Here are the highlights from that book ...
Page 43:
Page 56:It is the dream of every neuropsychologist to discover someone who has such an unusual view of the world that we are forced to reconsider our ideas about how the brain works. Two things are necessary to discover such a person. First, we have to be lucky enough to meet him or her. Second, we have to be clever enough to recognize the importance of what we are observing.
"I'm sure you're both lucky and clever," says the Professor of English. Not so. I was lucky once, but not clever. As a young research worker at the Institute of Psychiatry in south London, I was studying how people learn. I was introduced to someone with severe loss of memory. For a week he visited my lab every day in order to learn a simple motor skill. His performance improved in a fairly normal manner and even after a gap of a week he retained the new skill he had learned. But, at the same time, his memory loss was so severe that each day he would claim that he had never met me before and had never performed the task before. "How strange!" I thought. But I was interested in problems of motor skill learning. This man learned the skill I taught him normally and so I wasn't interested in him. Many others have, of course, recognized the importance of people like this. Such people can remember nothing that has happened to them even if it happened only yesterday. We assumed that this was because the events that happened were not recorded in the brain. But, in the person I studied, the experiences he had yesterday clearly had a long-term effect on his brain since he was able to perform the motor task better today than yesterday. But this long-term change in the brain had no effect on his conscious mind. He could not remember anything that happened yesterday. Such people show that our brain can know things about the world that our mind does not know.
Page 63:Even if all our senses are intact and our brain is functioning normally, we do not have direct access to the physical world. It may feel as if we have direct access, but this is an illusion created by our brain.
[...]
As a young research student, Hermann Helmholtz was told by his professor that it would be impossible to measure the speed of nerve conduction. It would be too fast. But, like all good students, he ignored this advice. In 1852 he was able to measure the speed of nerve conduction and showed that it was rather slow. In sensory neurons it takes about 20 msec for the nerve impulse to travel 1 meter. Helmholtz also measured ?perception time? by asking people to press a button as soon as they felt a touch on various parts of the body. These reactions times turned out to be even longer, being more than 100 msec. These observations show that our perception of objects in the outside world is not immediate. Helmholtz realized that various processes must be occurring in the brain before a representation of an object in the outside world appears in the mind. He proposed that perception of the world was not direct, but depended on "unconscious inferences." In other words before we can perceive an object the brain has to infer what the object might be on the basis of the information reaching the senses.
Pages 80 - 81:John Morris and his colleagues had previously found that when people are shown fearful faces (as opposed to happy or neutral faces), activity increases in the amygdala, a small part of the brain that seems to be concerned with detecting dangerous situations. Whalen and his colleagues repeated the experiment, but this time the fearful faces were presented subliminally. Sometimes a fearful face was presented followed immediately by a neutral face. At other times a happy face was presented followed by a neutral face. On both these occasions you would say, "I saw a neutral face." But when the fearful faces were present, activity would occur in the amygdala even though you were unaware of the fearful face.
Page 107:A group of suggestible, but otherwise perfectly normal, university students were hypnotized. They were then given a word association task.
The experimenter read out a list of words and the subjects responded with the first words that came into their heads (bed -- pillow, bridge -- river, garden -- lawn, etc., etc.). While they were still under hypnosis the subjects were told that they could no longer remember doing this task. Then the experimenter read out the same list of words and the subjects again had to respond with the first word that came into their heads. So this is the key question. If you had a "real" memory loss due to brain damage so that you were unable to remember having done that word association task just now, would you respond with different words or would you give the same words again?
"Obviously I would give different words next time," says the Professor of English. "Which words you give is just a matter of chance. There are so many different associations for the word tree that it is very unlikely that you would give the same word again."
"That's what most people think," I reply smugly. "Unless they have attended some neuropsychology lectures."
I know the Professor is wrong through studies of people with severe amnesia who really can?t remember doing the task. These people tend to give the same words that they gave just before. And they may give them a little faster.
