[SEEK-Taxon] Re: GUID

[Richard Pyle]

Hi James,

> I'd like to make clear my original question: What is expected
> as terget of GUID?  If a GUID is designed to be given to a potential
> taxon (name-sensu-citation triplet), why do we need yet-anotehr ID
> although we have the triplet?  Encoding mechanism of the triplet is
> rather detail issue.  If we think that GUID more than triplet is
> necessary, I guess that we are thinking of something beyond a
> potential taxon.  Rich's
>
> > I see the value in an arbitrary GUID system as metaphorically a
> common "flag
> > pole" around which all datasets can rally.
>
> sounds so at least for me.  I do not insist to concatinated text
> itself.  I don't think we can choose an appropriate  GUID system
> without understanding the target object to be identified by the ID.

Sorry -- I didn't focus on that aspect of your question.  I think that the
target of the GUIDs in this context is indeed the "name-sensu-citation"
triplet (couplet?), and within the SEEK context is intended as the surrogate
primary key for a concept instance.  At the meeting I attended, all agreed
that the scope/domain of such "name-sensu-citation" triplets/couplets
included minimally all original descriptions (what I call Protonyms), as
well as all "significant" subsequent usages of the name that represent the
"birth" (or anchor definition) of each concept/circumscription.  I got the
sense that in the SEEK context, the ideal would be to avoid enormous
proliferation of "potential taxa" (potential concepts), which means to
minimize duplication of concepts (i.e., minimize the number of times where
one concept instance maps to another concept via the "congruent" type of
relationship) -- except to record major published works that happen to share
congruent concepts.  My hope was that the GUID system, if established, would
be open for use on *any* instance of a name-usage (Assertion), even if SEEK
limited its interest only to well-defined "concept-bearing"
name-sensu-citation instances.

Of relevance here is also the discussion we had at the meeting of
establishing a separate GUID series for "References" (a superset of
date-stamped Agents, of which Publications are a subset); or adopting one of
the several such ID series that already exist.  If there was yet another
GUID series for "Names" (i.e., "Protonyms"), then each Concept could be
uniquely identified by the combination of Reference GUID and Name GUID, and
therefore would not need a GUID of its own.  However, the premise of the
Taxonomer model, and a philosophy that seemed to resonate at the UCSB
meeting, is that "Names" do not exist without the context of a concept; and
therefore "Names" might best be treated as a subtype of concepts.

I think all agreed that "Concepts" are characterized as the union of a
"Name" and a "Reference".  Names by themselves cannot stand as unique
identifiers because of problems of alternate spellings, and problems of
homonyms.  So a concatenated unique identifier for a name would need to be
qualified by details of the Reference in which it was originally described.
Concepts would need to be further qualified by details in which the name is
used as the representation of a concept.  The fundamental problem,
therefore, is creating a concatenated string formula that reliably serves as
a unique identifier for a Reference.  Even if the scope of "Reference" is
restricted to "Publications", there is still some problem with establishing
uniqueness by concatenating citation details.  Authors-Year-Page is probably
not going to be enough, so the question is what other details of the
Reference instance should be used to uniquely represent identify it.  But I
digress....

Of course, I do not speak for SEEK -- these are just my own impressions
based on my own opinions, and my interpretation of what was discussed at the
UCSB SEEK meeting earlier this year.  If I have misrepresented anything, or
if I have still failed to address your question, please let me know.

Aloha,
Rich

=======================================================
Richard L. Pyle, PhD
Natural Sciences Database Coordinator, Bishop Museum
1525 Bernice St., Honolulu, HI 96817
Ph: (808)848-4115, Fax: (808)847-8252
email: deepreef at bishopmuseum.org
http://www.bishopmuseum.org/bishop/HBS/pylerichard.html