330 Margaret Morrison
with another dimension measuring degree of fitness, then we can think of the
field as comprising several peaks or humps relative to the latter. These humps
are the result of factors that influence, either positively or negatively, fitness.
For Wright the problem of adaptive evolution concerned the mechanism whereby
species could find their way from lower to higher peaks in the adaptive landscape
under the influence of selection. If the maximum peak a species reached was
not the absolute maximum the only way to enhance fitness would be through
something other than the steady pressure of selection. Wright suggested four
factors that would contribute to this progress; the changing environment would
continually change the peaks; new mutations would add further dimensions to
the field thereby allowing new possibilities for advancement; random genetic drift
would produce stochastic jumps from one peak to another as would the division
of a large population into smaller ones.
We can begin to see here just how significant the differences were between
Fisher and Wright. For Fisher there is only one way to think about selection
— by taking account of the additive effects of genes, with its power limited only
by the availability of new variation in the form of mutation. Although both of
these are analysed statistically Fisher’s overall view of selection is deterministic
in the sense that only it and mutation are responsible for evolutionary change.
Wright, on the other hand, introduced drift (sampling error) and migration as
factors that also affect gene frequencies. He claims that there “must be some
trial and error mechanism on a grand scale by which the species may explore the
region surrounding the small portion of the field which it occupies. To evolve,
the species must not be under the strict control of natural selection” [1932, 359].
Wright’s emphasis on interactive genetic systems and the fact that selection is
altered by random variations points unequivocally to a causal understanding of
the evolutionary process; an understanding that was represented by his use of
path analysis in evaluating information relevant for evolving populations.
Although Fisher’s work on dominance indicated his awareness of the importance
of interactive genetic systems, his interest was in the long term where he thought
the effects of single genes would prove more important. So, except in cases where
we are interested in the probability of the survival of an individual new mutant,
stochastic effects are simply not significant; natural selection is considered the sole
agency in genetic evolution. He emphasised the main effects of single loci rather
than complexes of loci and the steady and essentially deterministic increase in the
frequency of each allele having a selective advantage. This is not to suggest that
selection is deterministic in the traditional sense but only that the action of natural
selection is not affected by the presence of random factors. However, if we have
small populations of the kind that Wright favoured then each tends to become
more or less homozygous. But, because of migration there is an infusion of genes
into different populations we get another level of selection that is interpopulational.
In that sense there are two levels of selection depending on the structure of the
populations with the result that selection can operate on either individual genes
or gene complexes. For Fisher, of course, selection at the level of gene complexes