Participants then studied drawings of objects--including pictures of a boat, telephone, bell, and clock--and learned to visualize them, after which they were shown boxes of three different sizes to standardize the notions of small, medium and large.

Following this preparation, participants were told to visualize certain objects at specific sizes while answering questions about the object. For example, the "anchored right higher?" question asked whether the right side of the anchor was higher in the drawing.

Several questions would be asked in this manner, and three sessions (one for each size) were performed. Thus, there would be four results: one for simply listening to auditory cues and three for visualization at each size.

It was found that there was indeed substantial brain activity in spatially organized regions even when the participants were resting.

More importantly, when comparing neural activity from the listening session and the imagery session, Kosslyn and his collaborators found evidence of activity in areas 17 and 18 of the brain, parts which are known to be spatially organized. Activity was found to vary depending on the size of the image, a relationship which parallels that found when eyes normally view an object.

The evidence of activity in area 17 is particularly interesting because this region is the first to receive input from the eyes.

"Visual imagery appears to affect processing in the brain from the start," said Kosslyn. "Thus, it is not surprising that people sometimes mistake having visualized something for having seen it."