|A book by|
William H. Calvin
UNIVERSITY OF WASHINGTON
SEATTLE, WASHINGTON 98195-1800 USA
The Throwing Madonna|
Essays on the Brain
Copyright 1983, 1991 by William H. Calvin.
You may download this for personal reading but may not redistribute or archive without permission (exception: teachers should feel free to print out a chapter and photocopy it for students).
Scanned, OCR'ed, and webbed -- but NOT proofread (14 Jan 97)
Probing Language Cortex: The Second Wave
The highest activities of consciousness have their origins in physical occurrences of the brain, just as the loveliest melodies are not too sublime to be expressed by notes.
W. SOMERSET MAUGHAMThe problems and findings at other levels are usually largely irrelevant for those working at a given hierarchical level. For a full understanding of living phenomena every level must be studied but . . . the findings made at lower levels usually add very little toward solving the problems posed at the higher levels. When a well-known Nobel laureate in biochemistry said, "There is only one biology, and it is molecular biology," he simply revealed his ignorance and lack of understanding of biology.
Working from the bottom up, neurobiology focuses upon ion channels, collections of channels and sheet insulation called membranes, diverse collections of membranes called nerve cells, collections of varied nerve cells called circuits, modular collections of circuits exemplified by the leech's ganglia and the visual cortex's hypercolumns--but then what next?
Working from the top down, we can distinguish linguistics, instinct, cognition, memory, hemispheric organization, cortical maps, but then. . . ? How do modular super-hypercolumns generate grammar, store the word "rabbit," recognize a fuzzy animal as a rabbit, pronounce the word "rabbit.~r set about throwing a stone at a rabbit? What are the missing levels that link our bottom-up and top-down approaches to the brain? Some idea can be gained by taking the language system of the brain and making some guesses about the missing links.
Not only is language a system of prime importance in our endeavor to understand ourselves, but it has visual aspects (reading), auditory aspects, memory aspects (as in naming, or recalling a recent word), emotional and selective-attention aspects, advance planning aspects, and motor aspects. Even a partial understanding of how it integrates these aspects could prove most helpful in approaching the frontal lobe role in strategy, the right temporal lobe role in emotional face recognition, or the parietal lobe role in keeping track of "extrapersonal space" into which one might move a limb.
The story so far is still monolithic enough to be told as a historical narrative. A century ago, there was much emphasis upon the size of the head, many attempts to find a bigger-is-better correlation (recounted by Stephen Jay Gould in The Mismeasure of Man). Since the cerebral cortex was obviously the part of the brain that was so much bigger in humans than in the brains of the great apes, it was considered the seat of higher functions such as language. The French neurologist Paul Broca had even identified some parts of the cortex in which stroke damage caused aphasia, and the identification of language with cortex was considered firm.
Language physiology, as opposed to the evidence from aphasia, started in the 1930s; Otfrid Foerster in Germany was a pioneer, but the most extensive observations, spanning three decades, were made in Montreal on epileptic patients undergoing brain operations under local anesthesia. Wilder Penfield, a neurosurgeon with training in physiology, used electrical currents to stimulate the surface of the brain as part of making a brain map for that particular patient, prior to deciding exactly which epileptic parts of the brain could be safely removed without causing more trouble than the epilepsy he was trying to cure. This technique was developed by the pioneering Boston neurosurgeon Harvey Cushing in 1909; what Foerster and Penfield did was to apply it to the study of language.
Penfield's researches, especially those done in collaboration with Herbert Jasper and Lamar Roberts, revolutionized our concept of brain maps in the 1950s. They showed aspects of both language and long-term memory. There was a region in the frontal lobe, corresponding to the "Broca's area" of the aphasia literature, where stimulation blocked language output: the right words would come out just as soon as the stimulation stopped. There were sites where the patient could talk but not find the right word (anomie: "That's a . . ., you know, the thing you put in a lock"). And there were sites where specific memories could be evoked: the so-called experiential responses. These "evoked memories" were sometimes musical when they occurred (which is in only a small percentage of patients tested). They always make for a good story: one young Seattle man heard rock music every time a certain right temporal lobe site was stimulated. And he was sure that it was music by a particular rock group, Led Zeppelin. The Montreal patients seemed to have more classical tastes. But in any case, it is a rare phenomenon and probably part of the patient's epileptic process rather than a normal memory process.
