"G"STOR?

I was reading this post at the Atlantic over the weekend, in which Alan Jacobs complains about the terrible user-interfaces of scholarly databases:

Librarians want students to specify their search terms, but that rarely happens: "Regardless of the advanced-search capabilities of the database they were querying, 'Students generally treated all search boxes as the equivalent of a Google search box,'" simply clicking into a text field and typing away. Problematic, yes -- but why shouldn't they be able to do that? Why shouldn't the software be able to help them out?

Not long ago I was using a research database to try to get a PDF of an article published in a journal to which my college's library has a digital subscription. I knew the title of the article, the author's name, the title of the journal, and the issue date. I plugged all those in to the appropriate text boxes, clicked "search" . . . and got hundreds of results. But the one that I wanted wasn't on the first several pages.

I sent an email to a reference librarian describing this event, and he wrote back saying, "Oh, see, you should have entered the journal's ISSN." Really? Exact title of article and journal, exact name of author, exact date of publication -- that's not enough?

I guess I think that's somewhat of a fair criticism. Jacobs goes on:

...And a Google search does some things that the research databases don't: for instance, when I'm searching for an article in the enormous JSTOR academic journal database, I don't use their search box, but instead go to Google using the "site" delimiter: site:jstor.org my-search-term. Why? Because if I spell a name wrong Google knows the correct spelling and asks me if I meant that. On JSTOR itself I just get inferior results.

So maybe our greater emphasis shouldn't be on training users to work with bad search tools, but to improve the search tools. Especially since serious research questions aren't as afflicted by spammy SEO as many other queries, by this point in the development of online life we ought to be doing a lot better than we are.

This seems right to me. When I was first in grad school, I spent a lot of time on the Philosopher's Index. But I never really got the hang of it, and if I'm honest, I'll admit that I haven't looked at it in years. I haven't checked to see if my current institution subscribes to it. And I can't remember the last time I tried to use the search utility on JSTOR. With either of them, if I'm looking for something specific, I can never find it; and if I don't have anything specific in mind, I'm not going to find anything I can use.

Jacobs alludes, on behalf of the Index/JSTOR search utilities, to the argument that this is because I don't know how to use them, and that if I understood the logic of these search engines I would get better results. But I guess I don't see that as a criticism of me, and I definitely don't see it as a suggestion of a way to improve myself. I see that as a fairly conclusive criticism of the logic of the search engines. If the search engines employed better logic, it would be possible to use them effectively without having to go out of your way to learn a complicated and unintuitive way of thinking.

The research database I make most use of is Google Scholar, and the two features I find come in handiest are the "Related Articles" and the "Cited By" tools. If you, like Jacobs, know the title of the article and the author's name, your search will be successful. It will work. And if you spell something wrong, it will be Google and help you spell it right. And if you want to see how other people have responded to the article, the "Cited By" results will tell you. You'll have to sift through the results to sort the articles that mention it in passing from those that mention it more prominently, but you'll know. And it helps you to see which articles made a big splash and which ones have had a more modest impact. And then it links you to the paper on JSTOR or Springer or whatever. I'm open to the possibility that I'm being really irresponsible with my research, but I sort of doubt it. Google Scholar works really well. If I'm using it to look for something I find it, and it has helped me discover I-don't-know-how-many papers I would never have found otherwise.

--Mr. Zero