[SEEK-Taxon] RE: LinneanCore Group Work

[Paul Kirk]

Just a few point to chew over ;-)

(1) - Taxonomy is a prerequisite to any nomenclatural novelty thus most
nomenclatural novelties are based on prior taxonomic concepts - a nom.nov.
to replace an illegitimate name in the same genus with only the holotype
representing the species has no changed concept (agreed?) - it might,
however, have a different set of characters used to circumscribe the taxon
or new characters may have been discovered which were previously overlooked.
Some might interpret this as a changed concept - perhaps it is, but there is
no part of the Code (ICBN) which says I must cite specimens, descriptions or
whatever when I introduce this nom.nov. so under those circumstances how
would a new concept be described? It would be a naked name usage, as most
name usages are, outside the narrow field of systematics.

(2) - Another 'event' to consider; see:
http://82.43.123.182/IndexFungorum/Names/SynSpecies.asp?RecordID=415945 and
http://82.43.123.182/IndexFungorum/Names/NamesRecord.asp?RecordID=462939

A comb.nov. and a misapplication in the same publication.

(3) - w.r.t. names based on single specimens being "effectively useless
constructs" - I'm told that half the names in zoology are based on single
specimens. Just an observation.

(4) - I'm now not sure whether the Glossary of Terms in Bionomenclature will
help our deliberation. However, I promised to make it available so here it
is - see:

http://82.43.123.182/LibriFungorum/SearchResult.asp?ItemID=9

please DO NOT advertise this unofficial site as it's not complete, the
external 'contributors' have not been contacted and there is some more
content to add.

Cheers,

Paul

-----Original Message-----
From: Richard Pyle
To: franz at nceas.ucsb.edu
Cc: Kennedy, Jessie; Jerry Cooper; ghw at anbg.gov.au; lblaine at atcc.org;
G.Hagedorn at BBA.DE; nozomi at biol.tsukuba.ac.jp; p.kirk at CABI.ORG; bti at dsmz.de;
dhobern at gbif.org; s.hinchcliffe at kew.org; garrity at msu.edu; Kukla, Robert;
C.lyal at nhm.ac.uk; yjong at science.uva.nl; weitzman at si.edu;
rlmoe at uclink4.berkeley.edu; cwilson at usgs.gov; sblum at CalAcademy.org;
seek-taxon at ecoinformatics.org
Sent: 15/11/04 09:15
Subject: RE: [SEEK-Taxon] RE: LinneanCore Group Work

> - that my example 1 wasn't strictly nomenclatural, but primarily
> a concept issue
> (let's see - no types, no diagnosis, no constituents - but still
> a concept);

I view classification issues as concept issues, because you are talking
about information associated with organisms (i.e., inferred/implied
relationships).  Nomenclatural issues involve information associated
with
Code-compliant name-strings.  Creating a "new combination" of
Genus+species
epithet is a Code-governed nomenclatural act (in botany), and would
likely
be treated as a separate instance in LC.  A partiucular re-usage of a
previously existing combination is a usage instance -- which means it is
not
"nomenclatural" in the sense that Code rules must be applied.  LC (I
don't
think) intends to treat repeated usages of the same combination as
different
LC instances.  Usages may or may not belong in TCS, but not in LC.

> - that most (!) zoolgists view "replacement names" and botanists
> view "nomina
> nova" as purely nomenclatural (in response to my question 2 which
> you later
> answer with "yes" [it's concept-stuff] as well - likely in
> disagreement with James);

Wait -- I'm confused. I thought I answered that "replacement
names"/"nomina
nova" were primarily nomenclatural acts (which might have concept
information in addition to nomenclatural information)????

