Simpson on randomness

Quote for today:

In the present synthesis adaptation, preadaptation (in a broadened sense), and non-adaptation all all involved, and all can be assigned immediate, if not ultimate, causes. That all basically depend upon mutation is partly a matter of definition. That the mutations are spontaneous and random, at least in the special sense elsewhere defined, is a conclusion warranted and, with some restrictions, demaned by the experimental data. Calling them "spontaneous" and "random" simply means that they are not orderly in origin according to the demands of any one of the discarded theories. The incidence in time and the individuals affected seem to be random or nearly so in the same sense---that they do not agree with hypotheses that assume a more specific incidence. That the direction of mutations is entirely random is certainly not true; but neither is it true that mutations regularly occur in one direction only. Given a certain hereditary type of development pattern, the changes that can occur in it and their effects upon the structures developed are strictly limited, and alternative changes are not introduced in exactly equal numbers; but in almost every case the change can be and is in at least two, frequently more, different directions.

George Gaylord Simpson, The Major Features of Evolution, New York: Columbia University Press, 1953, p135f.