Harvard Magazine
Main Menu · Search ·Current Issue ·Contact ·Archives ·Centennial ·Letters to the Editor ·FAQs

In Dr. Jeremy Schmahmann's hands, an inverted model of the brain, displaying the cerebellum--the darker lobes--front and center.Photograph by Paul Janovitz

The patient was a high-powered Boston executive. her job was making hard decisions, thinking on her feet, juggling several projects at a time. But one day she realized that something had gone wrong. She found her concentration wavering; decisions that were normally second nature came slowly; and handling many things at a time seemed impossible. She could think all right, if a bit slowly, but her high-level administrator's skills were crippled. She was having trouble managing her thoughts.

She had noticed something else. Her hands trembled slightly. That may have made her suspect a medical problem. In any case, she decided to seek a medical explanation by visiting Massachusetts General Hospital, where evaluation revealed she had experienced a stroke in a part of her brain called the cerebellum.

To old-time neurologists, this connection between the cerebellum and thought management would have seemed bizarre. Up through the 1980s, standard doctrine held that the sole job of the cerebellum (located under and behind the cerebral hemispheres) was coordinating movement. But recent knowledge has implicated the cerebellum in coordinating lots of things--including incoming data from our senses, visual orientation, language, and complex thought.

That has important scientific and medical implications. Among other things, our new knowledge of the cerebellum has revealed a definite pattern of medical symptoms arising from cerebellar troubles. That helps with both diagnosis and treatment. In the case of the Boston executive, for instance, she could be told that her symptoms weren't imaginary and that she wasn't going crazy. Her stroke could be properly treated; and she could receive a type of therapy designed to cope with cerebellar damage.

Beyond that, knowing what the cerebellum does is key to unlocking secrets of the mind. Brain science has made striking advances in recent times, revealing much about nerve cells, brain evolution, psychoactive drugs, and brain chemistry. But we didn't build the brain, so we don't have its working blueprint. To understand it, we must learn what its parts do and how they talk to one another. The cerebellum, both large and mysterious, plays a big role in the brain's work. So it is not surprising that what we are learning now about the cerebellum shows promise of shedding light on mental ills, general workings of the brain, and even the nature of human consciousness and thought.

Dr. Jeremy SchmahmannPhotograph by Paul Janovitz

A prime contributor to this new stream of knowledge is Jeremy Schmahmann. In his early forties, with jet-black hair and a ready smile, Schmahmann wears several hats. He is an associate professor at Harvard Medical School who lectures on neurology, practices neurology at Mass. General, and keeps a research office at the hospital. He is also clerkship director for the hospital's neurology department, which means he directly supervises every Harvard medical student working in neurology who comes to the hospital.

Schmahmann, who was born in Durban, South Africa, could prove apartheid's greatest accidental gift to American neuroscience. Belonging to a white liberal minority that viewed apartheid as morally tainted, he was repulsed by rising levels of violence and repression that made him feel his homeland would be a poor place to work and raise a family. He had been an American Field Service student in 1974-75 and had found the freedom and promise of America exciting. So upon completing his internship he left South Africa in 1982, did his medical residency in Boston, and joined the Harvard Medical School faculty in 1989--the same year he became a full-time staff member at Mass. General.

Soon after he reached Boston, Schmahmann began targeting the cerebellum. At the time, few would have found this research choice exciting. Brain anatomy is tedious; and the cerebellum, residing in the cranial basement, is hardly the brain's flashiest part. Indeed, in the 1980s some experts still questioned whether it was really needed. Thus, while the cerebrum appeared to be a mighty sorcerer performing incredible feats (like creating consciousness, reconstructing pictures of the world, and tying tenuously related ideas together in the mind), the cerebellum seemed almost menial, roughly equivalent to the sorcerer's apprentice.

