“Mental rotation and object recognition are indistinguishable from a behavioral viewpoint: You can’t tell them apart,” says the paper’s first author, Isabel Gauthier, assistant professor of psychology at Vanderbilt. “As a result, the field has been deadlocked over the question of whether the brain uses the same mechanism or different mechanisms for the two tasks.”
Michael J. Tarr, one of the paper’s co-authors and professor of cognitive and linguistic sciences at Brown University, had proposed in several papers with Steven Pinker at the Massachusetts Institute of Technology that the same mechanism must be involved in the two tasks. “There are parts of our brain that are involved in our ability to imagine the world,” he says. “The question is, ‘Are those the same as the parts of the brain that we use to know what things are?’ And the answer appears to be, ‘No, they are not.’”
Also collaborating on the study were William G. Hayward of Chinese University of Hong Kong and an fMRI team from the Yale School of Medicine headed by John C. Gore.
To find out what parts of the brain are involved in these two mental tasks, the researchers began with six unfamiliar geometric shapes that look something like the pieces from an unfolded Rubic’s cube. They used three of these objects – two of which were mirror images – for the mental rotation tasks and three which were slightly different but similar in appearance for the object recognition tasks. Next, they assembled a group of 15 subjects and used functional Magnetic Resonance Imaging machine to measure activity levels in different parts of their brains.
During the fMRI sessions, the researchers had the subjects perform two types of tasks using the objects. They displayed the objects in pairs on a screen and, in one case, they asked the subjects whether the two are identical or mirror images – a basic mental rotation task. In the other case, they asked them whether the two objects were the same or different – a basic object recognition task. The objects were shown at different angles from each other on the horizontal, vertical and out-of-plane axes and the researchers measured the time it took the people to answer.
When they examined the brain scans, the scientists found that the areas activated during the two tasks tended to lie on two different pathways in the visual system. These two pathways, called the ventral and dorsal, are sometimes called the “what” and “where” pathways. When asked questions about the identity of an object – for example, is it the same shape as a second object? – then the ventral pathway, which includes the temporal lobe, is activated. But when a person is asked where an object is located, the dorsal pathway, which lies in the parietal lobe, becomes active.
The first place the researchers looked was the parietal lobe because previous studies had shown that it is involved in mental rotation tasks. They confirmed these observations and found that when the difference in orientation in the mental rotation tasks was large, the amount of parietal lobe activity was greater than when the difference was small.
In the object recognition tasks, however, the researchers saw a much different pattern. They did see some activity in the parietal region. Surprisingly, however, the amount of activity in the parietal lobe decreased at larger orientation differences. In addition, they found that the brain area that did show an increase in activity with larger differences in orientation was in the ventral pathway. “This is the first indication we have that the brain doesn’t rely on the same processes to accomplish these two tasks, despite the fact that they appear to be so similar,” says Gauthier.
During the course of evolution, it seems as if the same solutions have arisen more than once for similar problems in the way our brains work, adds Tarr. “They look very similar behaviorally, but it turns out they use completely different neural circuits and the brain doesn’t know how to put them together.”
For more news about research at Vanderbilt, visit the online research magazine Exploration at http://exploration.vanderbilt.edu.
Journal
Neuron