-------- Original Message --------
Subject:        map cataloging on OCLC--statistics
Date:   Tue, 17 Jul 2007 09:10:44 -0500 (CDT)
From:   Christopher Winters <[log in to unmask]>
Reply-To:       [log in to unmask]
To:     [log in to unmask]


Colleagues,

In July of 2001, when I was preparing my annual report,
I decided to see how the University of Chicago was doing
in getting map cataloging records to OCLC in comparison to
other institutions. I did not know about 'dt="map,"' so I
just did keyword searches for "maps," "charts," or "views,"
limiting by format and institution. Since the results made
us look good, I naturally included the table in my annual
report. I've repeated similar searches in every July
since, adding a few institutions, and switching in 2005
to 'dt="map"' (which, unlike those keyword searches, picks
up records with no subject headings, foreign-language
subject headings, or subject headings like "remote-sensing
images"). I've been persuaded somewhat against my better
judgment to share the results. Click on

http://www.lib.uchicago.edu/e/su/maps/TABLE_9.html

Do read the footnote, and please remember:

[1] This started as an exercise for an annual report.
Exercise appropriate cynicism.

[2] It is _only_ a measure of libraries' success in
getting records to OCLC. Thus,

[a] The many libraries that have had trouble getting
records to OCLC are underrepresented. The University
of Chicago, for example, has an online catalog that
at one time or another was producing 007 and 008 fields
and/or degree signs that could not be interpreted by OCLC,
hence a large number (maybe 2000) of our records were
never noted by OCLC. I know other libraries have had
even more trouble.

[b] Libraries that have mostly reported to RLIN are
excluded. When and if the RLIN records now on OCLC
acquire holdings information, the numbers will be
very different.

[c] How many records a library reports to OCLC is
in part a function of how it treats sets. The U of C,
for example, never analyzes sets of topographic maps,
but we try to analyze monographic series. Libraries that do
make records for individual top(o) sheets will in effect be
overrepresented (aren't there supposed to be ca. 100,000
separate USGS topos, counting all editions?). Libraries
that do not make separate records for monographic series
will be underrepresented (something like 5000 of the
U of C's records are for USGS MF, I, and GQ series titles).

[3] I hope folks at smaller libraries will forgive
my big-library chauvinism. If I've left out any
major non-RLIN library that has done a large amount
of map cataloging, sorry, and please let me know.

[4] The table probably needs dozens of footnotes. I
know some of the large one-year jumps are due to recon
or other special projects; perhaps all these figures
should have explanations. The U of C's substantial
productivity considering that we are a collection with a
staff of FTE 1.5 is due in part to the fact that for many
years after our old homegrown catalog could (finally) deal
with map-format records, our cataloging department would
only let Map Collection staff "finish" matching LC records.
Eventually, restrictions on what we were allowed to do
were lifted, and we eventually upgraded these records.
It's a lot easier to upgrade an existing record than
to do original cataloging or even to download a matching
record. I suspect other libraries have similar stories that
explain what look like anomalies.

[5] I don't know what to do about the fact that atlases
get picked up in this search. Any ideas? The fact that
the word "atlas" is used in most western European and
Slavic languages as well as in Arabic (big exception:
Italian atlante) does allow one to do some exclusion,
as does the fact that "Atlases" can be a subject heading,
but plenty of atlases aren't called that and just have
"651  0 Place name--Maps" subject headings. Also, the U
of C treats some unbound "atlases" as sheet maps, as, I'm
sure, other libraries do, so the distinction isn't always
clear.

Hope you find this material interesting. Please don't
hesitate to let me (or everyone else) know about
mistakes. I realize that this sort of thing is
potentially a bit sensitive, and I hope it was okay
to share it.

Chris Winters
Bibliographer for Anthropology, Geography, and Maps
University of Chicago Library