CONTRIBUTIONS TO EDUCATION ►v< s* ^vv^vw-^i-v^nv^'^ - •: ':• IPm! ■'M ?'J*&2%-5jifei& »ss2£ >5; V'" • aga^&s^^ sSe? **•<£ 'U - ^a* - s ■€-''*'>■>-;■ ... $§&$»£ 0& €$& i* ..,< '•-> ".' £r;x Mi»^^ 5 t?ss»^ 2 aR€ 4 ®H^ •<*» "n.- ->-•. *>*v. v v'-vji ># - ' ~ >' -v^~*-.'‘V "-' J .-.' j.-.’• -:: m £ 3 36 *: 7.- •?i«BkJWi« >' ;;• *'*.«5wfc H tv. ^rT.-.% ..js£< S&''- a&o?. -If: »«• '&«a& k#$- 4. jg'S ' •»■>' *.'• ^•’Srewfwf 255.*!® £* j *:% w^;."^ **£< v ^ *§£18 ' ^•■',-v:Sr fvtir’.r^:"*'^ '*’’ ’ *"^ -.f J£^it 4 . ■ >.*•i;. v ..{ -b^?Sfe>sf»;s r\jsC'/i With the Writer’s Compliments CONTRIBUTIONS TO EDUCATION OF SCIENTIFIC KNOWLEDGE ABOUT THE PSYCHOLOGY OF LEARNING By J. F. DASHIELL Kenan Professor of Psychology University of North Carolina Chapel Hill, North Carolina Offprint from the Thirty-Seventh Yearbook of the National Society for the Study of Education, Part II. “The Scientific Movement in Education,” 1938 I !# y., , * Note: This offprint of Chapter XXXII (pages 393-403 of the Thirty-Seventh Yearbook, Part II) is made by permission of the Society for private distribution by the author. Copies of the entire yearbook (534 pages) from which this offprint is made may be secured for $3.00 postpaid in paper binding, or $4.00 postpaid in cloth binding, from the commercial agents of the Society • ;-V ? / . THE PUBLIC SCHOOL PUBLISHING COMPANY . Bloomington, Illinois All copyright privileges reserved by the Society CHAPTER XXXII CONTRIBUTIONS TO EDUCATION OF SCIENTIFIC KNOWL¬ EDGE ABOUT THE PSYCHOLOGY OF LEARNING J. F. Dashiell Kenan Professor of Psychology University of North Carolina Chapel Hill, North Carolina I. Education as Guided Habit-Forming Education is guided learning. Learning, in turn, is habit-forming, if taken generally enough; for even when in a particular case the aim is not that of forging some mode of thinking or acting that will operate always in a fixed way thereafter but is that of developing a more gen¬ eral attitude—that of independent inquiry, say, or that of making a student resourceful in solving ‘originals’—even then the teacher’s regard is toward the learner’s future, and he hopes that the present activity will set up in the learner an abiding tendency of some sort. Education is guided habit-forming. II. The Structure of a FIabit Within the time of the present generation and up until very recent years there flourished a beautifully over-simplified notion of what we mean by the word ‘habit.’ It is well illustrated by a classic experiment on the white rat, by Watson in 1907. In order to determine which of the sensory functions of that animal are finally essential to its learning to thread its way successfully through a complicated maze, he tested in the maze blind rats, deaf rats, rats with olfactory bulbs removed, and rats minus their vibrissae and with anesthetized feet, in each case the animals showed much the same capacity to learn as did normal ani¬ mals, and by the logical process of elimination W atson concluded that only the kinesthetic stimulations arising from the animals own moving muscles (also possibly the organic) were necessary for the building up 394 CONTRIBUTIONS OF THE PSYCHOLOGY OF LEARNING of a particular maze-running habit. And this apparent logical outcome bolstered the theory that what a habit essentially consists of is a chain of movement-aroused reflexes. When a man makes the first responses of a learned act, the very moving of the muscles and joints involved sets up sensory nerve currents that arouse the next responses; the currents from these arouse the next; and so on. A complicated performance—be it reciting a passage from Virgil or repeating the multiplication table or playing a Chopin nocturne—becomes a well-integrated habit to the degree that successive sensori-motor units become welded into a def¬ inite, unvarying sequence of little acts. But more recent experimental work by many (Gengerelli, Dennis, Hunter, Lashley, Macfarlane, Dashiell, and others) has exhibited the untenable nature of such an extreme view. Animals can be trained to take, say, a left-hand instead of a right-hand turn regardless of the length of the avenue of approach to the turn and of any new elbows in it. Once trained to run without error in an ordinary maze, they will swim the same maze without error if it be immersed; or again, if after the learning they be so operated that they can progress only by rolling and tumbling they will continue to negotiate the turns correctly. If given opportunity to take diverse routes to the