Richard Pyle deepreef at BISHOPMUSEUM.ORG
Mon Jun 21 10:26:18 CDT 2004

Richard Petit wrote:

> The
> reply to me indicates that my "feedback" was forwarded to the museum
> supplying the fossil name but there is no indication that  ITIS or other
> sources were notified of errors.

Was the fossil name, by chance, Cancellaria tholoensis Ladd, 1934?

The reason I ask is that when I downloaded my email 5 minutes ago, I
received a message forwarded from Donald Hobern of GBIF informing us of
feedback you had sent them about one of the records in our DiGIR specimen
resource (BPBM-MO 208256); providing the author and year when we lacked that
information in our specimen database.  For this, I am very grateful for the
feedback (as well as your correction of "Cancellaria bifusciata deshayes" to
"Cancellaria bifasciata Dshayes, 1830" -- to our credit, in that case we did
have "deshayes" in the Author field, even though we had spelled it
incorrectly and failed to capitalize it, and we had misspelled the species
epithet). I've made the necessary corrections to our DiGIR provider
resource, and if you go back to the GBIF site, I *think* you'll find that
the correction is already reflected there (unless they cache data, in which
case the correction will be reflected the next time they refresh the cache).

So, the good news is that the feedback system seems to be working the way
that it is supposed to.  The bad news is that yours are the first feedback
corrections we've received in the several months that we've had our nearly
600,000 specimen records available through our DiGIR provider.  I know for a
FACT that there are many, many more errors than this in our data; so it
would be nice to know that when experts discover these errors (as you have),
they would care enough to take the time send the corrections back to the
source.  We ALL benefit from that.

Now, to address some other comments in your email:

> On June 17th I posted a note concerning "authoritative sources" stating my
> opinion that ITIS could not be so considered.

I'd be interested in knowing how you define "authoritative" in this case.

> A reply was posted from Dr.
> Lane (GBIF) advising how to have errors corrected.  I then sent in four of
> five corrections and received a reply from ITIS (will forward to Dr. Lane
> off server but will send to anyone interested) in which I am advised that
> "it is important that the search interfaces continue to list these names
> since a user may recognize them as misspellings for names of taxa in which
> he or she is interested."  What about the multitude of people who will not
> recognize these as errors?  Why not show them to be errors if
> they must, for whatever reason, continue to be listed?  Why perpetuate

So long as the errors are flagged as such, and someone encountering those
errors is re-directed to the corrected information, I COMPLETELY agree with
the statement in the reply to you from ITIS.  It *IS* important that search
interfaces continue to list these names; both so they can be corrected by
someone who discovers them as erroneous (as you have done), *and* so that
they can be cross-referenced to the corrected version.  If the error entered
the system in the first place, it's quite likely that it did so because it
was transcribed from another source.  That means that the error likely
exists outside the context of the database itself, which means that someone
may encounter it in the "real world", and want to do a search on it, and
ideally thus be informed of the correction.  If the erroneous name is not
maintained in the database, the database cannot provide very helpful
corrective information to someone who might encounter the error in a
different context, not realizing that it is an error.

Going back to my earlier diatribe, the search database needs to maintain the
fact that Gray, 1831 used the text string "Holocanthus" to refer to the
taxon name "Holacanthus Lacépède 1802", so that if someone picks up Gray's
publication and sees the misspelled name and wants to do a search on it, the
database will be able to redirect the person to the correct information.

> The primary errors I sent notice of were a non-existent name; the type
> species of the genus Cancellaria being attributed to Pilsbry, 1940 instead
> of to Linnaeus, 1767; one species listed under three genera (one a junior
> objective synonym of another); and the listing of a Miocene species.

