Commentary

Spatial narratives

Jarrett Walker has another thought-provoking post up, this time on spatial vs. narrative navigators, linking to a magazine article that gives more background. In the comments, Russ links to a paper(PDF) about the effect that the Tube Map has on people’s mental maps of London.

The distinction between spatial and narrative navigation makes a lot of sense to me, but disagree with the way that Jarrett seems to set them up as a zero-sum game – either you’re good at mental maps, or you’re good at remembering narratives, or you’re mediocre at both. This portrayal is supported by the Maguire study discussed in the article that indicated that London cabbies who create too big of a mental map wind up with hippocampi that are shrunk in the front. But I feel like I’m relatively good at spatial navigation and not so bad at narrative navigation, so I think it’s possible to develop both – at the expense of whatever else you could be learning, of course, like how to pick up women.

An interesting point in that area: when I developed my prototype English to American Sign Language machine translation system (PDF), the biggest obstacle were phrases like “winds SSW 30 mph gusty near canyons.”

I asked a native signer and professional interpreter the best way to say “gusty near canyons” in ASL, and she drew a map of Albuquerque in the air, pointed to the canyons with a topic marking nonmanual, and made the sign for “wind” with emphasis. I asked her if it could be said another way, just with the sign for “canyon,” but she said that the map would be much more natural for a native signer.

To get a machine translation system to handle that would not only require a kind of sophisticated mapping subsystem that’s not usually found in these programs, but also a lot of background knowledge of geographic features – for every locale that has weather reports. It made me realize that not only is ASL much more complicated than I had given it credit for, but it’s a lot harder to write down than spoken languages. In the terms of Jarrett’s post, it seems clear that the National Weather Service was using narrative description, while the interpreter I consulted was using spatial description.

While I would agree with the comment left by Pantheon that spatial navigation is more effective than narrative navigation, particularly when it’s done by computers, I strongly disagree with his condemnation of narrative navigators, or of inferior navigators in general. Complaining about people who have difficulty finding their way across town is like complaining about bad spelling or malapropisms. In some cases you can make the case that people should develop these skills betters, but for a lot of them it’s too late to learn. There’s a range of ability along both axes, and a transit agency that only caters to riders within certain limits will run the risk of alienating riders.

grvsmth (Author)

Leave a Reply