same food-goal, they will oftentimes follow new alternative routes without making any turns in a wrong direction. In short, it has become clear that we need not go higher than subhuman species to observe that an organized habit is something vastly more than a chain of reflex arcs joined by the cement of kinesthetic impulses. Any well-learned performance remains a vari¬ able and adaptable performance. III. The Trial-and-Error Principle In the late nineteenth century, Lloyd-Morgan had invoked the prin¬ ciple of trial-and-error as the explanation of habit forming—“varied trial and error with the utilization of chance success.” This conception was then given powerful support by Thorndike’s well-known experi¬ mental studies on dogs and cats. His method was to put animals when hungry in enclosures from which they could escape by some simple act, such as pulling at a loop of cord or stepping on a platform. The animal was put in the enclosure, food was left outside in sight, and his actions observed. At first the animal’s behavior was decidedly random, consist¬ ing of mewings or barkings, scratchings, bitings, sniffing here and there, looking here and there, with no evidence of any system or directness. DASH I ELL 395 Sooner or later, however, lie chanced to hit upon the particular move¬ ment that secured his release; then, in the course of retrials on the same problem, lie came somehow to make the ^successful movements more and more promptly, until at last the definite habit would be set up of going directly to the lever or other crucial object. Thorndike’s work fashioned the mold in which the experimental studies of animal learning were to be cast for the next twenty-five years Many are the experiments in human psychology (by Bryan and Harter, Book, Starch, Bair, Batson, Chapman, Snoddy, Barnes, and others) leading to similar conclusions. I need mention only that of Swift in which he studied.the progress of learning the juggler’s act of keeping two balls in the air. He made the significant observation that a more effective technique in the handling of the balls was often hit upon without the learner’s being aware of the fact until he came to real¬ ize later that he had unconsciously adopted and established the new trick of throwing or of catching. Certainly it resulted from no analytic perception of just what to do or just how to improve. In consequence of experimental reports like these, the orthodox view of habit-forming, in human as in subhuman forms, leaned to an extreme emphasis upon the random, hit-or-miss, trial-and-err or, unanalytic character of the learning. IV. The Principle of Insight Within the last ten years a reaction has clearly set in. Adams, for example, repeated the Thorndike type of experiment, but with results that have to be interpreted differently. A cat placed in an enclosure appeared at no time to scramble about and claw indiscriminately; and in the course of 31 trials he found his way out, not by one stereotyped movement he had hit on, but by pulling the proper string in no fewer than ten different manners, pulling at it in widely different places and variously with teeth, chin, or paw. Maier, after familiarizing a rat with a table top and also, but separately, with an elevated maze pathway, then placed the animal upon the table with its food behind a barrier but approachable by the roundabout maze pathway. After a short period of vain explorations on the table, the typical rat would suddenly lace about, apparently seek out the maze entrance and promptly run this long-way-round to the food. It was, therefore, concluded that this ani¬ mal can combine the essential elements of two different experiences in a novel and adjustive fashion to reach an objective. 396 CONTRIBUTIONS OF THE PSYCHOLOGY OF LEARNING Most famous arc the researches on apes of Kohler, Ycrkes, Bingham, and others, which are so well known that a single sample will suffice. A chimpanzee, on finding a banana suspended from the roof at a point quite out of his reach, would be seen to make all sorts of ineffectual efforts to obtain it, jumping, running about, throwing objects at the banana, and climbing up and down over boxes left lying about. But at length, while touching a box he would hesitate, his movements would become slower; and then with eyes upon the food he would push his box over under it and mount to a point from which he could pull it down. In a word, then, the ape will solve a problematic situation, not by the method of repeated blind trials with