Recent knowledge has implicated the cerebellum in coordinating not only motor tasks but also the brief "working" memory that allows us to remember a telephone number from the time we look it up to the time we dial it. These brain-scan images show certain parts of the cerebellum activated when subjects were asked to remember up to six letters (left column). Merely reading and repeating the letters subvocally ("motoric rehearsal") produced less activation, while finger-tapping (right column) activated different areas.REPRINTED WITH PERMISSION: J.E. DESMOND ET AL., JOURNAL OF NEUROSCIENCE, VOL. 17, NO. 24 (DECEMBER 15, 1997): 9675-85

But there were mysteries. If the cerebellum's job was so modest, why was it so big? Why did it contain as many neurons as the rest of the brain combined? And why was it tied to the rest of the brain by huge nerve-cell conduits capable of transmitting roughly 40 times the flood of information carried from the eyes to the brain by the mighty optic nerve?

There were other questions, too. We humans pride ourselves on the deep folds and fissures of our cerebral hemispheres, believing this Nature's way of cramming great mental capacity into a small space. Why, then, was the cerebellum more deeply folded and fissured than the cerebrum? Why did the cerebellum seem to have grown relatively faster than the cerebrum over the course of human evolution? And why did parts of the cerebellum appear to have evolved in concert with parts of the cerebrum concerned with "higher order" thought?

The actual spark that got Schmahmann going was the medical problem of "neglect"--defined by doctors as a patient's unawareness of some spatial area or part of the body (commonly the left half) after the brain has been damaged by a stroke or other mishap. "A man with left-side neglect will dress the right side of his body or shave the right side of his face and ignore the left side, or hear something on his left side and look over to his right," Schmahmann explains. "What intrigued me was cases of neglect apparently caused by damage to a part of the brain deep in the cerebrum, a structure known as the caudate nucleus, that was supposed to deal exclusively with motion [motor tasks]. Since neglect clearly involves things other than motor tasks, the caudate nucleus seemed to be doing jobs other than those commonly ascribed to it. And if the caudate was doing more than motor tasks, what about the largest brain structure supposedly dedicated exclusively to motor tasks?" In other words, what about the cerebellum?

Schmahmann felt the cerebellum's true role might be assessed by exploring its connections with the cerebral cortex, the "gray matter" covering the surface of the cerebrum. Particular parts of that cortex tend to specialize in particular jobs, like thousands of small shops in a great factory. If one could show that the cerebellum was communicating mostly with motor "shops," that would tend to confirm a largely motor role; while if its connections were mainly with nonmotor shops, that would suggest it was doing other things.

But exploring brain connections isn't easy. If you slice below the gray matter of the cerebral cortex (which consists mostly of nerve cell bodies and receptor branches called dendrites), you will find that most of the brain matter below is white. This white matter consists mainly of the nerve cells' long impulse-transmitting axons, which are insulated by specialized cells that sheathe them in protective white coats. Many of these axons bundle together into the brain's equivalent of transmission cables. Separating the components of these cables and finding out just where they come from is time-consuming and challenging enough to try the patience of a saint.

Schmahmann knew that virtually all messages traveling from the cerebral cortex to the cerebellum went initially to a relay station in the brain stem, the portion of the brain that joins the spinal cord. This relay station is found in a bulge at the upper end of the brain stem, the pons ("bridge" in Latin). So Schmahmann asked Deepak N. Pandya, a pioneering brain anatomist at Boston University School of Medicine, for help in tracing pathways from nonmotor parts of the cerebral cortex to the cerebellar relay station in the pons.

Magnetic resonance images of the cerebellum showing greatest activation by essentially nonmotor tasks. Colored areas show relatively mild (green) to intense (red) activation when subjects (A) had their fingers rubbed with sandpaper; (B) compared the coarseness of different sandpapers; (C) raised and dropped small irregular objects; and (D) compared the shapes of different grasped objects. The study focused on the dentate nucleus, a key part of the cerebellum that appears as two small dark crescents in C and that was most highly activated by the two tasks requiring comparative analysis (see B and D). REPRINTED WITH PERMISSION: L.M. PARSONS AND P.T. FOX, "SENSORY AND COGNITIVE FUNCTIONS," IN THE CEREBELLUM AND COGNITION (J.D. SCHMAHMANN, ED.), INTERNATIONAL REVIEW OF NEUROBIOLOGY, VOL. 41 (ACADEMIC PRESS, 1997). THIS ILLUSTRATION WAS ADAPTED FROM J.H. GAO ET AL., SCIENCE 272 (1996): 545-547, WITH PERMISSION

