Tales from Both Sides of the Brain : A Life in Neuroscience (9780062228819)
Page 9
The telltale trials were when the right hemisphere saw a particular color, say “red,” and N.G. said the wrong color, say “green.” After a few flat-out mistakes, the patient started stating the correct color every time. Somehow she had learned a strategy that made it appear that the left hemisphere could name something only the right hemisphere had seen. She would start to say “gree . . .” but then she would stop and then guess correctly by saying “red.” What was happening was the left hemisphere was doing the talking and the disconnected right hemisphere was hearing the “gree . . .” being uttered by the guessing left hemisphere. The right hemisphere somehow stopped the speech emanating from the left brain by giving some kind of cue, such as nodding the head or shrugging the shoulders. The wily left hemisphere would pick up on the cue, which it had figured out during the first erroneous trials, and would change its response to the only other color! All of this happened in the flash of an eye.
I wanted to look into this self-cueing strategy more deeply. In a sense, this kind of cueing was taking place outside the brain. The patients were learning the equivalent of tango dance strategies: That is, one side of the body tapped the other to get communication going between the two halves of the brain. It could look like the two separate brains were unified by internal connections and communication, but in reality, it was external signals providing the communication that united the two. We also began to wonder if there was cueing actually going on within the brain. After all, the surgery only disconnected the cognitive and sensory systems housed in the cortex. There were still dozens of ways one side could connect with the other in complex, albeit more indirect, ways within the intact subcortical pathways of the brain. And, as I mentioned above, we wondered about the more ethereal aspects of mental life, such as emotions. It seemed as though monkey experiments could get at the emotion question.
So, in Caltech style, we just did it. This required building more special testing devices more animals, and more honing of my own surgical skills. Surgical procedures were taken very seriously and carefully planned. We all trained ourselves, first by attending surgeries performed by seasoned members of the lab. I was lucky, for the highly skilled Giovanni Berlucchi, visiting from Pisa, Italy, let me sit in on his surgeries. Sperry was also a fantastic surgeon. Once, when I was in the surgical room, watching him at a delicate point in the surgery, he looked intently into the operating microscope and softly said, “I can’t seem to see the anterior commissure.” I leaned forward to better hear him and in doing so, jarred the table, to which he calmly said, “Oh, there it is.” Always unflappable.
To carry out this experiment, the monkeys, once they’d recovered from surgery, were outfitted with goggles equipped with one red and one blue lens. The colored light filters allowed different visual images to be projected to the separated hemispheres. We wanted to know what would happen to the work pattern of one hemisphere if the other were suddenly exposed to an emotionally powerful stimulus, such as a snake. Would the emotionally provoked half brain dominate or subcortically influence the half brain that was engaged in the simple and emotionally neutral task of visual learning?
The answer was clear. The animals jumped back. The hemisphere that had seen the emotional stimulus, the snake, and experienced the emotion, fear, cued the rest of the animal: Something is wrong! With that gross and unmistakable cue, the animal became agitated and ceased working on the discrimination task and would not return to work, either. Cross-cueing of a kind was evident once again. In this instance it appeared that one separate and distinct mental system could be agitated and in that agitation not let the other mental system function in its normal way. The idea crept into our heads that the “mind” was a collection of mental systems, not just one. At the time, this was a new and important idea. It was absolutely crucial to the understanding of why the split-brain monkeys, as well as the patients, behaved as they did.
Testing different theories on both animals and humans continued. In the later 1960s, years after we both had left Caltech, Steve Hillyard and I were collaborating on a study. We were trying to figure out the language capacity of the third Caltech patient, L.B., when we spotted another variant of cross-cueing. We set up an easy test for the patient. All he had to do was name the number (1–9) that was flashed either to the left or the right visual field. Normally we would expect the right visual field stimuli to be named quickly. Thus, if a “1” or a “4” or a “7” flashed up in random order, the patient’s left speaking hemisphere would respond correctly. It did. Each number was named with about the same reaction time.
What initially surprised us, however, was that the right hemisphere seemed to be naming all the numbers, too. What was going on? Was this our first patient to show transfer of information between the hemispheres? Alternatively, was this a right hemisphere that could speak? (This possibility is always there and must always be investigated.) Or was the right hemisphere somehow cueing the left hemisphere again?
Hillyard plotted out the reaction times for each response, and the strategy L.B. was using became apparent. All the numbers flashed to the left hemisphere were named rapidly in about the same amount of time. When the same random list of numbers was presented to the right hemisphere, however, “1” was reacted to more quickly than “2,” which was reacted to more quickly than “3,” which in turn was reacted to more quickly than “4,” and so on all the way up to “9.” Another cross-cueing strategy revealed! The left, speaking hemisphere started counting using some somatic cueing systems such as a slight head bob, which the right hemisphere could sense. When the number of bobs hit the number that was presented to the right hemisphere, the right hemisphere sent a somatic stop signal that the left hemisphere could sense. At this point, the left knew that must have been the number that was flashed and said it, not the right!2 Unbelievable. Trying to outfox this cross-cueing system, we ran another series of trials. This time the patient was required to respond immediately. While the left hemisphere continued to respond correctly and quickly, the right hemisphere’s score dropped to chance. The brain was shifting strategies to accomplish the same goal.