the gradual selection of the successful movements in a manner not necessarily involving any atten¬ tion thereto, but by the method of attentive scrutiny of the field and a 'perceiving of certain crucial relationships within it. This technique of Kohler’s has been applied in America to learning by human infants and children and by defectives (Alpert, Matheson, Aldrich and Doll, and others) ; and the observations obtained on these subjects are consistent with those of Kohler. They are quite consistent also with other studies of analytical learning by human subjects. In general, then, it must nowadays be recognized that a habit is ac¬ quired in man and in many brutes not always by blind hyperactivity with occasional chancing upon adequate movements and then a slow process of unconscious selecting and fixating of these lucky movements; instead it is often a display of intelligent insights. Experimental studies of insightful learning have often been called studies in ‘problem-solving,’ and the term is a happy one. Immediately we see the relevance to many forms of educative training. ‘Originals’ in arthmetic, algebra, geometry, and physics are to be contrasted with the drill work of learning the multiplication table or the spelling of impor¬ tant words. So, too, with the ‘thought questions’ in civics, in literature, and in many other studies; the learning task then set the child is not that of fixing by repetition a given mode of speaking, writing, or acting, but that of bringing his resources and his experience—his stock of habits—to bear on a difficulty that is new to him in some ways; and he is forced to make some analytic examination of it to see whether he can discern the crucial relationships that obtain. More generally, we may say that a fundamental assumption of present-day education is that the child is learning in the truest sense when he is achieving new insights. DASHIELL 397 V. The Principle of Conditioned Response During the last fifteen years or so much attention has been given in the experimental laboratories to the phenomenon of the ‘conditioned response. The observations and claims of Pavlov and of Bekhterev have been subjected to increasingly detailed experimental examination and, with scarcely an exception, verified. Those Russian workers had found that a dog could be trained to salivate or a man to jerk back his foot involuntarily upon the receptiqn of any stimulation, however seemingly irrelevant, if only it be frequently accompanied or closely followed by a stimulation that was already effective. Thus there reappeared the centuries-old principle of learning by association, but now clothed in garments of high biological respectability. Many particular detailed principles contained in the phenomena of conditioning were unravelled by Pavlov’s genius—conditioned inhibition, differentiation, irradiation, and several others—and the bulk of experimental work since has been devoted to their successful verification (by Hull, Hilgard, Marquis, Bernstein, Garvey, Jones, Switzer, Wolfle, Razran, Skinner, and others). On the interpretative side, however, the views of Pavlov seemed not to have been fairly grasped until of late. It had been supposed by many (a) that conditioning was an ultra-simple process in which one specific stimulus was substituted for another specific stimulus or one specific motor response substituted for another specific motor response; and (b) that this substituting occurred in an impersonal, mechanical man¬ ner that might be ever so remote from what the man or animal might be interested in at the time—might be, in other words, unmotivated. But certain lines of experimentation (by Liddell, especially, also Hull, Zener, et al.) have led to a rereading of Pavlov, a rereading that has supported broader interpretations; and nowadays the two assumptions just mentioned would be considered naive, because ( a ) the phenomenon of stimulus-substitution (or response-substitution) is one that orig¬ inally involves almost the whole organism and only gradually becomes localized, and (b) it occurs only when the organism is in an alert con¬ dition, and its behavior from first to last reveals needs and tensions that seek relief. From the practical angle two points may be mentioned. First, the conditioned response has thrown into a sound biological perspective the simpler instances in which the student-pupil acquires a new stimulus- response, as when he comes to say ‘cat’ when seeing that word printed 398 CONTRIBUTIONS OF THE PSYCHOLOGY