The subjects in the hypnosis experiment gave different words when the word association task was repeated. Like the Professor of English, they thought that this is what happens if you can't remember doing the task before and they acted according to their belief. But they didn't know that this is what they were doing. So here's what your brain has to do in this experiment without you knowing anything about it. First, it must set up a general strategy for performing the word association task, "give a
different word from last time." Second, for this strategy to be successful, it must remember which words were given last time in order to avoid giving them again. Third, it must monitor each action to overcome the strong tendency to give the same word again.
So here we are near the top of a hierarchy for controlling action. And we find that our brain can set up and monitor a complex strategy for action with our knowing anything about it. My knowledge of my own body and how it acts on the world is not direct. There is much about me that my brain hides and much that it makes up. In which case, when I look in the mirror, why doesn't my brain show me as I truly am -- young, thin, and with abundant black hair?
From page 142:Using his box, Skinner demonstrated the arbitrary nature of response learning in a most elegant experiment on "superstition" in the pigeon. A hungry pigeon was put into the Skinner box and food was presented at regular intervals with no reference whatsoever to the bird's behavior. After a short time the pigeon was seen repeatedly performing some arbitrary action. One pigeon turned counter-clockwise around the box, making two or three turns between the appearances of food. Another repeatedly thrust its head into one of the upper corners of the box. A third developed a "tossing" response, as if placing its head beneath an invisible bar and lifting it repeatedly. The pigeons had learned to repeat whatever action they happened to be performing just before the appearance of food. Skinner called this "superstitious" behavior because the pigeons acted as if they believed that their behavior caused the food to appear when this was not the case. He suggested that superstitious behavior can arise in humans in just the same way.
[...]
A reliable informant from the Cambridge psychology class of '68 tells me that they were able to cause an eminent neuropsychologist to lecture from the far left side of the theater by yawning and dropping their pencils whenever he moved to the right. An interesting feature of such experiments is that they work only if the target is unaware that he is learning about reward contingencies in the environment. We don't have to be aware of associations to learn them -- in fact it helps if we are not aware of them.
From Page 132:In my brain, perception depends upon prior belief. It is not a linear process like that which produces an image on a photograph or on a TV screen. For my brain, perception is a loop. In a linear version of perception, energy in the form of light or sound waves would strike the senses and these clues about the outside world would somehow be translated and classified by the brain into objects in certain positions in space. It was this approach that made perception so difficult for the first generation of computers. A brain that uses prediction works in almost the opposite way. When we perceive something, we actually start on the inside: a prior belief, which is a model of the world in which there are objects in certain positions in space. Using this model, my brain can predict what signals my eyes and ears should be receiving. These predictions are compared with the actual signals and, of course, there will be errors. My brain welcomes these errors. These errors teach my brain to perceive. The existence of the errors tells my brain that its model of the world is not good enough. The nature of the errors tells the brain how to make a better model of the world. And so we go round the loop again and again until the errors are too small to worry about. Usually only a few cycles of the loop are sufficient, which might take a brain only 100 milliseconds.
A system that constructs models of the outside world in this way will use any information it can get to help it make better models. No preference is given to vision or sound or touch as long as they are informative. And the system will make predictions about how the signals coming from all the senses will change when I act on the world. So, when I see a glass of wine, my brain is already making predictions about what the glass will feel like and what the wine will taste like. Imagine the shock and horror of picking up a glass of red wine and discovering that it is cold and sweet.
From Page 134:My Perception Is Not of the World, But of My Brain's Model of the World
What I perceive are not the crude and ambiguous cues that impinge from the outside world onto my eyes and my ears and my fingers. I perceive something much richer -- a picture that combines all these crude signals with a wealth of past experience.17 My perception is a prediction of what ought to be out there in the world. And this prediction is constantly tested by action.
Now, any system makes certain characteristic types of error when it fails. Luckily, these errors are very informative. Not only are the errors important for the system to learn, they are also important for us when we observe the system for discovering how the system works. They give us clues as to what kind of a system it is. What kind of errors will a system that works by prediction make? It will have problems whenever there is ambiguity: when two different objects in the outside world cause the same sensations.18 This problem can usually be solved because one model is much more likely than the other. It is very unlikely that there is a rhinoceros in my room. But this means the system is fooled when the unlikely situation is, in fact, the correct one. Many of the visual illusions beloved of psychologists work because they trick the brain in this way.