But this early work in Montreal was essentially the first look at terra incognita, and the more detailed exploration and mapping did not occur until a quarter-century later. This "second wave" took place primarily in Seattle, and while it grew out of the Montreal tradition, it also retraced a fin de siecle argument of the Paris school of neurology and linguistics.
Language in the Depths of the Brain
A French neurologist, Pierre Marie, said that the deeper (and hence more primitive?) parts of the brain such as the thalamus also played a role in language. This caused another stressful argument in the Parisian academic circles about the turn of the century, reminiscent of the arguments that had caused the Linguistic Society of Paris to ban any presentations on the origins of language starting in 1886 (they reconfirmed the ban in 191 1). There the issue lay, for six decades.
In the early 1960s, many neurosurgical centers were treating Parkinson's disease by controlled damage to part of the thalamus in the brain's depths, essentially trying to bring the muscle stiffness system back into balance again. Arthur A. Ward, Jr., was building up a physiologically oriented neurosurgery department in Seattle at the University of Washington; he had trained in one of the leading physiology departments of the day and then taken his neurosurgical training with Wilder Penfield. One of Ward's specialties had become the treatment of Parkinsonism by sticking a long needle into the thalamus under x-ray control, identifying the sites physiologically, and then heating a small area to destroy it. At the time, he was training a young neurosurgeon, George A. Ojemann. They noted that when they electrically stimulated the thalamus at subliminal levels prior to making the destructive lesion, the patients had trouble talking. Language in the thalamus? Shades of Pierre Marie.
Being good physiologists, they began to plan how to take this surgical opportunity in awake patients, in order to learn more about what language was doing in the thalamus. George Ojemann went off to the Clinical Center at the National Institutes of Health near Washington, D.C. Arriving at the same time was a psychology Ph.D., Paul Fedio, and together they started up a project to test for language in the thalamus. They designed a series of tasks for the patients to perform that would test specific aspects of both language and memory for verbal material: the patients would watch a slide show on a movie screen in the operating room and name the objects they saw, then recall them from memory some seconds later after performing another task designed to keep them from rehearsing (counting backward by threes is such a consuming mental task). They were originally interested in naming: if the patient didn't respond during stimulation, was consciousness being altered? But as we shall see later, it was the memory aspect that proved to be the most interesting.
Two years later, Ojemann returned to Seattle to join the faculty and started thalamic experiments in earnest. Ward had designed a neurosurgical operating room (O.R.) ideal for physiological experiments, with copper shielding built into the walls, ceiling, and floor to prevent electrical interference (remember the screen wire cage in Jerusalem?). This was not as essential for the thalamic operations as for the epilepsy operations, as Ward had continued the Montreal tradition of doing epilepsy operations under local anesthesia, using physiological mapping of the exposed cerebral cortex to help make decisions and to further our basic knowledge of the human cortex. It was the combination of the two surgical procedures typically done under local anesthetic-- thalamotomy on Parkinson's victims and cortical removal for epileptics--that proved to be so important in the evolution of the language and memory studies, in what became the ~second wave" of language physiology.
Language Localization in the Cerebral Cortex
The slide show for testing Parkinsonism patients in the O.R. was gradually used during epilepsy operations as well. Neurosurgeons have often tested for language during epilepsy operations: the location of language areas cannot be judged from a standard map, as humans have proven to be quite variable, with many individual aspects to where language is located. But the testing was often informal, making it hard to progress past the point attained by Penfield and Roberts. With the slide show and stimulation protocols developed for the thalamic operations, a new era opened up in cortical language physiology. At the NIH, Paul Fedio and John Van Buren made a series of observations, beginning just before George Ojemann left to return to Seattle and continuing until about 1970. In Seattle, Ojemann tested his own patients as well as those operated upon by fellow neurosurgeons. But while the thalamic observations were frequent, progress was slow on testing cortical language organization, with about one patient per year tested up until about 1976, when the pace quickened. These observations established that there were sites in the cortex that disrupted short-term memory for verbal material (stimulation while the patient was counting backward may not have affected the counting, but it certainly affected later recall of the prior material).
Several things happened in the mid-1970s to speed the language research. First, there was the completion of another research project being done on epileptics in the O.R., where Ojemann, Ward, and I studied single neurons in the epileptic brain, comparing their firing patterns to those of induced epilepsy in research animals, trying to keep the animal research on the right track by comparing it to the real thing. Its completion meant that some time, always exceedingly valuable in the O.R., was freed up. But most important was a sabbatical year spent in Seattle by a neurolinguist, Harry Whitaker, editor of the journal Brain and Language.
