A scholar wants to present a groundbreaking working paper on teleportation to his colleagues around the world.
Instead of submitting the paper to a print commercial journal and waiting months for results to be published, the researcher can simply pull up MIT's Center of Teleportation Research Web page and instantly submit the paper and data sets online, for all his cohorts to review.
This virtual intellectual asset sharing is part of DSpace, a joint project between MIT and Hewlett-Packard to create a long-term, sustainable digital repository.
This fall, MIT plans to open the DSpace archive to all its professors. The project will also release a set of free software tools so that any college or university can create its own online repository.
"We see this kind of system as a kind of public good," said MacKenzie Smith, associate director of technology for MIT's libraries.
As universities churn out vast amounts of research in electronic form, many are building immense digital archives to capture, distribute and preserve intellectual output.
"We really have to start thinking about alternative models, so we don't lose a record of scholarship from this particular era," Smith said.
"Technological obsolescence is a huge problem with keeping material viable beyond the next 5-10 years. It is yet to be proven that publishers who electronically publish this material will succeed in preserving it."
Other institutional archives include the University of California eScholarship Repository, Ohio State University's Knowledge Bank and Caltech's Library System Digital Collections.
These archives may provide more efficient, open access to research than costly commercial journals, which scholars often rely upon to publish their work and establish prestige.
"Today, institutions are paying unconscionably high prices to buy access to the research they support that is published in profitable commercially published journals," said Rick Johnson, enterprise director of SPARC, the Scholarly Publishing & Academic Resources Coalition.
"Institutional repositories can stimulate emergence of new publishing models that are more efficient and effective in serving academe."
A number of successful open-access e-print archives already exist. Physicists and mathematicians publish drafts of unpublished work in the Los Alamos Physics e-Print archive. Cognitive scientists self-archive papers in CogPrints. Engineers and scientists post pre-prints in the U.S. Department of Energy's PrePRINT Network.
Individual authors can use free self-archiving software to build online repositories that are compliant with the Open Archives Initiative.
Institutional repositories build upon this idea, allowing faculty to post everything from research data and data sets to images and audio files, theses and dissertations, conference papers, listserv archives and other "gray literature" online.
Faculty members are often desperate to find a safe online repository to post research data and other "born digital" material, Smith said.
"The data produced for those (materials) is very much at risk."
But deciding what material to keep and how to store it is daunting, especially when much of it is undocumented or unstandardized, Smith said.
"There are easily petabytes of data on campus now. Over the next 20-30 years, it could become an unimaginable amount of data."
While scalability is an issue, getting faculty to actually participate in the DSpace project may pose an even greater challenge.
"It's not the technology that's the problem," Smith said. "It's getting people to use it and understand why it's important."
Costs for managing these archives can be lowered, thanks to free software like DSpace. Many universities already have the technologies, standards and protocols in place to implement systems immediately.
MIT is considering launching value-added commercial services to help pay for the expense of running DSpace. While institutions using DSpace software could sell access to their own archives, MIT plans to make its materials publicly available, free of charge.
Perhaps the most difficult issue that universities face in building their own repositories is in upending age-old processes such as peer review.
"The biggest challenge is the inertia of the traditional scholarly publishing paradigm," Johnson said. "The benefits of voluntary participation in institutional repositories must be clearly articulated to faculty, whose reactions will vary depending on discipline."
"We're talking about changing practices that have been built up over hundreds of years," Smith said.
DSpace directors are educating faculty on how to protect their digital rights so publishers will allow them to retain a copy of an article that is published in a commercial journal, in an online institutional repository.
While these repositories will open access to intellectual capital, universities will still have the ability to restrict access to internal information.
"The benefits of institutional repositories are obtained only if access is open beyond the individual university," Johnson said. "At the same time, policies need to accommodate concerns about pre-publication access or copyright restrictions, so there are legitimate circumstances where an institution might limit access to particular content to a specific set of users."
Despite these barriers, the potential for institutional repositories is limitless.
Universities could create a virtual library that would link high-energy physics research from different institutions across the nation in a seemingly seamless collection.
That could bring research beyond the lab to the general public.
"This will accelerate and expand communication, and bring down some of the barriers between developed and developing nations," Johnson said.
