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


        You'll all be glad to know I'll be out for a few days. Maybe the
volume of phylogenetic systematics-related postings will drop. We can only

Wed, 12 Aug 1998 16:18:49 PDT Matthew Troutman wrote (quoting me)
><<[...]If Aves == {Ratitae + Tinamoui + 
>Neognathae}, in which case _Archaeopteryx_ isn't even an avian, many 
>many birds are closer to the common ancestor than is _Archaeopteryx_.>>

>I have one problem with the latter 
>definition; what about the Galliformes?  Peter Houde (1988) has 
>published evidence that galliforms and tinamous are sister groups.  Does 
>this mean that galliforms belong in Tinamoui or Neognathae?
        I'll tackle the second question first: It depends on the definition
of Tinamoui and Neognathae, definitions I do not know. Personally, I hate it
when supergeneric taxa are used as anchors for phylogenetic definitions,
partially because of this sort of confusion.
        As for your first question, work it out: does this mean that
galliformes are descendants of the common ancestor of ratites, tinamous and
        Answer: yes, they are. "Sister group" implies that the two share a
more recent common ancestor with each other than with any other taxa.
Galliformes would share a more recent common ancestor with tinamous than
with ratites and neognaths. The MRCA of tinamous and galliforms is therefore
a descendant of the MRCA of ratites, tinamous and neognaths. Draw it as a
Venn diagram:
        ((tinamous, galliformes), ratites, neognaths)
        As long as galliformes are descendants of the MRCA of the three taxa
mentioned, they are included within Aves.

>We MAY have to change (yeah, right) our definition of Aves soon, though (now
>I'm talking crap)...
        Change is bad. That's why folks are trying to establish rules of
priority for phylogenetic definitions. If we change our definitions to suit
our concepts, we imply that a taxon has some import outside of describing a
natural group. You are in essence saying "I know what Aves is, and this
definition is wrong for it". That is a typological notion. There is nothing
in nature which *must* be called Aves. There are clades in nature, and the
one we choose to call "Aves" is arbitrary. We may wish to be careful in
naming it to ensure access to the literature. However, once it is named, the
name becomes associated with that definition, and typological connotations
should be abandoned. To continually "refine" our definition of Aves (other
than to promote access to the literature) is theoretically unsound. Given
the state of the literature today, we should be able to arrive at a
definition which works NOW, which gives access to the literature AS OF
TODAY, and start afresh for the future.

>"Gautheir (1986) claimed to use recency of common ancestry, rather than 
>similiarity, to determine classification-a claim with which I disagree.  
>Like all phylogenetic hypotheses, his must be based on specialized 
>similiarities (synapomorphies). The claim that his proposed phylogenetic 
>relationships are based on recency of common ancestry is not supported 
>by his methodology, which uses basic typological morphological 
>assemblages.  Recency of common ancestry implies a genetic relationship 
>inferred only from morphological analysis."  Tarsitano (1991); 545-546.  
        Tarsitano's claim that typological characters lead to a typological
phylogeny and thence to a typological taxonomy is inaccurate. Tarsitano
fails to appreciate that the phylogeny itself is not typological. There is a
true genetic phylogeny out there, and regardless of the methods employed in
the attempt to recover it, it is a real phenomenon. Taxon *definitions*
based on common ancestry are independant of the methods used to
recover/estimate phylogeny. They identify a real group present in the
natural system.
        Tarsitano appears to be confusing the phylogenetic hypothesis and
the phylogenetic taxonomy (a common misunderstanding). The hypothesis is
based on the analysis, but the taxonomy is independant of the analysis, and
may indeed be made a priori of the analysis. Therefore, a typological
approach (if cladistics is shown to be such) in recovering phylogeny does
not condemn this strictly non-typological taxonomy.
        Tarsitano puts the cart before the horse, stateing that Gauthier
implies that the phylogenetic relationships he found are based on common
ancestry. This is not the case. Gauthier ran an analysis, developed a
cladogram. Then he used that cladogram as a phylogenetic hypothesis, from
which he derived hypotheses of recency of common ancestry. These
relationships allowed the application of phylogenetically defined taxa.
Morphology->cladogram->phylogenetic hypothesis (consisting of heirarchical
map of propinquity of descent)->taxonomy. The analysis was not based on
recency of common ancestry, the latter was an hypothetical result of the
former. Again, phylogeny is confused with taxonomy.
        Tarsitano's claim that "recency of common ancestry implies a
relationship inferred only from morphological analysis" is confusing.
Hypothesis of recency of common ancestry may be derived from many sources.
Whether or not cladistics produces "basic typological morphological
assemblages" or not is irrelevant when those assemblages are used as a
phylogenetic hypothesis. I can use a randomly generated groups of taxa as a
phylogenetic hypothesis, and from that hypothesis will come relationships
based on recency of common ancestry. That's what a phylogenetic hypothesis
(in this sense) means. "Recency of common ancestry" does not imply any one
method. I certainly do not understand what magical "morphological analysis"
he is referring to, or why it is any more proper to the task of producing
hypothesis of propinquity of descent.
        Apart from the misunderstandings concerning phylogenetic taxonomy,
Tarsitano does hit home on the subject of phylogenetic hypothesis
generation. He states that Gauthier's phylogeny is typologically because it
is character-based. I have recently argued on this list that characters are
classes, and I have not altered that opinion. However, as we have seen, this
does not make phylogenetic taxonomy typological. Indeed, the result of the
analysis need not be considered typological either when it is used as an
hypothesis of the workings of a real process (rather than simply as a
classificatory mechanism). Just as we reduce an image to pixels to recover
it on our TV screen, we reduce morphology to units to analyize and reproduce
the phylogeny. The image is no less of a typological construct if we can
still watch the Daily Show.
        In conclusion, I believe it is fair to say that Tarsitano simply
misunderstood Gauthier's point (as has so often happened in PT discussion).
    Jonathan R. Wagner, Dept. of Geosciences, TTU, Lubbock, TX 79409-1053
                    "...To fight legends." - Kosh Naranek