[seek-kr-sms] OWL - taxonomy

Nico Franz franz at nceas.ucsb.edu
Fri Nov 4 12:18:29 PST 2005


Hi Serguei:

I think I understand this representation and also think that I like it. 
Systematists do in fact refer to the junction of Morph's 12, 25 & 27 as 
a "hypothetical ancestor". Certain characteristics (i.e. 12, 25 & 27) of 
individuals from extant species are single out to "approximate" what 
critical features the ancestor had. It's a historical science where you 
infer the past from the distribution of features in the present plus a 
model of how frequent and fow likely certain changes are.

So if these are valid intensional definitions of Cyclanthura and C. 
pilosa, do they in any way compromise the ostensive statement (which the 
real-life classification makes at the same time) that Cyclanthura 
(currently) contains 15 species named so-and-so (pilosa, carinata, 
uncinata, etc.)? In other words, can there be a representation in OWL 
where your intensional solution below and a perhaps more conventional 
ostensive tree of parent-child relationships can coexist and inform each 
other?

Also, would it make sense for us to further reduce the size of the 
Cyclanthura example, but nevertheless stay close to reality, so that we 
can see whether we can accommodate the most relevant issues? I could 
prepare another real-life example that doesn't have more than 5-8 
taxa...(not 22). I am assuming you and others are also continuing to 
think about what's on the table right now. I guess I am asking how to be 
helpful with next steps...

Cheers,

Nico

Serguei Krivov wrote:

> Here is much more neat owlification for Nico’s description of genus 
> Cyclantura
>
> Property hasMorphAncestor connect the species to what it was sometime 
> ago. It is inverse of what I earlier called geneticMutation (perhaps 
> hasGeneticVariation would be better???)
>
> The first line is a clean intensive definition of genus Cyclantura as 
> all set of all individuals that have necessary set of morphological 
> features (Morph12, Morph25 and Morph27) plus those species whose all 
> morphological(genetic?) ancestors had all those features. This 
> includes all those known now and those that will be known in the future.
>
> Second definition suggest that all ancestors of Cyclantura-pilosa 
> (perhaps ther is just one) had that set of morphological features 
> (Morph12, Morph25 and Morph27). Thus on account of this 
> Cyclantura-pilosa is subclass of Cyclantura (here we treat species as 
> classes). If in future someone proves that a species X has (direct) 
> morphological ancestor with features described by Morph12, Morph25 and 
> Morph27 that X would come under genus Cyclantura. In such 
> formalization we do not even need those shadowy species , although 
> they could still be here and the last line stating that property 
> hasMorphAncestor is inverse of hasGeneticVariation (which was 
> geneticMutation in my previous message)connects this formalization 
> with one I described earlier.
>
> …just wanted to put another idea on the table.
>
> serguei
>
>------------------------------------------------------------------------
>
>_______________________________________________
>Seek-kr-sms mailing list
>Seek-kr-sms at ecoinformatics.org
>http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/seek-kr-sms
>  
>



More information about the Seek-kr-sms mailing list