Some anecdotal experience on OCLC numbers..
As a person who used to respond to ILL requests that were mostly based on
OCLC#s, I can tell you that the percentage of them that are imperfect
identifiers is significant. This may be skewed in the ILL situation, but I
know I made a regular report of OCLC #s that did not match up between our
catalog and WorldCat records. (I don't know how this happens, I'm not a
cataloger.)
For example, search WorldCat for these OCLC#s (aka accession number in
their search form, for those unaware,) and see that the (accession) number
in the record doesn't match:
34356753
18870153
45748884
WorldCat is doing some kind of behind-the-scenes look-up, and those
cross-references are not publicly available; I know of no way to do a
reverse look-up of all numbers associated with one record. So, there may
or may not be issues there, depending on how you obtain your numbers. Get
them all consistently from WorldCat, fine, but mix in other sources, and
you have a problem.
Brian Stamper
The Ohio State University Libraries
Scholarly Resources Integration
610 Ackerman Road Rm. 5833
Columbus, OH 43202-4500
(614) 247-8415
stamper.10_at_osu.edu
On Fri, 08 May 2009 13:13:53 -0400, Jonathan Rochkind <rochkind_at_jhu.edu>
wrote:
> I am not a lawyer, but I'm pretty damn confident that you do not need
> OCLC's permission to use OCLC control numbers however you want. For
> reasons I explain in that blog post.
>
> Certainly OCLC numbers are imperfect as manifestation identifiers. Any
> service of bibliographic identifiers is going to be. Even ISBN is. But
> I doubt that any new system you try to come up with is going to be less
> imperfect than WorldCat at avoiding duplication etc. I'm curious what
> the "allowable" duplicates are in WorldCat though? If there are _many_
> of these, that might change my opinion about the suitability of OCLC
> numbers for this task. But mistakes in WorldCat don't, mistakes will
> happen in ANY system.
>
> We've already got OCLC numbers. It's an expensive thing to create. OCLC
> is already doing it. You do NOT need to be an OCLC member or have OCLC's
> permission to use OCLC numbers however you want (and this is in fact a
> _seperate_ issue from the OCLC Record Use kerfuffle. Even if maybe you
> DO need OCLC's permission to use the _records_, that still doesn't mean
> you need their permission to simply use OCLC numbers. I am quite
> confident you do not, although I am not a lawyer).
>
> Jonathan
Received on Fri May 08 2009 - 14:16:43 EDT