[tcs-lc] RE: LC/TCS - How many schemas?

[Richard Pyle]

> >How do you distinguish between an "Identification" and direct
> >reference to a
> >Voucher in the context of a Concept definition?  Are these
> >fundamentally
> >different things, in your mind?
> >
> yes - they are different in my mind.
> A direct reference to a Voucher in the context of a Concept
> definition means the taxonomist who defined the concept actaully
> used that Voucher in his work (now we may or may not have this
> information for legacy concepts apart from type Vouchers - but
> could encourage the practice for new ones).
> An identification may or may be have been by the same author of
> the Concept, but regardless the specimen but was not actually
> used during the process of defining the Concept.  If the
> identification was by the same author some people might say that
> he is extending the circumscription of the Concept and they
> should be included, but I'd say to keep things clean and avoid
> possible Concept drift they should still be counted as
> identifications and if and when he thinks it appropariate he
> might revise his definition of the Concept and include some of
> those earlier identifications in his new definition of the
> Concept. If the identification was by a different author then
> these should certainly be kept different.

I *think* I understand how you distinguish them....but I also see a lot of
fuzzy areas and subjective interpretations. All identification/voucher
events amount to "Person X included this specimen within a concept
circumscription that also includes the primary type specimen of this name."
Person X might know almost nothing about the group in question, Person X
might be the world authority on the group, or Person X might be anywhere
along the continuum of these two endpoints.  The identification might have
been made simply to enter a name (any name) on a specimen label, it might
have been made to unambiguously mark one edge of a taxon concept
circumscription within the context of a well-researched, highly detailed
published revision of the group, ir it might have been in any imaginable
context in-between. Hence, I don't see them as "fundamentally" (objectively)
different.

> >> I just think it makes a statement about how precise you are in
> >> your identifications - if you know names mean different things
> >> then you should say what you meant
> >
> >I COMPLETELY agree -- and the world would certainly be a much
> >better place
> >if all (or even a substantial number of) existing identifications were
> >"self-mapped" in this way.  But the reality is that they are not.
> >
> so if we don't know what they mean make that clear by using a
> nominal concept that can clearly be interpretted as ambigous in
> meaning rather than tagging a name and pretending the problem
> doesn't exist.

I know that this can be accomodated by the TCS approach, just as (I think)
all of the needs of TCS can be accomodated by some of the alternative
approaches being discussed.  In fact, I think it's safe to say that all of
our collective needs will be accomodated by any approach (provided enough
carful thought goes into it).  I think the question and/or source of the
different perspectives is more along the lines of which part is the "dog",
and which part is the "tail" (and therefore how the information should
optimally be referenced, combined, dissociated and modularized).

Rich