Click here for Table of Contents

Click here for home, brianrude.com

Skills

     When we think of skills we usually think of motor skills. Throwing a ball is a skill. Walking is a skill. Typing is a skill. Playing pool, unlocking a door, writing and drawing, pouring milk into a glass, turning a screw with a screwdriver, and cutting paper with scissors, are all skills. All these examples are movements or series of movements.

     Movement occurs in response to neural impulses sent from the brain. A muscle consists of a large number of muscle fibers, each supplied with a nerve ending. When a muscle fiber receives a signal from its nerve ending it twitches. When a lot of muscle fibers in a muscle get impulses at the same time the whole muscle contracts, its strength of contraction depending on how many muscle fibers got signals to twitch and how many did not.

     Each individual twitch of a tiny muscle fiber cannot be said to be a skill in itself. The skill consists of the pattern, over time, of a very large number of individual muscle fiber twitches. This pattern of muscle fiber twitches, of course, comes from a corresponding pattern of neural impulses sent out by the brain. A skill, then, is in the brain, not in the muscles. A poor pattern of neural impulses from the brain produces a poor skill. A good pattern produces a good skill. For example when a small child first tries to grasp a spoon he is very clumsy. His brain sends out messages to thousands of muscle fibers at one time and the resulting movement sends his hand in the general vicinity of the spoon. His brain sends out more messages and the hand opens. When the child's sense of touch tells his brain that the palm of his hand is touching the handle of the spoon, the brain sends out messages that cause thousands of muscle fibers to twitch causing a rough grasp of the spoon. The end result of all this is a very clumsy movement by the child. As time goes by however, the child refines this pattern immensely. The brain learns that some muscle fiber twitches can be delayed a few milliseconds, other muscle fiber twitches can be spread out over a second or two. Countless other muscle fiber twitches can simply be deleted. Other groups of twitches must be tied to the reception of key tactile or visual stimuli. Through all these adjustments the brain little by little learns to send out a better and better pattern of neural commands. Thus the action becomes more and more skillful.

     There are two general ideas about the development of motor skills that I want to address, before then going on to discuss perception and cognition as skills.

      First is the idea of "cortical lead" in skill development. The "cortical" here, of course, refers to the cerebral cortex, which we use for conscious thinking. Skill development occurs through a process of trial and error. We try, and try again, and get closer to the desired response the more we try. Should each trial be simply another attempt with no thought or plan? Or should each trial be consciously controlled? Should the cerebral cortex lead the other parts of the brain? I will argue that both of these are true, that both are just different ends on the same continuum.

     Consider this example: A two year old has a new toy, a game in which the central point is to whack a mechanical mouse with a swatter when it sticks its head out of the mouse hole. The child simply tries again and again, and over time develops some skill. Does the child consciously "practice"? Compare this example with this next example: An accomplished pianist is working on a new piece. At one point in this piece a string of six notes in the right hand requires a fingering pattern in which the seventh note must be played with the middle finger in order for the right hand to be in position for a subsequent chord. The pianist analyzes the notes and sees that the best way is for the thumb to twice go under the index finger and then fingers two and three to skip keys. She then practices this pattern slowly with only the right hand until the movement is reasonably proficient. Then she plays it faster. Then she adds the left hand, again playing slowly. Then she plays it with both hands faster. Then she starts a few measures before the difficult passage and tries to get it in context as it naturally comes. Then she practices from the beginning of the piece. In this example the acquisition of the skill is very carefully led by conscious thought. The pianist is using a cortical lead. Her cerebral cortex leads the parts of the brain that control her movements. She consciously decides each step in the practice and deliberately makes each movement in accordance with a plan. The child whacking the mouse, on the other hand, shows none of this conscious planning. He is not using a cortical lead.

     Should we ask the child to practice whacking the mouse with the meticulousness and deliberation that the adult pianist uses? Is a cortical lead the only way to practice? Or should we advise the pianist to quit thinking about it and just do it? Is there such a thing as too much cortical lead? Should both the child and the pianist use the same amount of cortical lead?

