[Taxacom] Semantic Web: What is a species?

Richard Zander Richard.Zander at mobot.org
Fri Jan 30 10:45:08 CST 2009


One of the reasons this thread is important, I think, is that
cladistics, and any method involving a data set alone, is one step
removed from the actual specimens. The data set (ignoring molecular for
the moment) consists ONLY of those traits the alpha taxonomist thinks
are nicely linked, organizing traits (organizing other traits including
those difficult to describe or measure) for different taxa he/she infers
from the specimens examined. The species is that inferred from all
activities of the alpha taxonomist, and may not include some or even
many important characters of look and feel, and, e.g. commonly
biogeography, that helped distinguish and characterize the species. Some
"good" traits are not good for other taxa, yet all traits are lumped
into the data set with the expectation that convergence will be dealt
with by convergent traits being snowed under by lots of good traits that
point the way. But the process of deciding on what traits are good
limits the number of traits in the data set.

Analysis of the data set allows a neat tree that helps organize one's
thinking about relationships, yet some traits are used for some taxa
that were Never Meant to Be Used There, because they were discarded by
an alpha taxonomist dealing with the full range of traits. "Many a thing
growtheth in the garden that wast not planted there."

Species are not diagnoses. Exemplars represent the alpha taxonomic
apprehensions of what is out there, and morphological analyses lump all
the basic organizing traits for each and every taxon and shake the
traits all together in a data set, then run a simple evolutionary
model-based non-ultrametric clustering program, and you get, well, much
the same thing that the alpha taxonomists got (because the traits are
actually linked) plus wrong relationships (because the traits are
treated as unlinked). 

Therefore, discussing species as real or not is quite different from
discussing a data set, and helps protect the mind set that species are
not rows in a data set. Data sets can introduce persistent and
problematic errors. Total evidence can be no solution when data is
distilled and extracted and treated as taxonomically unlinked.

I think reductionist dealing with traits rather than taxa is important,
but must be tempered by common sense.

*****************************
Richard H. Zander 
Voice: 314-577-0276
Missouri Botanical Garden
PO Box 299
St. Louis, MO 63166-0299 USA
richard.zander at mobot.org
Web sites: http://www.mobot.org/plantscience/resbot/
and http://www.mobot.org/plantscience/bfna/bfnamenu.htm
*****************************


-----Original Message-----
From: taxacom-bounces at mailman.nhm.ku.edu
[mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of Kenneth Kinman
Sent: Thursday, January 29, 2009 9:55 PM
To: taxacom at mailman.nhm.ku.edu
Subject: [Taxacom] Semantic Web: What is a species?

Hi Curtis,
       Semantics can be tricky.  If I read your post correctly, you
regard species as "special" and perhaps "real" as well, but not "more
natural".  I'm just not sure what you mean by the latter.  "More
natural" than what???   And I am tempted to also ask, is a monophyletic
(holophyletic) species any "more real" than the paraphyletic mother
species that might have given rise to it?  
           ----Cheers,
                      Ken Kinman                        




More information about the Taxacom mailing list