[Taxacom] Mona Lisa Smile

Richard Pyle deepreef at bishopmuseum.org
Mon Jul 31 00:27:08 CDT 2006

I've been offline since my last post on this thread, and I haven't finished
reading it all through yet, so this may be moot/old/addressed already,

Some time ago, Tom Lammers responded to my post:

> >>The "Classification Is A Hypothesis" approach you advocate 
> (to which I 
> >>do not personally subscribe), seems to
> imply that there is some sort of singular "natural" 
> classification "out there", against which these 
> classification hypotheses can be "tested".<<

... with:

> There is.  It's called "the actual historical facts of the 
> group's evolution."  

I would regard that as a hypothesis about phylogenetic relationships -- not
a hypothesis about classification.  How we choose to incorporate evidence
supporting or refuting phylogenetic hypotheses into our human-imagined
classifications is not about hypothesis-testing, it's about language
definitions of terms, and communication.

> These populations we see today ARE 
> derived from pre-existing populations through modification 
> over time.  

In many/most cases, they may well be (introgression being among the spoilers
of neat bifurcating phylogenies).

> I believe that 
> the more closely our classification reflects that underlying 
> "truth," the better it is.  

I think you mean that you believe that classification should closely reflect
hypotheses about phylogeny, and that ideally the classifications should
usually/always reflect our "best" hypotheses of phylogeny.  I think this is
what Ken Kinman would brand as "strict cladism" (or -- what was the word Ken
used -- "Monomania"? I prefer "monophilia" -- goes better with

In any case, many taxonomists share this belief, many others do not.  But
regardless, we're talking about "belief" here -- that is, an opinion about
the prefered human-defined "axis" about which our primary classification
scheme should revolve.  This is not a hypothesis to be tested, but rather a
convention to be defined.

> We can argue about whether 
> cladistics is the best method for inferring this historical 
> truth, or HOW our classification should reflect it (e.g., is 
> paraphyly okay?), but I cannot imagine anyone thinking it 
> would be a BAD think if our classifications reflected 
> evolutionary history.

I would agree that most of us agree that evolutionary history should be the
primary guiding principle of our Linnean classification schemes.  The
arguments are really about how much "wiggle-room" we should allow ourselves
in order to gain more functional classifications in particular
circumsatances, and/or more stable taxonomies.

But I think Pierre's original point is that perhaps we should have several
different such constructs within biology to emphasize different aspects of
biological affinity (and therefore allow more effective field-specific
communication).  But the meta-point is that classifications are
human-defined constructs, not hypotheses to be tested. Phylogenies are
hypotheses that can be tested; but the degree to which we map our
classifications to these phylogenetic hypotheses is obviously still a matter
of some debate.


Taxacom mailing list
Taxacom at mailman.nhm.ku.edu

More information about the Taxacom mailing list