No subject

Richard Pyle deepreef at BISHOPMUSEUM.ORG
Mon Apr 5 09:40:03 CDT 2004

> I'm completely confident that they will *not* be exactly the same.
> Evolution isn't that simplistic. (And to paraphrase a colleague,
> if working out evolutionary history were easy, it would already be done.)
> >How often do publications dealing with phylogenetic patterns identify
> >which one their evidence pertains to?
> Perhaps not nearly often enough. :-)

O.K., then so far we seem to be in close agreement.

> >Embryonic
> >development and subsequent organismal growth does indeed represent the
> >assemblage of matter -- doesn't it?
> Yes. (But I'm not sure why this is important to the issue at hand.)

Fair enough -- but there is a specific reason why I brought in the seemingly
irrelevant matter/entropy thing.  And that relates to why I didn't see the
forked river analogy as being directly representative of evolutionary
history (addressed in more detail in another post to follow).

> None of the phenomena you mention will give rise to a tree.
> Lineage-splitting is the additional requirement, and I and many others
> associate lineage-splitting with speciation. The evidence is IMO
> overwhelming that speciation is not completely dictated by
> recombination in
> populations (certainly it would be folly to assume that geological
> vicariance events resulted from gene flow).

We pretty-much agree here, so I'm not sure how to respond (I think it's
another one of the miscommunication things).  I wasn't sure which phenomena
that I mentioned, that you are referring to in the first line.  Do you mean
the molecules, genes, genomes, organisms, etc. phenomena?  Or something

Another one of the tricky words is "lineage".  We can all visualize a
lineage, but do we all visualize it the same way?  We seem to agree that a
gene lineage may well be different from a lineage defined as a series of
organismal reproductive events, which may be different still from a lineage
of homologous morphological characters.  So, when we're talking about
lineage-splitting, are we really thinking the same thing?

> That's not the only way to look at it. You could talk about population
> lineages.

But a population is defined by the set of individual organisms that it
contains (across time and generations).  So referring to a population can
only be thought of as a short-hand for referring to a collective set of
individuals.  The individuals are he ones who reproduce; not the population.
In the ideal world, those sets of individuals constituting populations would
be "discrete" (there's that word again), with distinct boundaries -- both in
terms of individuals and the gene sequences they contain.  But "populations"
are just another fuzzy zone on the continuum from "species".

> You could talk about coalescence of gene trees. You could talk
> about it as a hierarchy of apomorphies. My point is that none of those
> things solely (or even perhaps jointly) *determine* evolutionary history;
> they are simply different ways of looking at the pattern.

Wow!  We are DEFINITELY in full agreement here!

> >The danger, though, is that by only using the shorthand
> >version, and forgetting what it really is that we are saying, we risk
> >falling into the trap of thinking of the concepts of 'A', 'B', and 'C' as
> >"real" entities, with discrete boundaries.
> I suggest that the shorthand version is more accurate, because it makes
> fewer assumptions.

Maybe more accurate, but certainly not more precise.  As my wife says, it's
generally better to be vaguely correct, than precisely wrong -- so I see
your point.  But one of the downsides of accurate vagueness is that
different people might try to extract different precision specifics from it.

> And I object to your equating "real" and "discrete".

Sorry! If I did so, it was by mistake.  I think of "discrete" as being able
to define boundaries; whereas real is contrasted with imagined.

> I am convinced that I
> am real, and I suspect the same of you, but I'd be hard-pressed to
> demonstrate that I were discrete.

Well....yes and no.  Yes, I do indeed believe that I am real.  As for being
discrete, I would say that we, as individual organisms, are certainly more
discrete that populations or species are.  We are temporally discrete in the
sense that we have reasonably precise starting points (the union of sperm
and egg) and ending points (death).  We are spatially discrete in the sense
that we are mostly contained within the boundary layer of our epidermis
(plus some hair).  We may not be physically discrete because we're
cycling of matter, and claim "ownership" to less than 10% of the cells
our spatial (epidermal) boundaries.  So I think we can apply the word
"discrete" to ourselves, as long as we do so with

> The requirement of discreteness is IMO
> one of those Aristotelean leftovers that hinders modern science. Certainly
> the particle physicists have gotten over it.

I'm not sure if it hinders modern science, but it certainly seems to be
hindering me! :-)  I semi-jokingly made reference in an earlier post to
carrying on the argument all the way down to superstrings, so I know where
you're coming from on the particle physics thing.  But I think it's still
fair to use the word
"discrete" in the sense of, "individual organisms are more discrete than
populations or taxa".