Their work, done in rhesus monkeys, involved injecting a radioactive tracer into particular parts of the cerebral cortex to see whether the tracer traveled through the brain's white matter to the pons. Initial results showed direct connections to the pons from a part of the cerebrum dealing with sensory information and attention. Later results revealed inputs to the pons from all sorts of nonmotor areas--including parts of the brain specializing in memory, vision, spatial orientation, touch, language processing, planning, foresight, judgment, attention, motivation, emotion, and integration of higher-order behavior. What's more, Schmahmann and Pandya found that the wiring was specific--as though each tiny connected part of the cortex had its own specific telephone line, and its own specific receiver in the pons. As Schmahmann puts it, "We found a predictable topography organized in such a way that each cerebral cortex area had its own region of the pons with which it was talking."

Some years later, while Schmahmann and Pandya continued tracing these connections, a research group led by Peter Strick, of the State University of New York Health Science Center at Syracuse, found major nerve connections leading back from the cerebellum to various parts of the cerebral cortex, proceeding by way of an intricate gray-matter structure called the thalamus.

For anyone interested in the brain, the thalamus rates special attention. Centrally positioned at the base of the cerebral hemispheres, it is large, mysterious, complex, well-connected to other brain areas, and closely associated with human consciousness. We know from the famous case of Karen Ann Quinlan, as well as from other coma victims, that damage to the thalamus can leave someone alive but permanently deprived of consciousness--one reason why an array of theorists investigating human consciousness are now focusing on the thalamus and its connections.

As shall be seen, what we are learning about the cerebellum appears to be shedding light upon the thalamus. But the latter's complexity made it hard for Strick's team to figure out what happened to nerve impulses from the cerebellum after they reached the thalamus. Did those impulses proceed onward to specific parts of the cerebral cortex or did something else occur?

To resolve this question, Strick's group used a reverse tracer powerful enough to start in the cerebral cortex, pass through connections in the thalamus, and show up in the cerebellum. Their results showed (irrespective of whatever else might be happening) that nerve pathways starting in the cerebellum proceeded via the thalamus to various nonmotor parts of the cerebral cortex.

These findings, in combination with Schmahmann and Pandya's work, revealed a large information highway feeding messages out from the cerebral cortex to the cerebellum (via the pons) and feeding processed data back from the cerebellum to the cerebral cortex (via the thalamus). Thus while the actual brain geometry was convoluted, the arrangement could be drawn to look like a baseball diamond--with the cerebral cortex at home plate, the pons at first, the cerebellum at second, and the thalamus at third.

Does the cerebellum help turn our thoughts into words? Growing evidence suggests that it does. These positron emission tomography (PET) images show activation of experimental subjects' right cerebellar hemispheres after they were given nouns and asked to state related verbs. The left image shows relatively great activation at the start of the exercise; the middle image shows less activation after the subject got practice generating verbs related to the initial noun; and the right image shows renewed activation when new nouns were introduced. ADAPTED WITH PERMISSION: M.E. RAICHLE, J.A. FIEZ, T.O. VIDEEN, A-M.K. MACLEOD, J.V. PARDO, P.T. FOX, S.E. PETERSEN, "PRACTICE-RELATED CHANGES IN HUMAN BRAIN FUNCTIONAL ANATOMY DURING NONMOTOR LEARNING," CEREBRAL CORTEX, VOL.4 (1994): 8-26

Around the time when Schmahmann and Pandya started tracing connections to the cerebellum, other researchers began chipping away at the orthodox belief that the cerebellum did only motor tasks. Some related evolution of the cerebellum to evolution of parts of the cerebrum dealing with cognition (perception, thought, and memory). Others found links between autism (the Rain Man disorder) and cerebellar defects. Still others demonstrated that the learning of conditioned responses required the cerebellum--that without it Pavlov's dogs would never have learned to salivate at the sound of a bell.

