Subject: Re: Bird splits
Date: Feb 2 16:38:16 1996
From: David Wright - dwright at u.washington.edu


On Fri, 2 Feb 1996, Christopher Hill wrote:
> The convenience of standardization should not be underemphasized. Perhaps
> having a standard published list of "accepted" names is not critical for
> recreational types (I am happy to put any name on a dragonfly, I don't
> care if it's 30 years out of date!). Some people are obligated to keep up
> with current taxonomy, though, including, for example, anyone who wants to
> publish in formal scientific journals. As Dennis points out, not even the
> editors of bird journals are always up to date, and they have the
> "official" taxonomy readily compiled and available.

The point here is that having an "official" taxonomy for organisms whose
species-taxa status is in contention serves *convenience*, at the
expense of obscuring the true state of our knowledge (and recreational
types such as listers care at least as much as biologists whether a given
populations is pigeonholed as a "species" or a "subspecies"). Having a
committee decree which species-taxa names are "best" does not serve the
science of systematics, even if it does streamline things for people who
are not working in systematics or evolutionary history (e.g., listers,
many ecologists, many behaviorists, most medical types, etc.). Suppose
someone suggested creating a commitee to decide which of several alternative
hypotheses to explain bird song, or plumage colors, or a particular display.
The suggestion would be laughed out of the room as ridiculous (I hope).
Why is it less ridiculous for hypotheses in systematics? As behaviorists,
ecologists, etc., adopt phylogenetic methods to study evolutionary histories
of the traits they are interested in, systematics will become to them more
than just a system of naming. Right now, I think it is fair to say that
the majority of non-systematist ornithologists are only using the *names*,
not the *systematics* -- and it is *convenient* to have a committee to
decide which name to use. (Not to single out ornithologists -- the same
is true for mammalogists, herpetologists, etc.)

The obvious problem here is that we are using names both to identify
populations, and to express the phylogenetic/evolutionary relationships
of these populations to other populations. Because we use a binomial
nomenclature (genus + species), the name of a species is supposed to
reveal which other species are its closest relatives (i.e, members of the
same "genus"). Adding "subspecies" to the system adds another layer
of confusion, of course, by making the sytem trinomial. If we simply
used one name for each population, that name would remain unchanged even
if our ideas of its relationship to other populations changed. With the
current system, changes in either 1) relationships of a "species" to other
"species," which can result in a change in the "genus" part of a species'
name; or 2) changes in assessment of a population's status as a "species"
in its own right, or part of a "species" -- a "subspecies" -- which changes
the "species" part of its name (for example, "Icterus galbula bullocki" vs.
"Icterus bullocki")

Consider the "tree of life" metaphor for phylogenetic descent. Focus
your mind's eye on the end of one of the branches. It is made of
branches that bifurcate again and again, originally ending in little
branch-tips. The branch tips are the populations we are concerned with.
What are they? Under the Biological Species Concept (BSC), these
branch tips may be "species," or they may be parts of species --
"subspecies," depending on the degree of interbreeding between the
branches in question. You might wonder, "Why not just call each
branch tip a 'species' and avoid the ambiguity?" This is the
Phylogenetic Species Concept (PSC), and many cladists, who take a
"top-down" view of biodiversity advocate this as the simplest and most
objective way to do it. But an advocate of the BSC would counter that
some of the branch tips that appear separate today but interbreed might
merge with other in the future, if interbreeding becomes frequent enough
for the popluations to merge; to a BSCer, populations aren't "species"
until they have lost the capability of merging with other populations.
The PSCer would reply that the BSCer really doesn't have a clue which
populations will merge with one another in the future; why not just make
it as simple as possible and call distinguishable populations "species"?
Who is right? It depends on what you want a species to be.

OK, back to the branch on the tree of life. It is real, and
exists whether we recognize it or not (that is, the historical
pattern of descent is real). But the categorical names we apply to
the branches -- "class," "order," "family," "genus," and yes, even
"species" are not real: which branch gets which label is arbitrary. It
is true that the species can be labeled more objectively, but there is
more than one objective set of criteria on which to base this labeling,
and deciding among them is arbitrary. It is possible to recognize the
branches on the tree -- "clades" -- without assigning them to one of the
ranked categories (family, genus, species, etc.).