     I believe neither the child nor the pianist should be advised to copy the other. Both are doing right for their individual purposes. They should not use the same amount of cortical lead. They are on two ends of the continuum. On the child's end of the continuum little or no cortical lead is appropriate or desirable, or even possible. On the pianist's end of continuum, a great deal of cortical lead is absolutely necessary. Progress would not be possible without it. Only by this conscious planning will the pianist spend sufficient time concentrating on the small details that require intense and prolonged practice. Only by conscious planning will she figure out what those small details are. The two-year old might theoretically gain this same advantage if he could concentrate his thinking on the problem. But of course he can't. He is simply incapable of this conscious planning. That will come in later years. Perhaps more importantly, his conscious planning may be less effective than his mindless practice. Cortical lead might offer no advantage whatsoever. The cerebral cortex is only one part of the brain. It is indispensable for multiplying fractions and composing poetry, but other parts of the brain may be much better for whacking mice. Children do not learn eye-hand coordination because we teach them, or tell them to learn it. They learn it because they are given the opportunity, and some part of the brain automatically takes the opportunity.

     I will next introduce an example to show the middle ground in this continuum. Consider this example: A twelve year old boy wants to play football on the school team. He wants to excel, especially as a kicker, so he consciously decides to practice everyday for the rest of the summer. Should he just kick the ball everyday with no more planning than the two-year-old child applies to whacking the mechanical mouse? Should he practice with all the deliberation of the pianist I described? Or would he do best in some middle ground?

     For the twelve-year old football kicker neither the mouse-whacker model nor the pianist model is entirely appropriate. A twelve-year old can certainly benefit by conscious planning of practice. Coaches of children this age do coach. They tell their charges what to do, what not to do, what to change about what they are doing, etc. But also at this age a lot of progress is made just by doing. The twelve-year old can profit by a good deal of relatively mindless practice. An intermediate level of cortical lead seems appropriate for this, and many other, situations.

     The degree of cortical lead is very dependent on experience. The pianist I described used little cortical lead when she first took piano lessons at age seven. She couldn't. She had no conception of how to. By age twelve her practice was much more directed. She used much more cortical lead in her practice because she had been shown how to. As time went on her teacher directed her attention to more and more details and subtleties. When taking piano in college her teacher directed her attention to even more details and subtleties so that she learned to use even more cortical lead.

     The second aspect of skill learning that I want to discuss is the idea of "learning curves". I'm not sure I'm using the term in the same way others use it. I am concerned with the simple idea that in any given topic learning starts rather slowly, speeds up as the mind gets the basics needed for progress, and then slows down as diminishing returns set in. This is a very general description, but it often applies to many different topics. It does not apply only to motor skills, but perhaps is best illustrated by motor skills. And it has some important implications.

     I will use the two hypothetical graphs below to illustrate the general idea of learning curves. In each graph the curve starts with a low slope, then gets considerably steeper, and then levels off again. The difference between these two curves is in the details, not in the general shape.

     4 5 6 7 8 9 10 0 1 2 fluency in letter recognition accuracy hitting croquet ball In the first graph the horizontal scale covers ages four through 10. This is six years, a long time. During this time fluency gradually increases. The vertical scale of the graph would have to be something like the number of letters correctly identified per second when presented by flash cards.

     The second graph represents the following situation. A twelve year old girl is exposed to the game of croquet. Each evening in the summer the family sets up the wickets and they enjoy playing for a half hour or so. The object and rules of the game are quickly learned, but the motor skills are not so easy. The essential skill involved is the ability to accurately hit the wooden ball with the mallet. She starts out rather chaotically, clumsily whacking the ball with little effect. After a few days she has gained a great deal of sensory and motor information. She begins to take aim before she strikes the ball. She greatly increases her cortical lead. And of course she makes progress. By the end of the summer, she plays fairly well. She continues to improve, but improvement comes slowly. She is not interested in perfection, so she does not practice to any great extent. She just enjoys the game. She probably realizes intuitively that further progress would come only a the cost of a great deal of intense and tedious practice.

     The horizontal scale in this graph would cover about two months, a considerably shorter time than for the first graph. The vertical scale could be measured as something like "wickets scored per hundred strokes".

