Mapping the functions of various areas
                     of the human brain is difficult--and controversial

       In the 19th century, practitioners called phrenologists divided the surface of the human
       brain into 35 different regions, each of which was thought to contribute to a certain
       aspect of personality, such as "spirituality," "mirthfulness" or "conjugality." The
       phrenologists claimed to discern someone's character by the location and size of the
       bumps on his or her head. A protrusion over the "conscientiousness" area, for instance,
       meant that the person was punctilious to the degree that that particular brain region had
       grown from use, much as a muscle does after repeated exercise.

       Now, more than 150 years later, some researchers have begun to ask whether modern
       attempts to "map" the functions of various regions of the cortex--the brain's "gray
       matter"--essentially come down to using high-tech methods to do the same thing the
       phrenologists claimed to do. "There are people who scorn the idea that various areas of
       the cortex have unique functions," observes Robert Desimone, director of the National
       Institute of Mental Health's Division of Intramural Research Programs. "They call it
       ‘neurophrenology.'"

       And those who believe that fine functions--such as seeing colors or hearing certain
       sounds--can be attributed to small patches of cortex sometimes disagree strenuously
       over where to draw the margins of those patches. In 1998, for instance, a scholarly
       battle raged in the pages of Nature Neuroscience between Roger B. H. Tootell and
       Nouchine Hadjikhani of Massachusetts General Hospital and Semir Zeki and his
       colleagues at University College London. At issue was whether Tootell, Hadjikhani
       and their co-workers had identified a new area responsible for conscious color
       perception within the visual cortex, which is at the rear of the brain, or if they had
       simply "rediscovered" an area that Zeki had previously laid claim to. The issue still has
       not been settled.

       Part of the problem arises because some researchers analyze the brains of rhesus
       macaques, whereas others focus on imaging human brains or studying patients who have
       suffered injuries or diseases that affect only particular brain regions. Often areas that
       appear to have one function in monkeys do not play the same roles in humans. In
       addition, the brains of individual monkeys and humans can differ slightly, making it
       very difficult to be certain that researchers are looking at the same spots in two or more
       brains.

       Pinning down the function of particular brain areas has been made feasible by the
       development of functional magnetic resonance imaging (fMRI). Unlike other imaging
       methods, fMRI allows researchers to monitor local cerebral blood flow--a marker of
       brain activity--without administering radioactive materials or magnetic contrast agents.
       But fMRI machines are expensive to run, and so far relatively few neuroscientists have
       them.

       Josef P. Rauschecker and his colleagues at Georgetown University Medical Center
       have recently used the fMRI technique to create a detailed functional map of the
       auditory cortex, which is situated on either side of the brain. They have found that the
       auditory cortex is divided into separate fields that process sound information in a
       hierarchical fashion. Core areas at the center of the region analyze pure tones; so-called
       belt areas surrounding the core areas respond to several tones combined into a more
       complex, buzzlike stimulus.

       The idea of hierarchical processing--that the brain initially extracts from stimuli their
       most basic features and then builds them up again to reflect the complexity of the
       world--originated in the 1970s with studies of the visual cortex. But for many years,
       scientists favored the view that the auditory cortex decomposed sounds into many
       single frequencies and processed them in parallel.

       Rauschecker's new work should stir the pot. "There are people who think that pure
       tones are the best to map," he comments. "But you have to put the information together
       again to hear a voice or a complicated sound."
 

       --Carol Ezzell