Summary of Nomenclature and GUIDs

[Roger Hyam]

Yde,

This is not a problem. It doesn't really matter if some one tags data to
a TaxonName rather than a TaxonConcept. It just shows that they are
ignorant of which concept the specimen they are looking at belongs to or
ignorant of concepts all together. But what is important is that it will
be obvious what they have done and people can reason from their
decision. Probably in a large number of cases it simply won't matter
because there will only be one widely accepted concept anyway ...

If people really hate TaxonConcepts they will deliberately ignore the
whole thing and just link to TaxonNames (which will be bad) but at least
it would be a start and it would be clear what they are doing.

In the system I put forward earlier everyone can play and use it how
they like and it should be possible to work out what is going on. The
only things that will mess the system up is people not referring to
TaxonNames at all or a whole bunch of nomenclators issuing GUIDs for
TaxonNames in the same group and the community not then agreeing to use
just one of nomenclators. IMO neither of these things is likely to occur
to such an extent as to cause a problem.

Hope this helps,

All the best,

Roger



Yde de Jong wrote:
>>> [Yde] A detail I would like to discuss  within this context in the
>>> near future is the best possible way to link species names to
>>> genetic resources (e.g. DNA barcodes).
>>
>> [Sally] This is something that CBOL have been discussing. I think
>> that their current draft is that barcode sequences should be linked
>> to the best available name (ie a well defined concept if there's a
>> checklist, otherwise a name from a nomenclator, & only if there's
>> nothing available from a nomenclator should they go for just a name,
>> or even
>> an unpublished candidate name like sp A, sp. B ).
>> When CBOL were discussing the link to names earlier this year (in
>> May) things weren't so well advanced on the GUID front so their
>> approach may need to be reviewed if we do get a good system of GUIDs
>> in place for names or concepts. However I know that they are
>> including GBIF in the discussions so hopefully whatever they come up
>> with will be compatible with whatever we come up with ...
>
> Thanks Sally and Rod (for technical comments). Yes I agree that
> linking to the best available (checklist/index) concept or
> (nomenclator) name seems to be the proper way, however, here exactly
> my fear begins. I simply don't believe that because of the inherent
> instability of taxon concepts and because of the taxonomical
> under-education of most molecular workers this will provide a
> reliable model. I am afraid that when no very strict regulations on
> identification, voucher typification and naming of used lab-specimen
> will be established, we will not only end up with a similar bad
> situation as currently in GeneBank (30% of the sequenced taxa wrongly
> identified and no vouchers for control), but also an unworkable
> taxonomic framework because of disintegrating taxon concepts.
> Especially because the expected expansion of DNA barcodes will be
> much larger compared to current sequence data, which could enforce
> fuzzy working.
> Of course ideally DNA barcoding should proceed from type specimen
> (not found realistic in our museum) when possible and professional
> taxonomists should be closely involved. I know CBOL is working on
> guidelines for vouchering, and so on, but still...
>
> Sorry about my sceptics and have a nice weekend!
>
> Yde
>

--

-------------------------------------
 Roger Hyam
 Technical Architect
 Taxonomic Databases Working Group
-------------------------------------
 http://www.tdwg.org
 roger at tdwg.org
 +44 1578 722782
-------------------------------------