[Taxacom] More evidence turtles are diapsids

[J. Kirk Fitzhugh]

Whewell's consilience can't do an end run around evidential relevance 
(sensu Carnap) in non-deductive inference, and I've never seen any 
commentary on his concept that would rationally allow such. To say some 
explanatory hypothesis is most consilient, i.e. has greatest explanatory 
coverage, then this would entail one has not partitioned relevant 
evidence used to infer hypotheses. To do otherwise would mean having 
hypotheses that are not maximally explanatory, respectively. But just as 
critical, the disparate hypotheses are only relevant to the subsets of 
evidence in need of explanation, and as a consequence, drawing any 
comparisons between the hypotheses would be meaningless since they don't 
address the same sets of questions.

Kirk

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
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
http://www.nhm.org/site/research-collections/polychaetous-annelids
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~




Ian Stocks wrote:
> I've been following this conversation with interest- I agree that two
> different data sets are cannot be used as reciprocal tests of the
> veracity of the hypotheses being generated by the data, but I have to
> wonder about the role of Whewell's notion of consilience. Is it possible
> that if a data set is portioned (e.g. morphology and mnolecules), and
> they generate the same hypothesis (e.g., tree topology), then we gain
> confidence in the hypothesis through consilience. Not a test per se,
> more in line of the verificationist frame, but perhaps still better than
> nothing?
> Ian
>
>
> Ian Stocks, PhD
> Insect Diagnostician and Collections Manager, Clemson University
> Arthropod Collection
> 312 Long Hall
> Department of Entomology, Soils, and Plant Sciences
> Clemson University
> Clemson, SC. USA
> 29634-0315
>
> 864 656 5035
> ians at clemson.edu
>
>
> -----Original Message-----
> From: taxacom-bounces at mailman.nhm.ku.edu
> [mailto:taxacom-bounces at mailman.nhm.ku.edu] On Behalf Of J. Kirk
> Fitzhugh
> Sent: Tuesday, October 20, 2009 2:32 PM
> To: taxacom at mailman.nhm.ku.edu
> Subject: Re: [Taxacom] More evidence turtles are diapsids
>
> True, the differences to which you refer are between the nature of the 
> test evidence required for testing theories (universals) and hypotheses,
>
> as well as the relations between observer, causes, and effects. These 
> are matters that have been recognized since the 19th century, and can 
> even be found in Popper's writings, as well as other 20th century works.
>
> I've pointed this out at great length in my own work. But, the 
> differences are nothing terribly earth shattering.
>
> My point all along is that we continue not to correctly address testing 
> in lieu of meaningless congruence, which is no test whatsoever and not a
>
> rational means to assess the veracity of hypotheses.
>
> Kirk
>
>