WorldCat on the web

WorldCat is now available to all comers.

Over the years, I've had several conversations with librarians who wanted to ban the public from having access to multiple-library databases such as WorldCat. And I've encountered libraries, even large public libraries, which outright forbade such public access, reserving it to staff-only use. What an appalling display of arrogance! I could understand a policy of directing users to other sources such as the local library catalog and fulltext periodicals databases first, since materials found through those sources are more likely to be available locally. In libraries outside the US, I could understand a policy of preferentially directing users to interlibrary union catalogs that were less US-dominated, such as Canada's Amicus. However, an outright denial of WorldCat access to the public seems to me to reflect nothing more noble than the library staff's desire to reduce their workload by keeping their customers ignorant of useful materials that could be requested from the library if customers knew of their existence.

If you're unfortunate enough to be dependent on such a library, web-WorldCat will be a welcome bypass route around the librarians' roadblock. However, as we'll see, if you have access to WorldCat by way of FirstSearch or some other traditional gateway, there are still reasons to prefer it for certain types of searches.

Quick reactions:

Interface

The Google-like one-line search interface is familiar and unintimidating to web-denizens, and it will probably meet the needs of most users, but an advanced-search option enabling the user to make more precise searches is also needed. I don't like being forced to use a dumbed-down, one-line, undifferentiated keyword search when a more precise definition of search parameters is needed.

This type of interface is the best model for Google because the sources of Google's metadata -- the terms describing what webpages are about -- are so notoriously unreliable. Even so, Google offers an advanced search option, which the web version of WorldCat fails to do.

Unlike Google, OCLC has excellent, professionally-produced metadata available. Its data, derived from library catalog records, are clearly and reliably classified and labelled. Author information, title information, subject headings, publisher identification, etc., are separately identified in OCLC's data, and they need to be separately searchable. There's no other effective way to search for books *about* Henry David Thoreau, for example, without drowning in a sea of editions of books *by* Henry David Thoreau. Or to find, say, a list of books put out by a particular publisher, without also including unwanted materials whose titles, authors, or other datafields include terms similar to the name of the publisher. I'll refrain from supplying more specific examples. The creative researchers of the world can no doubt supply all the weird queries needed to demonstrate that an advanced search option should be made available as soon as possible if web-WorldCat is to be as useful as its older cousin.

Edit: It appears that it is possible to specify which type of search one wishes to do, but only in a way that is extremely unlikely to be discovered by most users. (Compare search results for "ti: leaky roof" with results for "leaky roof".)

Coverage

A quick couple of searches for fairly obscure titles suggest that the web-WorldCat service includes OCLC catalog records that are not widely held, including archival materials that are held by only one institution. This is an improvement over previous methods of open access to WorldCat, which excluded records of items held only by small numbers of institutions. However, web-WorldCat will not identify some holding libraries/institutions.

For example: a search for Paul C. Morris's Schooners and schooner barges of Cape Cod, a cassette tape of a verbal presentation by a maritime historian, will locate the OCLC catalog record for the item. But when one clicks on the "libraries" tab of the search results page, the web version of WorldCat claims that "no libraries with the specified item were found." In fact, according to a search of WorldCat by way of FirstSearch, there is one institution that holds a copy of the tape. (Clams, Inc., of Hyannis, Maryland. This appears to be their website; the tape appears in their catalog as a library-use-only item.)

Similarly, a search of web-WorldCat for Mahlon Neill White's The Leaky Roof : the story of a railroad, a history of a small regional railroad, offers access to the OCLC catalog description of the book and indicates that 13 institutions own copies. In fact, according to FirstSearch-WorldCat, there are 16 owning institutions. The ones excluded from the web-WorldCat results appear to be the State Historical Society of Missouri, SMU's DeGolyer Library, and the Wisconsin Historical Society. The first two are ILL nonsuppliers, but the WHS is listed by OCLC as an ILL supplier.

I initially thought that the criteria for exclusion from web-WorldCat's list of holding institutions might depend on whether the institution is an ILL-supplier, but the WHS example suggests otherwise. Perhaps holding institutions, or certain types of holding institutions, were allowed to opt out? Or perhaps the determination is made on an item-by-item basis, depending on whether the institution does or does not permit circulation of the specific item? Someone with more cataloging expertise than me would have to answer that.

I'd welcome any thoughts that other folks have on the subject -- this is just a quick preliminary set of reactions on my part.