THE POWER OF THE BEDSIDE EXAM: CASE D.R.
When studying neurologically disrupted patients, certain general principles emerge. For example, the patients nearly always strive to complete a goal that has been set by their examiner. One might think and hope they are solving a task in a particular way, even when in fact they are solving it in another. The challenge is to identify the way they are solving it. When you do, underlying mechanisms are revealed that are frequently surprising. When sorting through my hundreds of hours of videotapes of patients, I recently came across a particularly vivid example, which revealed how cueing can occur when a patient is simply trying to copy a gesture made by one hand, with the opposite hand (Video 4).
This was Case D.R., a split-brain patient from the Dartmouth series of cases. D.R. was also a college graduate and an accountant. After spending time in South America, she had moved to New England. Along the way she had become a Trekkie. She had all the Star Trek episodes recorded and also owned a rather expensive model of the Enterprise! After her surgery, she showed all the standard disconnection phenomena. Visual information did not transfer between the hemispheres, nor did tactile information. Her left hemisphere was dominant for language and speech; her right hemisphere functioned at a lower cognitive level, being able only to recognize pictures, but not to read. We wanted to examine her motor control capacity. With her eyes wide open, I asked her to hold out her two hands, fists closed; that was the starting position for each subsequent command. I then asked her to make the “hitchhiker” sign with her right hand. She did so instantly. I then asked her to do the same thing with the left hand. She also did that quickly. I then asked her to make the “a-okay” sign with her right hand. Again, she did so quickly. When asked to do it with her left hand, after a slight hesitation, she had no problem.
Here is where learning begins for the examiner when testing neurologic patients. One has
to make sure that the task a patient is trying to complete for you is being done the way you imagined it would be done. In this case, I knew, of course, that the patient had undergone split-brain surgery. By the time we worked with her in the 1980s, I knew that there was tremendous variation on how well a disconnected hemisphere could control the ipsilateral hand. Of course, there never was a problem in controlling the contralateral hand, as both the sensory and motor systems needed for such activity were all represented in the same hemisphere. Controlling the ipsilateral hand, however, was a very different story. How did her dominant left hemisphere, which had to interpret my spoken command messages, send them over to the motor systems in her right hemisphere, which controls her left hand? Those motor control systems for the left hand were unquestionably managed by her disconnected right hemisphere. How was information presented to one hemisphere being integrated for use in the opposite, disconnected half brain?
Recall that Case W.J. was remarkably unable to control an ipsilateral arm and hand, while having little problem controlling the contralateral arm and hand from a particular hemisphere. This was quite a dramatic situation. As I said earlier, many of the original split-brain stories about two minds being present in our skulls instead of one came from W.J.’s behavior. Nonetheless, as more patients were added to the study pool, many began to show good control over both the ipsilateral arm and the contralateral arm. Yet, even when patients had good control of the ipsilateral arm, good control over the ipsilateral hand seemed to elude them. Again, how did all of this work?
Back to Case D.R.: In the video, she was making hand gestures with both hands that appeared responsive to my verbal commands. I knew D.R. had split-brain surgery and that her dominant language hemisphere was disconnected from the motor control systems of her right hemisphere. I was eager to learn how she was completing the task of controlling her left hand so easily, given that her two half brains were disconnected from one another. What to do? Armed with this knowledge, I changed the exam ever so slightly, and this allowed the answer to emerge.
Instead of asking D.R. to make a “hitchhiker” gesture with her right hand first, I asked her to make the gesture with her left hand first. She couldn’t do it. After she failed, I then asked her to make it with her right hand, which she did instantly. Same story with the “a-okay” sign: She just couldn’t do it if the left hand had to do it first. Why was that?
Obviously, what was going on was that when the right hand (controlled by the left hemisphere) went first, it set up a model and an image for the right hemisphere to see and to copy. If a model was present to copy, then the right brain could mimic a gesture and perform that task easily. In essence, the patient had visually cross-cued the information from one hemisphere to the other outside of the brain, thereby trumping the fact that her brain hemisphere connections had been severed. If this were true, then what would happen if the patient were asked to do the task with her eyes closed? With bedside testing this was easily done. The exam continued.
I asked the patient to close her eyes and to make a “hitchhiker” sign with her right hand, which, again, she instantly carried out. Now, with her eyes still closed I asked her to make it with her left hand. Amazingly, she could not do it. The patient’s right hemisphere could not understand the spoken command, and with her eyes closed, the left hemisphere could not cue the right hemisphere by providing the model gesture of the right hand to copy. As a consequence, the left hand sat there frozen with inaction.
