than they had originally acquired it. Pages 231 and 233 present the two tables from Ulrich's work upon rats. Bas- set (p. 237), whose general method of testing retention was the same as that of Ulrich, states that normal animals after a 60-day rest required 5.76 days to relearn the circular maze (p. 100), whereas the animal with lesser brain weight required 12.6 days. The final time for traversing the maze on the training series was, for both normals and inbreds, 6 seconds. After the 60-day rest the average of the first trial for the inbreds was 31.8 seconds, for the normals 22.5 sec- onds. The normals required 4.68 to relearn the latch box, whereas the animals with lesser brain weight required 6.74 days. The final time for solving the problem on the training series was 4 seconds for both normals and inbreds: after 60 days the average time of first trial for inbreds was 81.5 sec- onds, for the normals 59.6 seconds. Allen shows that the guinea pig retains a very simple labyrinth without great loss for at least 63 days. The retention of the habit was almost perfect. The time for the first success of this animal was 1 minute. Sackett says that the Canadian porcupine re- tains the Hampton Court maze with some loss for a 10-day period. Its whole time on the trial for retention was 6 minutes and 30 seconds; its second retentive test was 1 minute and 5 seconds. The final time of the training series was 1 minute and 10 seconds. The test of the retention of fastenings showed that at 13 days there was no loss; while at 30 days there had been a slight loss which could be regained at the end of 1 trial; at 50 days there was almost a complete loss. Davis states that the raccoon's retention of simple fastenings remains practically undiminished for periods of no practice of more than a year. In the case of the combination lock (p. 99) the retention was imperfect after a period of 286 days, but the relearn- ing was rapid, only 24 trials being necessary to gain the facility that originally required 107 trials. Cole states that the memory for a combination of 7 fastenings was not perfect at the end of 147 days in the raccoon. Three ani- mals were tested on this box after that period of no practice. Only the first succeeded in w^orking all of the fastenings RETENTION and releasing himself; he undid the 7 fastenings and came out of the box in 34, 28, 131, and 182 seconds respectively. The other animals worked almost all of them but failed to complete the work. A cat tested by Thorndike on a simple puzzle box showed a perfect response at the end of 20 days. Another cat which opened this box at the end of the train- ing series in 8 seconds, opened the box 14 days later in 3, 9, and 8 seconds respectively in three trials. Colvin and Burford state that the dog retained the habit of reacting to standard color for 3 weeks without loss. In the squirrel 1 month's suspension of time brought considerable loss in the accuracy of the same kind of habit. ^^ The following table is taken from Kinnaman's study of the monkey: Last 10 tests of First series of tests on Apparatus training retraining 50 days later 1st combination 6.4 seconds 13.17 seconds lock Male 2iid combination lock maze 1st combination 5.1 seconds 14.2 seconds lock Female 2nd combination lock maze Thorndike shows that habits of manipulation in the mon- key suffer little loss after periods ranging from 67 to 240 days.
Birds. — Hunter states that the retention of the pigeon for a simple labyrinth is practically perfect for 4 weeks where there has been no intervening training whatsoever. The pigeon would thus seem to be somewhat superior to the cowbirds and sparrows tested by Porter. Rouse mentions incidentally that associations in the pigeon are fairly per- manent and remain practically unaltered for at least 6 weeks. There are no records of tests which would suffi- ciently establish this point. Porter states that the English ^® Johnson gives a rough statement on the retention of habits of manipulation in dogs (blind and normal). The loss in accuracy (in simple problem boxes) at the end of 60 days was only about 10%.
