History

The history of institutional repositories (IRs) is relatively short, with the first discipline based repositories being implemented in the early ‘90s. In a recent international survey of fifty three institutions by the Primary Research Group (2007) the average repository had been in existence for 2.85 years, with repository age ranging from ‘recently started’ to ten years.

It is important to note that some of the driving forces behind the development of repositories, such as the problems associated with the present system of scholarly publishing, have been discussed by academics for over twenty years Harnad (2006) and Odlyzko (2006).

1.  Current institutional repositories

Currently IRs are being implemented by institutions at a growing rate with the Lynch and Lippincott (2005) survey of United States based institutions finding that 40% of these institutions had in fact deployed an IR of some description. Of the universities and colleges that did not house an IR, 88% reported that they had plans to implement one in the near future.

Similarly the two main repository software platforms, EPrints and DSpace, have experienced a surge in usage; according to Lomangino (2006), the use of EPrints rose from 125 to over 200 repositories between 2004 and 2005. In late 2008 there are 286 known repositories using EPrints, with DSpace the chosen software platform for 373 repositories at the time of writing (Registry of Open Access Repositories). Lomangino (2006, p.7) found that repositories that comply with the Open Archives Initiative’s (OAI) interoperability standards have risen from 243 to 617 repositories since late 2003.

Although repository implementation has been increasing in the last few years, there are still a number of barriers that limit the filling of repositories with scholarly material. In addition, not all the contents qualify as scholarly. Lomangino (2006, p.8) comments on the University of Oregon IR, where only 18% of the 1,900 documents in the repository were actually authored by academic staff.

2.  Problems

Problems inherent in the scholarly publishing process have led to two main outcomes for libraries and researchers: important journals needed for research are becoming unaffordable, and the shift to electronic journals has meant that some journals are lost if the library cancels a subscription to a database or journal. Chan (2004) discussed these two problems in relation to IRs and reiterates the longstanding observation that subscription costs have been increasing, often above the rates of inflation, and at the same time as library budgets have been reducing. Journals have also been supplied through databases which have meant libraries do not have a physical copy. Chan (2004) argues that both of these problems result in a loss of access to research resources and outputs, and may, in fact, reduce research impact for scholars.

The implementation of repositories by institutions can benefit libraries in a number of ways. Firstly, an IR can be the most cost effective method of providing access to research findings and as a result improving research impact. Secondly, by collecting and preserving academic research the institution’s own teaching and research can be greatly improved cite (Chan:2004). Repositories have been suggested as a solution to the problems with scholarly publishing and but until recently have not been a viable option for institutions to implement.

3.  Early repositories

Although the implementation of repositories has been increasing rapidly over the past few years it is important to mention several early discipline based repositories that have been very successful. Jones et al. (2006) identified a number of early repositories of note, firstly the arXiv physics repository which was launched on the 16 of August 1991 Suber (2007) to cater to physicists and has since been expanded to include other closely related disciplines. This repository served a particular need in the physics discipline as researchers needed to have their work made available to the community quickly due to the speed at which new discoveries were being made. Jones et al. (2006) noted that arXiv has not served as a model for other repositories. Other early repositories were also discipline based, such as:

• EconPapers (Economics working papers)
• CogPrints (Cognitive psychology)
• PubMed Central (Biomedical and life sciences)

4.  Discipline vs institutional repositories

Jones et al. (2006) discuss the early examples of discipline based repositories and state that it was in response to such repositories that the more recent IRs were developed, since academics were reluctant to deposit their work in a discipline based repository or that one might not continue to exist. Institutions may then provide a more trusted repository for their staff to deposit their research. Jones et al. (2006, p.7) cite the creation of the Open Archives Initiative in 1999 as being a major factor in the rise of institution based repositories. This raises the question of whether an author should deposit their research in the institution or discipline repository. Green and Gutmann (2007) discuss the differences between both types of repositories and ways in which partnerships can be built between the two.

5.  Key events in the history of institutional repositories

• 1991: Launch of the arXiv Physics repository
• 1999: Sante Fe Convention which resulted in the agreement upon a framework for interoperable archives, now known as the Open Archives Initiative (OAI)
• 2001: Launch of EPrints by the University of Southampton
• 2002: Launch of DSpace by Massachusetts Institute of Technology
• 2002: Publication of Ryam Crow’s SPARC Position paper entitled: The Case for Institutional Repositories: A SPARC Position Paper

6.  Key resources in the history of institutional repositories

• Peter Suber’s blog entitled Open Access News: News from the open access movement
• Jones, Andrew and MacColl (2006) book entitled ‘The institutional repository’