Skip to content

Category: Marginalia

The Method of Multiple Working Hypotheses

This morning, a Geology professor presented one of her assignments to this workshop that I’m attending, and one concept jumped out at me. She said that one of the things that makes Geology difficult for freshmen is that all evidence can be interpreted in multiple ways (Sound familiar? It sure does to me from both my Literature and my LIS backgrounds). Turns out, this was articulated as a philosophy of studying Geology back in 1897 when T. C. Chamberlin published his ideas about The Theory of Multiple Working Hypotheses.

Holding more than one hypothesis in your head leads to “certain distinctive habits of mind,” particularly thinking critically about each hypothesis and being more thorough in looking at the range of potential evidence (845-846). The idea is that if you don’t get attached to one theory, you won’t fall into the trap of only seeing or “lingering with great pleasure” on the evidence that fits that theory (840). You’ll also always have to confront the issue that one or the other of your hypotheses might not stand the test of research, which forces you to constantly evaluate each one’s validity.

This strikes me as exactly what I’ve struggled to articulate to my students who come to me with research questions for which they’d really just like to find some articles from which to pull supporting quotations. Whether the questions are overly simple (how better to know that you’ll find supporting evidence?) or impractically difficult, my students really aren’t interested in evidence that complicates their projects. (The major exception being those who feel they should find at least one source that definitively and categorically disagrees with them, in order to tear it to pieces while demonstrating that they’ve considered counter arguments. But this is not really holding a counter-hypothesis in mind.)

If I can get students to entertain the possibility of at least temporarily holding another hypothesis in mind, perhaps they’d be more thorough in their research, and perhaps they’d be more able to take a critical look at what they’ve found.

Of course, I should note that Chamberlin identifies a couple of disadvantage to this approach. First, it’s not something that “young scholars” can do very well (848). (Hmmm… that seems to include my entire audience.) And second, you can’t express “more than a single line of thought at the same time” (857). In other words, in the middle of our lightning fast 10-week term, I’d be asking them to entertain the possibility of doing a bunch of research that wouldn’t contribute directly to their final product. That’ll go over like a lead pipe to the head.

Still, I wonder if there’s a way to be more conscious about this as a method of thinking while acknowledging that it is not entirely practical for my students.

Chamberlin, T. C. “Studies for Students: The Method of Multiple Working Hypotheses.” The Journal of Geology 5.8 (1897): 837-848. [Available via JSTOR]

Comments closed

The Pot Calls the Kettle Black…and Ends Up Looking Blacker

[Edit: If you’re reading this in an aggregator, be aware that after the first three paragraphs each section has a heading. The text is there, but the formating isn’t.]

Thomas Mann has some very good points in his rebuttal of Karen Calhoun’s report on the Changing Nature of the Catalog. I agree with him that subject control cannot go away. I agree that there are many levels and kinds of searchers and that “quick information seekers” need different things than scholars of, say, history. (Though I would argue that scholars of different disciplines each have different needs, and that the scholarly research examples Mann gave were overwhelmingly History and Literature oriented.) Relevance ranking is not necessarily the best option for all researchers, and new researchers will probably always need some training. I would also hate to see careful cataloging disappear in favor of whatever is faster and more automated. (This might work marginally well in the sciences, but I can’t see it ever working well for the humanities.)

But that’s all that I got out of his report (21 pages plus and appendix) except that he’s very, very angry at Calhoun. I think that he could have made his points much more effectively if he’d kept a tighter lid on his anger, sarcasm, italics, and multiple punctuation (such as when he says, “Pushed beyond-??” on page 15). This narrative tone is so deafening that it completely drowns out his argument. In fact, all the silent screaming and ranting that Mann did kept him from building a plan for the future that looks anything different from the library services of the past. (And in the spirit of full disclosure, let me say that I’m reading this from the perspective of a librarian at a four-year college library.)

His primary flaw is that he talks in absolutes, in “either/or” terms, in it’s-better-for-them-this-way terms, and with the idea that if we have something nobody else has people will come. I also disagree with his dismissal of the business model of evaluating our services.

Either/Or Thinking, so 1.0
Most of Mann’s rebuttals go something like this: “Calhoun’s argument advocates change; this change wouldn’t serve my scholars; things should stay the same.” For example, he bridles at the concept of increased reliance on keyword searching. He argues that Google-like searching is not as powerful as subject searching, and that it isn’t as good at collocating. Given. But who said catalogs would turn into Google? And his example that “Cuba — History — Invasion, 1961” collocates more items that are actually about the Bay of Pigs than “Bay of Pigs” would gather disregards the process of discovering the controlled vocabulary in the first place. I don’t know of many librarians who, hearing a request for items about the Bay of Pigs would say “Oh, search by subject for Cuba–History–invasion, 1961.” If we don’t know these things, how will our users? Staring at the blank search box to which Mann often refers, the vast majority of users will not come up with the proper controlled vocabulary. Instead, they will use keywords and (hopefully) use the results of this search to find the appropriate controlled vocabulary. I would urge both authors to think in terms of increased keyword accessibility AND access to controlled vocabulary.

Also, I agree that displays of LC subject headings help map the breadth and depth of concepts. I do not think, however, that left-justified lists are the only ways to do this. What if users were able to choose between a left-justified list and a spatial concept map?

There are many more examples, but I’ll only mention one more. One section of Mann’s argues against relevance ranked result lists because scholars need instant access to the most current research in their fields. Why oh why can’t we give people a choice between relevance and date ranking?

It’s Better for Them This Way
My absolute least favorite sentence in the whole critique is: “The fact is, no researchers — either scholarly or superficial — will ever do efficient searches in online resources without some prior instruction and education.” This cop-out is his answer to the problem that even though searching for the LCSH subheading “Personal Narratives” will bring back more consistent results than the keywords “eyewitness accounts” and the like, but that researchers will rarely if ever come up with this phrase without help. This cop-out is just like it’s cousin “We only have 50 minutes.” It’s simply an excuse for a broken system. So, what if we allow user tagging of subject terms to add to the (limited) “see” and “see also” records? We should not require users to learn esoteric subject headings just because that’s the only way to get consistent results. Let’s fix the system instead.

Again, there are lots of examples of this type of thinking in Mann’s work, but this post is getting WAY too long, so I’ll move along.

We Offer Something Nobody Else Does, So People Will Come
This doesn’t even need explanation. Some people might come. But just as users shouldn’t be satisfied with the first superficial results of a broad keyword search, we shouldn’t be satisfied serving only those with intense and complex research needs or those who’ve uses libraries before.

Libraries and the Business Model
Just because we don’t depend directly on our users for our funding or work actively toward making a profit doesn’t mean that the business model doesn’t apply to us. At my library, the fact that the librarians are working actively to customize our services to our population, that we experiment with new ideas and services, that we show up to campus events (even evening ones), and that we’ve put ourselves “out there” in the campus culture has resulted in drastically increased support from college administration and therefore significant financial commitment from the college. Second, we should be evaluating our services as rigorously as do those who have to do so in order to receive paychecks. And finally, just because we think of new ways of doing things doesn’t necessarily mean that we turn our backs on Mann’s “scholars.” It means that we find ways to serve the “scholars” and the “quick information seekers” and everything in between.

Oh, and by the way, implying that catalogs have a “life cycle” does imply that death is an option. That was not rhetorical sleight of hand. That was the whole point of the metaphor. (See Mann’s rant on page 5.)

That’s my rant in response to a rant. Constructive criticism and rebuttals welcome.

Comments closed