Unique and Stable Numbering Systems versus Nomenclature

[David Nicolson]

On Thu, 15 Mar 2001 12:26:32 -0500, christian thompson
<cthompson at SEL.BARC.USDA.GOV> wrote:

>But unfortunately a "unique & stable numbering system" probably won't ever
>happen  as the community will not support such. What David doesn't realize
>is it isn't numbers that is the requirement for "unique & stable.". It is
>the community acceptance of a system to make anything "unique & stable."
>Zoological Nomenclature is an International Standard which should give you
>"unique & stable" identifier (keys, etc.), but it fails because people will
>not follow it nor allow its modification to better provide the "unique &
>stable," etc.  Set up a registration system like the Bacteria people did,
>and you get "unique" names. "Stable" fails because of taxonomic progress
and
>classification paradigms.

Only if naming standards account for all groups (i.e., not just animals, not
just plants, etc.) can the hope of "unique" names be approached. I believe
the bacterial system specifies that you can't propose bacterial names
already used for plants/animals, but I am not aware of that elsewhere.
Stability is indeed a bugaboo, and not to be legislated re taxonomic (i.e.,
subjective) issues. The stability I am suggesting is not in what name is
applied (nomenclatural stability), but in how you can cross-reference that
name even if the name is in some way altered (changing terminations, fixing
other errors, etc.). As Chris suggests below, there is no way to be sure
someone does not have the wrong number, just as there is no way to be sure
someone has correctly applied a name (barring vouchered specimens &
significant effort). I don't think it is up to anyone to take responsibility
for someone else's misuse of names.

>
>Social Security Numbers are great unique and stable identifiers (keys) for
>people who live and work in the United States of America simply because the
>US Government forces everyone to get one from them.  In theory, ITIS
>taxonomic serial numbers (TSN) should also work, but given the traffic we
>had on TAXACOM about a year or so ago no one seems willing to accept ONE
>taxonomy set by ITIS.

I'm not suggesting ITIS is the end-all in this regard. Certainly until it
develops a multiple classification system that limitation will remain. Even
then I am sure reservations will be plentiful. ;-)

>
>The other point I wonder about David's remark is why  "keystroke errors"
>only happen to names, but somehow numbers don't have the same problems.
>Given the number of incorrect telephone calls I receive, I suspect people
>will also make mistakes with TSNs.

Actually I was trying to focus on the possibilities of inter-database
communication, where typing the numbers ought not to be occurring.

>
>David has also included the identification problem in his discussion. So
>let it be said, unique numbers do not solve in anyway the identification
>problem. If some one identifies a fly as Musca domestica Linnaeus (ITIS TSN
>150251), but it is really Fannia canicularis (Linnaeus) (ITIS TSN 151350),
>when they record their identification they will use number 150251. David
how
>does you "unique and stable" numbering system fix that? The only way to
>solve the identification problem is to require that every observation is
>vouchered by a specimen which has an unique code, etc. TAXACOM went thru
>that one too in reference to GENBANK, etc.

Understood. Still, I was focusing on the results of these realities for
lists of names... The reality of having to deal with a system that may
contain multiple records with identical names, and how those can reliably be
differentiated within and between data systems. Dealing with that reality
can be approached, even if solving the underlying factors that create it
cannot.

>
>Yes, over the past 250 years, we have made lots of misidentification,
>re-used the same name (homonymy), etc., but some how, names like Diptera,
>Muscidae, Musca domestica Linnaeus, have remain fairly stable and are
>largely unique. They have served successful (so far as you consider
>Systematics a successful science) to be "keys" to information. So despite
>ITIS's TSN they still are much more universal and useful for biodiversity
>information exchange.  AND once GBIF is done they will be!

Indeed, for hands-on purposes I suppose they will remain so... I'm not sure
about it in the context of reliable cross-database communication.

>
>Smile :-)

Agreed! Heck, the thread hasn't even gotten into the whole "concept"
problem...

Dave