248 STUDY OF HABIT FORMATION sparrow did not show any loss of skill in opening a simple food box after an interval of 8 days. He further states that the retention of a simple maze is surprisingly good for birds at the end of 30 days. The cowbird stands first. Her average for the first 10 trials in retraining equals that of the last 10 of the initial training. The vesper sparrow and the female English sparrow do not do quite so well. Retention of simple fastenings after 30 days by the cowbird and pigeon shows some loss; the male pigeon showed less loss than the cowbird. The female pigeon and the cowbird lapsed into a poor method which was used early in the training series. At the end of 120 days the cowbird and the pigeon were again tested upon the simple fastenings. While the loss was considerable the habit was not completely lost and was readily reacquired. Thorndike retested chicks on a simple labyrinth at the end of 20 days and found that retention was practically per- fect. Breed shows that of 9 chicks trained to the black- blue habit and tested 30 days later, 5 obtained perfect records. Breed presents evidence to show, too, that chicks which acquire the black-blue habit most readily also show the most perfect persistence tests. Thus of 3 animals that required 100 trials to perfect the black-blue modification 1 made perfect persistence tests; of the 2 that required 70 trials 1 made a perfect record; of 3 that required 60 trials, 2 made perfect records; and the chick that com- pleted its work in the shortest time also made a perfect record.
Amphibia and reptiles. — ^Yerkes shows that the green frog retains a simple labyrinth habit for at least 30 days. Casteel states that the turtle {Chrysemys marginata) re- tains a simple sensory habit — vertical vs. ho:"izontal line — for 2 weeks without loss.
BIBLIOGRAPHY Allen, Jessie, " The Associative Processes of the Guinea Pig," Jour.
Comp. Neu. and Psych., 1904, XIV, 293. Basset, G. C, " Habit Formation in a Strain of White Rats with Less than Normal Brain Weight," Behavior Monograph, Ser.
BIBLIOGRAPHY 249 Book, W. F.," The Psychology of Skill. Missoula, Univ. Montana, Breed, F. S., " Reactions of Chicks to Optical Stimuli," Jour. Animal Caer, H. a., and Watson, J. B., " Orientation in the White Rat," Jour. Comp. Neu. and Psych., 1908, XVIII, 27. Casteel, D. B., " The Discriminative Ability of the Painted Turtle," Jour. Animal Beh., 1911, I, 1, Cole, L. W., " Concerning the Intelligence of Raccoons," Jour. Comp.
" The Relation of Strength of Stimulus to Rate of Learning in the Chick," Jour. Animal Beh., 1911, I, 111. COLVIN, S. S., and Burford, C. C, " The Color Perception of Three Dogs, a Cat, and a Squirrel," Psychological Monograph, Ser.
Davis, H. B., "The Raccoon: A Study in Animal Intelligence," Am.
Jour. Psych., 1907, XVIII, 447. Franz, S. I., " On the Functions of the Cerebrum. I. The Frontal Lobes in Relation to the Production and Retention of Simple Sensory-Motor Habits," Am. Jour. Phys., 1902, VIII, 1. Franz, S. I., and Lafora, O. R., " On the Functions of the Cerebrum: The Occipital Lobes," Psychological Monograph, Ser. No. 56. Hicks, V. C, " The Relative Values of the Different Curves in Learning," Jour. Animal Beh., 1911, I, 138. Hicks, V. C, and Carr, H. A., " Human Reactions in a Maze," Jour.
HoGE, Mildred A., and Stocking, Ruth J., " A Note on the Relative Value of Punishment and Reward as Motives," Jour. Animal HuBBERT, H. B., " Time versus Distance in Learning," Jour. Animal Hunter, W. S., " The Delayed Reaction in Animals and Children," Behavior Monograph, Ser. No. 6. " Some Labyrinth Habits of the Domestic Pigeon," Johnson, H. M., " Audition in Dogs," Behavior Monograph, Ser.
KiNNAMAN, A. J., " Mental Life of Two Macacus rhesus Monkeys in Captivity," Am. Jour. Psych., 1902, XIII. Porter, J. P., " Further Study of the English Sparrow and Other Birds," Am. Jour. Psych., 1906, XVII, 248. Richardson, Florence, " A Study of Sensory Control in the Rat," Psychological Monograph, Ser. No. 48. Rouse, J. E., " The Mental Life of the Domestic Pigeon," Harv.
Sackett, L. W., " The Canada Porcupine: A Study of the Learning Process," Behavior Monograph, Ser. No. 7. Slonaker, J. R., " The Normal Activity of the White Rat at Dif- ferent Ages," Jour. Comp. Neu. and Psych., 1907, XVII, 342.