> >Let me re-state the long version, as well as two more variants of it.
> >
> >1. The pedigrees of any given individual in the scope of organisms
> >represented by the label "A", and of any given individual in the scope of
> >organisms represented by the label "B", share a more recent individual
> >organism in common than the pedigree of any given individual in
> the scope of
> >organisms represented by the label "C".
> I claim this as an *outcome* of lineage-splitting.

Hmm....are you saying that the lineage is the "reality", and the pedigree is
merely an outcome of that real lineage?  I tend to think of it the other way
around -- that the "reality" is that organisms have reproduced generation
after generation, and we are trying to see the evolutionary patterns (what I
think of as linage splitting) that itself is the outcome of this long trail
of reproductive events.

> >2. The genetic sequences of the Cytochrome-B gene of any given
> individual in
> >the scope of organisms represented by the label "A", and of any given
> >individual in the scope of organisms represented by the label "B", have
> >fewer differences from each other than either does with the
> sequence of the
> >Cytochrome-B gene of any given individual in the scope of organisms
> >represented by the label "C".
> This is meaningless as stated, because it makes no provision for
> autapomorphy or saturation, which could decrease sequence similarity
> between closest relatives (however you want to measure it). That's why
> "mere" similarity coefficients don't work for phylogenetic studies of DNA.

Fair enough.  I should have put a little more time and thought into that
one.  How should I have stated it to more accurately reflect current
methodologies in inferring relationships from genetic data?  Maybe something
along the lines of "...apomorphic substitutions and other alterations of the
Cytochrome-B gene..."?

In both cases, my use of the word "apomorphic" implies some cladistic
confidence based on outgroup comparisons -- but I was deliberately trying to
avoid opening that can of worms (though I think it is relevant, I didn't
want to further obfuscate the conversation...)

> >3. The sets of apomorphic characters of any given individual in
> the scope of
> >organisms represented by the label "A", and of any given
> individual in the
> >scope of organisms represented by the label "B", are more congruent with
> >each other than either is with the set of apomorphic characters
> of any given
> >individual in the scope of organisms represented by the label "C".
> Um, I think you are using "congruent" in a sense that I don't understand.
> Do you mean that "A" and "B" share more apomorphies?

Yes -- sorry. Re-reading my statement #3 above, I see that I garbled it
considerably.  Suppose we have a set of a dozen characters.  Suppose that
we've confidently polarized the alternative states of each as apo- or
pleisiomorphic.  Suppose that "A" and B" share 9 of the apomorphic states,
"B" and "C" share only 2, and "A" and "C" share only one.  Taken as a set of
12 characters, there is more congruency between "A" and "B" among
apomorophic states, than there is between any other pair-wise comparison.

> >My question is, do all three of these statements equally
> collapse into the
> >simplified statement, "Species 'A' and species 'B' are more
> closely related
> >to each other than either is to species 'C'"?
> I'd say no, but because of technicalities.

What about with my revisions above?

> My ramblings are based on the idea that we don't know that for
> certain, but
> that we have well-established theories of the ways that "traits" (genetic
> or morphological) can be transferred longitudinally or laterally. One
> approach to non-congruence is to throw up our hands and say "It's too
> complicated, so I'm going to stuck with my 'expert taxonomic opinion'."
> Another is to postulate unknown forces at work. I maintain that
> the best is
> to analyze the data in the context of accepted mechanisms and see whether
> we can make any sense of it.

A agree -- but how is the last sentence ("...make any sense of it") any
different from "expert taxonomic opinion"?

> >But my contention is that we really do not yet understand enough
> about how
> >to extract phylogenetic information from genetic data; and that we do not
> >yet understand how to reliably identify apomorphies, pleisiomorphies and
> >homologies from among morphological characters; and -- more
> fundamentally --
> >that we haven't even got our collective heads around what we mean by the
> >notion of a phylogenetic history.
> And this is different (in the context of the history of science) how?

Well....I think we have our heads around the periodic table pretty well.  Of
course, in evolution, we are talking about histories (things that have
happened in the past), and there is NEVER any certainty in that.  But there
are varying degrees of certainty.  And I guess when you cut through all the
verbiage in my long posts, my ultimate point is that the level of certainty
associated with most arguments about phylogenetic history is considerably
less than that expressed by many of the participants in such debates.

> I'm certainly not going to brand you a fool; yours are among the
> best-thought-out posts on the list.

Ah!!  We most definitely disagree on that last point!! :-)


More information about the Taxacom mailing list