"Whit" had for many years studied patients with various injuries to the brain (his studies with Maureen Dennis, of language moving from the left to the right brain in children with congenital left-brain damage removed by surgery, were mentioned in Chapter 10). And here was the opportunity to apply many neurolinguistic techniques to fairly normal patients in which the stimulation produced only a temporary blockage-- harmless and on demand, at the press of a button on an electrical instrument. The neurolinguists' tricks-of-the-trade began to be adapted to the neurosurgical and physiological setting.
The Bilingual Brain
Ojemann and Whitaker found that the standard maps of language were, in fact, only average: language maps were indeed quite individualized, just as are our faces. One of the particularly memorable studies initiated during Whit's sabbatical resulted in an article entitled "The Bilingual Brain," where bilingual epileptics were tested in both languages. Electrical stimulation of some sites affected performance in naming simple objects in both languages--but at other points, it would affect only English and not the other language, or only the other language and not English. Just moving 5 millimeters along the top of a gyrus could change the language that was disrupted.
Separate brain for different languages? Bilingual patients are not often seen in the U.S.; the definitive study on the subject will probably be done somewhere else in the world where people are typically multilingual. It would be especially desirable to compare closely related languages and widely different languages (the languages compared to English in Seattle happened to be Dutch, Spanish, and Greek--all from the Indo-European family of languages). A start has been made on a more diverse language comparison. Richard Rapport did his neurosurgical training in Seattle and then spent a year in Kuala Lumpur, where the three Malaysian epileptics on whom he operated were fluent in both Chinese and English. They too had separate sites for each language, plus the usual overlapping sites.
Sex and IQ
The individual variation from an average map is not just a random business; some of it is clearly related to sex and IQ, as Catherine Mateer, George Ojemann, and Samuel Polen (a professor of speech pathology and audiology on sabbatical from Western Washington University) have recently shown. The area of cortex where naming can be disrupted is somewhat smaller in women than in men. This had been suspected because left-brain strokes cause aphasia much less often in women than in men.
The maps also vary with the verbal IQ of the patient (in a manner independent of sex). Only one of nine patients with an above-average verbal IQ had naming disruption from the parietal lobe. Seven of nine patients with below-average verbal IQs (range 69-96) had naming disrupted at parietal lobe sites. The issue here is not the total amount of cortex devoted to language, but its extension into the parietal lobe above the Sylvian fissure. And the above-average-IQ patients might also use that cortex for language, with stimulation there simply incapable of disrupting the rest of a well-organized system elsewhere; it simply isn't known how to interpret the results yet. It seems likely that some arrangements of language cortex are more efficient than others.
Is Language Superimposed on Motor Sequencing?
Another important development took place when Catherine Mateer came to Seattle in 1977 as a postdoctoral fellow (she later joined the faculty). As recounted in Chapter 4, she and Doreen Kimura had established that left-brain strokes affecting language also disrupted motor sequencing, for both hands and for both sides of the face. Mateer and Ojemann then went on to test oral-facial sequencing in epileptic patients in the operating room. As expected, stimulation disrupted sequences but not individual actions. The big surprise was the 86 percent overlap in the sites where this occurred and sites where phoneme discrimination was disrupted. A purely motor task and a purely sensory task, both living at the same sites? Did imitation lie at the heart of recognition?
But actually this intimate relationship had been predicted by the linguists who noted that the categories for phoneme discrimination seemed to parallel the diffficulty of pronouncing them. This "motor theory of speech perception," established in an influential review in 1967 by A. Liberman and co-workers, suggested that we detect a speech sound (a phoneme) by comparing it to a motor schema or template for pronouncing it. Just as some people move their lips while reading, this theory suggests that we all subliminally move our lips while listening to someone speak. This intriguing interaction between the sensory side of language and the motor sequencing of the mouth has led to many new ideas about how language got started in the cortex surrounding the Sylvian fissure. Kimura was perhaps the first to suggest that a lateralization to the left brain of motor sequencing would provide an important alternative to the traditional theories of language building atop other communication circuits.
