thinking 14-year-old. There was one point, however, where my teacher, a gentle man truly concerned about our immortal souls, hit on an argument I found disturbing.
He pointed out that evolutionary theory had no good, or even very plausible, way to explain how life could have arisen from non-living materials to begin with. There are some simple probability calculations one can perform to guess how long it would take randomly interacting molecules of a prebiotic sort, even under favorable conditions, to spontaneously form the very particular building blocks of today's organisms. The results aren't pretty for a non-miraculous account of life's origins--the biosphere looks like an extremely unlikely accident.
It is not only at the origins of life that mainstream biology invokes accident as a central explanatory principle. As Stuart Kauffman, fellow at the Santa Fe Institute and winner of a MacArthur "genius" grant, points out in At Home in the Universe, "Biologists see organisms as tinkered-together contraptions, and evolution as a tinkerer. Organisms are Rube Goldberg machines; the jawbone of an early fish became the inner ear of a mammal. Organisms really are full of the strangest solutions to design problems. Biologists delight in discovering these and noting to one another, and particularly to those of us inclined toward theory in biology, `You'd never have predicted that!' Inevitably the assertion is correct." Kauffman cites a statement by Jacques Monod, a Nobel-winning biologist, that "Evolution is chance caught on the wing."
We and everything living, in this view, are entirely historical accidents. Kauffman's research program--one might almost say his obsession--is to demonstrate the possibility of providing better explanations than "accidents happen" when confronted with the intricate structure and behavior of complex evolving systems.
At Home in the Universe is a popularization of his 1993 tome, The Origins of Order, where Kauffman first set out a comprehensive statement of his vision. That vision takes in not only biology but also economy, culture, and society. It seeks a new synthesis between reductionism and holism, as well as chance and necessity; it tries to lay a mathematical foundation for predicting the occurrence of spontaneous order while considering the role of evolution and selection; it bids fair to bring forth a new technology of universal biochemical synthesis. It is not modest.
Kauffman's basic tactic is to admit that the specific manifestations of complex evolving systems are accidental, but to provide grounds for believing that something like them--something orderly, self-regulating, and complex--is statistically likely, even inevitable, given the right initial conditions. A second thrust explores the limitations of natural selection as a mechanism for generating order and fitness, showing how evolving systems can get sidetracked by mutational drift, internal complexity, and myopic adherence to local rather than global optima.
A third theme is the role of coevolution--interactions between adapting populations that affect the fitness of one another--in creating order and stimulating a better fit with the environment. And running through it all is the conjecture that complex systems have a tendency to evolve "to the edge of chaos," where most of the component parts have stable relationships, but there are also areas of instability which allow the system to respond to contingencies in the environment.
The origin-of-life issue is where Kauffman argues, on statistical grounds, that life is not a cosmically improbable accident but rather the most likely consequence of random chemical processes--that "there are compelling reasons to believe that whenever a collection of chemicals contains enough different kinds of molecules, a metabolism will crystallize from the broth." This is a view of life as a process, a pattern found amid the electron dance of chemicals, much as sound is a pattern of motion imposed on the collisions of gas molecules.
Think of a living entity as a set of chemical reactions (powered by a flow of energy from outside sources) that is "orderly" in the sense that the same kinds of chemicals, in roughly the same proportions, are produced over time from an external supply of basic "food" molecules. Let us suppose further that all the chemical reactions in the set must be catalyzed (facilitated by another molecule, a catalyst) if they are to proceed rapidly enough to keep the network going. The problem is where the catalysts come from, and the only possible solution is for them to be the products of some of the reactions in the network. Thus, for a set of interacting chemicals to look like a living entity, each chemical's generating reaction must have a catalyst within the set. Kauffman calls this property "catalytic closure" and refers to sets that have it as "collectively autocatalytic systems."
When put this way, it sounds pretty implausible. What are the chances that you could find a set of chemicals and reactions that just happened to have catalytic closure? Kauffman's big idea is to show that this intuition is wrong, that a) you have enough random catalyzed reactions in a set of chemicals, you get catalytic closure, b) the ratio of reactions to molecule types increases as the number of molecule types increases, and therefore c) if you gather up enough different chemicals, and each has a very small random chance of catalyzing any given reaction, then you are guaranteed to generate a collectively autocatalytic system.
"As the ratio of reactions to chemicals increases, the number of reactions that are catalyzed by the molecules in the system increases. When the number of catalyzed reactions is about equal to the number of chemical [types], a giant catalyzed reaction web forms, and a collectively autocatalytic system snaps into existence. A living metabolism crystallizes. Life emerges as a phase transition."
This is a great story. Is it true? The mathematics are fine, but whether the model is a good representation of chemical reality is outside my competence. A couple of things trouble me. First, if random collections of diverse chemical species, suitably confined and energized, have a propensity for coming to life, how come we haven't seen any in nature? As far as I know, all known self-sustaining chemical reaction networks are conventional living cells, complete with RNA, biological amino acids, and so on. No radically different chemical forms of life seem to exist, although Kauffman's theory seems to say that they could be forming regularly in various cracks and clays and crannies. Perhaps they cannot survive competition with "standard" life and so are destroyed soon after snapping into existence; perhaps we simply haven't been looking for them and so have missed them.
Second, if collective autocatalytic systems are so prone to being born, then why haven't the other planets of the solar system become infested with various forms of life (whose chemistry would differ from ours)? Kauffman argues that life of some kind, not necessarily the particular chemical patterns we see in our biosphere, is sufficiently likely to form that we should consider ourselves not "We the accidental" but rather "We the expected."
Just how expected are we? Perhaps it is unfair to consider it hedging when he says, "I shall not be overly surprised if in the coming decades, some experimental group creates such life anew, snapping into existence in some real chemostat, creating protocells that coevolve with one another...I would not be overly surprised. But I would be thrilled." And the Sunday newspaper supplements would be full of chin-pulling pundits worrying over the "troubling implications" under illustrations with Frankenstein themes. But I'm afraid that until the attempt is made and either succeeds or fails conclusively, the catalytic closure theory will generate plenty of goose bumps, but few thrills.