Cladistics and "Eclecticism"

Tom Wendt twendt at MAIL.UTEXAS.EDU
Mon Feb 4 08:23:10 CST 2002

Tom Benedetto says:  "Cladists are interested in character change to the
>same degree as anyone else, and have, in fact, a far superior protocol for
>presenting that information to the rest of the scientific community."

Well, OK, I got carried away when I said that, to cladists, "The
amount and type of evolutionary change that occurs (or doesn't occur) along
the way is apparently of no interest"--it clearly is of interest in
defining branching points, and of interest when plotted on a phylogeny to
see patterns of change.  But, when I said that they consider it "not worthy
of inclusion in a classification", I am supported by Tom Benedetto, who
says, as I quoted,
" 'Important' nodes, or what Denis refers to as 'significant gaps', are no
different from anyother node. Their 'importance' means nothing more than
that scientists find them to be interesting to discuss, for whatever
 Whether 1 character-change (however defined) or 100 occur between
branching point A and branching point B is totally irrelevant in strictly
cladistic classification unless a branching point is involved; whereas, in
eclectic classification, it is quite important.  And this brings me back to
the "strict cladist" conversation of earlier--I think that this IS a useful
term, because almost everybody is using phylogeny (however imperfectly
known) as a basis for their classifications; however, strict cladists use
just the branching points in making their classifications, while
eclecticists use the branching points and the amount and direction of
change.  (I'm talking about what we WANT to do, not necessarily what we CAN
or DO do--"we" meaning systematists in general.)  A quite separate question
involves the extent to which cladograms accurately depict phylogeny, but
that's a separate horse to beat...

Tom Benedetto says:
>What makes more sense to you Tom?
>Case 1-,A scientific literature which presents an accurate rendition of the
>lineage divergences with the character transformations mapped onto that (for
>example - we have taxon A, which is one of the subtaxa of higher taxon B,
>and we can see this relationship onthe cladogram-, and we have a series of
>marks on the cladogram indicating the number of character transformations we
>have discovered, and it shows that there were an unusually high number of
>transformations in the taxon A branch
>Case 2- A scientific literature that presents a classification in which
>Taxon A and Taxon B are ranked at an equal level because some systematist
>(who had private access to the information laid out in case 1) decided that
>the high number of transformations in the A line warrented a presentation of
>the two taxa as being equal, in some sense.

Everybody should be trying to do as in Case 1, for the reason I mentioned
above--we all are interested in the cladagram with the changes mapped onto
it, and all such information should always be presented.  This is a totally
different viewpoint from saying that the resultant classification should
reflect ONLY branching points.  A strictly cladistic classification (not
the cladagram itself, but the naming of clades) tells you nothing about
"how much" or "what kind of" change has occurred.  Why do we want a
classification that leaves out information?  (I completely agree that the
"how to" of eclectic classification is problematical and worthy of spirited
discussion, but I'm more curious about the theoretical question of what we
want from a classification.)


More information about the Taxacom mailing list