From Page 134:Color Is in the Brain, Not in the World
But all these ambiguous figures have been invented by psychologists, you might say. We don't come across such objects in the real world. True. But the real world also is inherently ambiguous. Consider the problem of color. We only know about the color of objects from the light that is reflected from them. The wave-length of the light is what makes the color. Long wave-lengths give red, short wave-lengths give blue, with all the other colors in between. There are special receptors in the eye that are sensitive to these different wave-lengths of light. So does activity in these receptors tell us what color the tomato is? There is a problem here. The color isn't in the tomato. It's in the light reflected from it. When illuminated with white light, a tomato reflects red light. That is why we see it as red. But what if the tomato is illuminated with blue light? It can't reflect any red light, so does it now look blue? No. We still perceive it as red. From the colors of all the objects in the scene our brain decides that the scene is being illuminated by blue light and predicts what the "true" color of the various objects must be. What we perceive is determined by this predicted color, not by the wave-length of the light striking our eye. Because we see the predicted and not the "real" color, we can create striking illusions in which patches which are identical in terms of the wave-length of the light seem to have quite different colors (see Figure CP6, color plate section).
[Note that I prefer the "finger sausage illusion" for identifying the eye's blind spot ... From http://wiki.answers.com/Q/What_is_the_eye's_blind_spot ...Our brains build models of the world and continuously modify these models on the basis of the signals that reach our senses. So, what we actually perceive are our brain's models of the world. They are not the world itself, but, for us, they are as good as. You could say that our perceptions are fantasies that coincide with reality. Furthermore, if no sensory signals are available, then our brain fills in the missing information. There is a blind spot in our eyes where there are no light receptors. This is the point where all the nerve fibers carrying the sensory signals from the retina to the brain (the optic nerve) come together -- so there is no space for light receptors. We are not aware of this blind spot because the brain makes something up to go into that part of our visual field. Our brain uses the signals from the region immediately around the blind spot to supply the missing information.
Put your finger straight in front of you and stare at it. Then close your left eye and move your finger slowly to the right, but keep staring straight ahead. There is a point where the tip of you finger will disappear and then reappear beyond the blind spot. But inside the blind spot your brain fills in the blank with the surrounding wallpaper pattern, not with the tip of your finger.
[...]
You might think that this tendency to hallucinate was too high a price to pay for our brains? abilities to make models of the world. Couldn't the system be tuned so that sensory signals always dominated our experience? Then hallucinations could not occur. In fact, this is a bad idea, for many reasons. Sensory signals are simply too unreliable. But more importantly, such domination would make us slaves to our senses. Like a butterfly, our attention would continually flit from one attraction to the next. Such slavery to the senses can sometimes happen as the result of brain damage. There are some people who cannot help but act on everything they happen to see. They put a pair of spectacles upon their nose. But then they see another pair and put those on too. If they see a glass, they must drink from it. If they see a pencil, they must scribble with it. They are unable to carry out a plan or follow an instruction. It turns out that these people usually have extensive damage to the front part of the brain. Their strange behavior was first described by François Lhermitte.
The patient . . . came to see me at my apartment. . . . We returned to the bedroom. The bedspread had been taken off and the top sheet turned back in the usual way. When the patient saw this he immediately began to get undressed [including taking off his wig]. He got into bed, pulled the sheet up to his neck and prepared to go to sleep.
Through its use of controlled fantasy, our brain escapes from the tyranny of our environment.
So How Do We Know What's Real?
There are two problems with fantasizing the world. First, how do we know that our brain's model of the world is true? This is not a real problem. For us to act upon the world it doesn't matter whether or not our brain's model is true. All that matters is that the model works. Does it enable us to make the appropriate actions and survive for another day? On the whole, yes it does. As we shall see in the following chapters, questions about the "truth" of the brain's models arise only when one brain communicates with another, and we discover that another person?s model of the world is different from our own.
It's the spot where the optical nerver goes into the back of the eye. There are no receptors in this place.
...
At arm's length, hold out hands so the index fingers are pointing towards each other -- tip to tip but not quite touching. Focus on a spot on a wall that is several feet away. Next, move fingers to about 12 inches from the eyes (this will block the view of the spot). A little sausage-shaped object floating between the fingers will appear.]
