Randomness, Chaos and Prediction in Evolutionary Theory

We often hear that evolutionary theory tells us that the history of life has been directed by the whims of randomness, or even that ‘we are here by chance’. At the same time, evolutionary theorists often construct models of evolution that are taken to be predictive, and geneticists and molecular biologists occasionally offer us extremely accurate predictions of molecular-level evolutionary change. How should we understand this interaction between prediction and randomness? I will explore here one particular kind of prediction – predictions on the basis of quantitative estimates of fitness – in light of both the data that we need to draw those predictions and some recent mathematical work on the impact of chaos on evolutionary models, with the aim of examining what we might still be able to say about the predictability of the future of life in an evolving world.