2^ In general the literature on human learning is summarized. See also THORNDIKE, E. L.. Educational Psychology, New York, Col. Univ., 1913, 2, " The Psychology of Learning."
250 STUDY OF HABIT FORMATION Smali., W. S., " Notes on the Psychic Development of the Young White Rat," Am. Jour. Psych., 1899, XI, 80. Thorndike, E. L., Animal Intelligence. New York, Macmillan, 1911. Triplett, N. B.. " The Educability of the Perch," Am. Jour. Psych., Watson, J. B., Animal Education. Chicago, Univ. Chicago Press, " Kinesthetic and Organic Sensations: Their Pole in the Reactions of the White Rat to the Maze," Psychological Monograph, Ser. No. 33.
Watson, J. B. and M. I., " A Study of the Responses of Rodents to Monochromatic Light," Jour. Animal Beh., 1913, III, 1. Yerkes, R. M., The Dancing Mouse. New York, Macmillan, 1907.
" Formation of Habits in the Turtle," Pop. Set. Mo., " Instincts, Habits, and Reactions of the Frog," Harv.
" Modifiability of Behavior in its Relations to the Age and Sex of the Dancing Mouse," Jour. Comp. Neu. and Psych.
Yerkes, R. M., and Dodson, John D., " The Relation of Strength of Stimulus to Rapidity of Habit Formation," Jour. Comp. Neu.
Yoakum, C. S., " Some Experiments upon the Behavior of Squirrels," Jour. Comp. Neu. and Psych., 1909, XIX, 541.
CHAPTER VII FIXATION OF ARCS IN HABIT ^ Introduction. — Historical outline. — Misconceptions and illogical pre- suppositions.— Enumeration of some of the problems. — Factors involved in fixation. — Application of the principle of frequency in motor habits. — Justification for the use of the principle, — Application of the principle of recency in motor habits.- — Sensory habits. — Substitution. — Repetition of movement in absence of original stimulus. — Conclusions.
Introduction. — We found in the preceding chapter that the most insistent question in the study of habit was one that dealt with the mechanisms by means of which useless movements are eliminated. It follows from our previous discussions that we must be able to account for the elimina- tion of useless movements upon purely objective grounds.
Historical outline. — The psychological literature shows that the problem under discussion has always been the hete noire of the parallelist. He avoids it altogether or else goes over either' into surreptitiously worded or more out- spoken form of interaction. All through the literature we find such statements as the '' conscious " guidance or " con- scious " control of movements, the effect of having a plan or purpose in mind in learning, etc. In the experimental literature we find the frank statement that all improvement and short cuts in learning come unconsciously: when, how- ever, they have thus accidentally appeared, consciousness steps in and selects them purposely. Consciousness is looked upon as exerting a positive influence upon the bodily mech- ^ In the preparation of this chapter the author has had access to a manuscript of Professor Harvey A. Carr, entitled " Principles of Selection in Animal Learning." This paper appears in the May number of the Psychological Review (1914). While the views ex- pressed in the present chapter are in many particulars different from those found in the above paper, -vve wish to call attention to Professor Carr's prior claims in the matter of authorship.
252 FIXATION OF ARCS IN HABIT anism. If a physiologist comes out and frankly admits that he is willing to accept interactionism, as Colvin does,^ we should not wish to quarrel with him for applying that principle in solving the problem under consideration. Whether he has a right to apply it anywhere depends upon many questions which we have no right to enter upon here. But most of them are not willing to accept interaction in this frank way. The most outspokenly worded form ap- pears from psychologists who speak of the pleasure result- ing from a successful movement as stamping an act in, and the displeasure resulting from an unsuccessful act as stamping it out. In order to bring these tendencies clearly before us, we give below the views of a few psycholo- gists.
