Phylogenetic epistomology (was cladobabble--here we go again)

["Augustus T. White" <augwhite@neosoft.com>]

King, Norm wrote:
> 
> Jonathon Woolf wrote (11/14/97; 6:37a):
> 
> >OK, time for another of those stupid questions for which I am
> >probably becoming infamous...

[Take it from me, a stupid question (which yours isn't) is orders of 
magnitude better than a stupid answer]
 
<snips> 

> I see this as one of the flaws in phylogenetic taxonomy.  There should be
> only one way to define phylogenetic taxa.  Forget the praticalities--I
> know the reasons why these three approaches to definition have been
> taken.  Conceptually, the system of phylogenetic taxonomy is flawed in
> that regard.  All taxa should be defined in the same way so they are
> comparable entities.  (I admit, however, that I'd even settle for all
> taxa within a phylum to be defined in the same way.)
> 
> As a teacher, my biggest problem is finding a way to be current, yet not
> snow my audience of lay people under a blanket of undecipherable
> cladobabble.  The best approach seems to be to ignore the current
> ("proper"?) definitions for reptile and tetrapod, for example, and
> explain what they are in the more traditional way.  This makes sense to
> everyone rather than just to the academic elite, the latter of whom seem
> to enjoy arguing about how many angels with Greek and Latin names can
> dance on the head of a pin.  (Froth, froth, froth!)

This debate seems more philosophical than scientific, although its not a 
trivial philosophical problem.  Definitions have no reality, unless you 
are a Platonist, Kantian or what have you.  From the point of view of an 
extreme empiricist (me), definitions are arbitrary boundaries placed on 
sense data.  They are "good" if useful, not because they reflect an 
underlying reality.  Since people have different utilities, the "best" 
definitions are likely to differ.

Paleontologists find useful definitions which are highly congruent with 
accepted ideas of phylogenesis and evolution.  This provides an excellent 
fit for their work and advances a self-consistent explanation of their 
sensory inputs.  This is an important utility and a good reason to use 
cladistic definitions.  The fact that we don't know where some of the 
pieces fit under cladistic analysis is all to the good.  The definition 
"works" because it points out where information is needed and our 
inability to classify under the definition is congruent with our 
ignorance of actual ancestry.  The inability to point with certainty to 
the exact node for definitional pusposes is congruent with our 
evolutionary ignorance of when and what, precisely, the evolutionary 
divergence occurred.

Now to Norm King's students, this is not a useful definition.  Their 
utility is to be able to recognize and classify organisms they observe 
without a lot of fuss and bother.  Their principle utilities are directed 
at other work, and evolutionary precision requires more time and energy 
than they want or have time to spend.  Even if they had it, they lack the 
background to make much sense of cladistic definitions.  Thus, the better 
definition for these students is one that allows classification on 
physical inspection, without a great deal of comparative knowledge.  

There can never be one definition that is best, so long as people 
approach the same sense data with radically different utility functions. 
 The *social* utility in communication tends to reduce the number of 
definitions that are in actual use, but it is neither surprising or 
unfortunate that more than one set applies.

  --Toby White