     In both of these situations learning starts slowly. It starts slowly because there is little to build on, because things seem chaotic and confused, and because one doesn't know how to practice. But as one gets beyond this initial stage the learning rate increases. Certain fundamentals get "nailed down" and one can get into a routine. The later stage of learning is again slow because the learning is virtually complete. Getting from "near perfection" to "perfection" can take

    forever. Time is the most important consideration of a learning curve. How long is one in the early stages in which learning is slow and confusing.? How long is one in the productive middle stage? How long is one in the last stage when progress is again relatively slow? Some topics go through the entire learning curve quickly while other topics go through the learning curve slowly, and yet other topics have a very long learning curve, measured in years. In the first graph above the time line is measured in years. It takes a long time to really become fluent in letter recognition. In the second graph the time line is measured in days. Yet other examples might have learning curves measure in hours, or in decades.

     When learning to walk the beginning stages may be recognizable not long after birth. This is a stage of slow learning while the child matures enough to really begin learning. Then comes a stage of much faster learning. Shortly thereafter, perhaps at age about two the learning of walking is all over The child no longer has to learn to walk, he simply walks. Learning is essentially over, for there is nothing left to learn. The child is at the very top of the learning curve, where returns in learning are almost zero.

     Compare this to the learning curve in calculus. One cannot even begin until one has a solid foundation in algebra. For most people the learning curve is cut off rather abruptly at the end of some academic year. For those who continue deep into calculus the total learning curve might cover ten years or more.

     It would seem desirable to try to stay on the productive middle sections of learning curves. This is when the learner receives the most benefit in relation to his effort. This is basically a restatement of what I mentioned in Chapter Four concerning the degree of fluency to be obtained before moving on to another topic. If one jumps to the next topic too early then one misses the productive middle stage, and doesn't have a good foundation on which to attach more learnings. If one aspires to an unrealistic degree of fluency then one spends a lot of time in the upper part of the curve, the part where returns on effort are more and more diminished.

     Both the early and late sections of the learning curve have a low slope. They give little result in relation to effort. Thus it is best to avoid these sections when possible. It is usually easy to avoid the late section of the learning curve, simply by refusing to work for perfection. Obviously this is not a strategy used by Olympic gold medalists. They must spend years trying to get from near perfection to perfection. But this does not apply to everyday teaching and learning. In everyday teaching and learning the choice of an appropriate degree of fluency is important.

     It would seem impossible to avoid the early section of the learning curve. You can't get to the middle productive stage until you have gone through the early less productive stage. However there is a way out of this. Very often the early stage of learning one thing can be embedded in the middle stage of some other learning. Indeed this is unavoidable, because after the very early stages of life very little is actually new. Most everything we learn after infancy is primarily recombinations and extensions of early learning. We encourage free play and exploration among young children. One very important reason for this, I think, is that we intuitively understand that embedded in this free play are the very early stages of a great deal of learning.

     I will now return to the basic definition of a skill that I began with and discuss perception as a skill. A skill, I have argued, is the brain's ability to send out a large number of neural impulses in a precise pattern. These neural impulses are partly in response to sensory input, and partly as a chain of motor-motor associations. A consideration of the stimuli that prompt these neural impulses is very important, of course, but for the moment I want to consider only the neural output. Can this idea be related to the idea of a perceptual or conceptual skill? What is a perceptual skill? What is a conceptual skill? My hypothesis is that perceptual and conceptual skills are very similar to motor skills in that they are a matter of the brain being able to send out a large number of neural impulses in a highly complex and precise pattern. They differ from motor skills only in that the ultimate destination of these impulses is other parts of the brain, instead of muscle fibers.

     In the rest of this chapter I will be primarily concerned with developing the idea that perception is a matter of skill. In the next chapter I will try to develop the idea that reasoning is a skill.