This one simple bedside test revealed so much. It revealed not only the dramatic disconnection effects of the surgery, but also a basic truth about goal-directed behavior. We are all eager to achieve singular, unitary goals, and we behave as we wish in specific circumstances. We somehow achieve this unitary output from a highly modularized brain with multiple decision centers. In human patients, when the normal neuronal pathways are disrupted, the goal may still be achieved through whatever alternate mechanisms and strategies remain available. In this instance, two things were clear: The right hemisphere, disconnected from the left, could not follow a verbal command, and it was the hemisphere with the major control over the left hand. The explanation might have been, however, that the left hemisphere could have governed the ipsilateral left hand through ipsilateral corticospinal pathways (a small number of neurons that are uncrossed) that we know exist. Yet we had already shown that that explanation could not be true, because the verbal command to gesture with the left hand could not be followed either when the eyes were closed or when the left hand was directed to respond before the right hand. What was going on?
Clearly the right hemisphere could only execute the command when it visually saw and imitated a model of the posture being requested. The overall system, with all of its separate modules, had cued itself into completing the goal. This cueing is a ubiquitous mechanism of goal-directed behavior.
NEW CASES, NEW FINDINGS
While these experiments on basic sensory-motor control were booming along both at Caltech and in subsequent years at Dartmouth, I grew fascinated by the idea that we could show what a separated right, nonspeaking hemisphere could do in terms of thinking, perceiving, understanding, planning, and all the rest. Getting anything out of W.J.’s right brain had proven difficult, even though he was clearly skilled at visuomotor tasks such as the block design test. He responded normally and easily to the flash of a picture or word to his left hemisphere. Yet flashing the exact same information to the right hemisphere generally provoked only minimal response. It was like pulling teeth. Driving to Downey every week in my old Studebaker was becoming a chore. Sometimes I went just to claim a $3.67 gas reimbursement, a rate that enabled me to run my car for the rest of the week.
It wasn’t until we began working with N.G., a pleasant young woman with an exceptionally supportive husband, that we moved beyond the fundamental sensory-motor integration tasks we had studied so intensely in W.J. She, like W.J., had been operated on for uncontrollable epilepsy and treated by Bogen and his neurosurgical mentor, Philip Vogel. She was agreeable to testing and, as with most of the patients, it became a big part of her life. After all, here we were showering attention on the patients and compensating them for their time. We all built lasting relationships that have continued over the years. Just last spring a relative of N.G’s husband called me after almost forty-five years of no contact, just to say hello.
Soon after N.G. came L.B., a young boy of twelve, another favorite and extensively studied patient. He had also been operated on to control his severe epileptic seizures. L.B. proved to be a remarkable case. Years later, again out of the blue, he sent me a yet-unpublished manuscript he had written about his personal experience as a patient and experimental subject. In writing his personal perspective, L.B. had been assisted by Caltech’s science writer, the wonderfully sensitive Graham Berry.3
These two new surgical cases brought real energy to the project. While they quickly confirmed the basic disconnection effects of W.J., they provided us with new insights about the function of the right hemisphere. Their right hemispheres responded to our tests with glee and vigor, even though their left hemispheres remained unaware of the content being processed by the disconnected and largely silent right hemisphere.
By this time, I was using my Bolex camera a lot. When testing N.G., I would place it on a tripod pointing down in such a way that one could view not only the patient’s face, but also objects placed out of view that the patient would sometimes have to touch as part of the test being run. In this way, it was easy to visualize the actual tests and the sometimes stunning results. First, objects held in the right hand could be easily named, but not those in the left hand. Second, pictures of objects flashed to the left hemisphere could enable the opposite right hand to find the matching object, but not the ipsilateral left hand. Third, and here was where the new era really began, pictures and even words flashed to the supposedly language-impoverished right hemisphere triggered the left hand to retrieve the correct object out of view.4 It was asto
nishing and remains so to this day (Video 5). We were looking at the first real evidence that the right hemisphere could be capable of cognitive activity and complex behaviors without the left hemisphere knowing anything about it.
After years of studying N.G. and L.B., Sperry and I concluded that the right hemisphere had a large vocabulary. It could react correctly to printed words as well as to line drawings of all types.5 There was even some capacity in the right hemisphere for simple spelling and an occasional written word, but rarely so. We pushed on in the hope of finding independent higher-level thought of some kind, and attempted several tests requiring simple arithmetic. Here we had an occasional success with simple addition, but none with subtraction.
We were always on the lookout for brain functions that did transfer from one separated hemisphere to the other. After the monkey work with emotive stimuli, I wanted to see if humans would react in the same way. Would a potentially emotive stimulus cue the opposite hemisphere in some way, where it was clear that nonemotive stimuli did not? This test required a stop at the back of the magazine shop, where, at that time, plain cardboard was placed in front of the risqué magazine covers. Magazines would have to be bought and pictures clipped out, photographed, then put into a slide carousel so they might suddenly appear in the left visual field as part of a sequence of pictures of more routine objects, such as spoons and coffee cups. I was nervous about this experiment. While I was a vigorous postadolescent, to be sure, I was also a practicing Catholic, and as you well know . . .