Re-imagining browsing

In a recent issue of Educause Review, Robert Kieft describes work at the Tri-Colleges outside Philadelphia to plan for adding tables of contents and sample text to a prototype library catalog. The impetus behind these experiments is to provide users with a robust browsing experience in a world where distributed collections make shelf browsing insufficient.

I like this idea, as when I do shelf browsing, I do open up books, read a random page or two, and just generally poke around to see if the book looks interesting. I also believe this is an area in which our current catalogs don’t remotely match the physical experience.

But I also think there’s more to browsing than having something catch your eye then explore it further. When shelf browsing, we don’t look at everything – we only look at select things. In a bookstore, a cover might catch one’s eye, but this is less likely to happen on shelves with plain library bindings. A title or an author might be the hook that causes one to pick up one book and not another. To some extent, this type of browsing activity is random.

But what if it were less random? Can we re-imagine (or at least extend) our notion of browsing to make it more targeted? I like to think of browsing both as the ability to “look inside” a resource to learn more about it before committing to it, and as a “more like this” feature that introduces me to resources that I didn’t previously know existed. Shelf browsing of course does this but is also obviously limited by physical constraints, so that only one aspect of a work can be brought out by a classification scheme that locates it on the shelf. This isn’t news to anyone—see for example, the much-blogged-about Everything is Miscellaneous by David Weinberger for a discussion of what he calls first-, second-, and third-order methods of organization.

What I’m interested in is the ability of our catalogs to bring out more flexibility in browsing. If I’m looking at a resource, I want to be able to note one feature of it, and instantly get other resources that share that feature. I’ll then want to be able to add or subtract features to exert control over the size and characteristics of my results set. Say I’ve just read The Godfather. I might want more mafia fiction after that. But that’s a long list that I might consider too broad. Perhaps I’d limit myself to novels about the Sicilian mafia, or set in the 1940s or ‘50s. Then after I select a work or two, I may want to move on to real-life mafia stories, but only from the mobster’s perspective (not law enforcement’s). I’d likely find Henry Hill’s story in Nicholas Pileggi’s Wiseguy, which could lead me to watching the movie Goodfellas, and start wondering what happened to Hill. From there I might start exploring the history of the witness protection program, then FBI training methods, then perhaps all the way to Thomas Harris’ Silence of the Lambs and the film that was made based on the novel! And now imagine we can do all of that in a few clicks, without having to type anything in, or to know any of those books or films exist. And imagine if this could happen across multiple databases of content (including those from outside the library sector, such as IMDB), without me ever having to know that.

This vision is also nothing all that new. Certainly the faceted browsing features that are becoming increasingly common are a major step towards the overall goal. Technical protocols for connecting distributed databases are similarly emerging at a fast pace. I don’t think that massive initiatives that attempt to solve all the connection and interface problems at once are the answer. Instead, I see that smaller initiatives that perform a proof-of-concept of one issue at a time is the most effective way forward. It’s a sort of survival of the fittest – different people develop a few different ways of solving one particular issue, and let others adopt the one they believe works best. Each of these smaller solutions can then be building blocks for larger solutions – solve a problem, undertake an initiative to glue some of the smaller technologies together, iterate. Focusing on the smaller problems doesn’t mean we have to lose sight of the larger picture, and can help us avoid the paralysis that’s possible when the problem looks so large as to be intractable.