[Taxacom] RES: south-west Australia

[Curtis Clark]

On 6/24/2011 7:50 AM, Robinwbruce at aol.com wrote:
> In systems where we are exploring generalities of  uncertain natures I 
> do think that use of statistical logic may be premature;
I think you sell statistics short here. Although parametric statistics 
has a lot of assumptions, there are many other statistical tools for 
pattern analysis (if I were a statistician, I could suggest some). The 
value of this approach is twofold: (1) the human mind is superb at 
seeing pattern, even when it doesn't exist, and statistical tools 
provide an independent check, and (2) the human mind in contrast often 
has trouble seeing actual patterns in multidimensional space, and these 
techniques can suggest patterns that we might not otherwise see.
> I do not claim to be a panbiogeographer, just a biologist; but Croizat 
> is, to my mind, a very important figure in 20th century biology, 
> others are free to disagree.
When I first learned of Croizat's work, I thought "this is probably 
important", and I still don't discount the basic ideas (assuming I 
understand them correctly), but, within the culture of science, 
panbiogeography is on a side track, no doubt due to Croizat's own 
eccentricities. I would venture that it will become mainstream only if 
its current practicioners reinvent it, or more likely when it is 
rediscovered in the future, much as genetics was.



-- 
--
Curtis Clark
Cal Poly Pomona