cf./aff. - indeterminate determinations, certain uncertainties...

[Julian Humphries]

Jim Croft says:
>In our database design deliberations we discussed the arcane runes that
>taxonomists feel compelled to inscribe on determinavit slips in what
>appears to be a strict formal ritual, but we could not discern any
>consistent difference in practice in the use of ?, cf. or aff.  .

Likewise.  The whole point is that despite the protestations of individual
systematists, there are thousands of these notations in catalogs for which
we have no recourse to the intent of original determiners.  In ichthyology,
it was common practice to use pencil rather than pen for "questionable"
identifications.  That was a real challenge to decide how to handle on data
entry!  Other forms of uncertainty include quotations (or braces) around
names, English equivilents of cf/aff, and quaint designations such as "blue
spotted variegatus" (for specific epithet).

>Accordingly we decided to transliterate all these to a single qualifier
>flag: '?',  which seemed to have just right nuance of doubt.  This
>logical and sane unilateral decision was greated with howls of derision
>by real taxonomists who insisted that *their* cf/aff had real, distinct
>and immutable meaning and it was just other taxonomists who did not know
>what they were talking about.

At the MUSE Project we came to exactly the same decision (and have met with
similar protestations).

>We did managed to convince people that this cf/aff thing was an
>attribute of the identification and quite distinct from
>agg./sens.lat./sens.str. which is an attribute of the taxon and that
>neither were an attribute of the name.  (Actually I am not really sure
>they understood or believed this...)

Again, our design decision matches exactly.  The field "Questionable" is
use to refer to the identification.  All questions about taxa (and/or
names) appear in the taxonomic authority files.

.
>Back to the cf/aff thing.  The argument we are being confronted with now
>concerns the rank or position in the name string this uncertainty flag
>if to be applied.  To simplify the database (and in my opinion, to reflect
>reality) we want to ally it to the lowest level of the name in the
>determination: to the family or genus if that is far as you could get,
>or to the variety or forma if that was appropriate.

One more time, (it is almost eerie how identical our design decisions were,
either we both converged on the right solution or we have similar cockeyed
views of the world) this is how we used the questionable identification
field.  It applies to a question of identification at the lowest level
assigned to the taxon.

However, we also felt that because the "real" identification of a specimen
is a binomial (minimally, of course varieties, forms, etc might apply as
well) that any identification that did not include a specific epithet was
questionable.  You might be certain that the specimen is Eucalyptus, but if
you don't know what species it is, then there is a question about the
identification.

Re Dick Jensen's comments (on the needs for rules vs chaos), he is exactly
right.  But, it is also clear that we have no way to know how those rules
were interpreted for catalog and label entries.  For text based database
systems it seems a consistent rule to view all such qualifiers as forms of
questions about identification.  Images of labels can be used for those
interested in interpretation of original determiners intent.

Happy returns, Julian

Julian Humphries
The Vertebrate Collections and The MUSE Project, Cornell University
Building 3, Research Park
83 Brown Road
Ithaca, NY  14850

Voice: 607-257-8143
Fax:   607-257-8109
Email: jmh3 at cornell.edu