[BOAI] Re: Comparing repositories - subject-based, institutional, research and national repository systems

[Jessica Perry Hekman]

On Mon, Nov 23, 2009 at 06:19:50PM +0100, Armbruster, Chris wrote:

> - Armbruster, Chris and Romary, Laurent, Comparing Repositories Types: 
> Challenges and Barriers for Subject-Based Repositories, Research 
> Repositories, National Repository Systems and Institutional 
> Repositories in Serving Scholarly Communication (November 20, 2009). 
> Available at SSRN: http://ssrn.com/abstract=1506905

Although you are soliciting responses about your proposed distinction, I 
was inspired to respond to this part of your paper:

  If the repository is to have any value over the long term, then a 
  quality control system must be implemented and the integrity of the 
  corpus preserved. One widespread misunderstanding is that repositories 
  are there to archive 'everything' when, in fact, users are ever more 
  concerned about the value and relevance of research results.

I am new to OA, and a student (though I worked in online publishing for 
15 years before returning to school), so I apologize if any of the 
following is misguided; I hope the members of this list will point out 
where I went wrong, if I did.

When I discovered OA, I envisioned a future world in which value and 
relevance of research results is maintained by entities which provide 
the same services as do journals today: facilitating peer review, 
organizing publications by topic, and providing a statement of 
legitimacy (the difference in how you value a paper that was published 
in a very high quality journal with a high impact factor, vs. a paper 
published in a journal which no one has ever heard of). These indexing 
entities could be what journals become, or they could be something new, 
or there could spring up a mix of both. Because the indexing entities 
would provide this service, repositories wouldn't need to do so, and 
would be free to archive everything. Users wouldn't have trouble 
navigating through this deluge of information, because the indexing 
entities would guide them. A user would follow the list of new articles 
in his chosen indexing entities, just as today he reads Nature or the 
New England Journal of Medicine. For hunting down a specific article, 
he'd use PubMed or Google Scholar. Actually browsing a particular 
repository directly for content would be a painful process that no one 
would waste their time on.

I naively assumed that the OA community was envisioning this as well, 
but given the bit of your paper that I quoted above, I'm now wondering 
if I'm wrong.

You have said in other places in your paper that repositories, 
especially institutional repositories, aren't particularly good at 
providing this kind of indexing service, and I agree with that. What 
repositories *are* good at is providing a place to maintain copies of 
publications (whatever "publications" come to be in the vibrantly open 
access future), and place for indexing engines to hunt for content. 
Isn't there value in maintaining everything? The long tail phenomenon in 
internet retail sales has allowed items of niche interest to be found by 
their audience. Isn't that useful in scholarly publishing as well?

Of course, repositories will have to clearly mark content as peer 
reviewed or not. Perhaps there is already a metadata standard for this 
out there which I have not yet found, as I am very new to OA.

I do think that the role of indexing entities is an extremely important 
one. Repositories are worthless if they can't be navigated. I just don't 
believe that providing sophisticated navigation, indexing, adding 
statements of legitimacy, etc., is the job of the repository. I believe 
a third party can provide those services much more effectively.

Thanks,
Jessica