As often happens in science, no one research project proved conclusively that the cerebellum was doing cognitive tasks. But the mounting evidence was approaching a critical mass. Researchers with pet theories about what the cerebellum might be doing began coming forward with their ideas. And new imaging techniques began confirming in a compelling and dramatic way that the cerebellum was indeed involved with higher thought.

Inspired by all this, in 1991 Schmahmann wrote an article for the Archives of Neurology titled "An Emerging Concept: The Cerebellar Contribution to Higher Function." There he reviewed the accumulated evidence and presented a new theory suggesting that just as the cerebellum regulates the rate, force, rhythm, and accuracy of movement, it also regulates the speed, capacity, consistency, and appropriateness of thought.

Put another way, Schmahmann and other theorists came to feel that the cerebellum's business was to detect, predict, and correct errors. It didn't matter whether the errors involved movement, perception, memory, thought, or a mixture of all four. The cerebellum was receiving a vast flow of data from parts of the brain dealing with all these things; and while nobody knew precisely how the cerebellum worked, they did know that it had a very regular structure with immense computing capacity, so it seemed likely that it was modulating and harmonizing many kinds of information--much as a skilled chef effectively blends and harmonizes many kinds of food. In the cerebellum's case, according to Schmahmann, "the aim is to take information, smooth it out, and make it harmonious with the intended goal--regardless of whether this is a motor goal or some other goal. This implies that when you lose the cerebellum's contribution to a given task, the task may still be performed, but will be performed poorly."

Exciting as all this seemed to Schmahmann, it created no great enthusiasm among senior neurologists at Mass. General. As he recalls, "When I presented the results of our work at grand rounds some years back, these venerable older neurologists said, 'Nice, but if patients don't have behavior problems from their cerebellar lesions, what's the clinical relevance? It's fine and dandy to speculate about this stuff, but if there's a cerebellar contribution to cognition, show us!' So I accepted the challenge, collected 20 patients with brain lesions involving only the cerebellum, and worked with Janet Sherman, a neuropsychologist, to define a behavioral syndrome resulting from cerebellar lesions."

Thirteen of the 20 patients had suffered strokes, three had cerebellar infections, three had cerebellar atrophy, and one was recovering from removal of a cerebellar tumor. Schmahmann and Sherman studied these patients over a period of seven years, giving them periodic neurological examinations and comprehensive psychological tests. The results showed a clear pattern of behavior changes that typically included reduced intellectual ability (reduced I.Q.); impaired "executive function" (reduced ability to plan, reason, make decisions, initiate actions, and change strategies); and impaired "working memory" (lessened ability to hold information in one's head briefly in order to manipulate it). The patients also had trouble with spatial organization, language, grammar, and emotions--the latter tending to be very flat and to be associated in some cases with uninhibited and inappropriate behavior.

As Schmahmann recalls, "The combination of flattened emotion and disinhibition that appeared in some patients was quite striking."

These drawings based on the work of Schmahmann and Pandya show connections between the rhesus monkey's cerebrum (left) and pons (right). Colored parts of the cerebrum send signals to correspondingly colored parts of the pons, from where they go to the cerebellum. Green parts of the cerebrum deal mostly with motor tasks; but the red, orange, blue, and purple areas deal with many other things--including thought, memory, planning, judgment, and attention. The areas shown are an inverted cross-section of the cerebrum (top left), a side-view of the cerebrum (bottom left), and three cross-sections of the pons.REPRINTED WITH PERMISSION: J.D. SCHMAHMANN, HUMAN BRAIN MAPPING, VOL. 4 (1996): 174-198

"One of the first patients with this problem was a woman in her early twenties recovering from removal of a cerebellar tumor. The surgery was very smooth and uncomplicated. But after she woke up and began walking around the ward, the nurses noted that she was behaving very strangely. She was distractable, inattentive, impulsive, giggly, disinhibited (to the point of disrobing in the middle of the corridor), cheeky to her parents (which was very unusual for her), and in general exhibiting a complete personality change. As the months progressed, it became apparent that she had difficulty making the simplest decision. She would ring her mother up at work and say, 'Mom, how do I make a chicken sandwich?' Fortunately, she recovered just about completely and is now leading a normal life."