     When a young child first starts to learn the letters of the alphabet he must not only build the association between the sight of a letter and the concept of the letter, but he must first learn to perceive the letter. Seeing is not automatic, his brain must do something. It must send out impulses to perception centers in the brain, and these impulses somehow enable perception to occur. This may sound strange at first. It might be argued that perception is a more passive thing, more a matter of just opening some kind of gate and letting the sensory information flow in. It would seem that one doesn't have to "learn to see" the shapes of the letters, one has only to engage in the more difficult matter of remembering which letter is which. However there are several lines of evidence that suggest it is not quite that simple, that perception itself is a matter of skill.

     The first line of evidence that perception is a skill involves the idea of attention. Consider this example: A man is deep in thought, staring at a closet in his house, trying to figure out how to remodel that closet into a small bathroom. His wife comes up, says a few words to him, then leaves. He replies with a few noncommittal words and continues his thought. A few minutes later he realizes he doesn't have to slightest idea what she had said to him. Suspecting she will be irritated if he doesn't listen better he then goes to her to get the message. This time he has no trouble. He receives the message. How did he completely fail to get the message the first time? He understands English, doesn't he? And there was no noise to block out her words. Why didn't he receive the message? The answer, of course, is that he simply failed to pay attention. The man in the example, as you might suspect, was me, and it was not the first nor the last time that my attention was found wanting.

      This example makes attention appear as a conscious and active thing, and that, I believe, is the case. One can direct his or her attention. One can hear, or one can be deaf. One can see, or one can be blind. It does seem that the willfulness of attention is not quite the same as for motor actions. Our attention is often directed rather automatically, and it can be hard to direct attention as we would like to. If one is listening intently to a weak voice on the telephone it is easy to be blind to a child munching on a stolen cookie. If one is looking intently for a certain item in a book it is easy to be deaf to another's question about the price of coffee this year. It can be very hard to direct attention to a boring lecture when one is more concerned with an exciting party planned for the evening. It can be very hard to attend to anything when a tooth is aching. However within broad limits attention is a willful act, just as movement is a willful act. Teachers demonstrate this every time they snap at a class to pay attention. It might also be pointed out that muscle movement, which we think of as totally voluntary, is not always as voluntary as we would like to think. When the dentist says "open wide" we may find it hard to do so. Both movement and attention are under voluntary control, but within limits.

     Thus I am led to believe that the brain sends out messages to perception centers when it wants to perceive something, and if these messages are not sent out then perception does not occur. This is what attention is. So to some extent attention is like a skill. The brain must do something. Still, it does not follow that the brain sends out a large number of signals in a precise pattern, as it does in a motor skill. However a second line of reasoning leads in this direction.

     This is the "data sieve" argument. At any one moment there is a very great amount of sensory stimulation impinging on a person's sense organs. The retina of the eye, for example, consists of thousands of nerve endings. In the normal light we live and work in practically all of these are being stimulated to one degree or another practically all the time. Yet the brain cannot attend to all of this input from the retina at one time. The words on the computer screen as I write this take up a small part of my retina. My brain focuses on this small amount to the exclusion of all the rest because it is what is significant. If I were to attend to all the unimportant visual stimuli that takes up the rest of my retina then I would not be able to attend to the small part that is important. The cochlea in the ear contains thousands of nerve endings. But again the brain can not attend to each and every impulse sent from every auditory neuron. Every muscle and every joint in the body has proprioceptive nerve endings which constantly send information to the brain. Again the brain cannot attend do to all of this at one time. Therefore the brain must have some means of limiting this flood of impulses and sorting it out, opening the gates for important information and closing gates to irrelevant information. There must be some sort of "data sieve", something to screen out unwanted information and allow in wanted information. Further, this data sieve must be instantly changeable and minutely controllable. The information that is totally irrelevant one moment may be very relevant the next moment. Thus it seems reasonable that the opening and closing of these many gates, the operation of the data sieve, would be a matter of skill. To get the information it wants, and to exclude unwanted information, the brain must act. It must send out a large number of impulses, perhaps mostly to close gates to unwanted information, in a precise pattern.

     The third line of reasoning that perception is a skill, and that it rests on the brains ability to send out a large number of impulses in a precise pattern, rests on the simple observation that perception takes practice. This is not at all obvious, and I will give examples at some length.

