How much of a lottery is evolution

How much of the history of life is a matter of chance? In some ways the answer must be "a lot". Alterations of DNA such as point mutations are random, so the generation of new variants in a population is a random process. Natural selection is emphatically not random - it is a very precise process, capable of favouring one variant over an infinitesimally different one - but it depends on environmental factors that are largely unpredictable. All these points are generally agreed among biologists.

However, opinions differ about the range of possible forms of life. Gould argues that the history of life is almost entirely contingent. Rewind the tape of evolutionary history and play it again as many times as you will; it will never be the same twice. If life on Earth began and evolved all over again, there would be no dinosaurs, no fish, possibly no eukaryotes, and certainly no humans or bananas. This is because DNA changes entirely at random, so body plans can change in a virtually unlimited number of ways at every evolutionary step. The overall process is completely unpredictable.

In contrast, Kauffman argues for a finite number of possible body plans; a finite number of solutions to the problem of being alive and reproducing. If the tape of history were rewound and played back, then it could look similar - though not identical - on each re-run. This is because an adaptive complex system, such as life evolving on Earth, produces order as it goes along. Kauffman's position here is consistent with the Gaia hypothesis. He reasons that the "phase space of evolution" (the set of all possible as well as actual living forms, past and present and future) is not fixed; it evolves in response to the actual organisms it encompasses. The self-organising nature of life at the organism/ecosystem level cannot be seen from the molecular biological standpoint, but it is undeniable: it is a mathematical property of complex systems. In practical terms, the "Kauffman mechanism" might work through symbiosis (causing "jumps" to higher levels of organisation), and through co-evolution with a continuously changing environment. According to Kauffman, it results in a much more limited range of possible body plans than Gould would allow, though what is "possible" changes over time.

In Gould's favour one can cite the succession of extraordinary organisms that once lived on the Earth but have vanished with all their kind; no subsequent species has ever resembled them. Evolutionary novelty in body plans is unpredictable. One can also cite the astonishing diversity of species extant today. In Kauffman's favour is the indisputable fact that organisms have become more and more complicated as time has gone on; hardly a random process, but probably the result of successive symbioses. There are also many cases of convergent evolution, such as the close anatomical similarity between mammals that have returned to the sea (whales, for example) and fish.

Biologists argue for a range of positions between the "extremes" represented by Gould and Kauffman. This is as it should be. Too much agreement is a bad thing; debate is the life-blood of science. No serious biologist today doubts the theory of evolution, but the details of how evolution works - what constitutes the environment, what the unit of selection is, how contingent the process might be - remain the subjects of healthy and productive argument, and promise to remain so.

Was this article helpful?

0 0

Post a comment