> - that you like Berendsohn's vagueness, but are yourself not
> vague enough to
> move on the "a type does not make a concept" position - good luck
> finding the
> answer in Berendsohn's papers;

I don't understand this statement. What does "move on the 'a type does
not
make a concept' position" mean? Don't read too much into this issue.
The
only reason I raised the issue of "concept consisting of only a single
specimen" as a "useless" concept, was in the context of justifying the
perspective that primary type specimen designations could be viewed as
name-object data with philosophical legitimacy.  This is kind of moot,
though, because so far two members of the "name-nerd" camp (myself and
Sally) have expressed preference for primary type designations to be
dealt
with outside of LC.  The point is, my only reason for suggesting that a
"concept of one specimen" was a useless concept is dangerously close to
becoming moot.

> - that your notion of concepts is more liberal than Jessie's,
> except for that
> bit about types which you consider exempt from the criteria
> sufficient to create
> a concept,

Read my answer above.  I get the impression that you misunderstood (or I
mis-represented) my perspective on the issue of "primary type specimen
as a
concept".  This point exists only in the context of explaining the
rationbale for why some might think that primary type designation
belongs
within LC.  I certainly have no problem with anyone who wants to
extablish a
"taxonomic concept" whose scope is limited to a single specimen.  I just
think that such concepts are effectively useless constructs.  They
amount to
the statement: "This specimen is the only individual that belongs in
this
taxon". For what biological purpose would one find such a taxon
circumscription useful for communicating ideas?

> but then you (correctly, in my view) back-paddle
> somwhat "I would
> slightly prefer to surrender this one point of universe overlap
> over to the concept side";

No back-peddling.  I still maintain my original points (however poorly I
might have communicated them):

 - One can make a legitimate case that the designation of a primary type
specimen is nomenclatural information, because it is governed by codes
of
nomenclature.

 - One can make a legitimate case that the designation of a primary type
specimen is concept information, because it involves an actual organisms
(not just name-objects); and also because all other applications of
name-objects to specimen-objects (including all secondary types) fall
into
the universe of concept information.

 - Together, the above two points underscore a point I have made several
times that the only real *potential* point of contention between
name-object
data and concept-object data is with regard to primary type specimens. I
emphasized "potential" because there doesn't necessarily have to be
contention.  So far, there has not been, as at least two of the
name-nerds
(myself & Sally) have expressed a slight preference and/or willingness
to
keep all name-object associations with specimen-objects in one place --
which would clearly be outside of LC.

> - that the LC should handle nomenclatural things within the TCS
> (even though -
> as I understand - the LC rejects the sec. annotation which is
> central to what I
> call TCS's goal 1);

UGHH!!!  LC does not "reject" SEC. annotation. Rather, SEC. annotation
falls
OUTSIDE the realm of LC (but still within the realm of TCS). LC should
deliver all of the information that the TCS group intended to include
within
the "NameDetailed" element, and should include NONE of the information
contained elsewhere within TCS.  LC should be a small subsection of TCS
(the
subsection that deals with Names).  LC and TCS should be *perfectly*
comlementary (i.e., there should be 0% true redundancy between LC and
TCS).
All apparent redundancy (e.g., dual references to the "Smith, 1949"
publication for an instance involving the name "Aus bus") should
actually
represent two separate pieces of information (one as an attribute of the
name-object "Aus bus Smith 1949", and the other as an attribute of "Aus
bus
Smith 1949 SEC. SMith 1949"). Other apparent redundancies (e.g. "Rank")
should either be clearly identified as non-redundancy (i.e., a
concept-Rank
has a different meaning than a name-Rank), or they should be collapsed
into
a single element within TCS (either inside, or outside, of LC).

