Paraphyly=mistakes? (There's the rub)

[Richard Pyle]

> So the previous argument for a "phylocode"-like scheme, which was
> "The need to accurately communicate phylogenetic hypotheses via
> nomenclature (which really boils down to the need to communicate using
> text, rather than requiring graphic images of
> cladograms)", is now abandoned and a new one is taking its place

Sorry -- Nope, no abandonment.  Just my poor communication skills at work.
What I meant was, we need the ability to *define* text strings that strictly
represent phylogenetic nodes.  Cladograms work well, but it can be
cumbersome to always have to refer to graphic figures when discussing
hypothesized phylogenies.  By unambiguously defining text strings to
represent nodes, those text strings can be used in place of graphic
cladograms when endeavoring to communicate phylogenetic hypotheses.
Unfortunately, Linnaean names are suboptimally suited to be *defined* as
nodes, both because of the inherent association with ranks, and also because
they are ultimately defined by (anchored to) a single point (i.e., the
Holotype of the type species, etc...)

...which leads to:

> Phylocode names are
> anchored at two places in the scheme of things, which somehow is supposed
> to enhance their stability.

It enhances the stability of their definitions, not necessarily implied
their contents.

> How that would be is unclear - obviously, such
> names are likely to be more sensitive to changes in this scheme than names
> that are connected only at one point to the scheme, picking up changes at
> two places.

No, not really.  The "Definition" of a Phylocode name remains constant:  it
is, and always will be, the most recent common ancestor shared by at least
two different species (or specimens, or nodes, depending on how the name was
originally defined).  In other words, the "definition" remains constant over
time (even though the complete scope of its contents may change over time as
phylogenetic interpretations change). When phylogenies are eventually
solidified, the names (by their definitions) also solidify.

In contrast, Linnaean names are anchored to only a single point (the
holotype of the type species, etc...).  Thus, they can slide up and down the
phylogenetic lineage, superceding or resurrecting other names of like-ranked
taxa, at the whim of any splitter or lumper looking to add a publication to
his or her CV -- all without a single branch change in the hypothesized
phylogeny. Their true "definitions" (taxonomic scope) are subjective, even
with a "known" phylogeny.

That is the key distinction between a name defined by a single point, versus
a name defined by more than one point. The exact position of Phylocode names
may move about as different phylogenetic hypotheses are asserted, but they
will always represent a "real" node -- independently of our ability to
accurately map where that node fits in the larger evolutionary context.
Single-point Linnaean names, on the other hand, aren't defined in any
evolutionary context, and thus are subject to change both with changes in
the intrpreted phylogeny, *and* in the absense of phylogenetic
rearrangements.

I hope that clears up the point I was trying to make, but I fear I've only
made things more confusing....

Aloha,
Rich