Verbal Memory Surrounding the Central Core
Just as for thalamic memory testing, the epileptic patient is shown a slide depicting a common object and responds, "This is a dog [or ball or telephone]." A second slide gives the patient another task to do for distraction, and then a third slide prompts the patient to recall the object on the first slide: "Dog." If stimulation during the second-slide distraction task causes errors on third-slide recall attempts, the site of stimulation becomes known as a memory site (under certain conditions, first-or third-slide stimulation can also produce discrete memory-type errors). A map of such sites varies quite a lot from patient to patient, but they typically form a rim around a central core of sites where phoneme discrimination and oral-facial motor sequencing are disruptable.
Recall errors can also occur from stimulation during the first or third slides. Sites causing third-slide recall errors from firstor second-slide stimulation ("input" and "storage" phases) tend to cluster in temporal and parietal lobes; sites causing recall errors during third-slide stimulation ("output" phase) are more common in frontal and parietal lobes. There are some sites, typically located near the interface between core and memory rim, at which stimulation disrupts naming on the first slide. The patient will say, "This is a . . . you know, it's a . . . thing you put a key in."
There are sites at which only reading is disrupted. Given a future-tense sentence to read with a blank near the end of the sentence to fill in ("If my son is late for class again, he'll the principal"), the patient substitutes jargon, or repeats himself, or may be unable to complete the sentence. (For example, a reading error is the response "My son will getting late today he'll see the principal," where the noun and verb stems are correct.)
Still more interesting are the sites where the only known effect is to disrupt the good grammar of a sentence. In a sentence completion task like the one above, the patient will read it correctly, but in filling in the blank will make minor errors in tense or word order, though not enough to destroy the meaning of the sentence (see Chapter 4 of Inside the Brain for examples). [1997 note: Try Conversations with Neil's Brain instead.]
Ojemann sees the overall maps as suggesting a central core for motor mimicry and phoneme discrimination (in frontal, parietal, and temporal lobe gyri near the Sylvian fissure; thus "periSylvian"), a surrounding rim of regions specializing in shortterm verbal memory (though with recall aspects more frontal, and input and storage aspects more temporal), and at the margins between these two major regions, a patchwork of specialized sites related to syntax, naming, and reading. Looking back at the extensive literature on aphasia, it is apparent that in most cases, part of the central core has been permanently damaged. Damage to other regions may cause only temporary deficits.
But this patchwork-quilt picture of language cortex says little about how the individual areas work together to produce language, to recognize and remember words. Language is surely a committee effort of these areas. At some point, our reductionistic approach must reverse, to incorporate the broader picture of interacting regions--just as a portrait of a legislative body cannot be merely a series of chapters on the individual members but must analyze the shifting coalitions that group the members together for some purposes but not others, and must analyze the emergence of leadership.
Selective Shifts of Attention
The most interesting outcome of the thalamic language studies that started the second wave was some memory effects. Stimulation during the first slide might not affect correctly saying "This is a ball," but it would increase the chances that "ball" would be the reply on the third-slide recall attempt after distraction. Stimulation during the third-slide recall attempt didn't always cause recall errors, but it increased the chances of an error beyond that expected without stimulation. Stimulating during both first and third slides might cause no change in the unstimulated error rate, as if the effects averaged out.
The simple interpretation of this result favored by Ojemann is the following: Stimulation taps into the brain's specific alerting system (a subdivision of the reticular formation's sleep-wakefulness system). Stimulation orients the patient to incoming information from the visual world (indeed not just any information; left-thalamic stimulation affects only information with verbal significance). Thus stimulation during recall attempts shifts attention away from internalized information in short-term verbal memory, making errors increase. Stimulation during the original presentation causes the patient to pay even more attention to what is on the first slide, thus increasing chances that the information will be recorded vividly enough to be recalled after the distraction.
The thalamus contains many pathways, not only to the reticular activating system, but connecting to the cerebral cortex. Such as language cortex. Might the thalamus be actmg somewhat like an orchestra conductor, tuning down the activities of some of the patchwork regions of language cortex while enhancing others, then changing things around as the tasks shift? Does the thalamic story fit with the cortical story?
