Re: A critique of the BBC aquatic ape programme and the transcript.



Pauline M Ross wrote:
> On 26 May 2005 10:31:46 -0700, "JAE" <jae@xxxxxxxxxxx> wrote:

[snip]
> >In essence, you're putting selection, a measurable force, and making it
> >the null hypothesis. This isn't appropriate.
>
> Yes, but *if* there is an measurable difference between species or a
> change over time within a species, that has to have happened because
> of something (selection, drift, whatever); I'm just proposing that *in
> this situation* that natural selection is the default mechanism.

You are thus letting "selection happened" be your null hypothesis and
this is splendidly inappropriate if you are interested in if and how
selection operated. The a priori expectations of what is most likely
does not make one a default or another.

Whether or not you consider it more or less likely to be responsible,
drift is a more appropriate null. It is much more a "rule of biology"
than any notion that that specific adaptation will occur. In anything
less than an infinite population, simple sampling error in meiosis will
lead to stochastic changes in allele frequencies. This is unavoidable,
a mathematical truism. It is *mitigated* by things like the strength
of a counter selection and a larger population size, but it happens
regardless. EVEN IF you consider that any particular trait or suite of
traits is more likely an adaptation, drift is a better default because,
perhaps definitionally, it has to happen. There is no similar
constraint on selection for any particular genetic system and thus it
is what needs to be positively confirmed. This has nothing to do with
odds and everything to do with creating an appropriate hypothesis.



[snip]

.

Relevant Pages