[tcs-lc] more TCS/LC interaction

[Richard Pyle]

Hi Nico,

> You keep making the point that whether the "sec." is added or not
> makes a fundamental difference.

It is my way of notating the difference between what I see as a name object,
and a concept object.  Without the "SEC.", the name is technically limited
in scope to the primary type specimen (dot), and otherwise represents a name
to which nomenclatural rules apply.  With the "SEC.", it implies the circle
that circumscribes more "dots" than just the type specimen.

> I think our nominal concepts transcend that gap.
> Take these two statements instead.

As I have said many times before, I think the "Nominal concept" is
functional equivalent to a "Name object" (1:1 correspondence), and I imagine
it as the primary unit of exchange for name-only data within TCS.  Stated
another way, I think that when we speak of "LC data wrapped in a TCS
burrito", we mean a TCS package that contains only Nominal concepts.

> "Aus bus Smith sec. [nothing filled in here] is the type species of
> the genus Aus Smith sec. [nothing filled in here]"  -- nominal concept
annotation

O.K., that's fine by me -- but let me ask this:  Where, in TCS, is the part
about "is the type species of" recorded?  I gather you would have two
nominal concepts (one for "Aus bus Smith SEC. nothing", and one for "Aus
Smith SEC. nothing") ... but do you then record the "is the sype species of"
relationship in a RelationshipAssertion?

If so, then my response is that I do not think this is the most elegant, or
even an appropriate use of RelationshipAssertion.  I believe that this
non-concept, purely nomenclatural connection between Name-objects "Aus bus
Smith" and "Aus Smith" belongs within LC -- not in RelationshipAssertion
(outside of LC)?

If I'm wrong about this assumed use of RelationshipAssertion to store "is
type species of" information, then I apologize for muddying the waters.

> "Aus Smith sec. Smith includes Aus bus Smith sec. Smith"
> -- original concept annotation

No problem here.

> With a bit of empathy, you could perhaps see that both statements
> describe a kind of broadly taxonomic relationship among the referents
> (i.e. the perceived taxa, a genus and a species) that are tied to
> the names "Aus" and "Aus bus". Clearly, the former statement is more
> immediately connected to the rules of nomenclature.

Right -- and I guess the problem we're having is that the nomenclaturalists
see the distinction as being profound enough that these two kinds of
relationships should not both be documented in the same way (at least I see
it this way).  One statement has to do with names, and the other statement
has to do with biological organisms.

> One might say: whether the sec. is filled in or not is a fundamental
> difference.

It sort of is, because I see a difference between "Aus bus Smith SEC.
[nobody]" and "Aus bus Smith SEC. [somebody, but I don't know who]".  The
first is not missing any information, because it applies to a name object.
The second *is* missing information, because it refers to a concept
circumscription (circle) drawn by somebody (inclusive of more than the
primary type specimen).  So, it's not clear to me which of these two things
an instance of "Aus bus Smith SEC. [nothing filled in here]" would apply to.

> But one could just as well look at the possibly very similar taxonomic
> implications of the two statements - even if the Codes don't demand
> such a view. Our perspective on the TCS would give merit to the two
> kinds of languages exemplified above (Code-prescribed relationships +
> concept relationships). It would also allow recognition of the
> difference of having a "sec." filled in or not, depending on the data
> available.

But does it make a difference between "Sec. [Nobody]" and "Sec. [Somebody,
but not sure who]"?

Here's why there is a distinction between these two:

When we want to record the fact that "Aus bus Smith" is the type species of
the genus "Aus Smith", there is no statement about any concepts involved.
It is a purely nomenclatural piece information, and it matters not what the
shape/size/position of any "circle" that would circumscribe a set of dots
is.

On the other hand, when we want to record the fact that some particular
specimen was identified by Joe Jones as "Aus bus Smith", there *is* an
implied concept circumscription (unless the specimen identified by Joe Jones
happened to be the type specimen of the name Aus bus Smith).

If I understand you and Jessie and others correctly, you would use a Nomnial
Concept TCS instance in both cases; even though in one case there is an
implied concept circumscription (we just don't know what it is), and in the
other case there is no implied concept circumscription.  I believe you are
correct that the information *can* be successfully modelled this way -- I
just do not believe that it is the optimal way to do it.

> That and why we SEEKsters won't encourage throwing around names
> and relationships without speakers attached to them is well established.

None of us wants to encourage that -- EXCEPT when there is explicitly no
implied concepts involved.  As soon as a name is applied to a specimen or
individual organism other than its primary type specimen, then a concept is
neccesarily implied.

> But, elegance aside, having nomenclature and taxonomy side by side yet
also
> sufficiently independent of each other is possible within the TCS.

If I didn't believe that statement, I wouldn't be spending so much precious
time on these emails!! :-)

Aloha,

Rich