I need to get a sense for what data resource you searched to find these.
Based on the "feedback" bit, I assume you were searching the GBIF DiGIR
portal.  If that's the case, then some explanation might be in order here.
That portal is a window on specimen data. As such, it is provided by various
Museums around the world, based on their in-house specimen databases.  It is
probably very safe to say that most of these data resources have NOT yet
been cross-checked against more "authoritative" sources (such as ITIS,
Species2000, IPNI, etc.).  Therefore, errors discovered on the GBIF portal
are NOT necessarily reflective of ITIS.  There is no doubt that errors exist
in ITIS and other authoritative resources.  In my experience, Bill
Eschmeyer's Catalog of Fishes represents one of the highest-quality
taxonomic resources on the web (in the world).  It covers over 56,000
species-group names and over 10,000 genus-group names, including details on
original descriptions, type locality, type specimens, and alternative
variations of how the name has been used over the centuries. And it's DAMN
hard to find a mistake in it.  But Bill himself will be the first to admit
that mistakes definitely exist.

The point is when you have a relatively small group working on building and
proofing a taxonomic database containing tens of thousands or hundreds of
thousands of names, it's simply not possible to eliminate all the errors.
Likewise for specimen data, and their correlated taxonomic determinations.
Speaking from nearly two decades of first-hand experience, it is a
MONUMENTAL effort to correct the taxonomy on specimen databases containing
large numbers of records.  The issues are much more complex than simply
correcting misspelled names and replacing synonyms with names in current
use.  So the dilemma that every natural history institution faces is this:
Do we hold back our data close to our chests until we've achieved near
perfection in the data?  Or, do we expose the data to the world, warts and
all, hoping that the sum of the benefit exceeds the sum of the harm.

For a variety of reasons, we've decided to go with the latter approach.  But
a big part of this decision assumed that taxonomists would be willing to
provide feedback to us whenever they encounter errors (Big kudos go to
Richard Petit for caring enough to do just that).  Some might let the fear
of embarrassment prevent them from risking posting a dataset that has not
been 100% verified. We think the benefit of enlisting the help and feedback
from potentially thousands of taxonomists to help correct errors in our
database only works to EVERYONE's advantage.

> The Academy of Natural Sciences of Philadelphia maintains two excellent
> databases of gastropods (one for the Indo-Pacific and the other for the
> western Atlantic).  These databases are corrected whenever errors
> or changes in systematic position or assignment come to their attention.

The folks who are building these databases at ANSP, as well as others like
them (Bill Eschmeyer, Chris Thompson, and many others) deserve a HUGE round
of applause, and tremendous amount of respect from every taxonomist on the
planet.  They are the ones who had the right vision long ago; back when
others were saying it was impossible. In my view, these people are
under-appreciated.  The same could be said for ITIS and Species2000 and
others who are least trying to establish authoritative lists, and get them
in the hands of whoever needs them.

To me, the next step is clear, and it comes back to the issue of GUIDs.  The
time has CLEARLY come to coordinate these efforts, synergize the work that
has already been done, and prevent the future duplication of effort.
Furthermore, for the benefit of scientific progress, the time has come to
transition these efforts from closed, in-house projects with the pie being
sliced up among various taxonomic groups and/or competition between
different groups; to a broadly collaborative and cohesive effort that the
entire taxonomic community can directly benefit from, and rally behind.
Central to such an effort would be a robust feedback system that would allow
essentially anyone to submit corrections (and get credit for those
corrections).  Sort of a global peer review, if you will. (Well, not sort
of -- *exactly* a global peer review.)

It seems to me that GBIF is clearly headed in that direction, and in my
personal opinion, we should all be rooting them along, and doing whatever we
can to transform the vision into reality.  My hope is that "critical mass"
will be achieved in the October meetings in New Zealand, where most of the
major players will likely converge. The time is right -- I can just smell
it.  Doug Yanega: where are you?!? Rally the troops!!! ;-)


Richard L. Pyle, PhD
Ichthyology, Bishop Museum
1525 Bernice St., Honolulu, HI 96817
Ph: (808)848-4115, Fax: (808)847-8252
email: deepreef at

More information about the Taxacom mailing list