[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]


        As per the request of some, we should continue this off-list. First,
I wanted to clear up one or two things publicly...

At 02:02 PM 4/1/97 -0500, Peter Buckholz wrote:
>I think that most people stick to genera (or should) nowadays.
        Yeah, but Gauthier went through and defined most of the important
stuff already...

>I'm not sure what you mean by this...  Could you clarify what you mean by
>'different types of anchor taxa'?
        There are two types of "anchor taxa" which are germaine to the
discussion here (there are one or two more, and I don't like them very
much). Again, this is my terminology, and here are some better definitions:

ANCHOR TAXA: Taxa used to define a phylogenetic taxon. A phylogenetic taxon
is defined by relationships amongst anchor taxa, which are either explicitly
named, or for which diagnostic attributes are given.
INCLUSIVE ANCHOR TAXA: An anchor taxon which is a member of the taxon being
defined. Some ancestors (and all of their decendants) of an inclusive anchor
taxon will also be members of the taxon being defined.
EXCLUSIVE ANCHOR TAXA: An anchor taxon which is not a member of the taxon
being defined. Taxa more closley related to an exclusive anchor taxon than
to any inclusive anchor taxon are not members of the taxon being defined. No
ancestor of an exclusive anchor taxon is a member of the group being defined.

        Thus, in a node-based clade, inclusive anchor taxa (+) are those
whose most recent common ancestor with all other inclusive anchors is the
common ancestor of a node based taxon being defined. With a stem based
clade, allk members are animals which are more closely related to the
inclusive anchor taxon than to any exclusive anchor taxa.

>I was however thinking
>that the slash symble "/" could be used to denote multiple taxa, like: { A >
>B / C };
        Mickey notes that this use of "/" is inconsistant with mathematics.
        I STRONGLY resist this sort of complication, as it further blurs the
distinctions I am trying so hard to point out. The implication of "or",
specifically interchangeability, has no place in the context of phylogenetic
taxonomy. We are trying to explicitly note the groups whose
interrelationships define our taxon (ie. anchor taxa). We want to precisely
codify the criteria for membership, and ambiguous statements like "or" just
don't cut it (see below). I firmly believe that by specifying each anchor
taxon and it's relation to the taxon being defined, you can provide a
clearer, more explicit (perhaps a bit more complicated, but simple != good)
framework to work in.
>{_Corvus_/_Rhea_/_Archaeopteryx_ > _Ornithomimus_/_Troodon_/_Tyrannosaurus}
        This is an excellent example of why I prefer to explicitly name
anchor taxa and note their type (inclusive, exclusive, etc...).
Superficially, this looks fine, we're just reading "all critters more
closely related to crows, rheas or archy than to _Ornithomimus_, _Troodon_
or _Tyrannosaurus.
        But, does this require that the first three form a clade and so do
the second three? Does it include any animal more closely related to *any
one* one of the first three than to *any one* of the second three? Do the
second three form a clade within the overall common ancestry, and the
"birds" differing branches leading to that clade, essentially defining a
parapyletic group? If my animal is more closely related to the "birds" than
is _Troodon_, but not more closely related to the "birds" than is
_Tyrannosaurus_ is it a member of the group (it meets one of the criteria,
does it have to meet the other two?). The "or" just doesn't really seem to
carry a specific meaning. Monophyly is about common ancestry and common
decent, which are concepts which do not allow for alternatives ("or"), only
exclusion of a sister group and common decent from a common ancestor.

>"All those animals that share
>a more recent common ancestor with _Corvus_ or _Rhea_ or _Archaeopteryx_ than
>with _Ornithomimus_ or _Troodon_ or _Tyrannosaurus_."

        This doesn't even sound like a necessarily monophyletic taxon...
which brings me to another point. In sitting around playing with my system,
I discovered something very simple. You can't have more than one inclusive
anchor in a stem based taxon without risking paraphyly. Work it out, you can
demonstrate this to yourself.
        Another cute bit you discover rapidly is that phylogeny is
commutative. You can reduce a suprageneric taxon to its definition and
substitute it into the definition of a more inclusive taxon:
        Taxon A = { + Q, + R }
        Taxon B = { + A, - S, - T } ergo:
        Taxon B = { + Q, + R, - S, - T }
        Ergo, Taxon B is potentially paraphyletic.
        Would I have figured this out without using the system I use now?
Sure. Would Peter's system have made this easier to understand or explain
(even with an "or")? Unlikely.

>I have no idea how I would write an apomorphy based clade in shorthand (and
>even less of an idea of why one would want to...), perhaps something like {+
        Yes, but the relationship between anchor taxa and the apomorphy is
very different in a traditionally defined apomorphy based clade.

>Anyways, no system is set in stone... 
        I don't mean to be picky here, I do think your system is pretty
cool, but I strongly believe that any such system needs a firm conceptual
framework. I don't see any modification of your system that could produce
such a result, certianly no modification that won't end up just being
basically like mine. Not trying to be pompous here, but I just don't think
that making up a format for each new case beats a generalized ("modular")
approach. :)
      Jonathan R. Wagner, Dept. of Geosciences, TTU, Lubbock TX 79409
            Web Page:  http://faraday.clas.virginia.edu/~jrw6f