Spurious Tuples

I am old enough to have used the Reader’s Guide to Periodical Literature, in print.

There. I said it.  As a matter of fact, I used print indices of various sorts right through my undergraduate degree and my first graduate program.  (Ah, those print volumes of RILM, eventually supplanted by the CD-ROM version that ran on the DOS-only computer. Good times, good times.)

But I do have a point here:  I’m actually very, very glad that I am this old, and that I have this experience under my belt.  Because now, whenever I use a database to search for articles, I have a very clear mental model for what I’m doing: what exactly is contained within the database, how it’s searching, what it’s finding.  And quite frankly, I don’t think you can use a print index and not have a very clear mental model for the process of indexing the periodical literature.

My friend and former colleague Kim Duckett talks and teaches a lot about the two processes of scholarly research: discovery and access.  Discovery is what happens when you’re searching an index: you’re discovering what has been written about the topic. Access is getting your hands on the full text of whatever it is that you’ve found.  The advent of full-text content in online bibliographic databases has elided the distinction between the two processes somewhat; sometimes so much so that students are unwilling to pursue articles that aren’t in full text in the database they’re searching (or don’t realize that they can pursue those articles).  And I’m not entirely sure that’s an entirely good thing.

When you’re working with a print index (this is the part of the post to which the title refers), the research process is pretty straightforward:

1. Identify articles/items of interest.  Collect them into a list.
2. Identify which of those items you have immediate access to, and go get them.
3. Of the remaining items, prioritize which ones you want to pursue.  This generally involves a cost/benefit analysis: is it worth waiting for Interlibrary Loan? (how fast is your ILL? is there a cost to the user? etc.)  Is it worth going to another library to get it? (how far is that other library? will you be going there anyway? etc.)

It’s that third step that’s important here: you’re constantly thinking critically about your articles/items and their relative value to your overall research project, constantly re-evaluating your priorities, etc.  And this, I think, is a step that gets largely lost in the process when it’s facilitated by full-text access.  The tendency to grab the first full-text items you can find (what I think of as a “pillage and plunder” approach to building a bibliography) seems to be much, much more overwhelming in this environment.¹

And this, finally (FINALLY) brings me to my point — or, part of my point — which is that recently I re-discovered a fascinating article by Martin Gordon called “Article Access — Too Easy?” published in a book entitled Serials Librarianship in Transition: Issues and Developments in 1986.  Yes, you read that correctly: nineteen eighty-six. Nearly a quarter-century ago.

And Gordon makes essentially the same point, though some of his language and ideas are charmingly quaint when viewed from this perspective (due primarily to the fact that online searching at that time was largely mediated by librarians).  His concern is largely that the tremendous leap in accessibility of material via online searching will lead to research papers becoming “an exercise in seeing how many citations they can append to their essay in the hope that quantity will either add to its substance or hide the lack thereof” (170).

And lo, it came to pass.

His other concern is that the greater accessibility of the journal literature will lead to undergraduates over-using it, in situations where books might be more appropriate:

Unlike monographic sources that tend at the undergraduate level to provide overall views of a topic, periodical articles are apt to be as pieces in a landscape of possible sources that require careful selection and placement in order to be of value. … How well they mesh with one another as well as their ability to update or expand the monographic choices are of primary importance in selecting them. (171)

Which leads me to my other point (FINALLY), which is that his concerns have, at least to some degree, come to pass.  And this is not entirely a bad thing: undergraduates should work with the primary research literature in their field.  But so often, the primary research literature is either written at a level that’s far beyond their comprehension, especially in their first couple of years; and/or it’s exactly as Gordon describes above: one very narrow slice of a much, much larger picture.  And that’s not always the best kind of information for what undergraduates need.

Now, thinking about the kinds of information that undergraduates need, got me thinking about Barbara Fister and her colleagues’ recent study of the contents of aggregated multidisciplinary databases (PDF), and librarians’ assessments of the value of those databases.  They found, unsurprisingly, that aggregated databases tend to pad their offerings with journals of dubious quality, and/or highly specialized or technical journals whose value to undergraduates is questionable.  They also found, however, that librarians were generally satisfied with these databases and didn’t want to see them restricting or reducing their contents to better meet undergraduates’ needs.

I’m not precisely certain, but I think I may have taken the survey that Barbara and her colleagues administered as part of the study.  In any case, I probably would have sided with most of the librarians in the study for one simple reason: I don’t want vendors making the decisions about what to include and what to exclude; I want librarians making those decisions.  And this is the wonderful bit of history that I learned from Barbara’s article: at one time, we did.  The periodicals to be indexed in general indices² like the Humanities Index and the Social Sciences Index from Wilson were voted on by the subscribing libraries (275).

Partly this was for practicality’s sake: in an age when interlibrary loan was much more cumbersome, libraries wanted their own holdings to be foremost in an index’s contents, and before online catalogs, it was much simpler for librarians to report this information than for the vendor to assemble it it/themselves.  But there was also a pedagogical/collection development component at work here: librarians understood which journals were more appropriate for their undergraduate students’ work, and prioritized those journals for inclusion in the indices.

Maybe this makes me hopelessly old-school and out of touch (see above re: “used a print index”), but I’m not sure that such a thing isn’t such a bad idea after all.  Maybe instead of an “undergrad” checkbox, we need a whole separate “undergrad” database?