OF LEARNING on a card. Second, it has become even more fruitful in furnishing a key to those peculiarities in one’s emotional make-up in which how one feels about a thing or a person or an idea seems utterly inappropriate and irrational. The early work of Watson in exhibiting how a senseless fear can be implanted in a child by conditioning, and again how it can be eliminated by conditioning, has had extensive influence in the fields of mental hygiene. For example, an unreasoning phobia for the sound of church bells or for the sight of a locomotive has become understandable enough if we see its genesis in some intense or repeated experience of the child when bell-ringing or when a locomotive were incidental accom¬ paniments of a situation in which a child experienced pronounced fear or dread of some other and entirely adequate incident or factor. VI. The Role of Reward and Punishment It has been implied in a preceding paragraph that contemporary research appears to demonstrate that a person or animal will form a new habit, even on the level of the conditioned response, only if in some way he is motivated. But ‘being motivated’ is an ambiguous term; for it may refer to the guidance of learning by incidental pleasures and pains, or to the general energizing of the learner, or to the part played by his interest in some objective goal. The pleasure-pain doctrine is of ancient vintage, but its modern psychological form makes more modest claims than did its ethical and economic predecessors. The scientific problem of recent times, by con¬ trast, is: What part do the incidental pleasantnesses and unpleasant¬ nesses (or rewards and punishments), play in the directing of learning? The Law of Effect, advanced in the 9Q’s by Thorndike, held that sat-, isfying (agreeable, pleasant) results tend to repetition of the act pro¬ ducing them, while annoying (disagreeable, unpleasant) results tend to omission of the act. The law was thus a supplement to the trial-and- error description of learning. This well-known law has, however, been repeatedly challenged. How can the result of an act, runs one query, work in backward direction upon the act itself? Or, runs another, how can the mere conscious awareness of pleasantness or of unpleasantness affect a physical process? Experimental inquiries have taken different forms. A recent one is that in which a person learns on a punch board or a stylus maze to make the correct punch or correct turn at each of some thirty choices arranged in irregular serial order (see, for instance, studies by Tolman, DASHIELL 399 Hall and Brctnall, Muenzinger, Crafts and Gilbert, Bunch, Hulin and Katz). In different experiments the right responses were variously ac¬ companied In no signal, by buzzer, or by shock; and the accompani¬ ments of wrong responses were equally varied. One general outcome is that the attaching of physical punishment to a wrong response is no more effective than attaching it to a right one; nor is it any more effec¬ tive in either case than attaching a sound signal. Accordingly, the idea has been advanced (a) that punishment has little if any direct guiding (eliminative) value in its own right, (6) that it is valuable only be¬ cause it is informative, or (c) because it speeds up or energizes the person generally. It must be noted also that if the punishment be too severe, it is likely to disrupt the whole process of learning. Meanwhile Thorndike has been repeatedly demonstrating with sim¬ ple experimental materials that rewards do lead somehow to the repeti¬ tion of the response rewarded, even though punishment seems to have no value. It .may be added that the other classic 'secondary laws’ of associa¬ tive learning—frequency, recency, primacy—have not been shown to be important. VII. The Importance of the Goal-Seeking Attitude • What is to be made of the relative insignificance of incidental re¬ wards and punishments as directors of learning, especially with human beings? Recall that either a reward or a punishment may, if attached to a right response, serve very well for the fixing of that response; clearly it gets its importance not from its own intrinsic character but from its serving as a helpful cue to the learner in obtaining his more ultimate objective. There is eloquent testimony, then, in the experi¬ ments referred to in the section just preceding, that one cardinal in¬ gredient in a learning process is that the learner is pursuing some objec¬ tive, is goal-seeking. This is the central and emphatic