     Most of our perceptual skills, I believe, are developed very early in life, before we are even ready to stand up and take a step. An example of this is learning to perceive pictures. Very young children do not automatically see the relation between a picture of a chair and a real chair. They learn to see the relation through a process of learning. Perceptive parents interpret pictures for children when they are very young. They point out what is in the picture, even though it may seem too obvious to need pointing out. By school age interpreting pictures is such an advanced and automatic skill that we forget it was ever otherwise.

     A few years back I got an inkling of what would be like to be without this skill of interpreting pictures. I had been helping my young son with his reading about every day. We usually sat so that the book he was reading from was upside down to me, while being right side up to him. I expected not to be too fluent in reading the upside down words, but I didn't expect trouble in interpreting the pictures. One day, however, we were reading a book about dogs. Suddenly I realized from the text my boy was reading aloud that the dog in one of the pictures must be a puppy, not a grown dog. I turned the picture right side up and was astonished to see that it was obviously a puppy, though I had not perceived it as a puppy in the half a dozen times that we had previously read that particular book. After that experience I began to be aware that when looking at the pictures upside down I invariably miss a good deal of what is in them. The logical explanation is that I had just not had a great deal of practice in interpreting pictures upside down, and so my perceptual skills in that way were not highly developed. As a way to test this idea I suggest the following: Choose a magazine that you have not read before. Every day for a week look through it upside down, not trying to study the pictures carefully, but only to become familiar with them. Then after a week look through the same magazine right side up. If you are like me you will notice a good deal that you had missed in the upside down looking. This supports the idea that perception is a skill. It must be learned, and it takes a great deal of practice.

     Why is it hard to read words upside down? Again my explanation is that perceptual skills develop only with practice, and we don't usually get much practice reading upside down. When listening to my children read, as I described, and especially after the puppy picture incident, I decided I would learn to read upside down. I might as well, I figured, for I would be listening to my children's reading for a few years to come. I made some progress, to be sure, but I never became fluent. I was not as resolute in practicing as I first planned, but I made enough progress to definitely feel that practice was producing results. I interpret this as at least some support for the thesis that perception itself is a skill.

     A similar example of learning to perceive is learning to make sense out of what one sees under a microscope. I well remember my first few weeks in my first zoology course in college. The laboratory exercises involving the microscope caused me a substantial degree of frustration. I simply was not seeing what I was supposed to see. I scraped by, though, and forgot about the problem after a time. I didn't realize how much I had learned until a year later when I chanced to overhear a student asking the professor for help. It seemed that he just couldn't make head nor tail out of what he was supposed to be seeing under the microscope. That got me to thinking about how hard it had been a year earlier for me, and how easy it was now. I had developed a skill over that year.

     The skill was not totally absent at the beginning of the year, of course. I had used microscopes in high school biology. But the development of the skill was rather dramatic. Since it developed gradually, as skills generally do, I had given it little or no thought during the year. And, of course, we cannot easily quantify perceptual skills. We seldom give them any thought. Only a bit of reflection, prompted by a chance overhearing of another student's problem, revealed this development of a perceptual skill.

     A refinement of microscope perceptual skills is to learn to keep both eyes open while looking through the microscope. One eye gets the image from the microscope, while the other eye gets the routine image of the table on which the microscope is sitting. I believe with a little practice one can learn to effectively and efficiently attend to the important microscope image and neglect the unimportant image from the other eye. A similar phenomenon is found, I believe, by some people who have cataract surgery and wear corrective contact lenses. They have one contact adjusted for distance vision and one adjusted from near vision. The brain can learn to use the image that is in focus and neglect the image that is not. I think the very same thing occurs naturally when a person needs optic correction for one eye but the other eye does not. The brain automatically attends to the better focused image.

     I believe that a substantial part of learning art is a matter of developing perceptual skills. I recently began carving violins out of hunks of wood. This illustrates the idea of visual perception as a skill. I had looked at violins plenty of times throughout my life. Why would I need to look at violins any more. Yet after doing some carving I continually had to look back at a real violin for guidance on just what the shape of the violin is. Now I have become much more sensitized to the shape, and automatically see nuances that before I was blind to. I have heard that learning art is a matter of educating the eye more than educating the hand. In other words visual perception must be learned. This certainly makes sense.

