Paraphyly=mistakes? (There's the rub)

[Richard Pyle]

> We've been here before, and I've said this before: Which do you
> prefer - a system in which the definitions of names can change when
> the hypotheses of relationships are altered, or a system in which the
> definitions never change, but require creation of new names when the
> hypotheses of relationships are altered?

I prefer the former if I'm concerned with the same things that most
biologists have been concerned with since the time of Linnaeus (with the
caveat of very compelling evidence of a real reason to change the inferred
relationships, and in the absence of compelling reasons to nomenclaturally
recongize paraphyletic lines); and I prefer the latter if I'm concerned with
arguing with my colleagues about how to draw cladograms.  Two sandboxes. Two
sets of needs. Two nomenclatural systems.

> I'll give a hypothetical example: Rhipicerid beetles are presently
> placed in the superfamily Dascilloidea. Suppose someone then
> discovers that they are actually a branch within the family
> Artematopodidae, in the superfamily Elateroidea. Under the present
> system, this is handled simply by redefining what we mean by
> Dascilloidea, Elateroidea, Rhipiceridae, and Artematopodidae. Things
> like this do happen frequently in insects, where taxa jump from one
> superfamily to another. It's a MAJOR shift cladistically, but causes
> very few problems.

...except when you want to map current definitions of those names, to old
definitions of those names (which I concur can be tracked by computers more
efficiently than they are now....but there are various limitiations in the
current state of data models standards, as well as propblems of interpreting
how the definitions actually changed over the course of taxonomic histoy --
but we already had thayt threat recently, so lets' not go there...)

> Under the Phylocode approach, however, the
> definitions and constituencies of those taxa are FIXED, which creates
> a nightmare when branches jump that far. For example, since
> "Dascilloidea" would be defined as "the most recent common ancestor
> of Dascillidae, Karumiidae, and Rhipiceridae" we are faced with a
> substantial dilemma: we must make a new name for the new resulting
> "quasi-Dascilloidea" (now Dascillidae + Karumiidae), and the name
> "Dascilloidea" (if one decides to retain it) jumps up the cladogram
> to occupy a new position, since the most recent common ancestor is
> now a lot farther back than had been originally supposed. [I'll note
> that my recollection of the Phylocode principles is that each name's
> definition explicitly lists all subtending names, and any change
> therein necessitates the creation of a new name, so "Dascilloidea"
> would have to be scrapped - you imply otherwise, so I'll go with your
> concept for the moment]

I have to admit that it's been over a year since I read the draft Phylocode,
and although I am sucribed to the email list, I auto-file all messages to a
folder for later reading -- I'm more than a year behind....

So anyway, I went back and skimed some of it quickly (no time for a thorough
re-read right now), but I did come across the following
(http://www.ohiou.edu/phylocode/art9.html):


*******************
Note 9.4.1: Examples of phylogenetic definitions are node-based, stem-based,
and apomorphy-based definitions. A node-based definition may take the form
"the clade stemming from the most recent common ancestor of A and B" (and C,
D, etc., as needed) or "the least inclusive clade containing A and B" (and
C, D, etc.), where A-D are specifiers (see Art. 11.1). A node-based
definition may be abbreviated as Clade (A+B). A stem-based definition may
take the form "the clade consisting of Y and all organisms that share a more
recent common ancestor with Y than with W" (or V or U, etc., as needed) or
"the most inclusive clade containing Y but not W" (or V or U, etc.). A
stem-based definition may be abbreviated as Clade (Y<--W). An
apomorphy-based definition may take the form "the clade stemming from the
first species to possess character M synapomorphic with that in H." An
apomorphy-based definition may be abbreviated as Clade (M in H). Other
wordings and other kinds of phylogenetic definitions are possible.
*******************

The fundamental point here is that the Phylocode names aren't defined by
their complete contents; but rather, a minimum of two "endpoints".

Now, to be fair, it's a bit more involved than I have made it out to be in
previous posts, and when I keep emphasizing "names applied to nodes", the
Phylocode states it in terms of names applied to "clades".  From this
non-cladist's perspecive, there seems no difference (for what is the single
defining point of a clade, other than it's root node?) Maybe I'll be
brutally corrected for my inappropriate use of the terms "node" vs.
"clade"....but I can accept that.

> Exactly in what way do you see this as an improvement in stability?
> To me, it's six of one, half dozen of the other: there are
> practically the same number of changes necessitated in BOTH systems,
> they're just different *kinds* of changes.

Yes, exactly -- different *kinds* of changes. Those different kinds of
changes have more or less impact on general nomenclatural "stability",
depending on how you actually use those names.  I see the current problem as
different sets of people trying to use the same set of names to represent
different things.  Stability on both sides will increase, as far as I can
tell, if both sets of players retreat to their different sandboxes.  The
important component of "stability" for Linnaean names is the complete set of
lower-rank names included within it.  The important component of "stability"
when discussing clades is the fundamental definition of each name. From my
vantage point, it seems that both sides stand to gain by utilising different
(and non-interfering) sets of nomenclatural tools.

> For my money, I think it's
> easier to use old names and keep track of changes in definitions.
> Otherwise, the frequent occurrence of rearrangements such as the
> above will ultimately and *inevitably* create a "ratchet" effect
> where all the old, familiar names fall by the wayside, one by one, as
> their original definitions are violated and they need to be replaced.

This strikes at the heart of why I think two separate systems are the best
way to resolve this dispute. The Linnaean names can be preserved as long as
needed (maintainin stability), until very compelling evidence is ammassed to
warrant a change.  Researchers who have a need to make textual reference to
nodes (clades) in an hypothesized cladogram can do so with an alternate set
of names, without disrupting everyone else in the biological sciences.

If I did not believe that the end-game for systematics will be a highly
stable, universally accepted cladogram reflecting something very reliably
close to the "true" historical relationships (perhaps when whole-genomes
can, within seconds, be sequenced and placed int context with a massively
complete databank of all other whole-genomes) -- then I can't imagine why a
phylo-nomenclatual system would have much value other than as a temporary
tool for communication among bickering cladists, buffering their "noise"
from influencing the otherwise semi-stable Linanean nomenclature.  However,
anyone who has paid even remote attention to the curve of technological
progress in recent decades must surely be able to see the writing on the
wall. Maybe it will take a hundred years (maybe much less), but the endgame
is all but inevitable. When we arrive, I think the taxonomists of the day
will see a phylogeneticly-based nomenclatural system as the most useful for
communication amongst themselves.

In the mean time, let's sit back and watch what happens when Phylocode (or
whatever comparable alternative) starts going.  Let's hope the practitioners
will exersize some prudence, and remain respectful of the Linnaean names.
Let's hope they test-drive the new system on clades that are very-well
supported, and widely accepted.  Let them decide if it is a superior tool
for their communication needs.  If it is, it will survive; if not, it won't
(or will be modified accordingly). But the important thing is that a greater
sense of stability (in the traditional sense) will be restored to the
tied-and-true Linnaean names.

O.K., enough already....

Aloha,
Rich