Musings from a Mouse

Continued from page 1

Instead, the curving mouse trajectories seemed to reveal an ongoing process, one that Spivey sometimes compares to a state of quantum superposition -- in this case, a mind existing in a "grey" area.

While Spivey, who co-authored the study with Marc Grosjean of the University of Dortmund, Germany, and Gunther Knoblich of Rutgers University, believes their latest work could eventually lead to a new model for understanding cognition, he admits that it's a small step.

And a hot-button topic.

 “It’s quite a big debate right now [over the best model of language acquisition],” he says. “There are a lot of traditional cognitive scientists who are essentially digging their heels.”

Jim Magnuson, an assistant psychology professor at the University of Connecticut and a researcher at the Haskins Laboratories in New Haven, CT, which specializes in biological bases of speech and language, thinks their work shows promise. But he also lays out the position of some critics: that the mouse movement "is potentially much more under conscious control than an eye movement.”

Measuring eye movements is the traditional method of cognitive studies -- and a costly and complex one. As Magnuson points out, humans average 2 to 4 eye movements per second, “and you are not aware of most of them, but with a mouse movement, it’s quite intended.”

Rather than discounting Spivey's use of a mouse, though, Magnuson feels that it could be a “methodological innovation,” as well as leading to less-expensive experiments.

“Eye tracking has been an important tool in usability testing for websites,” Magnuson says. “Now you could end up using mouse tracking much the same way.”

Finally, besides hinting at new understandings of human cognition and new kinds of computer-assisted research and design, Spivey's study might have implications for a field somewhere in the middle: artificial intelligence. As Spivey points out, biological neural networks might be a better model for creating AI applications, such as language-recognition systems, than binary-based computers.

 “If you want to invent a mind, you probably don’t want to be using a computer format," Spivey says.