[tdwg-content] What is an Occurrence? [what about the "token"]

[Richard Pyle]

> This was a bit of a "straw man" but I think what we would 
> both agree that annotating the identification to the "concept" 
> "as described by the key" would more accurately represent 
> the assertion that was made. It is as if there is pressure 
> to make documenting the identification process more "code 
> compliant" than making it accurately reflect what happened.

Yes, exactly!  The *SINGLE* most important thing we can do to reduce the
taxonomic ambiguity in our databases is to get people in the habit of
recording what field guide/monograph/key/whatever was used in making the
determination of the specimen's taxonomic identity.  Even if an expert
pulled the identification out of his/her head, s/he should document the best
published representation of the taxon concept that matches what the
identifier (person, not GUID) had in mind when making the determination.

Jim Croft once told me that he tried to get his users to do this many years
ago, but he simply couldn't persuade them to do this. (I think it was Jim
who told me this.)

There's such a huge difference in informatic value between "This specimen is
Aus bus", vs. "This specimen falls within the species concept of Aus bus as
circumscribed by Jones, 1950".  The latter sounds like a lot of extra work,
but in fact, all you need is one field labelled "sec", or "in the sense of",
with a drop-down list of publications that treated "Aus bus".  For most
field surveys & collections, you can probably find a single default
reference that would apply in 90% of the cases, and then tag only the
remaining 10% with a different reference, as needed.

The biggest problem I have with dwc:identificationReferences
(http://rs.tdwg.org/dwc/terms/index.htm#identificationReferences), is that
it allows many.  What to do, then, if two listed references present two
different concept circumscriptions (e.g., one sensu lato, and one sensu
stricto)?  Always fall to the strictest sense?

Personally, I think a "best practices" approach to this term would be "list
only the best Reference, unless it's absolutely necessary to indicate more
than one reference, from which a composite concept can be established".

> In my experience with my bugs and and some of the mammals, 
> the original descriptions and subsequent revisions are not 
> as informative as some in the community portray them. They 
> often do not serve as good guides as to what specimens are 
> instances of that concept and what specimens are not.

....so what, then, are the guides following?  Or are they presenting
original taxonomy within the guide itself?  If you can anchor the
identification to the field guide, that's 90% of the battle right there.
Later we can map the field guide to a mopnograph, or some other source for
the full concept definition.

> Also, that perhaps the Code should be revised to fit the 
> biology, rather than trying to get the biology and 
> related databases to fit the Code.

I don't follow.  Can you give me an example of what you mean?

Are you saying that the Code(s) should make rules for defining taxon
concepts, rather than just rules for establishing names?  I hope not! But if
so, then you might want to check out the Phylocode, which basically does
exactly that (to the extent that a clade is also a form of defining a taxon
cocnept).

Aloha,
Rich