A small object lesson about the scholarly communication ecosystem

Yesterday I started following links and ended up at the supplementary material for the article “Evaluating Big Deal Journal Bundles“, which reminded me that I want to read it in full. And while PNAS has OA content, the thing I want is not yet available. So I wrestled with our discovery layer for a while, realized it was never going to find an “early access” article indexed there, and submitted an ILL request by filling out the Illiad form manually.

Today, I got one of our standard ILL replies from our Collection Building staff. As I started reading, and saw that they cancelled my request, I thought, “Damn, did they miss the embargo?” but I should have more faith in my staff.

A request you have placed:
Title: Evaluating big deal journal bundles
Author: Theodore C. Bergstrom     Paul N. Courant     R. Preston McAfee, and     Michael A. Williams
Journal: PNAS, Proceedings of the National Academy of Sciences
TN: 180308
has been cancelled by the SUNY Potsdam Interlibrary Loan staff for the following reason:
Article is available on the web.
It can be found online with a “Google Scholar” search for the article title. Please see the library’s Reference Desk for assistance.
I have included a link to the article.

Because, hey, look at that. Turns out the primary author has it up on his website, and Google Scholar has indexed it.**

The takeaway? My very expensive discovery layer that gives access to our very expensive databases which we purchase in a Big Deal model cannot find this early access article in a reputable journal, but a library employee with access to Google can find the author’s archived copy of this article about the cost of the journal Big Deals, and thereby found the website in which much of Bergstrom’s supplementary material is also housed. And it probably took her less time to find it than it took me to fail.

Nah. Nothing’s broken here.

**I am not including the link, because it appears that while Bergstrom has put the .pdf on his website, he doesn’t appear to have actually linked it openly, so I don’t want to then openly drive traffic to something he may or may not yet have legal rights to post publicly. Google, however, has no such scruples. Do what you will with that knowledge, if you’re looking to read this thing. [Edited to add: I take it back; I didn’t read closely enough. It’s here. Go for it.]


  1. There are a bunch of issues here, of course, but your comment about the failure of your discovery service to find an OA version of the article caught my attention. Some of the discovery services claim to incorporate IR holdings into their indices, but as this circumstance indicates their is also an OA channel that falls outside of library- and vendor-managed channels. I believe this OA channel is larger than the one that libraries, publishers, and vendors have created, but I’m not sure there are real measurements about its size. In any event, it points to one of the fundamental shortcomings of the discovery services vs. Google Scholar, which in this case is more inclusive in its indexing.


  2. Exactly. Suggests something problematic about curatorially building an index as our discovery services do vs. algorithmically building it based on signals of trust and authority, as Google Scholar appears to do within the .edu domain. I can also imagine coiunter-examples where curatorially building the index is advantageous.


    • I think fundamentally, library discovery services are building indexes at a less granular level. Generally, we are curating based on the journal title level or higher or in the case of open access usually pulling it in at the collection level (eg DOAJ index, OAIster index), while Google Scholar bots are harvesting at the individual article level.

      So we can’t include a Scholar’s homepage say because who knows if everything on there is “Scholarly”. It will only be in our index, if the Open access collection we index from such as DOAJ includes it.

      I am starting to feel the Google Scholar method of crawling is superior for open access and free articles, because it can directly verify that an article is really available helping fight link rot – a big issue for open access material and one of the reasons many Summon libraries don’t add say DOAJ.

      If you are talking about comprehensiveness as is the case of a known item search, Google Scholar’s method will always win out, as collections profligateand curation by collection will always fall behind as you need to add collections manually.

      For discovery searches, curated collections might give better results, in so far it makes the job easier for the ranking algorithms.


  3. Stupid thing we’ve added to our SFX page – if we don’t have electronic access, you have an option to launch a Google Scholar search from there. It’s an advanced search for title and author using the first author name. It works amazingly well (in the sciences, of course) even when the metadata are horrible. Also, from that findit page, we can link through to ILL so that the ILL form is filled out already. I’m pretty sure it’s open source software running on top of SFX if your systems people are interested.


  4. So, on the ILL front, why is this a cancellation? I count these as Document Delivery filled so that I have a metric of how many we’ve found for patrons. If interested have your wonderful libraryfolk email me.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s