[Taxacom] More precise sound bite

[Mike Dallwitz]

Barry Roth wrote:

> I'm not sure why I would want to make a classification unless I thought I 
> could get something more out of it than what I put into it. And I think this 
> boils down to predictivity. Because of the fact of organic evolution, the 
> classification that best serves this need / desire will be one strongly 
> grounded in phylogeny.

One thing that you get out of any classification that includes names is the 
ability to communicate. All common nouns correspond to classifications, 
which are presumably chosen (or evolve) for their usefulness for 
communication, which probably depends on our ability to mentally picture and 
remember the concept of the noun, which is probably related to its 
predictivity in some sense.

Have classifications based on cladistic methods been shown to have better 
predictivity than other classifications? Predictivity could (for example) be 
defined as in

Gower J.C. 1974. Maximal Predictive Classification. Biometrics 30: 643-654.
Colless D.H. 1984. A method for hierarchical clustering based on
     predictivity. Systematic Zoology 33: 64-68.

> This also makes me more look charitably on monophyletic (i.e., holophyletic) 
> groups than paraphyletic groups.

Does this mean that belonging to a paraphyletic group such as reptiles 
usually has less predictive value than belonging to any of the monophyletic 
groups to which reptiles belong?

-- 
Mike Dallwitz
Contact information: http://delta-intkey.com/contact/dallwitz.htm
DELTA home page: http://delta-intkey.com