Statements from various authors: ^ JuDD states that " the mechanism involved in a habit which has resulted from selecting from among instincts is relatively easy to explain. We need only to assume that the stimulation which is given at the first experience has two theoretically possible lines of discharge, either one of which would be through a well-defined instinctive tract. The conditions of the first encounter carry the stimulation in question into one of the two instinctive channels and thereafter this selected channel becomes the natural and easy path of discharge for the stimulus whenever it occurs." (p. 218)..." Turning now from the habits which are developed through the •selection and modification of instincts, we come to the habits which cannot properly be traced to any single instinct or group of instincts. Let us suppose that a stimulus or a combination of stimulations is introduced into the nervous system of the child hut finds no specific channel of discharge open to it through inherited organization." Judd insists that the stimulus " will be widely distributed through the whole nervous sj^stem because it has no specific channel of dis- charge and because, as free energy, it must be transmitted through the nervous system until it finds a discharge into the active organs."
He cites the diffuse distribution in mature life when one is startled by an unexpected noise. As a case of habit formation with " little or no instinctive background " he describes the method the child employs in learning to write. Once again he states (p. 222): "...The nervous impulses which excite the muscles do not follow definite channels. In the case now under consideration the channels are not yet developed..." The mechanism by means of which the diffuse activity is molded into more definite form is described as follows: "...There is a tendency for all parts of the nervous 2 The Learning Process, p. xxii. • Italics ours except where noted.
HISTORICAL SUMMARY 253 system which are active at the same time to become related to one another in their activity; and...there is always a process of selection going forward by v/hich those combinations which attain the end toward which the individual is working are preserved and the others eliminated.* The first of these two principles is not difficult to understand when the close structural and functional interrelations of the various parts of the nervous system are recalled. The second principle is much more difficult to explain. This principle may be restated in the following terms: If an animal goes through a certain mode of activity and derives advantage ^ from the success which attends this performance, it is likely to repeat the activity; whereas, if the activity does not attain success, the animal wnll not repeat it."
Angell's statements of the general problem, while seemingly clear, on analysis are found to be very hard to state in mechanical terms. He takes the formation of the eye-hand coordination — the child learning to reach and manipulate a ball. He too starts with the idea of "diffusion" of nervous impulses. "...There are few or no preformed reflex pathicays over ivhich such neural excitement may he effectively discharged.^' In the exercise of these diffuse acts the child's hand comes in contact with the ball. What happens? "To begin with the mere shock of surprise and (generally) pleasure makes the connection of tactual motor sensations from his hand with the visual sensations from his eye extremely vivid."..." We may be sure that the child's organism is extremely likely to retain the memory of the highly vivid connection between the visual sensations of the ball and these tactual-motor feelings which accompanied the successful grasping of it."..." If the nervous system were an inanimate mass, we might liken that which occurs to the process by which a path is made across a meadow. The first wayfarer may have selected his special route for any cause whatsoever, and his course may have been devious, like those of the cows which are said to have laid out the streets of Boston. But he has left a mark in the down- trodden grass which the next person to cross the field is likely to follow. Presently the grass is wholly worn away, and thereafter every one follow^s the beaten path." He does not accept the metaphor whole-heartedly, however. " But the moment we recall the fact that the nervous system is part of a living organism, in which processes of nutrition and repair are constantly going forward, and when we remember, furthermore, that the organism itself can in large measure decide whether a stimulus shall be experienced again and whether a movement shall be repeated or not, we see that the metaphor of the pathAvay in the meadow must be abandoned in favor of some idea in which the vital processes of the organism are recognized and the living tissues treated as something other than so much static, plastic clay, which the accidents of the external world can mold to their own exclusive purposes."..." But the point which we must em- phasize is that the organism itself largely decides which pathways * I.e., he frankly assumes the very thing for which we are trying to provide a mechanism.
^ To be found on p. 22.3. I.e., those acts are selected from which the animal derives advantage. Yet on p. 22.5 he has a paragraph heading " Pleasure a result of organization, not a cause."
254 FIXATION OF ARCS IN HABIT shall in the first instance become thus established." He goes on to say that such multitudinous connections are possible, " that if the child were left to the accidents of the first external stimulations and to the vagaries of merely passive nervous centers, the chances would favor the acquirement of insane and harmful habits of reaction." ^ Angel 1 gives no more definite account of the mechanism of habit for- mation than the above. In the paragraph on the inhibition of use- less movements he states: " The more firmly the connection becomes established, and the more deeply the pathway is cut between the visual sensory centers and the hand-arm motor centers, the more do the irrelevant movements of face, legs, and body tend to drop away! " This inhibition is largely due to the fact that the newly formed channel is increasingly able to carry off all the neural excitation and in consequence less remains to overflow into other channels.
