Ex Libris MELvyl Tells (EL Mel Tells)!
Volume 2, Number 6, October 2003
In this issue:
Karen Coyle To Retire
[Karen Coyle will be retiring from the University of California on December 1, 2003 after twenty-four years of amazing contributions to worldwide research and scholarship through her work on the Melvyl Catalog and other projects. She is now on vacation until her retirement. I had the chance to interview Karen while sitting by Lake Merritt in Oakland, California on September 16.]
Ask Karen Coyle anything about the innards of the Melvyl catalog, and eventually, she’ll rise out of her chair, mosey on over to one of the many white boards scattered throughout the California Digital Library, and start drawing you a diagram to explain the question at hand. If your face continues to look blank (mine frequently does), she patiently gives another lucid and intelligible explanation to your question until the light goes on and you “get” it.
When a seemingly unanswerable question comes up about why a certain record displays in the catalog, the complexities of cataloging, the intricacies of the MARC record, the workings of Unicode, issues of copyright or privacy in libraries, the cry is always, “Ask Karen!” When Karen retires from UC at the end of this year, and it’s hard to imagine life at CDL without her.
One of the brains behind the Melvyl Catalog, Karen related a bit of pre-CDL history to me during our interview. The earliest version of what was to eventually become CDL was ILR—the Institute of Library Research. Begun in the 1960’s, one of ILR’s earliest projects was to make a book catalog for the UC libraries, so the idea for a UC-wide union catalog goes back quite a while. (By the way, while working on this article, I discovered a Finding Aid for the ILR http://www.oac.cdlib.org/findaid/ark:/13030/tf3j49n6zm at the Bancroft Library at UC Berkeley—some fascinating early history of library automation there!) Eventually, ILR morphed into ULAP, UC University-wide Library Automation Program, where the first automated catalog began—its output being in the form of hundreds of frequently updated fiche. Here’s where Karen comes in (twenty-four years ago): she was a continuing graduate student (already having an MLS form UCB), and had a lowly student job at ULAP when she got her fulltime position at ULAP. Karen worked on analyzing the OCLC and RLIN input, which were the only streams coming into the early Melvyl catalog. ULAP was renamed the Division of Library Automation (DLA), the last incarnation before CDL. The computer originally used for Melvyl programming was at UCLA; turnaround time for the programmers, who were located in Berkeley, was two days!
There were great challenges in building a union catalog, and it was
an exciting time—UC staff was working on quite a unique project.
Around 1980, when Melvyl was being built, there were few online catalogs.
All of the campuses had card catalogs and campuses were very nervous
about DLA losing the machine-readable data in the catalog. No one
had managed to apply keyword indexing to such a large catalog database
before. It was an unusual and tricky thing to do, and there were few
other pioneers out there with whom to share information. Another challenge
was loading Library of Congress headings into Melvyl. An attempt was
made to link bibliographic headings to LC headings but the vagaries
of author headings caused some odd matches to be made, demonstrating
that some human intervention is required to true authority control.
One of the major successes in Melvyl, however, was its merge algorithm,
which brought all copies of the same work together, from all of the
campuses. This algorithm has been incorporated into the Ex Libris
union catalog product and is being offered to their other customers.
There were some surprises in how the catalog came to be used. Intended primarily for students, faculty and staff at the University of California to find materials at all of the UC campuses, Melvyl’s creators were surprised to discover that an impressive portion of its use came from far beyond UC. More than a local or systemwide catalog, it was a bibliographic tool for the world. Second, staff was surprised (and looking back, maybe this shouldn’t have been a shocker…) that Melvyl became such an important tool for staff doing cataloging and acquisitions, in addition to its reference use.
At the beginning of Melvyl’s history, there was a certain amount of resistance on the part of library staff to this online catalog, but users took to it from the start. For many, it was their first introduction to an online resource—there were no computers in offices at that time, recall. Users flocked to those computers in libraries like geese to Lake Merritt. People developed personal relationships with Melvyl. In the initial two years of its existence, in the first flush of being able to type things into a responsive system, curious undergraduates typed in profanities, which showed up in the usage logs. But within the next couple of years, they were proposing marriage to Melvyl. Or as Karen put it, they moved from “anonymous sex to commitment” in two years. (“Melvyl, I love you!” was not an infrequent comment found in the Melvyl Comments feature.)
Which brings me to the formation of a whole subculture and an early public chat group that developed in telnet Melvyl, the only available version at the time. Melvyl had what you might call one of the earliest FAQs in existence in the form of its Comments feature. Users could type a Comment or question that would be responded to by the Melvyl team. A record of answered and unanswered questions would be available for users to browse. Spontaneously across the campuses, students started using the Comments feature as an early online discussion group by typing in a code word--Dilcue (Euclid backwards)--and then typing in their messages. They would search the database using the “Added since date” command, looking through the Comments for the word Dilcue, and respond to each other’s messages. While not speedy by any means, since comments and answers were only posted every few weeks, Melvyl Comments provided a long-distance, subterranean way of communicating for those in the know.
Along with Melvyl’s very successful merge algorithm, Karen is most proud of the work that was done to make available in usable format the varied records from the abstracting and indexing journal article databases mounted at the DLA, then CDL. Karen also (rightly) has a sense of pride because data quality and user service were always the key goals of Melvyl development. Every change made to the catalog was always well tested before its release, with the bottom line being to serve the user above all else.
In addition to her contributions to UC, Karen has been involved in and made international contributions to a number of other activities surrounding computing and the Internet and their impacts on society. Privacy, free speech, civil liberties, the role of women in technology, and copyright are all issues Karen is passionate about.
In thinking what she would like to do for libraries, if anything were possible, Karen thought about helping libraries to “finally get over the card catalog”. Do we really still need inverted headings, for example? Second, she would like to see libraries better integrated into mainstream computing; and finally for libraries to achieve the recognition they deserve for their contributions to technology, many of which the computer world has ignored.
After retirement from UC, Karen will do consulting on standards and metadata, privacy and library systems. She will have more time to continue her activism and rabble rousing, and hopes to return to Eastern Europe and other less developed areas of the world to facilitate access to digital materials.
Among the things Karen has really wanted to do, but has been unable to do as a full time employee, is to get a puppy. But even that may have to stay on hold with all the other things pulling her away. Have I made it sound like Karen is champing at the bit to leave the CDL? Well, there is one thing she’ll miss. And that’s the people she’s worked with most of her working life. Karen, we’ll miss you, too.
To keep track of Karen, visit her at her Web site: http://www.kcoyle.net