Overall, the study group exhibited behavior changes consistent and pronounced enough to constitute an identifiable disease syndrome. Imposingly dubbed the "cerebellar cognitive affective syndrome" by Schmahmann and Sherman, the discovery is significant. As Schmahmann points out, "Knowing about the syndrome can aid diagnosis, reassure patients and their families, and help determine whether the patient needs mental exercise therapy suited to recovery from cerebellar damage."

Nevertheless, it was not this discovery so much as brain imaging that made the cerebellum's expanded role credible. "Imaging was the giant," says Schmahmann, "because with clinical cases people were never quite sure what might be going on elsewhere in the brain. But when functional imaging came in, you could see the cerebellum working....In 1994, when I went to the First International Conference on Functional Mapping of the Human Brain in Paris, a wide range of imaging study slides were presented on nonmotor functions like attention, or working memory, or language, and it was phenomenal how often you could see cerebellar activation on the screen. Once this became apparent, people began to look specifically at the cerebellum's cognitive role."

But brain anatomy is tricky, and there was no formal guide explaining which parts of the cerebellum were doing what. So the imaging experts were winging it, using their own words to locate what they saw, and producing descriptions that were often vague or misleading.

Therefore, Schmahmann decided to devise an "atlas." Working with investigators in Canada and California, and with several of his own students, he developed a basic model and then mapped the results of 47 published brain-imaging studies onto this model to refine it. That work yielded the first topographic map of the cerebellum's "higher order" functions, a crude guide relating different cerebellar regions to different jobs.

"While the approach was fraught with problems," Schmahmann says, "it looks like there is indeed a motor cerebellum and a cognitive cerebellum. When you wiggle your finger, or your foot, or your tongue, different sites in the motor portions of the cerebellum will show the greatest relative activity. Likewise, in the cognitive regions, language tasks tend to activate certain lobules, while other tasks requiring working memory, or attention, or imagined motion most strongly activate other areas."

By then a wide range of related theories had emerged. Not all agreed with Schmahmann's theory of modulation, but the points of real difference tended to be small. That seemed reasonable, because it was known that people with severe cerebellar damage could move, think, and function. They just did so less well than if their cerebellums were intact. Therefore, it seemed clear that the cerebellum fine-tuned these functions, coordinating a vast range of activities without departing from its basic role as the sorcerer's apprentice.

Meanwhile, as animal and clinical research continued, brain imaging by many scientific teams started to define the cerebellum's role in such acts as identifying objects by touch; exercising conscious thought; generating words; processing music and other sounds; mentally rotating abstract objects; influencing emotions such as sadness, depression, and fear; learning repetitive skills (singing on key or serving tennis balls, for instance); and storing the information needed to exercise those skills. By 1995 some 30 teams in diverse places, from Harvard to New Zealand, were publishing papers on the cerebellum's role.

Seeing all this productive effort, Schmahmann decided to compile a book that would report what was happening. So he got the leading investigators in the field to submit articles on chosen subjects. The resulting book, The Cerebellum and Cognition, published in 1997 as a single volume of the International Review of Neurobiology, does not answer all prevailing questions about the cerebellum. But it leaves no doubt about the cerebellum's cognitive role, gives a broad overview of current research, and announces clearly that this field has come of age.

Still, one is tempted, like an elder neurologist, to ask, "so what?" We see now that our new learning can help deal with medical matters relating directly to the cerebellum. But the cerebellum is still the sorcerer's apprentice. Is there any reason to suppose that this fresh knowledge can help us to explore major issues like the origin of common psychoses or the nature of human consciousness?

While nobody knows, the most likely answer is that it can. To begin with, our new view of the cerebellum shows promise of proving important for psychiatry. Right now we don't know what causes many of the most devastating disorders like schizophrenia, mania, and depression. But they all involve failure to relate to reality. "So," as Schmahmann explains, "if the cerebellum is involved in cognitive and emotional behavior, and if a cerebellar lesion can produce cognitive dysmetria [a mismatch between reality and perceived reality], you are now frankly in the psychiatric realm."

