Sunday, April 10, 2011

Beautiful and Pointless Numbers about Poetry

Defending poetry with numbers seems as promising as defending a business plan with an interpretive dance. But David Orr’s book Beautiful and Pointless attempts the feat.

From his review in the New York Times Book Review, David Kirby summarizes Orr’s approach:

[W]hat makes “Beautiful and Pointless” different from thousands of other defenses of poetry is that, according to its author, poetry differs from music and stamp collecting in that people’s love for poetry is measurably greater than their love for any other activity. Poetry fans don’t just love poetry a little; they really love it.

To test that hypothesis, Orr went to Google and conducted two different searches, one for “I like X” and one for “I love X,” with X being represented by baseball, cooking, gardening and half a dozen other activities, including movies and poetry. Admittedly, the science behind this research is slightly less complicated than that required to make a lemon meringue pie, but the results are noteworthy. In every instance except two, more people “like” an activity than “love” it; for example, readers of romance novels like that art form 3.36 times more than they love it. The exceptions are poker, which splits 50-50, and — of course — poetry, whose partisans “love” it twice as much as they “like” it.

Reading that, my first thought was, I hope he searched with those phrases in quotes. With the quotes, Google would find documents containing the phrase as a whole. Without the quotes, Google would find documents that contained all the words but not necessarily in order or together. The difference matters: Much poetry, and commentary about poetry, is about love. The same cannot be said of poker. And it’s debatable where love fits in the world of romance novels. ;) So if Orr did not use quotes around his queries, poetry might have had a big advantage against the others.

Since it was easy to do, I asked Google myself. The answers were surprising.

I tried poetry, without quotes. I got a love-to-like ratio of 0.92. With quotes, I got 0.93. Hmmm. I was expecting Orr’s ratio of 2. Instead of getting double the love, I got no effect.

How about romance novels? With quotes, the ratio was 3.75—which sounds like Orr’s result except it is reversed; Orr’s 3.36 ratio was for like-to-love, whereas my 3.75 was love-to-like. Ouch. But without quotes, I got a love-to-like ratio of 0.43 for romance novels—much closer to Orr’s 3.36 ratio when it is inverted into love-to-like (0.30).

Poker? With quotes, the love-to-like ratio was 1; without quotes, 0.96. That was in line with the 50-50 split (as Kirby put it) that Orr found.

I checked poetry again. For the record, the results were:

  • 74,000,000 for love, no quotes
  • 80,700,000 for like, no quotes
  • 2,750,000 for love, with quotes
  • 2,960,000 for like, with quotes

Maybe, from Orr’s book to Kirby’s review to my interpretation, something got disconnected. But for me, the numbers failed to add up.

No comments:

Post a Comment