History [of biological complexity]

The origins of biological complexity have been debated since antiquity. For a long time it was assumed that the magnitude of the complexity was so great that it could never have arisen from any ordinary natural process, and therefore must have been inserted from outside through some kind of divine plan. However, with the publication of Charles Darwin's Origin of Species in 1859 it became clear that there were natural processes that could in fact shape features of biological organisms. There was no specific argument for why natural selection should lead to the development of complexity, although Darwin appears to have believed that this would emerge somewhat like a principle in physics. In the century or so after the publication of Origin of Species many detailed aspects of natural selection were elucidated, but the increasing use of traditional mathematical methods largely precluded serious analysis of complexity. Continuing controversy about contradictions with religious accounts of creation caused most scientists to be adamant in assuming that every aspect of biological systems must be shaped purely by natural selection. And by the 1980s natural selection had become firmly enshrined as a force of practically unbounded power, assumed—though without specific evidence—to be capable of solving almost any problem and producing almost any degree of complexity.

My own work on cellular automata in the early 1980s showed that great complexity could be generated just from simple programs, without any process like natural selection. But although I and others believed that my results should be relevant to biological systems there was still a pervasive belief that the level of complexity seen in biology must somehow be uniquely associated with natural selection. In the late 1980s the study of artificial life caused several detailed computer simulations of natural selection to be done, and these simulations reproduced various known features of biological evolution. But from looking at such simulations, as well as from my own experiments done from 1980 onwards, I increasingly came to believe that almost any complexity being generated had its origin in phenomena similar to those I had seen in cellular automata—and had essentially nothing to do with natural selection.