PiLLSBUEY makes voluminous reference to the nervous system but makes no eff'ort to apply the theory of synaptical connection in detail. He states that Sherrington believes "...that the synapse opposes the passage of an excitation in much the same way that a membrane opposes the transfer of fluids. At first this resistance is very great, but it is lessened wdth each act."..." The change that comes with action is a lessened resistance at the synapses, w^hether it be due to the coming closer of the processes or to a chemical change that makes more permeable the relatively impermeable membrane that at first separates the neurones." He then likens the synapse to a valve: " At first the valves are stiff and open only to strong cur- rents; with use they work more easily."..." At the beginning certain of the neurones constitute a path for an impulse from sense- organs to muscles. These original paths are few and make possible only the most essential activities for the continuance of the life of the individual. Additional paths of connection are formed by each activit}^ physical and mental. Whenever any two neurones chance to act together a connection is formed between them, the original gap is bridged and they come to form part of a new pathway from sense-organ to muscle. Sometimes the most important of the neurones that are connected lie within the cortex, and the learning that results is primarily learning of ideas rather than of movements, but the principle is the same as before. Learning, whether of new movements or of new ideas, is a process of making easier the passage of an impulse from neurone to neurone and is fundamentally the same everywhere," Again on p. 55: " If we turn from a discussion of the omnipresence of habit to the question of what goes on in the nervous system when a habit is acquired, we get back to our problem of the synapse and its relations. Each time an act is performed, it matters not how, there is some change in the synapses between neurones. The effect of the act persists and becomes stronger with each repetition. After several repetitions the connection between the cells becomes so close that wiienever the particular sensory cell is excited the impulse spreads to the motor cells that were active ^ Surely this vitalistic terminology is not insisted upon by Professor Angell. He would admit doubtless that were it not for the parents or other attendants the child would speedily destroy itself and that it is these same attendants that force the habits wliich enable it to exist when removed from the society of adults.
HISTOEICAL SUMMARY 255 before, and the neurones grow more and more to constitute what is l^ractically a single structure. The formation of habits is thus a process of decreasing the resistance of the synapses in the different possible paths of transmission."
Book's statements are contradictory and unclear. On p. 95 he says: "...All adaptations and short cuts in method were uncon- sciously made, i.e., fallen into by the learners quite unintentionally on the good days while practising under strain. The learners sud- denly learned that they were doing parts of the work in a new and better way, then purjjosely adopted it in the future." On p. 171 he makes the following statement after first reiterating the unintention- ality and unconscious character of short cuts: " When the advantage of the new method had been noticed it was generally, thereafter, made use of purposely, though even there consciousness seemed to be more of a hindrance than a direct help." In general he speaks of " extreme effort carefully applied to the work on a good day " as being the precondition to a forward step. " Fluctuations in attention and effort " are the factors invoked to explain large variations, lack of improvement, etc. Xo effort is made by Book to analyze the condi- tion which makes it possible for the short cuts to be fixed. He speaks in several places somewhat as follows: " Continued attention or application is assured by the pleasurable feelings and favorable attitude which always accompany success. The organism as a whole naturally tends to continue the activity as long as improvement is made. When progress stops the activity (mental and physical) be- comes disagreeable and attention goes elsewhere."