To tell the story requires another traveling-neurobiologist tale. In the aftermath of the 1973 Yom Kippur war, a young Israeli, Itzhak Fried, went to Tel Aviv University and studied physics. He decided to do graduate work in psychology and was admitted to the doctoral program at UCLA. UCLA has long been a center for the study of the reticular activating system. Fried not only learned to study the minute voltages seen on the scalp from brain activities preceding a response ("event-related potentials"), but he got mixed up with the neuroanatomists studying human brains, trying to spot anatomical differences between the left language area and the corresponding points on the right side of the brain. One of his mentors, Arnold Scheibel, a professor of anatomy and psychiatry, was familiar with the work being done in Seattle on language cortex, and so George Ojemann came to UCLA in 1978 to lecture on how variable the naming sites were between patients. Fried promptly arranged to come to Seattle and work with Ojemann for part of his Ph.D. dissertation in psychology at UCLA. (If this sounds roundabout, be assured that it is almost traditional in neurobiology--many of us have such a story.)
Besides stimulating, one can electrically record directly from the exposed cortical surface. This is routinely done with an EEG machine for helping to define the epileptic region. But it can also be used to examine a site's responses to a slide being flashed on the screen. The slides were simple black-and-white sketches of common objects, as usual, but with a bold slash across the whole picture at one angle or another.
When a region of brain is just idling, the EEG waves are larger and slower than when it is particularly active (we say that the EEG is "activated" when its bumps become smaller and more frequent; it is rather like how an automobile engine smooths out when accelerated from a rough idle). When the patient was asked to name the object shown on the slide (silently, to avoid complications from muscle commands), certain areas of brain would be activated. This was not general but in a patchwork; when stimulation mapping was done later in the operation, such EEG activation sites demonstrated anomie ("You know, it's a . . . a . . ."), whereas adjacent sites where activation did not occur did not exhibit stimulation-evoked language changes.
Furthermore, this selective EEG activation occurred only if the patient was paying attention to the verbally codeable aspects of the picture; if the patient was asked to pay attention to the angle of the slash across the picture of the object, the left-brain sites would not be activated (presumably right-brain sites were, but one cannot record from both sides of the cortex at the same time in the same patient). The same slide produced different responses, depending upon what the patient was set to do.
There are many aspects of the nervous system over which we have little voluntary control, such as autonomic regulation of heart rate and blood pressure. But selective attention for verbal versus pictorial aspects of our visual world would seem at the opposite end of the voluntary-involuntary spectrum, near the heart of higher brain functions. Here we have seen such selective attention mechanisms at work in language cortex. Self-- what makes you and me different--is surely all tied up with how we use such selective attention mechanisms to engage ourselves with different aspects of this multifaceted world.
Is the Temporal Lobe Really Temporal?
It is obvious why the frontal lobes were named "frontal," but the etymology of "temporal lobe" is less apparent. Is it because the temporal lobe has a clock? Perhaps, but that's not why it was named that.
The temporal lobe was named after the temple: the spot we often rub between eye and ear has the tip of the temporal lobe just behind it. But Robert Efron says that the temple was named for time (via the Latin tempus) in an amusing sense: as we age gracefully, the temple is the first place that the gray hairs appear. So the temporal lobe is, in a round-about way, named for time. If that piece of brain has something to do with time, however, it probably operates on a much shorter time scale, milliseconds rather than decades.
The function of the tip of the temporal lobe has always been a bit of an enigma. The reason that neurosurgeons can remove it to treat seizures is because so little happens when it is missing. If the removal does not damage the language sites farther back in the temporal lobe, the deficits are quite subtle (a minor diffficulty recalling proper names is one complaint of the patients). Thus our usual method for assigning function (deficits from strokes and similar damage) fails us entirely. Even stimulation mapping usually fails to show functions there, at least with the tests used thus far. Neurosurgeons have been known to suggest that its function is to cause temporal lobe epilepsy, as that is what changes when the temporal tip is removed.
This has suggested that the temporal tip forms some sort of redundant system. A passenger plane has several backup systems for lowering the flaps in case the primary system fails; remove one and you'd never notice it was gone unless the two other systems failed first. There are two special sets of data that indicate that the temporal tip has the capacity to function as part of language's motor system.
A patient with a slowly growing tumor near the traditional Broca's area of the left frontal lobe was mapped; the usual response of Broca's area to stimulation is arrest of all speech output (the patient tries to talk but cannot even say "This is a . . .a...), but stimulation had no such effects in this patient's frontal lobe. Slow damage can cause rearrangements even in adults, where another area takes over the function, and since the patient was only mildly aphasic, this seemed likely. In this patient, Broca's area responses were found below the Sylvian fissure, 3 centimeters from the temporal tip, showing that this enigmatic region has sufficient motor system connections so that it can, if necessary, function as the motor output specialization of the language system. In the second special patient, only finger spelling was disrupted at the temporal tip; this woman could hear but had learned to finger spell as part of her work with the rehabilitation of the deaf. This again shows that the temporal tip can function as part of a motor sequencing system.