Researchers led by Nancy C. Andreasen, A.M. '59, at the University of Iowa have taken this concept and applied it directly to schizophrenia. Their work, dealing with connections linking the cerebellum, the thalamus, and the cerebral cortex, suggests that disturbances in these connections can cause poor mental coordination, which in turn can account for schizophrenia's diverse symptoms. Should this theory prove correct, our new insights into the cerebellum could be putting us on the right track for discovering the cause of humanity's most notorious, severe, and pervasive disorder of the mind.

Our growing knowledge could provide insight into other brain mysteries as well. Consider, for instance, the enigma of human consciousness. In the past, certain experts tended to shrug their shoulders and say that consciousness is "distributed" about the brain--that when you get a lot of neurons together, it just "happens." The main problem with this is that the brain is so specialized: different parts of the cerebral cortex do different jobs. And your brain is not like a computer, where impulses travel at the speed of light. Brain messages plod along relatively slowly, typically at about 60 miles an hour; so if one brain site wants to send a complex message to another site, it needs to send a lot of information at one go. That requires a substantial neuron cable. And that would mean, if many areas had to talk back and forth with many other areas to coordinate things in consciousness, your brain would consist of cables and little else.

But things do get together in consciousness. We know that. Our conscious minds see a world in which touch, vision, hearing, language, motion, memory, higher thought, attention, speech, and emotion are all included. That suggests that there may be something resembling a "seat of consciousness" within the brain where messages registering in consciousness are sent.

Even so, we still need to explain a towering coordination problem. As we all know from experience, the innumerable bits of information getting into consciousness don't just "register" there but seem smoothly coordinated. Our view of the world appears seamless. Vision seems smoothly coordinated with touch and hearing, thought with memory and attention. Likewise, diverse movements seem smoothly coordinated with one another as well as with sensory inputs, attention, and higher thought. Indeed, everything seems coordinated in minute detail with everything else--a major hurdle for theorists trying to figure out how any seat of consciousness could work.

Enter the sorcerer's apprentice. If, as now seems likely, the cerebellum is doing most of the brain's fine-tuned coordination, not just for our muscles but for our senses, memories, emotions, thoughts, and attention, then the best place to look for such fine-tuned coordination has been found. This doesn't mean the cerebellum coordinates everything, because other brain centers communicate and coordinate with one another outside the cerebellum. But input from the cerebellum seems key to the sort of consciousness we know.

That could be important--because, as we have seen, most processed information from the cerebellum goes first to the thalamus before being sent elsewhere. And, as already noted, the thalamus is mysterious, sizable, centrally located, and well-connected. If anything, "well-connected" is an understatement, because nearly all incoming sensory messages--on everything you see, hear, touch, or taste--go first to the thalamus before being sent elsewhere. In fact, virtually everything needed to establish consciousness as we know it comes to the thalamus--not just new data from sense organs and processed data from the cerebral cortex and emotion centers, but also a vast flood of finely coordinated data received from the cerebellum. Seen this way, our new view of the cerebellum appears to provide a reasonable explanation of why our conscious view of the world is so well coordinated, and also supports evidence suggesting that if a seat of consciousness exists, its probable headquarters is in the thalamus.

Of course, one must be careful here. The evidence is circumstantial. We still don't know how consciousness works; and despite the brain's apparent need for a seat of consciousness, nobody has ever proved that such a thing actually exists.

Even so, Jeremy Schmahmann and other students of the brain have clearly come a good way down the road to full discovery. Around the turn of the last century, the brain seemed so mysterious that questions about consciousness and thought were relegated to philosophers and religious thinkers. Today, on the eve of the millennium, we have learned volumes about the brain; we have even started to learn the truth about how the brain's complex array of substructures interact; and while we still lack suitable answers to most "big" questions, we may well be approaching a point where such questions can be sensibly considered, and where much or even most of the brain's work is understood.

Contributing editor Jonathan Leonard '63 wrote "Dream-Catchers," published in the May-June 1998 issue of Harvard Magazine.

Main Menu · Search ·Current Issue ·Contact ·Archives ·Centennial ·Letters to the Editor ·FAQs
Harvard Magazine