contention of the more active members of the original Gestalt school, as well as of others not to be so identified. Experiments on children, apes, and other species (Kohler, Yerkes, Bingham, Lewin, Alpert, Matheson, Mac- Dougall, Wheeler, and many others) have been built around the moti¬ vation of the subjects toward some objectives—a banana or a toy hung vertically out of reach or laid beyond the cage bars too far to be grasped, so that the subject can obtain it only by adopting some more or less original way of laying hold of it. 400 CONTRIBUTIONS OF THE PSYCHOLOGY OF LEARNING In some of the animal laboratories this motivational phase of learn¬ ing has been analyzed (by Tolman, Honzig, Elliott, Blodgett, V illiams, Beeper, and others) into the external ‘incentive* and the internal ‘drive’; and the presence of both of them has been proved to be distinctly more effective than the presence of only one, while the absence of both has been attended, for the most part, by no learning. These results of these experimental investigations reinforce an edu¬ cational doctrine that has been preached ever since Froebel; namely, the view that enlightened teaching looks upon the pupil-student neither (a) as a passive absorber, nor (6) as a reactive marionette, nor (c) as a machine equipped with propeller but no rudder, but instead (d) as a desiring, seeking, interested biological organism who will learn best when provided with objectives. VIII. Principles of Economy in Learning Ebbinghaus, in his pioneer experimental study of memorizing in the 80’s, brought to light several helpful principles on how to learn. They include such well-known points as that it is more effective to distribute practice over separate intervals than to try to learn completely at one sitting; that meaningful material is more easily memorized than rote; that attempts actively to reproduce what one is memorizing are highly helpful; that to strike a rhythm in the reading or the repeating is often of value; and the like. All the foregoing have been abundantly sub¬ stantiated in the volumes of findings by other experimenters (Muller, Meumann, Reed, Pyle, Winch, the McGeochs, and a great many others); but one of Ebbinghaus’ practical principles continues to be challenged, namely, that learning a thing as a whole is more effective, than learning it piecemeal. This has turned out to be true only under such special circumstances that it no longer deserves recognition as a general law (Pechstein, Reed, W. Brown). Many sorts of modifica¬ tions of the whole-part principles are to be recommended—more, in¬ deed, than we have room to enumerate. IX. Transfer of Training The pioneer clearing in this field by Thorndike and Woodworth has been greatly extended (Ebert and Meumann, Fracker, Judd, Ruger, Dallenbach, Bagley, Crafts, Bray, and numerous others). That transfer does take place is frequently apparent, but why it takes place in one situation and not in another is difficult to determine. Of the several DASH I ELL 401 theories that have been offered—identical elements, generalizations, ideals, and so on—none has won universal adherence. But a general agreement would be obtained on some such proposition as the following: a beneficial effect of training in a given school subject upon one’s work in learning other subjects resides not in any peculiar and inscrutable potency within the subject matter itself, nor in any strengthening of the student’s power of memory-in-general or attention-in-general or rea¬ soning-in-general; it resides solely in those habits of responding in this or that way to this or that aspect or detail that are common to both situations. Some of the experimental studies have shown that a habit that has been learned may not after all, have a favorable effect on the learn¬ ing of another: it may have a hindering effect—the phenomenon of 'interference’ (Miinsterberg, Bergstrom, W. Brown, Culler, Poffen- berger). This seeming contradiction in the two phenomena has not been well resolvetl, unless it be in a recent study (Bruce) wherein it has been demonstrated that if in changing from one learning task to another the stimulus is varied but the response is unchanged, a trans¬ ference effect is noted, whereas if the stimulus is unchanged but the response is varied, an interference is noted. X. The Neural Bases of Learning Psychologists, it seems, have always had a leaning toward the physiological in their explanations of experiential and behavioral phe¬ nomena, and there is today a persistent