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

Re: STEM V. NODE BASED CLADES



Peter Buchholz wrote:
>I believe (someone correct me if I'm wrong) that a crown clade is defined as
>"all the descendants of all the living members of a clade."  For example, the
crown clade Mammalia is defined as {Monotremes + Marsupials + Placentals};
        Note that this does not *Explicitely* define the group as a crown
clade. If some of the animals currently alive turn out not to be decendants
of the common ancestor defined here, te clade is no longer a crown clade,
it's just another node based clade with no particular significance.

>It is basically the way the are defined, but there is a difference.  For
>example, Dinosauria is a node based clade {_Triceratops_ + _Corvus_}, so some
        I applaud Peter's insistance on avoiding suprageneric (above the
generic level) taxa in taxon definitions. It should be noted, however, that
the definition he cites here does *not* necessarily equal the definition we
have come to know and be uncomfortable with around here,  Dinosauria = { +
_Triceratops_, + Aves }.

Office <dinosaur@interport.net> wrote
>Below is a good example of
>what I mean about "personal" things that inadvertantly confuse the
>reader. Clearly, "Tetanurae" (P. Bucholz) is a knowledgeable individual,
>but until the subjective shorthand was explained, the previous material
>was less comprehensible than it might have been otherwise.
        Which brings up the point that I have *not* recently gone into how
my own shorthand works. I don't like to feel like I'm repeating myself, my
apologies to anyone who has been confused. It's a simple concept, but
somewhat more involved than Peter's. Note that this is my terminology, and
should be used with caution.
        Every (well, pretty much every) clade definition has "Anchor Taxa"
which delimit it's membership through aspects of their common ancestry.
There are two type of anchor taxa (that we are concerned with now):
inclusive anchors, which are members of the taxon being defined, and are
designated with a (+), and exclusive anchors, which are taxa explicitly
outside of the group being defines, and are designated by a (-). Curly
Brackets {} invoke the concept of common ancestry and decent, and a three
bar equals sign (unreproducable in ASCII?) denotes definition (as opposed to
synonomy, a regular equals sign).
        What it boils down to is:
        Node based taxon = { + A, + B }
                Read: "The most recent common ancestral population of A and
B, and all of its decendants."
        Stem based taxon = { + A, - B }
                Read: "All animals more closely related to A than to B."

        I have considered Peter's system, and although I like it, I find it
wanting in several respects:
[I don't mean to flame or berate Peter in any way, I think some aspects of
his system are ingenious. I do not think that folks should adopt it, owever,
for the following reasons.]
        1) While it uses curly brackets, the general form of Peter's system
bears more of a resemblence to several ad hoc (and confusing) attempts to
derive a shorhand from other sources (eg. Novas 1997), and to the notations
which spawned them (eg. Newick Notation for tree topology). The most
visually distinctive style is the one less likely to inspire confusion
amongst shorthand methods.
        Example: compare (_Allosaurus_ + _Sinraptor_), "_Allosaurus_ and
_Sinraptor_" denoting that the two are more closely related to each other
than to any other taxa *in the study*, a modification of Newick Notation
commonly used to specify a group in a particular tree topology, to:
        Allosauroidea = { + _Allosaurus_, + _Sinraptor_ }
        Allosauroidea  { _Allosaurus_ + _Sinraptor_ }
        Which of the above is less likely to be confused with the modified
Newick?
        2) It does not specifically distinguish between types of anchor
taxa. This is not vital to a shorthand, but promotes clarity of
understanding. Peter's system simply invokes taxa, without specifying the
relationship of that taxon to the group, and then adds a functional symbol.
My system specifies the role of the taxon included in the definition, and
omits the functional symbol per se, as this aspect is covered in the use of
{} to denote common ancestry, and may potentially be confused with other
notations.
        3) While it may be minimally sufficient to use <> symbols (and I
think it's quite clever, really), this does not allow for the clear and
convenient incorporation of multiple exclusive anchor taxa. "( A > B > C )"
is not really very easily deciphered.
        4) While Peter's system is certainly less complex than mine, mine
has a certain flexibility resulting from the fact that it is derived from
what I hope is a reduction to the essentials of taxon definition in PT,
rather than merely a quick way to abbreviate a definition.
        5) As a result of the above, my system has an advantage in terms of
other forms of clade definition (eg. apomorphy based) which, although
perhaps undesirable, are currently accepted ways of defining clades. I did
not reproduce this portion of the system above, as it is largely irrelevant,
and still in dev.
        6) One of the hidden advantages of a slightly more complex system,
especially an explicit one, is that it forces the definor and the reader to
address fundamental issues in the definition process: what am I naming, how
am I naming it, what will be the ramifications of further changes in tree
topology on the taxon concept I am addressing, will my taxon be lost
inadvertantly due to a fundamental error in how I chose to define it, etc.
----------------------------------------------------------------------------
      Jonathan R. Wagner, Dept. of Geosciences, TTU, Lubbock TX 79409
            Web Page:  http://faraday.clas.virginia.edu/~jrw6f