Future Scenario (Duncan Aldrich)
With this in mind, trying to think of a program that would be viable given information and user trends, I made up the following future scenario of the FDLP. Perhaps this is 8 to 10 years in the future. Three things I think we need in particular to think outside the box are metadata, intermediation, and the impact of the FDSys. The top part below on collections is more whimsical than well thought. The remainder I think may well be accurate – if indeed there still is an FDLP based on depository libraries. If you find some of my content annoying or downright dumb, remember, this is put forward to poke holes in. My basic thought is if we can glean a sense of what the future will be we can better prepare to get there.
While there are many libraries that continue to have collections of FDLP materials on their shelves and continue to receive them as depository libraries, there are far fewer than there were in 2005. Perhaps we now have 300 libraries in the program dealing with more than nominal digital or tangible collections. Ten collections are full (or almost) - these ten are called Full Regionals and are actually distributed collections shared by 4 or 5 libraries covering five or six states each. These are circulating collections having some but not a lot of preservation activity. Two full collections are housed monolithically in single institutions - light archives. These two are partially supported by federal dollars in order that they can afford to meet the light archive criteria. Their collections are more archival than circulating – they operate more like special collections departments than like circulating collections. A GPO / NARA partnership maintains a full collection as a dark tangible archive.
Most access to federal information is done over the Network - which of course is currently true today as well. The virtual collections are housed in the FDSys and its mirror, and in two libraries that have made a deliberate effort to mirror the e-collections in FDSys. When library users request materials in libraries that don’t own an item, the usual procedure is to grab an e-copy from one of these sites and either give the patron the file or print it out (at either their or the library’s expense). Sixty libraries participate in a LOCKSS partnership which further provides for redundancy, though these are mostly not complete collections. LOCKSS is fed by the FDSys. 100 to 200 libraries actively retrieve and load individual files from the FDSys for public access - but these are mostly transitory postings along the lines of items currently selected and disposed of after 5 years - no guarantees of ongoing maintenance. These items mostly support special local collections and projects (Hurricane relief, Martin Luther King’s birthday, burial of nuclear waste, etc.). These libraries have the option of having these materials pushed to them from FDSys via a profile, or of selecting and downloading individual documents.
Items included in the FDSys are harvested from federal sites with some level of intervention by GPO staff to help direct the crawler and categorize them (by agency, subject, whatever). Librarians specializing in subjects help identify fugitive documents.
A question on which I have no prediction - how many libraries will remain in the program to serve as service centers with little or nothing in the way of collection? Will there be a class of libraries known as service centers and what is it they will be?
There is simply too much information out there to provide full MARC cataloging. Additionally, the public is not interested in MARC cataloging, a system designed mostly to inventory library holdings. The public generally prefers using full text indexing along the lines of Google so there is a decreasing use of OPACs, AND full text search technologies have evolved to include limited (often machine generated) searchable fields so are much better at precision searching. GPO has adopted a high end full text software application that provides this fielded indexing - human intervention is minimal, primarily for quality control. Fields are either inserted by agencies in the original XML document (XML is standard for text items by this time) or are generated by the indexing software. Because everything is xml libraries collecting these objects can create their own access tools. Some fuller cataloging (MARC perhaps) is done for materials deemed important enough for tangible distribution, particularly legislative materials.
This is the hardest part to figure - how do we insinuate ourselves into the digital arena to assist users in locating materials relevant to their information needs. I am particularly concerned about this because I think the Web is the primary place we will find our users (decreasingly at our ref desks) so how do we intermediaryize things there? The national online service along the Illinois model will develop. I particularly liked a scenario Walt Warnig described to me in a sidebar conversation describing an FDLP network of experts/specialists:
"Technologically, one single librarian with special expertise could be accessed by everyone in the country or beyond, but physically, no single librarian could handle such a load. We must have gatekeepers. A gatekeeper might be this: an expert FDLP librarian could be accessed by other FDLP librarians (and local patrons), but not by just anyone anywhere or by non-FDLP librarians."
Will libraries be able to cooperate to develop very useful Network based tools for assisting users? We already have a variety of handouts up there, but can we get some more elaborate user aids that fully incorporate technologies available on the Web?