[Taxacom] PhyloCode & ICZN to "duke it out"?

[Richard Pyle]

Hi Laurent,

My previous note was written in a state of mind having had very little
sleep, so I apologize if it was rambling.

> Although the content associated to a valid Linnean name is 
> indeed not defined, it is NOT free to be extended at will 
> either; it is, in practice, directly limited by the Principle 
> of Priority. A name cannot validly apply to a taxon that 
> includes the type of any older name of the same group.

Yes, agreed -- when the scope of a name expands to include more than one
primary type, the only legitimate name to apply to the concept is the one
attached to the type with highest priority.  The Phylocode (I thought) had
analagous system (called "Precedence" -- Art. 12).  However, because those
names apply to clades, rather than types, there really isn't as much
potential for homonymy or synonymy.  Rather, names will usually (almost
always?) have different clade definitions, so they will both be "valid" in a
sense.  It's just that, without ranks, one name simply gets nested within
another in the sense that one clade is contained within another.  In the
case of Linnaean nomenclature, the same logic actually applies -- it's just
that we don't often think about it that way because we have a finite number
of ranks that names must "slot" into in order to be treated as valid.

For example, consider the two species:

Aus bus
Xus yus

...each of which represents the type species of its respective genus.

Some later worker decides that the type species of "bus" and the type
species of "yus" are congeners, and hence must choose among the two generic
names based on nomenclatural priority as prescribed by the code.  If "Aus"
has priority, than the lumper is not free to arbitrarily use "Xus" as the
genus for both species.  In other words, the Code requires that "Aus" be
regarded as the senior synonym, and "Xus" be regarded as the junior synonym.
In Phylocode terms, each "genus" name (not a genus in Phylocode) would be
defined as a clade, with at least two specifiers.  In the case that both of
these names have exactly the same two specifiers (e.g., the type specimen of
each of the two epithets), then they would be true synonyms.  However, it's
likely that they would not have the same two specifiers, in which case one
of the "genus" names would nest within the other, depending on which clade
was broader (more inclusive), and they would not (necessairly) be thought of
as synonyms.  At least this is my understanding of how it would work (again,
it's been a few years since I read the full Phylocode).

But my point is, the same is really true (or at least possible) within
Linnaean nomenclature.  As I said, a lumper would regard "Aus" as the senior
synonym, and "Xus" as the junioe synonym.  But said lumper might also choose
to recognize subgenera, yielding:

Aus (Aus) bus
Aus (Xus) yus

Thus, although "Xus" is still a junior synonym at the rank of genus, it is
"valid" at the rank of subgenus.  Without ranks, Phylocode names are
essentially all valid.

Two observations on this:

1) Unlike the Linnaean system, Phylocode names can invert positions on the
hierarchy.  E.g., whereas in the Linnaean system, a genus name could never
contain a family name within it; such reversals are entirely possible when
names are defined as clades, rather than by type specimens.  I can certainly
see this as causing confusion in cases where pre-existing Linnaean names
were adopted by Phylocode, then by their cladistic definitions found to be
be inverted in the hierarchy.  I doubt this will happen often, though.

2) Another common criticism of Phylocode is that it will lead to a
proliferation of names because every node could presumably be named.  But in
practice, this won't happen because: a) not every node will be named -- only
those for which a name assists in communicating ideas about inferred
phylogenetic rlationships; and b) many names, though perhaps technically
"valid" in the Phylocode context, will not be used because they have little
or no communicative value.

But I digress....

> If you think taxa should be monophyletic, which as you say 
> most taxonomists do in practice, this results in the Linnean 
> system being very densely filled with implicit phylogenetic 
> constraints, that are like stem-based definitions of the 
> maximum scope of each name. The PhyloCode definitions do not 
> (and probably could not) integrate all these constraints.

Perhaps not -- but thankfully, it's not my problem because I am more
interested in discovering and cataloging Earth's biodiversity than I am in
making inferences about phylogenies. This is a problem the Phylocoders and
the cladisticians and people who suffer from "monophilia" ("monophiliacs"?)
that try to force-fit Linnaean nomenclature on their ideas about
specific/precise evolutionary patterns have to deal with.

Aloha,
Rich