Which is smaller, a mouse or an elephant? What color is the sky? What does a car look like?

Obtaining an answer to such questions may seem easy, but defining the mental processes involved in such visually-based judgments have been a source of debate among philosophers and scientists for centuries.

The central question has been whether the brain actually forms a mental picture of objects, "looking" at them with the metaphorical "mind's eye," or whether it makes such determinations solely based on non-visual means.

After about two years of work, Harvard researchers in the departments of psychology, neurology and radiology have confirmed that the same area of the brain is indeed stimulated for both sight and mental visualization. The results were published in last week's issue of the journal Nature.

"We show that even when your eyes are closed, visual mental imagery activates visual areas [of the brain]," said Dr. Stephen M. Kosslyn, lead researcher of the study, in an e-mail message. "And it activates them in different ways, depending on the size of the visualized object--in a manner that is similar to what occurs in vision proper [normal sight]."

Kosslyn, who is a professor of psychology, said the results indicate that people who have more activation in the early visual area perform an imagery task faster.

"This might imply that brain scanning can be used to validate tasks that in turn can be used to select people who are good visualizers, and thus might make good observers, good architects, etc.," Kosslyn said in an e-mail.

In the past, the intangibility of visualization and doubts over its existence prevented any breakthroughs in the field.

But according to Kosslyn, three advances have overcome these difficulties.

The first was the development of tasks which require imagery.

The second was the discovery of spatial organization in specific parts of the brain. At the surface of the brain, called the cortex, sheets of tissue preserve images from the retina of the eye in patterns of activity in neural cells.

The third was the development of brain-scanning technologies, specifically PET (positron emission tomography), which can determine if patterns of neural activity in the brain are present during certain tasks. Using such technology, researchers have found that spatially organized parts of the brain are active in different places depending on the size of objects being viewed.

Kosslyn, along with researchers William L. Thompson, Irene J. Kim and Nathaniel M. Alpert, used these advances to determine whether tasks involving imagery would activate parts of the brain where spatial patterns of activity occur during perception.

Specifically, participants in the study engaged in various visually-based activities while a PET machine scanned their brains. Unlike other studies in Sweden and France, where spontaneous visualization of individuals was not taken into account, the Harvard researchers took careful steps to include this effect in their data. As a control, they scanned the brains of subjects told to rest with their eyes closed in a dark room.

Four tests, based on the same scheme, were used in the experiment. Participants first heard phrases such as "anchor...right higher?" with no knowledge of the imagery task.