Careful testing also shows that temporal tip removal does cause some changes in performance on an auditory judgment. Suppose that you listen to a pair of tones, one high and the other low except that the order is sometimes high-low, sometimes low-high, and the patient has to judge which. When the tones are spaced far apart ("Beep Boop"), the task isn't too hard. But if the second follows close after the first ("Beep-Boop"), the task gets hard. If the tones are presented in the ear opposite to where the temporal lobectomy occurred, the judgment is more difficult than usual and they have to be more widely separated ("Boop . . Beep") than usual before the judgment becomes reliable. If you think this is complicated, I am actually oversimplifying the results--but I did say that the deficits were subtle; it is amazing that Ira Sherwin and Robert Efron discovered the deficit at all. There had, however, been clues that the region had a role in various fine timing judgments; Paula Tallal and coworkers had suggested that the left brain processed rapidly changing auditory signals (of which speech is but one example) to a greater extent than the right brain.
So damage to the temporal tip suggests that the region has some redundant function and that it might be associated with timing. If you read the literature on heart cells beating away in culture dishes, or on models for circadian rhythms in sand fleas, it all starts to make sense. As mentioned in Chapter 4, redundancy allows for more precise timing: if one uses four times as many identical timing circuits in parallel, a cell which averages their outputs will double its accuracy (i.e., its standard deviation will halve). So loss of temporal tip might degrade some fine timing capabilities to merely good.
Readers of Chapter 4 perhaps will also remember an evolutionary argument for why precision timing circuits might be exposed to natural selection pressures: If redundant timing circuits can be used for motor sequencing as well as auditory processing (not unlikely, considering the evidence for the motor theory of speech perception), then they could help time the release of the rock during throwing. To double the throwing distance for a rabbit-size target requires an eight-fold narrowing of the launch window. Timing redundancy circuits would allow that to be done with a sixty-four-fold increase in timing circuits applied to the problem. A hominid brain that could temporarily borrow timing circuits from elsewhere, or had more to start with because of brain enlargement, would be a better brain for controlling throwing.
The whole periSylvian region probably is involved with timing and sequencing, as telling one phoneme from a closely related one is a matter of making fine timing judgments on a sequence of tones. Commanding the pronunciation of a phoneme sequence presents exactly the same problem of fine timing within a sequence. The temporal tip may merely be the cortical annex for supplementing the timer array for the really precise jobs.
Redundant timing neural circuits provide an attractive theory for relating motor sequencing and phoneme discrimination, for tying together throwing and language. It even suggests one reason why bigger brains might be better.
"Natural " Maps vs. the Time Window
Someone could do a fascinating doctoral dissertation on the cross-fertilization roles of the traveling scholar in neurobiology. The itinerant graduate student, the postdoctoral fellowship, and the faculty sabbatical have opened up many new avenues, permitted expertise in one area to be brought to bear upon new questions, and prevented many cases of parochialism and wasted effort by stirring the pot.
But language physiology is still a young subject and it should perhaps be emphasized that, though their backgrounds have varied widely, only several dozen people have had major involvement in it so far, even counting the first wave from Montreal. Contrast this to the traditional source of brain maps, the aphasia from stroke victims. This anatomic pathology version of language investigation has had contributions from thousands of investigators, whose backgrounds ranged the spectrum from the prepsychoanalysis Sigmund Freud to the post-physics William Calvin. But most of these investigators have been "observers" rather than "experimenters," because nature's accidents perform a messy experiment and we merely document the outcome and try to make some sense out of similar accumulated evidence. The first two waves of language physiology have given a glimpse of a detailed landscape which the stroke documenters only occasionally suspected.