     I mentioned a few paragraphs back that it might seem reasonable to think of perception as a simple opening of gates to allow the sensory stimulus into the brain. If this were the case then it would seem that a perceptual skill should come very quickly, rather than coming slowly with practice. I will hypothesize that perception works by opening gates, as I mentioned, but that the number of gates to be opened, or kept closed, is very large. This would fit with the argument I gave above about massive sensory input to many sense organs, most of which must be ignored at any one time.

     I will also hypothesize that there are many levels of gates. The data sieve is not one mechanism somewhere in the brain, but the workings of many parts. For example some part of the brain attends to a certain sector of the retina. At times the perceptual gates would be closed so that no perception takes place there. At other times perception may take place there, perhaps a letter "o" is perceived, but that information is sent no further, being irrelevant to whatever the brain is concerned with at the moment. At a higher level, the awareness of the letter "o" may be received by the brain, but given no more relevance. At yet a higher level the brain may be aware that the letter "o" is part of the word "now", but anything more is irrelevant. At still higher levels the words become parts of sentences, and the sentence become parts of thoughts and so on.

     A large number of gates, and a hierarchical level of gates, would require a large number of neural impulses to be sent out from the brain, and the pattern of the impulses would have to be precise. Perhaps there are millions of gates that can be opened or closed in the perceptual centers of the brain, and any given task, such as perceiving a single letter, requires that a certain precise pattern of several thousand of them be opened and the others remain closed. This would fit closely with the idea of skill that I described with regards to motor skills.

     This idea of innumerable gates would seem to apply to this following example of visual perception. The other day, for what reason I cannot remember, I sat and looked at a closet door. This consisted of two narrow louvered doors that open outward, finished in varnish over stained wood. Many times I had used those doors, but only in the moment of careful observation did I notice a number of things. First, the wood pieces did not match well in color. Some pieces were considerably darker or lighter than others. Had I assembled that door from its pieces, I thought, I surely would have done a better job in matching shades of color. Next I noticed the ornate knobs on each door. Then I noticed the style of hinges. And for a few minutes more I just noticed one more little detail after another. If visual perception consists only of opening a few general "vision gates", then it would seem that everything would be noticed at once, like the shutter of a camera opening. I hypothesize that it is more complicated than that, that each time I noticed another detail it was because I had opened a new combination of gates and closed down other combinations. By this perspective it seems quite reasonable that a cabinet maker who works everyday with wood and knobs and hinges and varnish would notice in a flash what I noticed over the course of minutes. His perceptual skills would be such that he would open and close different combinations of gates quickly and efficiently, not slowly and clumsily as I would do.

