schultz at unixg.ubc.ca
Sat Nov 2 17:32:29 EST 1991
[When I entered this thread in midstream, I mistook what appears to
be a semantic debate for a biological debate. I'll leave the more
biological points for another post.]
Rather than further muddying the waters with requotes, I'll briefly
summarize the semantic disagreement as I see it, then add my
I see no controversy here: an allele is any of the different forms of a
gene occupying the same locus on homologous chromosomes; hence
alleles undergo meiotic pairing and can mutate to one another. I
know of no biologist who defines the term in any substantively
Here's a common definition of evolution: "Evolution is any change in
the genetic composition of a population."
Arlin Stoltzfus apparently objects to this definition on the grounds
that it fails to "include," "circumscribe," or "refer to" the process of
isolation, which is an integral component of speciation and hence
evolution (although he has yet to explicitly state his own definition).
This criticism is unfounded because it confounds process and
mechanism, and hence confuses cause and effect.
By the above definition, evolution can occur (for example) by
mutation, drift, selection, or migration; these are different
mechanisms or causes of a population's changing genetic composition.
To say that a population is isolated is to make a claim about
migration: the rate of genetic influx or immigration to the population
must be near zero. In other words, isolation can occur only through
the agency of one of the mechanisms or causes of evolution. If the
above definition is incomplete because it fails to explicitly include
mention of potential low migration rates, then it is also incomplete
for not explicitly mentioning mutation, selection, or drift (or any
other mechanism), which are all at least as integral to the
evolutionary process as migration. If this sort of criticism were valid
in general, it would seem to imply that any definition of any process
would be incomplete unless it enumerated the entire suite of causes
or mechanisms of the process, which is absurd to say the least.
Although the above definition excludes isolation as sufficient for
evolution, it does nothing to preclude isolation or any other
phenomenon as a cause or effect of evolution. Let's say a railroad is
built between two herds of bison, completely blocking gene flow that
previously was frequent. Reproductive isolation, but no evolution,
has occurred, even if subsequent evolution is affected enough that
two different bison varieties eventually arise. In fact, if the simple
erection of a migration barrier in and of itself constitutes evolution,
then evolution can occur without any change whatsoever in members
of a population, a claim that is again absurd.
Similarly, the appearance of a physical barrier to migration hardly
constitutes speciation (if speciation is by definition a kind of
evolution). The barrier, however, may allow accumulation of genetic
factors that cause sterility or inviability in hybrids formed when
the barrier is later lifted. In this case evolution has certainly
occurred as a _result_ of the isolation, and perhaps speciation as
I see nothing wanting in the above definition in any biological
context; in fact it's a standard definition in wide usage, and certainly
does nothing to contradict any biological truths about the importance
of reproductive isolation in evolution and speciation.
University of British Columbia
schultz at unixg.ubc.ca
More information about the Mol-evol