Books, the idea: Words that read you back

Posted by Rob Walker on May 13, 2010
Posted Under: Things/Thinking

Waaaaay back in March, Elliot Van Buskirk of posted about “Text 2.0.” A video from a Swedish technology company demonstrates what this means: Basically, eye-tracking technology monitors how you read, and the text itself reacts. Somehow it senses if you want to know how a word is defined, or pronounced, or translated, and it tells you. If it concludes that you’re just skimming, it “fades out irrelevant information” in order to “streamline your reading.”

I have zero interest in this as a reading person. I envision triangulation against databases that cross-matches my eye movements (and thus, by implication, my mind) against some “norm,” thereby sacrificing idiosyncracy and individuality to the fabled CloudCrowd yet again. I also have zero interest as a writing person. I happen to think textual communication with an engaged reader is a truly singular thing, and the idea of some tech firm’s supposed expertise monkeying around in the middle of that connection is depressing. Moreover, in both scenarios, I’m fundamentally skeptical that the real-life quality of the technology will be remotely proximate to the sales pitch. Stuff like this makes me think of a future in which some sort of Clippy equivalent informs me that I don’t actually like Milan Kundera after all.

That said, as someone compiling an occasional series on the idea of the book, I’m very interested in this concept. If something like this somehow compiled individual reader data, straight from the eyeball, it could be another layer on the stuff that Amazon is tracking and broadcasting with its “Kindle highlighting” information. People who believe they need a mass of personal-behavior data to tell them who they are might want to know which passages this tracking software has concluded they like (as opposed to which passages they thought they liked, by highlighting, or simply by, you know, thinking.) Those who like the results might then have an interest in “sharing” them — Text 2.0 could automatically tweet its conclusions (“Turns I out I don’t like Kundera”), per your settings.

As Van Buskirk notes, the appearance of something like Text 2.0 is highly plausible, as the iPad and other e-readers evolve to include cameras; his reporting indicates that many of the relevant patents and business deals already seem to be in motion. (Here’s another link to his writeup.) He also makes the case that eventually this sort of technology could “reinvigorate the written word,” basically giving ebooks a fresh way to compete in the attention economy. Maybe he’s right, and in any case he’s saying what you would probably expect a writer to say. But he also says something you probably wouldn‘t expect: At the moment, he’s not really into ebooks:

I should admit that I have yet to make the leap to an e-reader. Having grown up a bookworm in the ’70s and ’80s, I prefer the feel and even the smell of paper books — and the fact that I can just chuck one in my bag for whenever I have a spare moment, without worrying about batteries, theft or breakage.

Probably there are rationally convincing response to these points that any e-reader enthusiast could make — surely paper-book-smell technology is in the offing!

But that’s just the thing — the idea of the book isn’t strictly a rational thing. If it were, there wouldn’t be much interesting about it, now would there?

Further diversion may be found at MKTG Tumblr, and the Consumed Facebook page. Tags:

Reader Comments

I wrote about this for eFiction Magazine.

The implications of this technology are many:

First, think about if advertisers could see what parts of ads consumers were looking at. Online advertisers could pay-per-look instead of per click. I hope this goes mainstream. Advertising will have another golden age.

Second, once we have this technology, imagine pairing it with MRI. If you’re wearing a magnetic resonance imaging helmet and your eyes are being tracked, then a computer could link the different electric signals in your brain for each word. Once it’s recorded most of the words you use, you would only have to think them for it to detect those same electric signals in your brain, and voila thought-to-text writing.

Written By Doug Lance on May 13th, 2010 @ 9:22 pm

I hadn’t thought about what this would mean for advertisers, but that’s a good point. (I’m not sure I think it would lead to better advertising, but that’s another story.)

And on the second point, again, I’m just a lot less optimistic about promise vs. results on that sort of thing. But maybe you’re right.

Written By Rob Walker on May 14th, 2010 @ 11:18 am

Add a Comment

required, will not be published

Next Post:
Previous Post: