Is evolution predictable?

See allHide authors and affiliations

Science  16 Feb 2018:
Vol. 359, Issue 6377, pp. 738-739
DOI: 10.1126/science.aas9043

You are currently viewing the summary.

View Full Text

Log in to view the full text

Log in through your institution

Log in through your institution


Forty years ago, Anderson (1, 2) and Poole (3, 4) debated the value of predictive models for understanding population dynamics in a temporally varying environment. Poole argued that he could predict the future with high confidence if he had a good record of the past. Anderson (1) expressed doubts. He pointed out that long-term predictions will hold only if past conditions persist into the future. He also argued that although sophisticated statistical models may be able to predict the future, they do not tell us why they succeed or, more pointedly, why they fail. Without deep biological understanding of the system under study, predictive models are not likely to offer much insight into either the past or future. On page 765 of this issue, Nosil et al. (5) apply the same scrutiny to evolution.