[Nhcoll-l] collector & determiner identities

[Doug Yanega]

On 2/18/16 6:46 AM, Shorthouse, David wrote:
> I am writing to inquire if anyone knows of any best practices guides
> on how museum staff *ought* to record the names of collectors,
> determiners, and other agents. Do any of you have an *agents* table in
> your database?
We have a database of nearly 500,000 insect specimen records in 
FileMaker Pro. How we deal with our data depends VERY heavily on the 
nature of insect specimen data, which has qualitative and quantitative 
properties different from most other natural history collections.

The database has grown "organically" since its creation in 1999, with 
new fields and functionality added depending on changing demands. The 
"determination history" field was integrated at inception, while the 
"collector" field was one of the very last additions (it was a complex 
rationale, not meaningful for this discussion), so only a percentage of 
the most recent records contains collector names.

For collectors, the software automatically builds a picklist based upon 
existing entries in that field, and they follow a fixed format. There is 
no *external* file that resolves names in the list with identity; if we 
have labels such as "J. Pinto" and "J.D. Pinto", the picklist gives only 
one matching option - "Pinto, J.D." - and the data entry person has to 
resolve it themselves (so identity resolution is performed at the moment 
of data entry). The list of collectors contains an *overwhelming 
majority* of people who have no known "identities" to resolve - 
laypeople and students comprise the majority of names, virtually all 
unknown and/or forgotten. A single entomology class could have specimens 
collected by 20 students, virtually none of whom ever collected any 
specimens outside of that one class, and then the next semester there 
might be 20 more such students - and so on, dating back decades. Trying 
to figure out who all those people were is not only a hopeless task, but 
also of no value in terms of scientific uses of our specimen data. 
Moreover, only a VERY tiny number of names in the list of collectors 
overlaps the people who have determined specimen identities in our 
collection; for the ~500,000 specimens so far, there are only about 20 
names that overlap!

More to the point of your question, the "determination history" field is 
also formatted in a completely different manner, so it cannot be indexed 
or cross-referenced with the collector field, even if we decided it had 
some utility to do so. We opted for a single-field approach to 
determination history because (1) the alternative would have required 
the use of a bare minimum of 15 fields (determination, name of 
determiner, and year determined, each iterated 5 times) and (2) only a 
VERY small percentage of our specimens have a physical label indicating 
the name of the person who identified it (and only a tiny percentage of 
those indicate the year) - so nearly every one of those 15 fields per 
record would have been blank, and even if not, typically only 2 of the 
15 would contain data. For the small amount of extra effort involved, 
and to minimize proliferation of empty fields, it seemed simpler to do 
manual searches on a single field if and when anyone actually thought 
there was a need to make use of the data therein (you can sort and THEN 
parse, if needed).
> Have you attempted to link people names in your
> specimen databases to their unambiguous identities and, by extension
> to their scientific outputs like datasets & papers published?
As noted above, we have not done this, and I'm not sure it's especially 
meaningful or helpful, also as noted above. About the only such link we 
might ever wish to make would be to the names of the determiners 
(virtually all of whom were publishing scientists or at least known 
entities), and that could be done manually rather than trying to 
automate it.
> Last,
> does the ordering of determiner or collector names on labels contain
> any semantic meaning as it does for papers? That's something I have
> not yet considered & quite frankly scares me if this is important.
>
Certainly neither has any meaning for insect specimens, with the sole 
possible exception of types, as type labels SOMETIMES include the author 
names as they appear in the published description (even this is not 
standard practice).

Peace,

-- 
Doug Yanega      Dept. of Entomology       Entomology Research Museum
Univ. of California, Riverside, CA 92521-0314     skype: dyanega
phone: (951) 827-4315 (disclaimer: opinions are mine, not UCR's)
              https://urldefense.proofpoint.com/v2/url?u=http-3A__cache.ucr.edu_-7Eheraty_yanega.html&d=AwIC-g&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=CLFZJ3fvGSmDp7xK1dNZfh6uGV_h-8NVlo3fXNoRNzI&m=ZSRwCk-5uzKFAvT5xNzu95-fcCc4sCld4V9BGRfTzEU&s=ydrCZlugM_cgDWMYGCf6o72avagBNYl6HUwk_fBcYfo&e= 
   "There are some enterprises in which a careful disorderliness
         is the true method" - Herman Melville, Moby Dick, Chap. 82

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.yale.edu/pipermail/nhcoll-l/attachments/20160218/7d11d900/attachment.html