Core Integration
During 2000-2001, with encouragement from the NSF, the six demonstration projects attempted to develop a joint proposal to create the Core Integration effort, but despite valiant efforts the six teams were unable to overcome technical and philosophical differences on how the NSDL should be developed and how the funding should be divided among the groups. Eventually, UCAR, Cornell, and Columbia broke away from the other three and submitted a proposal that was accepted by the NSF. This proposal included substantial subcontracts to three other universities.
Seven years later, the proposal still reads well, but some central themes have not passed the test of time. By avoiding making hard decisions, the group was able to present a united front and succeed in being funded (which is the purpose of any proposal), but it created problems that lingered for many years.
The first problem was that the Core Integration team promised to do too much. The team committed itself to extensive efforts in library building, technology, outreach, and evaluation. In the opening section of the proposal, three models for the NSDL were discussed with varying degrees of centralization; the proposal offered a framework that embraced all of them. The proposal used two appealing phrases, “one library, many portals” and “a spectrum of interoperability”, but made little attempt to measure the effort needed. At least eight metadata standards were to be supported and there would be a broad set of authentication and authorization services. The proposal envisaged extensive work with commercial publishers.
With encouragement from the NSF, the proposal was over-reliant on consensus building among the NSDL-funded projects. “We envision a strong NSDL community that sees itself as owning the program and having major influence on the character of the library.” In retrospect it was naive to expect effective collaboration among hundreds of projects many of them having little in common except being funded by the NSF. External events aggravated the problems of collaboration. Because of the trauma of September 11, 2001 the NSF postponed an initial meeting of all the grantees, including the services, collections, and research projects. When they were brought together several months later, the fragmentation into a range of independent projects was well under way.
Even within the Core Integration team, little though was given to effective management. In practice, the universities who were subcontractors to the Core Integration team used the funding more to support their general research programs than to provide specific services to NSDL. Organizationally, the history of the Core Integration team since 2001 is a progressive tightening of the management structure. The loose oversight in the early days allowed conflicting visions of the goals, even within the Cornell team. For several years, I combined part-time leadership of the Cornell team with the usual academic duties of teaching and other research. When Dean Krafft replaced me as a full-time principal investigator in 2005, he immediately reorganized the work at Cornell to give it more focus.
Finally, the proposal was very weak on how to support the library in the long term. Basically it stated that sustainability is important and the team would think about it. Nobody had recognized the urgency of sustaining the smaller projects, which would be a serious concern within very few years.
Comments on this entry are closed.
Bill writes that NSDL has been faced with two challenges: supporting education at all levels and mixing the missions of innovation and implementation. He adds that the “aim was to implement a digital library, but the funding program was more suitable for a broad program of research and development.”
Given the broad and mixed mission, does NSDL need to focus its resources in particular areas to ensure longer term sustainability? What would those areas of focus be, and can digital library services be created that provide value to some set of users or funders?