Sunday, February 24, 2013

Scientizing Art

I've always been fascinated with the way the eye moves around a piece of art.

Andrew Wyeth's "Christina's World" (or as I looked up "that painting of a girl in a field looking at a house")

This piece by Andrew Wyeth is an obvious example of an artist completely controlling your gaze. There are pretty much no options here. You look at the girl and then you follow her gaze to the house. You probably then take a quick glance at that other house/barn to the left, and then maybe follow the edge of the light circle around the houses. (It's my opinion that that is how the eye should go on this painting, but I have no eye tracking data to support it.)

A paper last year in PLoS One really tries to "scientize' this process by testing what factors determine the eye movements, and the 'clusters' where the eye tended to fall. Massaro et al., (2012) compare dynamic and static images and images that contain human subjects or nature subjects. Their cluster analysis overlaying classic paintings makes for quite interesting images:

The next installment at MoMA

This one is a dynamic human image. Each patch of color shows where the parts of the painting where the eye lingers (face, hands, ....crotch...). The authors do all sorts of interesting analysis on this and other paintings, having participants rate the painting for 'movement' or for 'aesthetic value' and since the paper is open access, it is free to people who may not have university access to journal publications. Anyone can read the whole thing here.

One interesting thing that the authors find is that pictures containing humans have fewer clusters than pictures of nature. I expect this is because certain aspects of humans (faces, hands ...crotches...) are so salient and the brain focuses directly on them, while all the branches of a tree for example have about equal 'meaning' for a person.

science creates modern art
 Another great image from this paper. The authors show how much gazing was done at different parts of a painting through a heat map. This one is a human static image. The end result is actually quite haunting because the place that you want to look is blanked out (sort of like a Magritte painting).

So here are my questions: If someone looks at a blank page, where does their eye naturally go? Is there some sort of common pattern that most people use just to scan an area? Do chimpanzees use a similar pattern to scan a blank page? Does everyone have their own unique scanning pattern? Or is it just pretty much random? 

And here's an idea for artists: Buy yourself an eye tracker and have customers come use it and stare at a blank page. Trace their eye movements and then create a dynamic painting (or T-shirt, or napkin drawing) that follows the person's natural scanning patterns. This would be the ultimate in commissioned custom art! (Then give me one for free, because I think this sounds like fun.)

© TheCellularScale
Massaro D, Savazzi F, Di Dio C, Freedberg D, Gallese V, Gilli G, & Marchetti A (2012). When art moves the eyes: a behavioral and eye-tracking study. PloS one, 7 (5) PMID: 22624007


  1. That's a pretty cool idea of scientizing art.

  2. I think another element in eye movement is determined in part by the development of written language in a culture. Left to right reading is usually followed by left to right reading of a canvas for example.
    I need to find out how much an eye tracker costs! Can I get one a Wal-Mart you think?

    I think all scientists should study art and all artists should study science. Artize and Scientize together.