Real Science => Sarcasm?

[Richard Zander]

The rhetoric from cladists is definitely heating up in public forums. Here
on TAXACOM, below, anyone not following cladism's nonsensical philosophy of
science is NOT DOING REAL SCIENCE. At the Smithsonian Symposium on the
Phylocode vs. Linnean Nomenclature just last week, a speaker said, right out
loud, without shame, that "using ranks leads to BAD SCIENCE."

My (lengthy) reply to him, during the comments period (in front of God and
E.O. Wilson) was (approximately):

Speaking of "bad science." Every other scientific discipline has some
standard method of ensuring rigor and significance of the results. Not once
has the ess word, statistics, been mentioned at the Symposium, nor is it
much refereed to except deprecatingly in the cladistic literature over the
past 30 years. We systematists are, for some reason, proud of our
statistical innumeracy, but this makes us easy marks for those promoting bad
methods.

Although optimization methods commonly demonstrate accepted relationships of
"uncontested phylogenies," we expect that better resolution of questionable
relationships would now be more reliably resolved. It is, however, commonly
accepted that optimization methods cannot retrieve the true tree if there is
extreme branch length heterogeneity or extreme convergence. Even assuming
that extreme branch length heterogeneity and extreme convergence is rare,
the philosophy of cladism that the least falsified, the simplest, and the
best explanation are acceptable as "results" is ridiculous. Scientists only
follow parsimony when the alternatives are few and clearly comparatively
unrewarding as theories - - this is not the case in phylogenetic estimation.
I submit that the absence of acceptable and understandable probabilistic
branch support measures in cladistic publications over the past 30 years
renders cladistic results no better than alphabetization for classification
or flipping a coin for absolute branch order.

Imagine convincing an actuary or an ecologist to abandon statistics and use
maximum parsimony and hypothetico-deductivism as method and justification.

The response of the speaker to my asking what "well supported" means in
cladistics was to confound methods of alpha taxonomy (every taxonomist must
judge the degree of support for a classification) with scientific,
statistical reliability of phylogenetic estimation (what is the chance of
being wrong?).

Want more? Check out Deconstructing Reconstruction at
http://www.buffalomuseumofscience.org/BOTANYDECON/moweb.htm

These comments are directed largely at students, of course, since those who
have bought into the philosophy gimmick are locked into cladism for better
or worse.


---------
From:
Richard H. Zander
Curator of Botany
Buffalo Museum of Science
1020 Humboldt Pkwy
Buffalo, NY 14211 USA
email: rzander at sciencebuff.org
voice: 716-896-5200 x 351



----- Original Message -----
From: "christian thompson" <cthompson at SEL.BARC.USDA.GOV>
To: <TAXACOM at USOBI.ORG>
Sent: Friday, March 30, 2001 8:14 PM
Subject: Real Science => Sarcasm?


> People have asked whether I was being sarcastic when I used the words REAL
> science. For those of you who don't know me, permit me to say that I have
> always clearly stated that nomenclature IS NOT SCIENCE. I will now go
> further and state: Those of us who use Linnaean nomenclature without
> applying Hennigian grouping and ranking criteria (synapomorphy, age of
> origin), need to warn their users. They should add the following to their
> classifications:
>
> Phylogenetic Warning:
> Comparing Linnaean Taxa above the species level may be injurious to your
> Science. Always consult a Systematist first.
>
> However, as with cigarettes I suspect those addicted to traditional
> classifications will pay no attention to our warnings! Despite that,
making,
> for example, biodiversity assessments based on the number of families or
> genera or whatever, isn't sound Science. The only really valid comparisons
> are those among sister-groups/clades (ala Mitter et alia). And Brent
Mishler
> may also be right about the species level too, as clearly species concepts
> across the full span of life, from microbes to humans, are not comparable
> either.
>
> So, in short, one of the positive aspects of the PhyloCode is that it
> attempts to incorporate REAL science into its nomenclature. And that
should
> force the others to admit that Linnaean nomenclature is an information
> retrieval system based on unique keys (names), but beyond that beware of
> what you attempt to derive from it.
>
> So to repeat my original warning:
>
> We should not be framing this issue as another paradigm war in which
> Linnaean nomenclature is going to be massacred* and the Phylocode is going
> to be triumphant. I fought in the evolutionary/Phenetic/Cladistic paradigm
> wars of the 1960s-70s. As a community we all lost. So, today we need to
work
> together, recognizing our strengths. Linnaean nomenclature is the
foundation
> of a universal information system for biology. It has changed since
Linnaeus
> to adapt to Darwinian** view of the World and will change more, but it is
> only an information system. Cladistic information is critical for our
> Science. The PhyloCode is attempt at improving how we communicate that
> information now (which is cladograms with Linnaean names). I believe it is
> too soon*** to decide even if the PhyloCode is best way to communicate
> cladistic information, let alone be deciding whether it should REPLACE our
> current information system. However, there is already a positive aspect.
It
> focuses attention on Linnaeus' original idea of SEPARATING SCIENCE
> (taxonomy, diagnoses, etc.) from Nomenclature, words acting as unique
> information keys for effective communication. How much Science should be
> built into our  nomenclature? And how?**** Let's address those questions,
> rather than throwing our lives down in "Linnaeus's Last Stand."
>
> * for the non-Americans reading taxacom, Pennisi' title, "Linnaeus's last
> stand," was a take-off on "Custer's last stand," a reference to a battle
in
> the Dakotas in 1876 where American Indians slaughtered a troop of American
> soliders. And what she was really trying to say with that allusion I won't
> even try to quess.
>
> ** Also, don't try to label us as "creationists" because we use "Linnaean
> nomenclature." "Linnaean" is used out of respect and because of priority,
> not because the system has remained unchanged, etc.
>
> *** Remember it was almost 100 years after the 10th edition of Systema
> Naturae that the Zoologists finally wrote our first Code (Strickland Code,
> 1842).
>
> **** I will conclude by answering my own questions. The more Science that
> is built into informational system, the more instability there will be as
> Science changes. New characters, intepretations, taxa, etc., will be
> discovered, etc.  So, let's let the ornithologists and the others who work
> on well-known groups worry about the PhyloCode, and we entomologists can
> names new species. No need to throw stones at each others (except for fun
> :-))
>
>
> F. Christian Thompson
> Systematic Entomology Lab., ARS, USDA
> Smithsonian Institution
> Washington, D. C. 20560-0169
> (202) 382-1800 voice
> (202) 786-9422 FAX
> cthompso at sel.barc.usda.gov
> visit our Diptera site at www.diptera.org
>