Recommended Reading, Part 1

The recent appearance of the “Biblioracle” created a stir on the bookish part of the Internet. John Warner, a contributor for The Morning News (home to the Tournament of Books, an annual favorite of mine), spent a few hours making book suggestions. Given the last five titles you read, he recommended one book, just author and title, that you should try next. Warner was quickly overloaded and gave up his prognostications, but clearly he hit a nerve with the twitterati. Before comments closed, the site received over 1000.

As a librarian specializing in readers’ advisory, I found the Biblioracle underwhelming. By failing to get more information from the reader, by providing no explanation for his selections, and by suggesting only one title, Warner violated most of the tenets of what an educated librarian would consider good readers’ advisory. He reminded me of an earnest friend that I now avoid because he always pushes strange books into my hands, tells me I “have to try this,” and then asks me for feedback a few days later, as if I have nothing better to do than immediately devour his suggestions.

The Biblioracle produced at least enough idle curiosity for hundreds of readers to overwhelm him and his website, but I hope most readers will take the advice of the most popular comment on the page: “Go talk to a librarian. They’ve been doing this sort of thing for decades.” I don’t mean to be hard on Warner. His suggestions were made in the spirit of fun, and I enjoyed browsing the titles he was given and trying to guess why he made the decisions he did. Deep down, I react to his exercise with a  mixture of professional pride and misgivings. My library and many others provide a more complex, thoughtful suggestion service that remains open every day of the week, but even then, one always wonders if book recommendations are really hitting the mark.

In discussion of the Biblioracle on other websites (such as this one), commenters debated the difficulty of building a computerized recommendation engine that works better than those we’ve seen online to date (for instance, Amazon’s ubiquitous recommendations). While Pandora and Netflix succeed to some (debatable) degree for music and film, book recommendations are tougher for several reasons. More books are produced than films or albums and describing their contents, then matching them to the thousands of reasons that individuals have for preferring one book to another will require a much bigger database of information than anyone has produced to date. The day will come when we build such an engine, but for now, a well read human who has thought about reading preferences beyond his or her own has a better chance of producing good results.

Reading a book requires a much greater time investment than listening to an album or watching a film. Once one’s reading progresses to the point where casual suggestions of “good books” based on the preferences of others are no longer enough, the costs of that investment deserves better than knee-jerk intuitions or profit-motivated marketing tools.

I’m a heavy reader, managing about 200 books a year, but ever since I thought one day about how many shelves it would take to hold the books I can reasonably expect to finish in my lifetime, I’ve become more protective of my reading. I don’t take or give reading suggestions lightly.

All that said, however, pointing others to good books is one of the great joys and privileges of the thoughtful reader. Sharing our reading delights is one of the best reasons why we participate in book groups and other more casual forms of book discussion. I’ll follow up on this post next week with some basic principles for how (and how not) to make good book discussions.



About the Author:

Neil Hollands is an Adult Services Librarian at Williamsburg Regional Library in Virginia, where he specializes in readers’ advisory and collection development. He is the author of Read On . . . Fantasy Fiction (2007) and Fellowship in a Ring: a Guide for Science Fiction and Fantasy Book Groups (2009).

Post a Comment