Google Busts Up the Ivory Tower — So It Says

Google Busts Up the Ivory Tower — So It Says

By Nathan Siegel



Because a great revolution is happening in ivory towers — and Google Scholar is leading the charge, according to Google Scholar.

By Nathan Siegel

Google Scholar is the academic proletariat’s greatest weapon, according to the company that now runs most of your life. 

To mark the academic search engine’s 10th anniversary, a team of Googlers published an analysis of journal articles from 1995 to 2013. Their findings suggest a democratization of academic work: Of the 1,000 most-cited articles each year, a greater number came from “non-elite” journals — the proportion grew by 64 percent. Total citations from non-elite journals, not just the top 1,000, doubled over the same time period. 

Of course, if you ask Ronald McDonald, the Big Mac is the most groundbreaking thing to happen to fast food since the deep fryer, so Google’s study should be taken with a grain of salt. And some caveats, like the unwillingness to flesh out core assumptions and release the actual number of articles analyzed, leave skeptics questioning the study’s validity. But experts not affiliated with Google have corroborated the findings. And the co-author of the study and co-inventor of Google Scholar, Anurag Acharya, maintains there’s nothing controversial about the study. “It’s just math,” he says.

Google Scholar now archives approximately 88 percent of all English-language scholarly work on the Internet.

Well into the 1990s, most research was done in the library, and since even the fastest speed-reader could peruse only so many hard copies, the journals best known and most widely distributed were the ones that got cited, according to the study. Times have changed, and Google Scholar now archives approximately 88 percent of all English-language scholarly work on the Internet. That’s up to 160 million documents, according to some estimates — Google doesn’t make those figures public. 

One hundred and sixty million pieces of academic brilliance?! Er, not exactly. Google Scholar will index most anything and everything, including a host of non-peer reviewed articles. “The limitation to Google Scholar is that you get everything, but there is no transparency” for how many or which articles get indexed, said Vincent Larivière, professor of information science at the University of Montreal and author of a similar study. “It all appears to be magic.” 

Acharya’s response: “If you make it easy for everybody to find something, they will find it. There is no magic there.”

And yet, two core assumptions for the study were not clarified, forcing readers to take Google on faith, says Phil Davis, who studies and consults on academic publishing. First, the amount of published research has increased since 1995, he says. And second, many elite journals are publishing fewer studies. It’s like musical chairs where there are more children and fewer chairs; there could just be an increase of non-elite publications, having nothing to do with programs like Google Scholar, he told OZY. “They may be right, but they clearly violated some core assumptions that most people shouldn’t take for granted.”

The system isn’t hack-proof, either. In fact, manipulating the citation metrics are “simple, easy and tempting,” according to a group of researchers that created a fake author, Ike Antkare (“I can’t care”) and six ghost papers. The plants garnered 129 existing papers, an additional 774 citations and no small bump in Scholar rankings.