I was interested to see Heather Christenson describe The Hathi Trust as a collaboratively sourced web scale research library in a recent article (Hathi Trust: a research library at web scale. PDF).

This reminded me of an entry I wrote a while ago about sourcing and scaling (which is referenced in the article). In a shared network environment, one of the most interesting issues facing libraries is how appropriately to source and scale activities.

A few years ago, this activity would have been sourced within the institution: each library would have developed its own infrastructure, user interface, local community outreach, and so on. Now, such an impulse is questioned. It makes sense to source something like this collaboratively. And it is provided at the level of the network: its target user population is the population of web users.

Questions about sourcing and scaling are becoming much more common as the logic of the network reconfigures patterns of information production and use. What does it make sense to do at institutional level? What does it make sense to source elsewhere (repository services in the cloud, for example, or institutional email services from Google)? And what should be left entirely to other providers? At what level, or scale, is it best to do things? Locally, or within a consortium, or ....?

Think of four sourcing options: Self (provide it locally), collaborative (provide it within a group), public (provided through state or national activity), or third party (provided by another commerical or non-commercial entity).

Think of three scaling options: local or institutional, group, and web scale.

These can be put together to give a variety of options. So, for example, Tripod, the shared catalog of Swarthmore, Haverford College and Bryn Mawr, is a collaboratively sourced group solution. PubMed is a web scale public offering. And, as already noted, Hathi Trust is a collaboratively sourced web scale service.

An interesting contrast between the US and many other parts of the world is that often what is done collaboratively in the US may be done through a public agency elsewhere. For example, Christenson contrasts the HathiTrust as a collaborative activity with something that the JISC, an activity of the public higher education funding councils, might provide in the UK. It is also common in many countries outside the US to have publicly supported union catalogue and related activities.

We can observe two trends. First, there is a trend towards externalisation: libraries are looking to collaboratively source activities or to outsource them to third parties. Think of collaborative activities around managing down print collections here, the West project for example, or the growth of shared library systems (the Orbis Cascade Alliance, for example, recently issued an RFI about a shared integrated library system). Think of the growing interest in cloud-based sourcing of systems and services.

Second, there is a trend to 'move up' in the network, by doing more things at group level within consortia or public contexts (think of OhioLink or Summit), or by leveraging network level services (think of social networking sites, for example).

The current economic environment further encourages these trends. Institutions look for economies of scale through collaboration. And they also want to focus attention on high value areas, and outsource routine or shared activities.

I was reminded of these issues while reading a very interesting internal report of the University of California on library services. This is the interim report of the systemwide Library Planning Task Force, convened under the auspices of the Systemwide Library and Scholarly Information Advisory Committee (splashpage, PDF of report and related material).

A stark environmental picture is presented:

  • The libraries will experience budget reductions of as much as $52M or 21% of their current budget base over the next six years. "To put this into perspective, this cut is greater than the total library budget of any single UC campus, and roughly equivalent to the budgets of three of our mid-sized campuses, all AAU members."
  • The libraries will likely lose the equivalent of $17M in buying power in the same period given publisher price increases. "This is equivalent to the current library materials budgets of two mid-sized campuses, and means a reduction in the systemwide acquisition rate of about 200,000 items per year."
  • Existing facilities will run out of space for new materials over the next 5-7 years, at the same time as "demand increases for extended hours and services and technologically well-equipped and flexible learning environments in the libraries' prime campus locations".

They go on to observe that the impact of these factors can be mitigated through collaboration. They propose four strategies:

  1. Expand and collectively manage shared library services.
  2. Support faculty efforts to change the system of scholarly communication.
  3. Explore new sources of revenue.
  4. Improve the existing framework for systemwide planning, consultation, and decision-making.

Of course, the University of California is an unusual institution, bringing together some of the world's major universities in a shared organizational framework. One result of this shared framework has been the California Digital Library, which concentrates operational and innovation capacity for the whole system. CDL has been responsible for some major services, and is an active partner in the Hathi Trust. Another is the Regional Library Facilities, north and south, for managing print collections. A major recommendation is that the range of such shared services should grow, whether sourced within the UC universities or externally.

Cooperation is difficult. Especially where money flows, and impact needs to be seen, at the institutional level. However, given the existing level of shared services, the organizational framework, and the pressures described in this report, it will be interesting to watch what services the UC libraries move to a shared environment over the next few years.

p.s. The report describes the Worldcat Local-based Next Generation Melvyl described in these terms:

The Next‐Generation Melvyl (NGM) initiative moves the discovery of information for researchers and students to the highest networked level. The initiative takes access to the highest level of aggregation and is vital for the most effective provision of information access and services. Strategically, NGM also positions the UC Libraries to provide aggregated access to a significantly increasing array of full‐text information resources: e.g., the millions of digitized books in the Google Books Project and the HathiTrust.

Comments: 0

Leave a Comment