     One might interpret this example as more a matter of attention than perception. By this interpretation one is not engaging in different perceptions, with different perceptual skills. Rather one is using the same perception and perceptual skills applied first to one detail and then to another. This interpretation may be valid, but it seems less applicable to the next example. Consider this figure. The cube on the left can be showing its bottom, like the middle cube, or it can be showing its top, like the cube on the right. It depends on how you look at it. This is nothing new of course. I have seen this example a number of times in different contexts. Here again it might be interpreted as a matter of attention. However one can certainly say he is giving it full attention whether he sees it as bottom facing or top facing. It definitely seems more than a matter of attention to me. I am doing something different when I see it bottom facing than when I am seeing it top facing. This fits very well with the idea of perception as a skill. I am opening different gates to see it the two different ways. Attention is involved, but attention is not all there is to it. I have never learned Braille, but now and then I have had access to a few words in Braille. For example the plastic lid on a soft drink paper cup sometimes has Braille markings. I have carefully felt such markings, trying to feel the pattern of dots that I can easily see. I am amazed that those tiny dots can actually be perceived and be meaningful, for I seem to feel only a general roughness. I can direct my attention to my fingertips, or at least I think I can, but that does not seem to be enough. Does that mean I could never learn Braille? I presume it does not. Were I to become blind I would probably learn Braille about as quickly as anyone else. An important part of that learning, I believe, would best be interpreted as learning to perceive. Perception seems to be more than just direction of attention. I have argued that perception takes practice, and that this is evidence of the skill nature of perception. But perception must work together with knowledge. A perceptual skill by itself would be as useless as a motor skill by itself. Any skill must be guided with knowledge. The association of knowledge with perception is worth exploring to some depth. I was told by a friend many years ago that when he was in the military, apparently during World War II, they were trained to instantly identify aircraft at a glance. They were first given pictures of the different types of aircraft, told what to look for to distinguish among them, and then drilled on them. Eventually, I was told, they could instantly identify any aircraft when its silhouette was flashed on a screen for a small fraction of a second. This makes sense when perception is viewed as a skill. A skill takes practice, and practice can produce amazing results if carried far enough. Of course a very important part of this is knowledge, rather than perceptual skill, but I think perceptual skill is also very important. There are many mundane examples of this kind of perceptual skill developed through practice and associated with knowledge. As a teenager I could instantly recognize a wide variety of models of cars. I didn't realize this till one day my mother shocked me by misidentifying one. I thought a Volkswagen and a Renault were as different as night and day, but to my mother they were both simply small foreign cars. Identification of cars, or aircraft, or cows or houses, may include a large component of knowledge and this knowledge can be very important. But the quickness and fluency with which one can tune in to small details suggest skill, not just knowledge. I will relate another example that illustrates this idea of perceptual skill and knowledge working together. When taking botany in college we had to identify trees by their physical characteristics. This contained a substantial degree of knowledge. Some trees have alternate simple leaves, for example, as opposed to opposite simple leaves, or alternate compound leaves, or opposite compound leaves. This is a matter of knowledge, and is easy to distinguish when looking at a specimen. However identification of the leaves and twigs is also dependent on things such as color. Some leaves are distinguished by their color being a little different than the ordinary green. When one is attuned to such subtle difference in color they may seem very obvious. But when one is not attuned to such subtleties distinguishing the difference in color may seem impossible. It involves subjective judgments based on sensory data. One must develop perceptual skills along with the knowledge. Another feature to look for when identifying trees is the texture of the leaves. The written description may describe a leaf as having "rough" texture. To the novice this is so indefinite as to be useless, but to the experienced person it is useful information. It is useful because the perceptual skills needed to make use of this information have been developed. Judgments may be made, that while still being subjective, are educated and meaningful. Using the example of identifying trees I have described three criteria, twig structure, leaf color, and leaf texture. These are in order of their subjectiveness, and their dependency on perceptual skill. I would add a fourth criteria, which is even more subjective and dependent on perceptual skill. That is the tree's profile, or distinctive shape. Elm trees are said to have a characteristic "Y-shaped" pattern of growth. Looking at an elm from a half a block away, and comparing it to other species of trees this description seems to fit. But trying to pin it down is difficult. Many other trees can be called "Y shaped" in the sense that the branches grow out an acute angle to the trunk. When looking closely at a tree the "Y-shape" is not apparent. One cannot look closely and say just what constitutes the "Y" profile. Thus no text book on trees can strictly define the characteristic. Yet after one looks at a lot of trees and develops an "eye" for their shapes, the "Y-shape" is very meaningful. It is meaningful because one develops the perceptual skill to make it meaningful.

     There is a continuum between learning to perceive and learning to direct one's attention once one already knows how to perceive. There is also a continuum between perceptual skills and knowledge. In the later stages (and "later stages" in perception could be by age two or three) most learning is entirely a matter of direction of attention and acquisition of knowledge. But in the earliest stages of learning, and occasionally thereafter, one must learn to perceive, and this, I believe, is matter of skill.

     I have argued that perception is best interpreted as a matter of skill. Next one must ask what difference this might make. Is it only academic, or are there some practical consequences or implications? I think there are many consequences and implications. But I think they are best investigated in the context of teaching particular subjects or topics. I have mentioned, for example, the pronunciation of sounds in foreign languages that are not found in one's native language. This involves perception as well as knowledge. I believe this perceptual component is best investigated in the context of language learning. Geometry, as another example, certainly involves perception and perceptual learning, and this, I believe, is best addressed in the context of teaching geometry.