> - your breaking-down of my examples into nomenclatural vs.
> concept information
> is completely redundant! there is not a single string of
> information in the
> nomenclatural part that you don't also represent in the concept
> part;

EXACTLY!  Because the "redundant" part is the part that is within LC.
LC is
(in my vision) within TCS, so everything that is within LC is, by
definition, within TCS.  So, for the TCS instance of "Aus bus Smith 1949
SEC. Smith 1949", the "Aus bus Smith 1949" falls within the LC
sub-section
of TCS (i.e., within the Name element); and the "SEC. Smith 1949" falls
outside the LC subsection of TCS.  The only reason they appear as
redundancies in my previous email is that I deliberately included the
Name
information *within* the Concept information to make it easier for a
Human
reader to understand what I was talking about.  If you got the
impression
that I was suggesting that the "Aus bus Smith 1949" part of  "Aus bus
Smith
1949 SEC. Smith 1949" occurs separately within LC and within TCS, then I
obviously failed in my attempt to communicate.

So, one more time in an effort to help clear the muddy water:

In the concept-object represented by the human-readible notation "Aus
bus
Smith 1949 SEC. Smith 1949", the information represented by the "Aus bus
Smith 1949" part would exist in TCS only once -- INSIDE the "Name"
element
(i.e., within the LC sub-section of TCS); and the information
represented by
the "SEC. Smith 1949" part would exist in TCS only once -- OUTSIDE the
"Name" element (i.e., NOT within the LC sub-section of TCS).

> I assume
> you understand that two concepts can have multiple kinds of
> relationships to
> each other (some type-related, others set-theory related)?

Of course!!  I've made reference to this fact several times in email and
on
the LC Wiki.

> I see your
> distinction here as one of the stronger arguments to ABANDON the
> nomenclatural/concept-related distinction; well done, I suppose;

I don't understand what you mean by this. Can you please explain?

> - "Stated another way: Aus bus Smith 1949 is a fundamentally
> different unit of
> information than Aus bus Smith 1949 SEC. Smith 1949." I had to quote
this,
> because it has a bit of a PhyloCodian ("philosophical
> correctness") ring to it;
> the fundamental difference to you is "just the type - no concept
> yet"; that is
> only YOUR view Rich, not James', not mine, not Jessie's, not
> Berendsohn's.

Yes, now I am certain that I have failed in communicating my point.  I
will
try a different approach.

When I go into our entomological collection, I see trays with pinned
objects.  Each pin skewers two objects.  One object on each pin is an
insect
specimen, and the other object is a paper label.  I would not have any
difficulty seeing the difference between these two classes of objects
(specimens and labels).  If I am correct in presuming that you, James,
Jessie, and Walter would all share the same perspective in seeing two
unambiguously different classes of objects (specimens and labels), then
we
are all of like mind.  Just as a label exists as some sort of paper-like
foundation with ink characters imprinted on it, so too does a scientific
name exist as a string of characters (represented as ink on paper, or as
digital Unicode characters stored electronically).  Neither the label
nor
the scientific name is manifested as a biological organism or set of
organisms.  The specimen on the pin is a biological organism, and a
taxonomic concept is a circumscription of 1 or more (usually "more")
biological organisms.  Neither the specimen nor the taxonomic concept is
manifested as ink printed on paper, or as digital Unicode characters
stored
electronically.

I think the issue you seem to be hung up on is "just the type - no
concept
yet".  This relates only to the question of whether the designation of a
primary type specimen is represented in the schema within the LC (=Name)
subsection of TCS, or whether it's represented in the part of TCS
(outside
the LC subsection) where other kinds of specimen-concept linkages are
established. Don't read too much into that part of the discussion --
this
conversation has become more complicated than it needs to be. I don't
know
exactly what "PhyloCodian ('philosophical correctness')" means, but it
doesn't seem to have any relation to this discussion.

> On
> the surface, Aus bus Smith 1949 doesn't exist outside of a
> publication context.
> Berendsohn thought that was fundamental to him, in order to
> identify Smith's
> 1949 mention of Aus Bus uniquely. Since you already above
> conceded that without
> mentioning types (my examples 1 & 2) one can have concepts, but
> then maintain
> that solely mentioning types doesn't imply concepts (as if those
> mentions didn't
> also exist only in print), I would venture that few will follow
> your position;
> it is actually quite idiosyncratic;

It is very clear from the ideas you have ascribed to me that I failed
miserably in communicating effectively -- because all of these
statements
that you claim represent my views, absolutely do not represent my views.