From Page 137:
From Page 156:Imagination Is Extremely Boring
We have already seen how visual illusions reveal how the brain models reality. The Necker cube, mentioned above, is a well-known visual illusion (see Figure 5.8). We may see it as a cube with its front edge pointing to the left and down. And then our perception suddenly changes and we see it as a cube with its front edge pointing to the right and up. The explanation is simple. Our brain sees it as a cube rather than as the two-dimensional drawing it really is. But, as a cube, it is ambiguous. It has two possible three-dimensional versions. Our brain randomly switches from one to the other in its continuous attempts to find a better fit for
the sensory signals
But what happens if I can find a naïve person who has never seen a Necker cube before and doesn't know about its tendency to reverse from one form to another? I show him the figure for a short time so that he does not see it reverse. Then I ask him to imagine the figure. Will it reverse as he inspects it in his imagination? I find that, in the imagination, the Necker cube never reverses. The imagination is utterly uncreative. It has no predictions to make and no errors to resolve. We don't create in our heads. We create by externalizing our thoughts with sketches and doodles and rough drafts so that we can benefit from the unexpectedness of reality. It is this continual unexpectedness that makes interacting with the real world such a joy.
In this chapter I have shown how our brains discover what is out there in the world by constructing models and making predictions. Our models are built by combining information from our senses with our prior expectations. Both sensations and expectations are essential for this process. We are not aware of all the work our brain is doing. We are only aware of the models that result from this work. This makes our experience of the world seem effortless and direct.
From Page 156:Science can attempt to explain how we can understand other minds. This is no different from explaining how we, as individuals, understand the physical world. This is much of what the science of psychology is all about. And, as we have seen in the last chapter, our knowledge of the physical world is essentially subjective. What I know about the physical world is captured in a model of that world created by my brain. This model is created from my prior knowledge and the cues provided by my senses. My brain creates a physical world of trees and birds and people. My knowledge of the mental world, the world of other minds, can be created in exactly the same way. From the cues provided by my senses my brain creates a model of a mental world of beliefs, desires, and intentions.
From Pages 166 - 167:It is precisely when we are not being agents, when someone else is moving our arm, that we are most aware of these internal signals. When we are being agents, these private signals are suppressed. And this means that we perceive ourselves as being agents in the same way that we perceive others as being agents: we note the relations between actions and the effects they cause. We take into account what we know about prior intentions. But we don?t take account of the physical sensations experiences by agents. It is precisely because we don't have any direct connections with the physical world, even the world of our own bodies, that we are able to enter the mental worlds of others. The mechanisms that evolved within our brains to understand the physical world also enable us to enter the mental worlds of others.
From Pages 179 - 183:Many different meanings can lead to the same words. So how do we choose the best meaning? The key point is that this is the same problem that our brains have solved long ago in order to perceive the physical world. The meaning (in this case, the cause) of the signals that strike our senses is ambiguous in the same way. Many different objects in the world can lead to the same sensory signals. What looks like a complex pattern of lines in two dimensions could be a simple cube in three dimensions (see Figure 5.10). As we have seen, our brain solves this problem by using guesses about the world to predict what will happen next as we act upon the world. The errors in our predictions enable us to refine our guesses until we have a good model of what is out there in the world. In the same way we (or rather our brains) guess what someone?s goals may be and then predict what they will do next. We guess what someone is trying to communicate to us and then predict what she will say next.
Prior Knowledge and Prejudice
So how do we start with our guessing? Making guesses about what people are like before we have any information about them is prejudging them. It is prejudice. Prejudice might be a dirty word these days, but it is in fact crucial for our brains to function. Prejudice enables us to start our guessing -- and it doesn't matter how accurate the guess is, as long as we adjust our next guess in response to the error. To use an innocuous example from Chapter 5, when we perceive objects in the physical world our brain always expects the light to come from above (see Figure 5.7). This is a prejudice that has been built in by evolution. When our brain watches people moving, it expects them to achieve their goals with a minimum of effort (remember the studies of imitation I described in Chapter 6). This too is an innate prejudice. These prejudices enable us to start the cycle of guesses and predictions through which our model of the world becomes more and more accurate.