Similar confusion occurs in Swift's statements (p. 184): "It is interesting that all the subjects improved by hitting upon better ways of working without any further conscious selection. There seems to be a competition of methods. Just how this selection occurs without conscious interference is not easy to say. Consciousness dis- covers modes of action already in use, and selects some of them for survival because of their success." While it is obvious that if any one has such a vitalistic principle as this within him he needs no physiological mechanism to account for selection, yet nevertheless on p. 185 we find the folloAving: "The question then arises, how are the necessary coordinations brought about? It does not seem difficult to bring the matter into line with phenomena already pretty well known. Let us suppose a successful toss and catch are made. This is followed by a double effect: it leaves, as every action does, a trace in the nervous system which facilitates later repetition of the same action, and the successful adaptation also gives rise to a feeling of pleasure. The effect of pleasurable sensation is a heightening of muscular tonicity or a general tendency to muscular discharge, which in the case of an action just performed — one whose neural effects are still lingering — is equivalent to a partial reinnervation of the same coordinated group of muscles, which again deepens the existing trace. The next actual effort finds the nervous mechanism a little readier to react in this favorable way. In case of an effort that does not lead to success, the slight displeasure at failure exerts its natural depressant effect upon the whole neuro-muscular system, and thus does not deepen the neural trace left by the original movement, and even, perhaps, breaks up the incipient coordinations that gave it 256 FIXATION OF ARCS IN HABIT its particular form. In any case, whatever its mode of action, it has not the reinvigorating effect upon the original neural trace exercised by the pleasurable sensation. In the long run, therefore, the success- ful movements, and the coordinations upon which they depend, tend to persist, while those that are unsuccessful tend to fall away."
Thoendike in his earlier monograph speaks in a highly figurative way of pleasure stamping in an act and displeasure stamping it out. In his book {Animal Intelligence, p. 244) the two laws of learning are stated as follows: "The law of effect is that: Of several responses made to the same situation, those tohich are accompanied or closely folloiced by satis- faction to the animal ivill, other things being equal, be more firmly connected with the situation, so that, ivhen it recurs, they ivill be more likely to recur; those ichich are accompanied or closely folloiced by discomfort to the animal ivill, other things being equal, have their connections with that situation weakened, so that, tvhen it recurs, they will be less likely to occur. The greater the satisfaction or dis- comfort, the greater the strengthening or weakening of the bondJ " The law of exercise is that: Any response to a situation will, other things being equal, be more strongly connected with the situa- tion in proportion to the number of times it has been connected with that situation and to the average vigor and duration of the connec- tions.'' ^ Misconceptions and illogical presuppositions. — It would seem that there must be something radically wrong with the statement of the empirical facts, since it seems so hard to arrive at even a satisfactory theoretical solution of our problems. Examination shows that three misconceptions have crept into the situation: (1) It has been assumed without further ado that the successful act is pleasant, and that the unsuccessful act is unpleasant. Without attempt- ing to venture into the realm of introspection, we shall advance some slight evidence to show that unsuccessful acts can be eliminated in cases where every movement in the situation as a whole is one which is ordinarily avoided by the animal. In Glaser's experiment, where rats were dropped into hot or cold water and allowed to find an en- trance, it was shown that the animal formed the habit of T It is obvious that the first law is no law at all, if by law we mean a generalization from a large set of empirical data. It is merely a state- ment of conviction. Furthermore, he calls it a law of effect, whereas it is obvious that it is a law of cause if it is supposed to be a statement of the working of the mechanism of fixation, as the second law un- doubtedly is. In regard to the second law it is clear that when this is stripped of its unnecessary implications of bonds or connections it is the same principle we have adopted on p. 262.
s All italics are Thorndike's.
CERTAIN MISCONCEPTIONS 257 turning correctly in the hot or cold water and finally reach- ing the exit. Suppose, e.g., we complicate the situation even further in the following way: instead of having an open vessel filled with water in which the animal can turn at will, we arrange partitions in the water, which the ani- mals may not clamber up, in the form of culs de sac. Furthermore, the exit from the maze as a whole should lead into a chamber which is hot but not hot enough to injure the tissue. The rat would ordinarily avoid the water, would ordinarily avoid the hot box, etc. It is ex- tremely probable that the habit would still be formed. The animal is so constructed that it must move when dropped into the water. If it moved at all, it would neces- sarily move so that certain acts would occur in accordance with the principles we have assumed to be adequate to account for habit formation (p. 262). It is our aim to com- bat the idea that pleasure or pain has anything to do with habit formation or that harmfulness or harmlessness has anything more to do with the situation. It is perfectly natural in unreflective minds that the idea of good or bad or harmful or harmless should be called in to explain the habits we force upon animals and children. It is a bit