Now We Know What’s Open:
Google Scholar Identifies Open Access Versions of Research Articles

When I first noticed this new aspect of Google Scholar, well over a month ago, I initially didn’t get it. What were these new tags marking up what are already pretty busy entries in the search results? Some entries had a green marker pointing at the title of the article. Others had the marker pointing at a URL that followed the title. And before or after the title or URL came a square bracketed tag that read [PDF] or [HTML].

Of course, this was Google. No explanation, no announcement. Then it hit me. Google was identifying a version of the article that was available as a PDF or HTML. And available in the sense of being able to click through to this version without facing either a password or credit card request. Of course, the arrow had to be green for go.

With little fanfare, Google had taken yet another step in its efforts to organize all of the world’s information. It had now organized–-and by that I mean it had clearly identified-–the research and scholarship for which a free, open access, copy had been made immediately available to readers all over the world.

Here are two examples from a search on “open access” in Google Scholar. In the first instance, the search results indicate, with the green arrow, that the version of Gunther Eysenbach’s article published in PloS Biology is open access, as it is archived in PubMed Central (although it is also available in this open access journal itself):

In the second instance, an article by Stevan Harnad et al. has been published in Serials Review and is available from Elsevier’s ScienceDirect for $31.50. At the same time, the mighty green arrow indicates that an open access copy of the authors’ final draft can be found at “soton.ac.uk.” This leads to the institutional archive for the School of Electronics and Computer Sciences at the University of Southampton.

It takes a couple of clicks to get to the Harnad paper, while other examples I tried led directly to the open access version. It is hardly surprising that automatically processing massive amounts of non-standard data by however clever algorithms is bound to turn up a few wrinkles. Such is the case with various aspects of Google Scholar, and not just this new feature.

Also notice that in both cases multiple versions are available (“All 12 versions”) and with the Harnad paper, the open access copy is to be found within that dozen. But now readers can immediately realize that there is an open access copy, often a click away, if not in this case.

Of course, most searches will turn up many articles that have no green arrow. The proportion of the literature that is open access differs across fields but is approaching 20% of the annual output, according to one recent study by Bo-Christer Björk, Annikki Roos and Mari Lauri. It will only increase, all the more so with the spread of open access mandates across funding agencies and institutions.

Still, at this point, Google Scholar’s identification of open access versions alleviates the frustration of seeing a system so adept at harvesting the location of every mention and citation of a paper, and then not making readily apparent this one extremely valuable piece of information–-can I read the work, if I don’t happen to belong to a library with a subscription or cannot bring myself to spend $31.50 to read it?

Having said all of that, I have to add that Anurag Acharya, chief Google engineer on Google Scholar, disagrees. He explained to me that the green marker indicates materials that are “available to a specific user and NOT what is open access (i.e., it is customized for each user and is not a global marker).” So, if I have library privileges to an article, it should get a green marker. If I am working in a low-income country, journals that have policies offering access in such circumstances should have green markers. And if a journal or a repository offers open access to its content, it, too, should have a green marker. “For example,” he added, “if you were testing from, say, India or Ghana or Kenya, and were searching for articles from PNAS, you would see a different set of links marked. Also, it as yet depends on the number of publishers that make accessibility information available to us. We expect this number to grow.”

You be the judge. It sounds like a shout out to open access to me. This new public exposure of open access materials certainly seems likely to result in their being read and cited more often. In the case of repository copies, it is true, it could well be the author’s final peer-reviewed and revised draft (which is all the publishers typically allow authors to post). But the number of open access journals is in the thousands and growing. And consider the difference a little green marker can make for an amateur astronomer or environmentalist, a high school student working on a science fair project, a Wikipedian struggling to build the best researched entry she can in Wikipedia, and then the reader who follows her links to learn more about the topic.

Google Scholar’s latest step in organizing all the world’s information can make all the difference for a whole new set of users, the difference, that is, between grasping at a glance what knowledge can be readily pursued and having to wonder if one’s feeling lucky.