We are innately predisposed to be prejudiced. All our social interactions begin with prejudice. The content of these prejudices has been acquired through our interactions with friends and acquaintances and through hearsay. I talk quite differently with my work colleagues than with the non-scientists at the party. There are so many things I expect my brain imaging colleagues to know already, so much shared knowledge. I can use all that jargon about stimulation, and BOLD5 signals and response suppression. But the Professor of English understands BOLD and suppression in quite a different way. I must be careful what I say -- she undoubtedly thinks that all psychologists are Freudians. Our prejudices begin with stereotypes. The first clue I can get about the likely knowledge and behavior of someone I know nothing about is from their gender. Even children as young as 3 have already acquired this prejudice. They expect boys to play with trucks and girls to become nurses.
Social stereotypes provide the starting point for our interactions with people we don't know. They enable us to make our initial guesses about the person's intentions. But we know that these stereotypes are very crude. The guesses and predictions we make from this limited knowledge will not be very good. Once we notice that someone is different in some way from our friends and acquaintances, then our brain expects that communication will be more difficult. We will have less in common. Our brain is less certain about what knowledge we share. So it is more difficult to predict what the other person will do and say. Of necessity, the way we communicate will be subtly altered when we try to communicate with someone different from us.
The Truth
In the very distant past our ancestors too were alone, constructing their models of the physical world, but unable to share them with others. At that time truth had no relevance for these models. It did not matter whether the model was a true reflection of the physical world. All that mattered was that the model worked by predicting what would happen next. But once we can share our models of the physical world, then we discover that other people's models are slightly different from our own. Some people are experts who clearly have better models of some aspects of the world. By putting together the models of many people, we can construct a new model that is better than any model produced by a single individual. And our knowledge of the world is no longer derived from a single lifetime -- knowledge passes from one generation to the next.
Can false models also be shared? A disordered brain can make a false model of the physical or mental world. Such a brain can create visions or the sound of voices when no one is speaking. But false models of the physical world are not so easy to share. I am not going to hear the voices created inside your brain. If I have a strange experience, I may check it out by sharing. "Do you hear a strange ringing noise, or is it just me?"
False models of the mental world are not easily so checked. And sometimes these false models are successfully shared with others. In cases of folie à deux, two or more people share the same psychotic delusions.
As long as this false model of literary world stayed within the family, the "normal" members believed it to be true. But once they discussed their beliefs outside the family, the lack of truth became immediately apparent.A 43-year-old housewife-writer was admitted to the hospital in a severely agitated state. Her history revealed a delusional state of 10 years duration regarding a conspiracy in the literary world. Her husband and three adolescent children shared these beliefs. Her primary diagnosis was paranoid state with a schizophreniform psychosis. The patient responded quickly to drug treatment. The children and husband agreed after two visits that they had mistakenly gone along with the patient's "over intense imagination."
But when false beliefs are shared by larger groups, truth becomes more fragile. This seems to have been the case in the tragic "Jonestown massacre."
[...]
Our brains' ability to communicate ideas from one mind to another can bring horror as well as benefit. We all know how is easy it is, at least briefly, to be deceived by false beliefs.14 Our mental currency consists of beliefs created by our brains. But I am optimistic. Whole communities rarely embrace false beliefs so whole-heartedly as the people in Jonestown. And beliefs are not as arbitrary as something like money. Our beliefs are models of the world, and the real world out there is a gold standard for
our models. In the end false beliefs can always be discarded because they make bad predictions.
I believe that the truth is out there. As long as we have ways of showing that one model of the physical world works better than another, then we can aspire to developing a series of better and better models. At the end of this series, although it is infinite in the mathematical sense, lies the truth -- the truth of how the world really is. Reaching this truth is the program of science. Science progresses by making models of the world, making predictions on the basis of these models, and using the errors in these predictions to construct better models. Now science is revealing that our brains use the same principles to acquire knowledge about the world. We are also beginning to understand how our brains can make models of the mental world. It is by sharing these mental models that the program of science becomes possible.