I agree with the statement "Aus bus Smith 1949 doesn't exist outside of
a
publication context", because I believe that "Smith 1949" is intended as
a
proxy citation of a publication object (authored by someone with the
last
name of "Smith", and published sometime during the 1,949th revolution of
the
Earth around the sun following the puported birth of Jesus Christ). So,
by
your comment above, it seems that Walter & I are of like mind on this
point.
The rest of your words quoted directly above make no sense to me, and
certainly bear no resemblance to my own views.

> - that because some aspects of Smith's 1949 "Aus bus" fall under
> the Codes'
> rules whereas none of Pyle's 2000 "Aus bus" would, PARTS of Smith
> don't receive
> a "sec.", but Smith 1949 still receives a "sec."; again - ALL
> NOMENCLATURAL
> NOVELTIES ARE PUBLISHED BE SOMEONE - that's the TCS's criterion;
> the purity of
> nomenclature is threatened in no way by Berendsohn's concept approach;

Which part(s) of the above is a quote from me, and which are your words?
If
it is all a quote from me, then I don't understand what I was trying to
say
(I'll go back and see if I can find it in my email, and see it within
context).

> - that you cannot follow my goal 1 / goal 2 distinction; in a
> nutshell: if we
> adhere only to goal 2 (communication about nature) then some experts
will
> definitely not consider certain types of information to be
> concepts (you're one
> of them, Rich), yet the greater significance of achievieng 1 (precise
> information management - tying name uses to authors) forced the
> TCS to ignore
> those objections;
>
> - then you do a 360 on your previous view: "I am very sympathetic
> to this point
> of view, because in my own data model (Taxonomer), I effectively
> take the stand
> that Name+Publication=Assertion, and Assertion=Potential
> Concept;" THEN you
> reserve again: "there are strictly nomenclatural connections between
> Name+Publication (as governed by Codes of nomenclature) that have
> absolutely
> nothing to do with Taxonomic Concepts (which I define here as abstract
> definitions of sets or circumscriptions of organisms);" so within
> one paragraph,
> you promote concepts (1) as "name sec. reference" (that's goal 2,
> btw) and then
> (2) as "abstract definitions of sets of organisms [more than just
> the type!, I
> suppose]"  (that's goal 1, btw); which is it, Rich, and which
> should it be for
> the TCS?;

Yes, indeed it is clear that somewhere in the stream of information flow
from my brain to my fingers to the internet to the monitor on your
computer
to your eyes to your brain -- something went terribly awry.  I suspect
it
happened between my brain & my fingers.

> - that "Indeed, if [I] dissect [your] arguments in this context,
> many of them
> run counter to what [you] express (and still believe) through
> Taxonomer;" - this
> is because in Taxonomer you are open to Berendsohnian vagueness,
> but when it
> comes to the LC/TCS interaction you are not;

Not correct.

>    In summary, you have not convinced me at all (or yourself in
> Taxonomer, for
> that matter) that there is anything "inpedendent" about
> nomenclature that would
> take a serious hit if represented as a concept. Or about
> nomenclatural relations
> that as represented as "sec." relatioship assertions. Look at
> Linnaean taxonomy
> as "concepts 'light'," and at the Codes as "concepts 'light'
> manuals." Forget
> the objective/subjective distinction, it is irrelevant here (and
> probably wrong
> too). Then your position about concepts will become consistent
> with your own
> practice, and consistent with the TCS.

Nico, I hope that my attempts above to clarify my position have been
more
successful than my previous emails evidently have.  Most of the things
you
seem to attribute to me have not been correct. I'm willing to blame
myself
for the misunderstanding.  Let me know where we stand after you read
this,
and whether I've been any more successful this time in communicating my
points.

Thanks,
Rich