[Taxacom] global names index gni as dead end?

Donat Agosti agosti at amnh.org
Mon Mar 7 04:10:49 CST 2011

Dear Rich
 I am glad that I am wrong. At the same time, from a users point of view, I
don't care what's underneath, and thus it stays a dead end. I just see the
dead end and wonder about the purpose for this.
As you point out, why to expose something that has a particular purpose, but
not the one to be shown to the public. You point this out below already and
I agree with your position.

IMHO GNI contributes, as is seen online, only to the increasing flury of
name servers. Why not have GNUB as the place where Google can get names and
thus avoid this dead end? I never use really any of COL and other services,
because they are notoriously behind the actual situation and normally, if
there is a something on COL, there is a database that is accurate - I just
have to find it and keep it delicioused, or let our machines do it.

Good dives


-----Original Message-----
From: Richard Pyle [mailto:deepreef at bishopmuseum.org] 
Sent: Monday, March 07, 2011 12:49 PM
To: 'Donat Agosti'; taxacom at mailman.nhm.ku.edu
Subject: RE: [Taxacom] global names index gni as dead end?

Hi Donat,
> I always hoped, that especially institutions like GBIF, one of the main
> in the field of biodiversity informatics, would push that names are linked
to a
> publication in which links are provided to the materials examined that
> to understand the species concept used in this particular usage.
> But no.
> I am hoping, I am wrong.

Your hopes are realized!  You are wrong.

First of all, name-strings on GNI are not dead-ends, they are linked to the
source records that provided the names.

But more fundamentally, you seem to miss the purpose of GNI. I would have
hoped that you, of all people, understood the purpose of GNI, given all the
conversations we've had on the topic.

GNI was deliberately established as an index of text strings purported to
represent scientific names of organisms.  It is an "index" in the strict
sense, like what you find in the back of a book.  In a book, an index is not
the same thing as a glossary -- definitions are not provided -- only page
numbers to where in the book the text string appears.  In the case of GNI,
the internet is the "book", and a URI to a database containing the text
string purported to represent an organism name.  GNI actually does more --
it includes a parsing service that can pull apart name components, such as
genus, species, authorship, and other qualifiers. 

Within GNA, GNI is only one component.  Another component -- GNUB -- does
not yet have a public interface (except in the form of services, such as
ZooBank).  GNUB is more like the glossary of a book -- meaning that the
names are accompanied by "definitions" (links to usage instances, such as
published treatments -- including original descriptions as well as other
subsequent usages).  GNI will serve the function of a bridge between the raw
text strings (as they appear in online databases, including things like
OCR'd text in BHL), and the well-defined and "curated" taxon-name usage
instances in GNUB.

When we originally discussed GNI, I advocated not even having a website with
a user-interface, because people would undoubtedly misunderstand the purpose
of the GNI.  Although my prediction has come true (case in point), I have
actually come to realize that the UI for GNI is useful, even without the
GNUB being functional yet. I use it as a tool to cross-link my text-string
names against other datasets - like ITIS.

Aloha (on expedition in Maui),


__________ Information from ESET Smart Security, version of virus signature
database 5931 (20110306) __________

The message was checked by ESET Smart Security.


__________ Information from ESET Smart Security, version of virus signature
database 5931 (20110306) __________

The message was checked by ESET Smart Security.


More information about the Taxacom mailing list