interest in the neural bases of learning. According to the older conception, differences in the amount of resistances at the different connecting points, or synapses, directed the flow of neural impulses set up by stimulation at a receptor, and the effect of learning was to change these differential resistances (by some wearing-down process) so that thereafter the impulses would be steered by these new lines of least resistance into other motor channels. Such theories imply that a human action is based in its neural 1 aspect upon a stream of impulses that follows one specific route or set of routes, and that a learned modification is, neurally speaking, a change from one specific route to another specific route. But this ultra¬ simple notion of the achitecture of the nervous system has been well- nigh demolished by numerous experimental researches. The facts of equivalence of stimuli, of transference, of variability in a maze habit, to name a few, are difficult or impossible to interpret in terms of such 402 CONTRIBUTIONS OF THE PSYCHOLOGY OF LEARNING pictures. So, too, with facts of everyday schoolroom observation; c.g., when a child learns a new word, and then—if properly taught!—is able to use the word in correct grammatical sequences in a hundred different situations, or, as when he has acquired ability to multiply, can carry on this operation in innumerable connections. Tremendous weight has been added to these objections by operative experiments in neuro-psychological laboratories. Franz and Lashley have led the way; and in latter years perhaps a dozen laboratories have provided opportunity for work of the sort (by Cameron, Loucks, Maier, C. W. Brown, Liggett, Jacobsen, K. U. Smith, Krechevsky, Ghiselli, and others). In outline, the procedure is to train an animal on a learn¬ ing problem, then operatively to deprive it of some region or regions of its brain or other structures of its central nervous system, and then to test the animal for its retention of the. habit that had been learned previous to the deprivation. A variant procedure is to test an animal for its ability to learn an entirely new habit after sitch operative loss of nerve structures. Certain highly specific functions, such as ability to learn to choose a signal that bears a detailed visual pattern, may be found dependent upon the integrity of very limited parts of the brain. Yet when anything like complex behavior is to be learned, such as the true pathway in a maze, no single part of the cerebral cortex seems absolutely essential, but such learning is a function of the cortex as a whole or of as much of it as the operator’s knife or cautery has left intact. Thus, a complex habit is not dependent upon the particular neural connections in any specific area of the brain but upon the total quantity of brain tissue that is intact. To generalize: Learning is not the building up of highly specific and fixed pathways, but the reaching of some new equilibrium in that dynamic field called the nervous system. One thing is sure: the conventional schemata with which the proc¬ esses occurring in the nervous system have been represented on paper and blackboard are as likely to impede as to facilitate a fair under¬ standing of how learning does actually occur. For all practical pur¬ poses, then, the less we think of the learner in physiological terms and the more we observe what and when and in what manner he learns, the more profitable will our educational efforts be. DASH I ELL 403 Some Classics in the Psychology of Learning Ebbinghaus, H. Memory, 1885. (Trans, by Ruger and Bussenius, 1913). Bryan, W. L., and Harter, N. “Studies in the physiology and psychology of the telegraphic language.” Psychol. Rev., 4: 1897, 27-53. (Reprinted in large part in Thorndike, Educational Psychology, 1913, v. II.) Thorndike, E. L., “Animal Intelligence.” Psychol. Monog., 2: 1898, No. 4. (Re¬ printed in book form, 1911, now out of print.) Book, W. F. Psychology oj Skill. 1908. (Reprinted, 1925.) Kohler, W. Mentality of Apes. 1925. Pavlov, I. P. Conditioned Reflexes. 1927. Lashley, K. S. Brain Mechanisms and Intelligence. 1929. Thorndike, E. L. Human Learning. 1931. Surveys of Experimental Literature Bills, A. G. General Experimental Psychology, Chs. IX-XIX. McGeoch, J. A., and others. Reviews appearing in Psychological Bulletin since 1927. Murchison, C. (ed.) Handbook of General Experimental Psychology, Chs. IX- XI (by Hull, Lashley, Hunter). FOR USE ONLY IN THE NORTH CAROLINA COLLECTION IJNCPS 52962