Google for Internet Database / unique classification?

[pierre deleporte]

A 11:30 22/03/2006 -0800, Doug Yanega wrote :
>Pierre Deleporte wrote:
>
>>Only the extreme terminal taxa would be universally defined, while
>>the classes and hierachies would fit particular needs. What would be
>>wrong with this, who would complain? (besides some "inside"
>>taxonomists I can bet)
>
>Let's consider practical examples:
>
>(1) (...) A taxonomist identifies a species the ecologist is studying as
>belonging to Genus/Family X, so the ecologist assumes that it
>possesses characteristic Y.

This would not be sound: a universal characteristic in a known taxon says 
nothing certain about undiscovered species (except in the trivial case when 
the feature of interest belongs to the explicit definition of the taxon). I 
think we should never rely on classifications to be "predictive" in this 
way (by the way I never understood the notion of "predictive" 
classifications, which can just summarize present knowledge.. any hint?).
The ecologist studying one species has to check for the presence / absence 
of characteristic Y in this species. The ecologist should control that 
snakes are legless tetrapods.

>  That ecologist would certainly complain.

The reviewer of his submitted paper will complain instead: this ecologist 
should have checked by himself. The only point is that taxonomists should 
teach how to use, and not to use, classifications. And also favor 
information retrieval, which an updated data base could do (but not a 
unique classification).

>(2) A biogeographer decides to use the distributions of the member
>species within Higher Taxon X to test a hypothesis. However, there
>are four different extant classifications each of which proposes
>different lists of member species, each of which leads to a different
>result relative to that hypothesis. That biogeographer would complain.

I think a sound biogeographer will not complain. He should be aware that he 
needs a phylogenetic tree, and pick a clade on biological grounds, possibly 
including some ecological or ethological sorting (dispersal abilities...), 
and not use any classification blindly. Anyway nobody should ever use a any 
classification blindly.

>(3) A museum curator receives a request for a loan of all unidentifed
>material of Higher Taxon X, and spends a week packing 14,000
>specimens and then ships them. A week later, they get back a package
>with 13,500 specimens and a note saying "These specimens are in
>Higher Taxon Y, so I don't need them". Meanwhile, elsewhere in the
>collection there are another 5,000 specimens that DO belong in what
>the requestor calls Taxon X, which they will never see because the
>curator places them in Taxon Z. Both the curator AND the requestor
>would complain.

The scientific community could complain against such curators and 
requestors. Hopefully they will correct themselves and agree next time on a 
common classification *before* mailing anything.
I could summarize my point like this : just think before you class (anyway 
you may class and for any purpose). Classification should be an adaptable 
tool for communication, not a burden or a trap.
Note that I insist on a clear specification (i.e. a basis for your 
"consensus") for any classification, so that people can agree on what they 
mean by "taxon Y in classification Z" (and not just "taxon Y").

>(4) The papers, including the recent ones, contain a total of more than a 
>dozen
>functionally different circumscriptions of said Taxon X, many of them
>directly contradictory - and, moreover, there are many references to
>species that belong in Higher Taxon X (in some of these
>classifications) that simply don't show up in library or online
>searches, because at the time they were published they were NOT
>included in Taxon X. Such a person would certainly complain.

Arn't nomenclatural codes for species supposed to help tracking such 
information (through revisons / synonymy)? So you are pleading for 
nomenclatural code for higher taxa...

>Like it or not, a classification is more than just a way to declare a
>hypothesis about which branch is related to which in the Tree of
>Life

Agreed. A phylogeny is not a classification per se. Just a possible 
explicit reference for some classification.

>  it is a tool for the organization of information ABOUT life,
>and we can only retrieve information correctly if everyone is using
>the same tool the same way.

I'm just saying that this is true for any number of different classifications.
When the ecologists are talking of "predators" and "preys", they better use 
a non ambiguous and "consensus" (explicit and useful) ecological 
classification.

>I am not saying that a consensus classification would necessarily
>have to be static, nor even necessarily *stable* - science does,
>after all, march on -

Understood

>  but it DOES need to be standardized (so anyone
>using, say, the name "Nymphalidae" in 2007 will be talking about the
>same thing as anyone else using the name "Nymphalidae" in 2007).

Here I think you mean a nomenclatural code for higher taxa... I was talking 
of choosing the classification you need for a specific use.

All consensus classifications must be standardized (consensus requires 
standards, including standards for updating).
Ideally any classification should be proposed with its explicit logics, 
nomenclatural standards, and possible uses according to some specifications 
in some contexts.

Can we agree on this...? Maybe we just disagree on "unique" classification, 
which will never optimally "answer to all questions" / fit all possible 
uses in my view....

Best,
Pierre



Pierre Deleporte
CNRS UMR 6552 - Station Biologique de Paimpont
F-35380 Paimpont   FRANCE
Téléphone : 02 99 61 81 63
Télécopie : 02 99 61 81 88