Bibliography project
Jane McGary (Mon, 27 Oct 2003 10:11:36 PST)

Thanks to Boyce Tankersley for his thoughtful remarks on the bibliography
database project. My reaction is that Boyce is envisioning a reference
aimed more at the scientific user than at the amateurs I was thinking of
serving. I believe there are existing bibliographic sources on the Web for
botany, are there not? No sense in duplicating their effort; what I had in
mind was a parallel tool indexing books and journals not covered by
existing sites.

Boyce wrote, >In the bibliographic references I encourage a simple Yes/No
field to identify if this literature citation represents the first time a
scientific name was published.

That's probably a good idea.

The plant names table could be a single table with fields for taxonomic
fields.

Yes, if we decide to use tables.

Alternatively, the plant name table could be relatively elegant with a
link to a plant family:///genus/ index table to autofill the plant family name
when the genus name is selected (many to one relationship in order to
support more than one plant family (when will the plant family
relationships/names settle down?)). A second table would be linked in a
similar way to hold recognized synonyms. ... A third table could
cross-reference common names (a quagmire) but important for many people
attempting to find information.

I think the first suggestion in the paragraph above is too refined for our
purposes, and also it requires that the person doing the data entry make a
decision about family membership, which, as Boyce mentions, is more than a
lot of botanists can agree on. As for a second table of synonyms, I would
rather include the synonyms in the same table, or document, as the current
names, with a convention to direct the user to the current name, e.g.
"Cyclamen neapolitanum = hederifolium". As for common names, that is a
whole other project, albeit an interesting one for a linguist; I won't
volunteer for it, but I think we could run a fascinating data collection
through the PBS forum and perhaps Alpine-L. In any case, you could include
the common names in the same format as synonym cross-references, as many
indexes to books do.

Which field is chosen to be required depends upon what level of 'data
mining' is anticipated. Based upon a quick review of some of the
literature, I would recommend the genus name field.

Undoubtedly right. However, I think the data mining is more likely to be
genus+species. I'd be most likely to use such a tool to find information on
obscure species names that come attached to seeds I buy. (And that means
that we have to index the relevant genera in the Flora of the USSR, which
is where most of these unknown names come from.)

This brings me to another topic I've been thinking about: Where to draw the
line on inclusion. I suggest that the database be limited to bulbous (in
the broad sense) monocots. That would eliminate, e.g., Anemone and
Ranunculus, and part of the genus Iris. I'd also exclude the Agavaceae,
unless presented with a persuasive argument to the contrary. However, I
would like to include the Alstroemeriaceae, because I like them so much....

I am seeking expert help on this database problem and will report when I
have more information.

Jane McGary