Paraphyly and names

Pierre Deleporte Pierre.Deleporte at UNIV-RENNES1.FR
Wed Jan 30 18:27:22 CST 2002

At 12:45 28/01/2002 -0500, Thomas DiBenedetto wrote:

> >>From: Pierre Deleporte
> >>... I don't think that "Nature" classes in any way.
> >>Rather, Human beings class for their convenience, so I can hardly grasp
> >>what a "natural classification" could be.

> >I DO think that nature has formed groupings "on its own". (...)
> >I really dont understand the line of reasoning that you raise. (...)
> >I find it to be competely unproblematical, and in fact rather obvious, to
> >refer to classifications based on the discovery of lineages as "natural
> >classifications" for they recognize groups that have historical reality.
> >Eclectic classifications, to the extent that they fashion groupings on the
> >basis of subjective human preferences, are artificial.
> >Tom DiBenedetto

Agreed that a classification based on synapomorphies may be said "natural". 
My point is rather that "naturalness" is not a very useful line of argument 
in favor of a classification, because classifications are not "natural" 
things but human constructs for some purpose. Nature did not form a 
classification of its own, even if humans can define "natural groups"... on 
diverse possible criteria.

See e.g. the following post of Thomas Pape, arguing in terms of efficiency 
rather than naturalness:
 >>"We may choose to refer to (use names for) monophyletic groups - or even 
for paraphyletic groups. Yet paraphyletic groups are less efficient for 
communication relating to origin of evolutionary novelties. (...)  Taxonomy 
is a communicative tool and much more than fitting the Tree of Life into a 
set of hierarchical ranks."

There is also a parallell thread on this list ("How to arrange a new 
Herbarium?") where people compare the merits of alphabetic order versus 
systematic order, on the basis of compared convenience of the two systems, 
not "naturalness". And I think this is a useful ground for discussing the 
point: if alphabetic order were more practical for collection management, 
why not? (I don't mean I agree with the criterion, I just agree with the 
way the debate is argued).

As for naturalness, I think that a symplesiomorphy is quite "natural", as a 
homologous character state. It is a synapomorphy at some level, and it 
changes into another character state, i.e. a synapomorphy at a lower level, 
and nature produced this "on its own".

As for artificiality, deciding where "birds" should start is no more 
natural than deciding where a possible paraphyletic "reptiles" should end. 
The problem is exactly the same when considered at the level of character 
states: a human decision of where something in nature ends / begins, i.e. 
"significantly differs". Your recent posts seem to indicate that you admit 
this artificiality anyway.

As you also stated (Tom DiBenedetto in "Paraphyly, aves, man etc."):

 >>" Perhaps it is an "arbitrary" decision to
 >>decide to classify according to phylogeny, as opposed to by color, or
 >>alphabetically, or eclectically. That is an issue for a higher level of
 >>epistomology." (...)

Yes, agreed ! But eclecticists simply did not decide to classify according 
to phylogeny (only), they decided to class according to two criteria (at 
least) :
- synapomorphies (indicating phylogeny), and
- symplesiomorphies (indicating some notion of "grades").
Thus I think that the conflict effectively stands at what you call this 
"higher level of epistemology".

But your immediately following argument is like this :

 >> (...) The issue here involves the standards that one adopts after
 >>one has made this higher level decision. If one decides to classify by
 >>phylogeny, there are explicit principles and standards that one must follow
 >>- essentially to recognize as taxa those groupings that emerge from an
 >>empirical analysis of the character evidence. End of story."

I think this argument, addressed to eclecticists, misses the target 
precisely because it does not stand at the relevant "higher level of 
epistemology", but completely inside the non-eclecticist line of reasoning 
(clade = synapomorphy = unique criterion = good for you ... = tautology, as 
underlined by Zdenek Skala). Tom Wendt also pointed at this "relevant 
level" in his last post (what do we want to convey as information(s) in the 
first place ?). If we can at least face this reality, maybe progress can be 
made in mutual understanding (I don't mean agreement).

Now I completely agree with your crucial question addressed to 
eclecticists: do they have a universal criterion ? But I take it as 
practical argument, not some "requirement of naturalness". I'm waiting for 
the answer with much interest, because if there is no satisfying answer, 
then the approach would not be able to overcome the difficulty raised by 
the very choice of eclecticism in the first place: setting an "external" 
universal rule for combining two mutually incompatible criteria (however 
"natural"). The elusive question of "where should birds start?", 
increasingly elusive with new fossil discoveries "filling the gap" with 
intermediary character combinations, will likely not help fixing "where do 
paraphyletic reptiles should stop?" with a universal criterion (vanishing 
"gap", and what else?...).

Naming some (not all) clades involve some arbitrariness (as well as fixing 
a classic "birds" some place, even if we name all clades). But moreover, 
choosing some places where things should be looked at "backwards" 
(symplesiomorphy / paraphyly) just adds arbitrariness, unless a universal 
decision rule is provided. The problem is not arbitrariness in itself, it's 
multiplied arbitrariness (unless...).

My point is simply that this may be argued on the grounds of convenience 
for conveying some kind of information, convenience for reaching scientific 
consensual agreement and thus stability, convenience for managing 
collections and what I know, rather than "naturalness / non naturalness".
Symplesiomorphy is natural. But is it convenient ? For what purpose ? 
Nature will not answer the question, but "systematics users" could.

If I remember well, the last view of the movie "Jurassic Park" is a flight 
of birds against a blue sky, without comment. Well, I explained this the 
following way to non-scientist friends: dinosaurs are still flying around 
there. They were not shocked at all by this "AMNH interpretation" (Ken's 
definition), and rather interested in learning an amazing bit of natural 
history, which they will remember quite well. Of course I was deliberately 
charging 'dinosaurs' with a new content: I choosed not to forge a new term 
like "eudinosaurs", simply now dinosaurs included birds (great news!), and 
the classic paraphyletic "dinosaurs" immediately became "non-bird 
dinosaurs", just like we currently talk of "non-human primates" when 
necesssary. In the particular case this cladistic naming survived pretty 
well the practical test of convenience for communication with non 
scientists. Maybe it could work with some scientists as well.

Why not let users decide what is needed as optimal classification(s) ?


Pierre, "hairy actinopterygian" (just a test)

Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 66
Télécopie : 02 99 61 81 88

More information about the Taxacom mailing list