[Taxacom] Reproducibility of phylogenetic analysis

J. Kirk Fitzhugh kfitzhugh at nhm.org
Sun Jan 24 17:07:31 CST 2010

It's amusing the reviewer would use Ghiselin's book as the reason. 
Personally, I found his book intriguing for the numerous references to 
Darwin's use of hypothetico-deductivism and being a master of testing 
his theory of natural selection. I've seen nothing in Darwin's works 
where he engaged in such testing, especially considering that he thought 
selection operated at such a slow pace. But then, I also deny that 
cladograms are ever properly tested by the established mechanics of 
testing explanatory hypotheses.


J. Kirk Fitzhugh, Ph.D.
Curator of Polychaetes
Invertebrate Zoology Section
Research & Collections Branch
Natural History Museum of Los Angeles County
900 Exposition Blvd
Los Angeles CA 90007
Phone: 213-763-3233
FAX: 213-746-2999
e-mail: kfitzhug at nhm.org

Richard Zander wrote:
> Bob Mesibov has made a good point. Regarding subjectivity and intuition
> in systematics, I here paste a pertinent passage from an article of mine
> recently rejected by a Very Good Journal:
> " 	A third method of scientific analysis is intuition, long
> lambasted as illogical and subjective though often defended as "common
> sense." The well-known psychologist-statistician G. Gigerenzer and his
> group have pointed out recently (Gigerenzer, 2007; Gigerenzer & Selton,
> 2001; Hutchinson & Gigerenzer, 2005) that intuition has a clear,
> describable and amazingly effective methodology even though largely an
> unconscious feature of human intelligence. Basically, intuition is an
> insensible mental implementation of genetic algorithms for rule-set
> production (GARP). The GARP method may include anything from
> trial-and-error of simple solutions until one is successful to
> sequential Darwinian evolution of more and more complex formulae until
> one is most successful according to some stopping rule. Formal genetic
> algorithms (Stockwell, 1999; Stockwell & Peters, 1999; Stockwell &
> Nobel, 1992) for scientific use are well known to floristics specialists
> and ecologists in the GARP software DesktopGarp (Scachetti-Pereira,
> 2002) and certain other software that solve problems fast with brute
> speed. Gigerenzer's main point is that intuition involves both long-term
> (e.g. lifetime of a culture) and short-term (such as solving an
> immediate puzzle) unconscious solutions to problems involving
> uncertainty or paucity of data. The process develops heuristics
> (rules-of-thumb) that rapidly provide solutions to difficult or
> intractable problems that cannot be easily or rapidly solved by exact
> methods (deductive or inductive, falsificationist or verificationist).
> Hutchinson & Gigerenzer (2005) give many examples in the field of
> biology. 
> 	Gestalt or omnispection methods of the past have a large element
> of intuition. The closure or past experience principles of Gestalt
> psychology demonstrate the remarkable ability of the human mind to
> complete patterns (notes in a melody, a picture from a partial drawing),
> but this power must be trained, and doubtless heuristics as described
> above are involved.
> 	I here suggest that over the past 250 years of Linnaean
> taxonomy, heuristics have been developed, accumulated, and used to
> create classifications (the more modern based on evaluations of
> evolutionary relationships) that are fully successful when judged by
> consilient analyses using molecular data. Given the faults and
> inadequacies of both morphological and molecular exact methods of
> evolutionary reconstruction, intuitive systematics as an integrative,
> evolving method is in fact a triumph of human scientific endeavor."
> The main references are:
> Gigerenzer, G. 2007. Gut Feelings: The Intelligence of the Unconscious.
> Viking Penguin, New York.
> Gigerenzer, G. & Selten, R. 2001. Bounded Rationality: The Adaptive
> Toolbox.  MIT Press, Cambridge, Mass.
> Hutchinson, J. M. C. & Gigerenzer, G. 2005. Simple heuristics and rules
> of thumb: Where psychologists and behavioural biologists might meet.
> Behavioural Processes 69: 97-124.
> Scachetti-Pereira, R. 2002. DesktopGarp, ver. 1.1.6. University of
> Kansas Biodiversity Research Center, Lawrence, Kansas. 
> Stockwell, D. R. B. 1999. Genetic algorithms II. Pp. 123-144 in: A. H.
> Fielding (ed.), Machine Learning Methods for Ecological Applications.
> Kluwer Academic Publishers, Boston.
> Stockwell, D. R. B. & Peters, D. P. 1999. The GARP modelling system:
> Problems and solutions to automated spatial prediction. Intern. J.
> Geogr. Inform. Systems 13: 143-158.
> Stockwell, D. R. B. & Noble, I. R. 1992. Induction of sets of rules from
> animal distribution data: A robust and informative method of analysis.
> Mathematics and Computers in Simulation 33: 385-390.
> The main objection of the reviewer was that " the "triumph" of
> heuristics in evolutionary biology has been quite well demonstrated by
> previous authors, most eloquently by Michael Ghiselin (1969; many new
> editions) in his classic "The triumph of the Darwinian method" and my
> paper apparently fell short. Strangeness prevails since our subjects are
> clearly different. I shall, however, cite this robust book in my next
> resubmission, to a Different Very Good Journal.
> *****************************
> 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
> Modern Evolutionary Systematics Web site:
> http://www.mobot.org/plantscience/resbot/21EvSy.htm
> *****************************
> -----Original Message-----
> From: taxacom-bounces at mailman.nhm.ku.edu
> [mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of Bob Mesibov
> Sent: Sunday, January 24, 2010 4:28 PM
> Subject: [Taxacom] Reproducibility of phylogenetic analysis
> This article
> http://arstechnica.com/science/news/2010/01/keeping-computers-from-endin
> g-sciences-reproducibility.ars
> is relevant to systematics, as it highlights an aspect of
> algorithm-based phylogenetic analysis which troubles some systematists.
> In the past, some Taxacomers have worried aloud that phylogenetic
> inference is a house of cards, implying (incorrectly, IMO) that if even
> one of the analytical steps is flawed, then the whole process is
> worthless. I and others have suggested here that algorithm-based
> phylogenetic inference is a closed, self-consistent process whose
> results are not tested by other means. Richard Zander has proposed an
> explicit statistical procedure to clarify the nature of Bayesian
> results, but I suspect his paper has been widely ignored.
> I'm not sure that any of this skepticism will make systematists any less
> enthusiastic about filling the literature with analyses, because trees
> are very useful and systematists are keen to build them using whatever
> tools are acceptable to their peers and their editors. What's worth
> keeping in mind is that most (imagine your own percentage here) of
> animal and plant taxonomy has been built by morphologists who made
> 'non-reproducible' and subjective judgements about similarities and
> relationships.
> Some of those judgements were wrong, and it's easy to feel superior to
> systematists who didn't have today's analytical tools. It would be a
> mistake, though, as the Ars Technica article implies, to try to escape
> responsibility for bad judgements by saying, 'Well, we used these data
> and these methods and got these results, so if they're wrong it's not
> our fault.'

More information about the Taxacom mailing list