[Taxacom] Why character-tracking doesn't happen?

Richard Pyle deepreef at bishopmuseum.org
Sat Sep 13 20:28:31 CDT 2008

Just a couple more comments on Neil's post: 

> My point was that the more distantly related organisms are, 
> the *more* that hierarchical rather than reticulate patterns 
> must be relevant to understanding their relationships. 
> Extraordinary events aside, there really is no gene flow 
> between extant oaks and palm trees, and yet these groups have 
> continued to diversify. 

I have no disagreement with this. But we were talking specifically about
"species", which I often see characterized as the only "real" unit of

> Usually the species boundary 
> represents a fairly profound shift between the relative 
> dominance of these patterns, but obviously in some groups 
> this is less so than in others. 

I do have a disagreement with this -- particularly in regards to the use of
the word "usually".  A lot depends on what you mean by "boundary".  At any
given snapshot in time, you will certainly find clusters of organisms that
have much more recent shared ancestry than other clusters; and very often
(perhaps even "usually"), you will find very few or perhaps no individual
organisms that fall in the space between the clusters.  So in that sense,
it's easy to see a boundary.  But in the context of time, if you trace the
ancestors of those two clusters backward, generation after generation, they
will eventually meet at the same cluster. And in this sense, I would argue
that there is NOT usually (I would guess practically never) an unambiguous
boundary (even if you argue from the perspective of vicariance) between
species.  Although we certainly see clustering of individuals at any moment
in time; such a slice through time also intersects many instances of the
"fuzzy" zone.  Sure, if you're doing a random pair wise comparison between
any two random species on Earth (e.g. humans and palm trees), you will find
VASTLY more cases of clear boundaries than you will find fuzzy patches.  But
coming from the other perspective, if you look at any random single species,
I suspect you will find VASTLY more cases that have some sort of population
structure/geographic variant/subspecies/etc. (i.e., fuzziness problem) than
you will find cases of species with unambiguous boundaries on all sides.

> Actually, before I read Richard's comment I was going to use 
> the arguable "unreality" of the individual to counter 
> Thomas's arguments about the unreality of the species! If you 
> accept the reality of the individual, I don't see how you 
> can't accept the reality of at least
> *some* species, unless you make a special case for existence 
> within a single continuous physical space as being a 
> prerequiste for reality. 

I agree that these two realms of "reality" (the "reality" of an individual,
vs. the "reality" of a species) do not differ in kind.  But they DO differ
rather dramatically in degree.  The proportion of organisms for which we
would have arguments about where the individual's boundaries lie
(birth/death in time; cell membranes in space) is vanishingly small compared
to the proportion of organisms for which reasonable people would disagree on
where to draw species boundaries. For practical purposes (which, if you get
right down to it, is what all of these arguments boil down to), it's much
more defensible to acknowledge individuals as "real", than it is to
acknowledge species (or any other taxonomic clustering) as real.

Time now to go play with my son, who has been patiently waiting for dad to
finish up on the computer this rainy Saturday afternoon....


More information about the Taxacom mailing list