Consider the following three claims that could be made about the mean value of a particular trait in a biological population:

1. Natural selection is the most important evolutionary force acting on the trait, and there is abundant genetic variation for that trait in the population.

2. The trait value is adaptive, in the sense of resulting in higher fitness than many other potential values of that trait.

3. The trait value is optimal, in the sense that it results in fitness that is greater than or equal to the fitness of any other value of that trait within some specified range.

Each of these claims has played a major role in debates regarding the truth and usefulness of adaptationist and optimality approaches to understanding the current characteristics of living organisms, and they underlie much of the discussion that follows. Most evolutionary biologists are interested in understanding the evolutionary forces that produce the traits observed in present-day organisms and in the fossil record. The three preceding propositions are important because they bear on the validity of approaches that have commonly been used to try to discover these underlying evolutionary forces.

This chapter is divided into two sections. In the first section, I address the common belief that the predominance of natural selection combined with the continued presence of abundant genetic variation will imply adaptation or optimality (i.e., proposition 1 will imply 2 and/or 3). I argue that, in many situations involving frequency-dependent fitnesses and/or environmental variability, neither implication is justified.