will show how much it has in common with the uninstructed learning of man. If an animal is confined in an enclosure with winding paths, all except one of which end in blind alleys, he will run about until by accident he finds the path that leads to the exit where food rewards success. Now, if again placed in the enclosure, the animal will not at once follow the direct path to the exit and food, but he will repeat many of his former errors. By degrees, however, he will finally learn to take the direct course to the food. We say that the animal has learned to get out of the enclosure. What actually has happened is that a connection-an association-has been formed between a situation consisting of confinement in an enclosure of definite appearance and following a certain path to food. The situation favored or permitted various responses, the possibility of which was provided for in the instinctive equipment of the animal. The responses were made at first, without selection, and one of them happened to meet the requirements for reaching the exit and securing the food. Gradually the useless actions were eliminated, and the correct associations attained sufficient strength to produce the series of effective responses without error or delay.

It

This same association can be established between actions that have no necessary relation to one another. is well known, for example, that dogs can readily be taught to sit, or stand on their hind legs, to obtain food. In the same way they may be trained to go to the corner of the enclosure farthest from the exit and turn around an exact number of times to secure food. The elements of these associations may also be increased so as to form a long series. Several connected enclosures may be made, and the animal may learn to pass from one to the other in quite different ways, pulling a string to open a trap-door in the first case, pushing up a lever in the second, climbing a ladder and going through a hole to reach the third enclo

sure, and finally returning to the point of departure by a different route to find food awaiting him. In all of these cases the response is made directly to a given situation, without the intervention of ideas. A certain situation is associated with a given response.1

The absence of ideas means absence of thinking. No inferences are drawn. Change the setting to any great extent and the responses are likely to be disturbed, though the essential element in the situation-the one to which a definite response must be made to secure the desired result may be quite as conspicuous as before. In other words, the result is not connected with a certain action upon a given object. There is no recognition of cause and effect. It is interesting to observe, however, that some of the more highly developed among the lower animals seem to select and pay especial attention to those parts that are connected with the essential element, and changes in its location do not always disturb them. But this is probably due to persistent and strengthened association rather than to any discernment of the relation between cause and effect.

Now, the prevalent opinion that man's method of learning is commonly different from that of the lower animals is unjustified. Animals learn by association. When one event follows another they assume that the succession will always occur. In some instances the response to such an occurrence becomes settled as an instinct. Chickens going to their roosts in the middle of the day during an eclipse of the sun is an illustration. For all the practical purposes of life such responses assume that the first event is the cause of the second. Now, it is not difficult to show that man employs this same associative learning more

1 Those who wish to make a more detailed study of learning among the lower animals may consult Animal Behavior, by C. Lloyd Morgan; Behavior, by John B. Watson, and numerous reports of investigations in the journals for general and comparative psychology, especially the Journal of Animal Behavior.

commonly than is supposed. Recent investigation1 of the way in which he solves a maze problem has shown that "the rational processes reported were unsystematic and seemingly futile. Adequate interpretations were suggested to the learner as the result of prolonged exploration, rather than reasoned out. Cues which logically should be utilized for correct inferences were disregarded, and ideas were acted upon in an uncritical manner until they were proven by trial to be incorrect." If the word "rational" were omitted from this account the description would be such as might be written of animal learning.

In those cases in which this process of associative learning seems to carry the learner forward in his thinking, he assumes that he is reasoning. The difference between this pseudo-reasoning and the real thing is that reasoning, whatever else it means, requires that the essential element in a situation be selected from the mass of unessential factors, and that its necessary connection with the result be discerned. "Associative reasoning" assumes that an action or other event which precedes the second has some necessary connection with it, as that of cause to effect.

One need not go back to the belief in a necessary connection between phases of the moon and sowing crops for an illustrative example of this sort of reasoning among human beings. The summer of 1915 was cold and rainy throughout the central and eastern parts of the United States and, as will be recalled, the belief was quite prevalent that the unusual "weather" was caused by the firing of cannon on the battle-front of Europe. When this idea was first promulgated by the "science editor" of certain newspapers the unintelligent accepted it and those better informed smiled. Later, however, when the low temperature and rain continued, even intelligent people began to say: "After all, there may be something in it." This belief is, of course, the persistence in another form of the old superstition 1 Fleming A. C. Perrin, Psychological Monographs, no. 70.

that rain can be caused by producing an artificial atmospheric disturbance. Naturally, "reason" did not give the same conclusion the following summer (1916) when an unusually high temperature was recorded in those localities which were cold the year before, and when, though the firing was continuous, the rainfall was considerably less than usual. Of course the error is forgotten, and right here we find another characteristic of human learning common to that of the lower animals. The latter quickly forget an event unless it has been worked into their nervous system by repeated experiences or has become instinctive through natural selection. Were it not so, new devices would have to be frequently invented to trap them. In man, also, the same methods of "trapping" continue effective.

The "Spanish swindle," which is a good illustration of repetition of the same deception, is so old as to have its origin almost lost in the past. Yet it still serves its purpose. In its simplest form it consists of information that a large sum of money awaits one in Spain. But a "small" amount is needed to secure the legacy. Again, the tricks of "wire-tapping" to obtain advance information about the winning horse, and other swindling devices, continue to succeed, even with those who know of them. Were it not for forgetfulness, swindlers would need to be geniuses at invention. But, as with the lower animals, the old tricks are usually as good as new ones. In man speedy forgetfulness is both a misfortune and a boon. The sorrows of yesterday are submerged in the joys of to-day; and an important element in the psychology of the stockmarket is that the pessimism of last week is forgotten in the present elation, though to-day's rise may be made to order. This human tendency to forget is of supreme importance in the psychology of learning, so far at least as what one learns through experience or otherwise is to be used in reasoning.

In speaking of animal learning, it was observed that the particular action which produced the desired result was one of many. The others either opposed success or were indifferent to it. The successful movement, or series of movements, was not planned. It was not even consciously selected. Given a continuous succession of more or less random movements, all directed toward a definite end, and some of them will be more effective than the others. Indeed, if the effort is persistent, the desired result will finally be attained. A fly is biting the cheek of a very young baby. Many random movements are made in the struggle to relieve the discomfort. At last one of these movements succeeds and the fly is brushed away. It is the same with the lower animals, and so it is with man in learning any act of skill.

Now, a matter of the utmost importance in the psychology of learning has been noted in connection with the random movements to which reference has just been made. The successful movement is produced without any further conscious guidance or selection than is supplied by the general effort to attain a desired result. Learners, in acts of skill, suddenly find themselves employing definite methods to meet certain difficulties. Quite commonly they do not know that they are using these methods until they notice that the difficulties are disappearing. There seems to be a competition of methods. Just how a selection of the efficient method occurs without conscious interference is not easy to say. "What happens in such cases," according to Thorndike, "is that the response, by being connected with many situations alike in the presence of the element in question and different in other respects, is bound firmly to that element and loosely to each of its

1

1 Those interested in the theoretical discussion of this question may consult, among other writers, the following: Edgar James Swift, American Journal of Psychology, vol. 14, p. 201; Edward L. Thorndike, Animal Intelligence: Experimental Studies, p. 264, and John B. Watson, Behavior, a Study in Comparative Psychology, chap. VII.