Our existing physiological maps are, however, inevitably biased by the backgrounds of the lucky investigators. As the tests used bear the stamp of their particular scientific upbringing, so they bias the maps made from the results of those tests (consider what might have happened to the maps if their backgrounds had emphasized Freudian concepts rather than Sherringtonian physiology!). Neurobiology is still a small enough world that we can trace genealogies, of who was influenced by whom. One of the auditory tests used in the O.R. was developed by C. E. Seashore at the University of Iowa, where George Ojemann grew up as the son of a psychology professor; later, he happened to acquire Seashore's desk and studied at it all during his time in medical school, only to later discover an interesting use for the Seashore test. The emphasis upon motor sequencing grew out of Doreen Kimura's neuropsychological interests via Catherine Mateer's involvement in test design. My theoretical emphasis upon precision timing buffers, which I have superimposed on the Kimura-Mateer-Ojemann notion of a primary specialization for motor sequencing, grew out of my doctoral dissertation with C. F. Stevens fifteen years earlier on why neurons are noisy. Interesting facts from the past are always popping up, so teachers never know where their influence reaches.
As more people get the opportunity to study language physiology, the emphasis will inevitably change as different backgrounds cause new methods to be brought to bear, new maps made. The answers you get depend upon the questions you ask, the methods you have and what's actually there in the brain. We must constantly ask whether our classifications are as natural as possible, and not overly biased by our approach.
This second wave of human language physiology has been extraordinarily exciting. The mosaic of functions demonstrated by the present stimulation maps show some of the natural physiologic subdivisions of language with a clarity diffficult to perceive in linguistic studies or in stroke impairments on a naturally variable population. As the functional relationships between such areas are further investigated, along with their ontogeny and functional plasticity, additional windows upon the subspecialties of language cortex will be gained. We may expect our definition of functional roles to become even more natural than those growing out of the present youth of language physiology.
Unlike other areas of science, however, this one could abruptly freeze in place. We must hurry, because our opportunities for access to the working human language cortex could disappear. Skulls are opened and language cortex mapped with electrical stimulation only when there is an important medical reason. That occurs now because anticonvulsant drugs are unsatisfactory in about 30 percent of patients; those who do not respond to the available drugs become candidates for surgery only if their epilepsy is of a particular type. For the others, there are no good answers.
A better anticonvulsant would improve the lives of millions of people. Yet opportunities to study language physiology would shrink at the same time, perhaps to the point at which research teams would disperse. It is true that other opportunities are opening up, such as MRI techniques and positron-emission tomography (PET scan), but they do not allow manipulation of small regions of brain in the manner that has made the electrical stimulation technique so powerful: stimulation is a harmless probe, but it cannot be delicately applied through the thick skull. While it might be possible to figure out the workings of a clock by ingenious ways of looking through the walls of its case, the brain is far too complex a computer to analyze only by observation of its undisturbed spontaneous activities. Just as it is most useful to poke a f~nger in the clock's works to move certain gears in figuring it out, so it is with the neurophysiologist's use of electrical stimulation in figuring out language cortex.
Thus we may well have a restricted "window in time" during which our civilization can ethically study the more detailed workings of human language cortex, a window caused by knowledge in one area (neurophysiology) temporarily exceeding that in an adjacent area (neuropharmacology). By making the most of our opportunity to peer through that window and probe with harmless tools, however, we may greatly improve our understanding of part of that "human essence" of language and thought.
The Throwing Madonna:
Essays on the Brain (McGraw-Hill 1983, Bantam 1991) is a group of 17 essays: The Throwing Madonna; The Lovable Cat: Mimicry Strikes Again; Woman the Toolmaker? Did Throwing Stones Lead to Bigger Brains? The Ratchets of Social Evolution; The Computer as Metaphor in Neurobiology; Last Year in Jerusalem; Computing Without Nerve Impulses; Aplysia, the Hare of the Ocean; Left Brain, Right Brain: Science or the New Phrenology? What to Do About Tic Douloureux; The Woodrow Wilson Story; Thinking Clearly About Schizophrenia; Of Cancer Pain, Magic Bullets, and Humor; Linguistics and the Brain's Buffer; Probing Language Cortex: The Second Wave; and The Creation Myth, Updated: A Scenario for Humankind. Note that my throwing theory for language origins (last 3 essays) has nothing to do with the title essay: THE THROWING MADONNA is a parody (involving maternal heartbeat sounds!) on the typically-male theories of handedness.
Many libraries have it (try the OCLC on-line listing, which cryptically shows the libraries that own a copy), and used bookstores may have either the 1983 or the 1991 edition.
- Powell's Books in Portland lists used copies in their web database.
Email || Home Page || The Calvin Bookshelf || Table of Contents || End notes for this chapter || Continue to read next chapter