[Fwd: Re: Probabilities on Phylogenetic Trees]

Doug Yanega dyanega at MONO.ICB.UFMG.BR
Thu Sep 18 17:28:58 CDT 1997


Another brief (relatively speaking) comment from the back of the classroom -

James L-W wrote:

>> So tell me, on a conceptual level, how do you imagine quantifying
>> something like the expected congruence of a set of morphological
>> characters? The image of them showing distribution patterns caused
>> "by chance" is a bit too abstract for my li'l head.
>
>        On a conceptual level?  OK - conceive of this..
>        imagine that a majority of your hypotheses of homology are
>        wrong.  You will find congruence, nevertheless,
>        by chance alone.

If you've included 500 "informative" bases in your analysis, each is a
separate hypothesis of homology, right? What proportion of those hypotheses
do you think are likely to be correct, given the nature of the "screening",
as compared to morphological data? I'm not saying we *know* one way or the
other (there is certainly some variability in the homoplasy levels of both
types of data), but it's not quite fair to throw stones at non-sequence
characters when the criticism is just as applicable to sequences, if not
more so.
        I *do* see your point here (I think), and appreciate your lack of
comfort, but my fear of homoplasy is not quite equal for both types of
data, even if only because I think one has many perfectly valid back-ups to
help spot and explain true homoplasy in morphological data (e.g., an
example dear to my heart, that all parasitic bees lose their
pollen-collecting apparatus, and about 5 other inevitable character
transformations associated with cleptoparasitism). You are worried about
"ad-hocness", and I'm not ashamed to admit I have less trouble with an ad
hoc statement like "these features arise on the tree in several places
because the taxa involved are all parasitic" as opposed to "about half of
the base transformations appear several places on the tree, and I have no
evolutionary explanation for any of these evident homoplasies". One is an
ad-hoc explanation with some teeth, at least.

>        I have NEVER heard Joe F. or Edwards, for that matter
>        state publicly that morphological data are not potentially
>        informative, or that sequences are the only way to go.

No, but I actually do think Joe once admitted (on sci.bio.systematics) that
he saw no easy way to assign probabilities (e.g. of transformation) to
morphological characters, and therefore they were not likely to ever be
included in statistical phylogenetic analyses. Saying you can't use the
data *is* essentially the same thing as saying the data is not potentially
informative.

>        On the notion of ad-hocness, here's a quandry.  Why,
>        if the max pars tree results in the fewest ad-hoc
>        hypotheses, are the jounral pages FILLED with crap
>        about why the researcher didn't get the tree they
>        expected?

Funny, I don't see much of this when I read papers on insect phylogeny (and
when I do, it comes with an explanation that is NOT "crap", as above). I'm
tempted to suggest that maybe some groups of organisms have a longer and
more confounding taxonomic history, and that this extra burden of
well-entrenched past classification is hard to revise without having to be
overly apologetic ("Gee, I can't understand why Reptilia is paraphyletic in
my analysis...").

Sincerely,

Doug Yanega    Depto. de Biologia Geral, Instituto de Ciencias Biologicas,
Univ. Fed. de Minas Gerais, Cx.P. 486, 30.161-970 Belo Horizonte, MG   BRAZIL
phone: 031-448-1223, fax: 031-44-5481  (from U.S., prefix 011-55)
                  http://www.icb.ufmg.br/~dyanega/
  "There are some enterprises in which a careful disorderliness
        is the true method" - Herman Melville, Moby Dick, Chap. 82




More information about the Taxacom mailing list