[tcs-lc] Taxonomic Product or Taxonomic Data

[Richard Pyle]

> Can some one raise a case from this story for why it would be better to
> have names separate from taxa in the schema. I am worried that I don't
> have enough arguments against it to hold a balanced view at the moment.

There are two ways to look at this.  The first is the straw-man schema I
presented (inspired by James) that puts Name-Objects as a top-level element,
completely separate from TaxonConcepts.  I'm not sure if James is proposing
this, but I am certainly not (yet) proposing it.  I just put it out there to
test where it falls short of meeting the needs on technical grounds.

Another way to look at it is having a "compartmentalized" (modular) Names
subunit within TaxonConcepts (more modular than represented in the current
version of TCS).  My favored view of this solution (which I'm warming up to
more and more as I continue to think about it) is to embed all the name
elements within TaxonConcept in such a way that they *only* exist for the
subset of TaxonConcept instances that are of type "Nominal".  Structured
name information for non-Nominal concepts would be accessed via pointer back
to the appropriate Nominal Concept instance (with a "NameVerbatim" element
to capture the exact orthography used in the "AccordingTo" publication).
The "pointer" could come in one of several forms (e.g., a dedicated element
of this sort more or less repalcing "NameDetailed"; or a "refers to Name"
Relationship).  That way, concept-less name data could be easily and cleanly
be passed via a subset of concepts filtered by "Type='Nominal'".

So, my question is: do you want me to provide reasons why I favor a
"modular" Names subunit?

Also, you seem to share my view that name data should be attached to Nominal
concept instances, rather than Original concept instances (as Jessie seems
to favor).  Is this correct?

Rich