light has been thrown upon the relative efficiency of various forms of instruction in modifying the learner's progress.

Learning to juggle balls a typical example of experimentation. A good example of the investigation of learning to perform a complex act is E. J. Swift's work on the tossing and catching of balls. Five persons, technically known as subjects,' practiced for a good many days, and the results of the practice were definitely measured. Each subject juggled two balls; that is, he tried to keep both going with one hand, catching and throwing one while the other was in the air. There were three principal aspects or elements in the acquisition of skill in the performance: namely, (a) acquisition of general control in throwing; (b) sureness in catching; (c) use of some special method of throwing so as to avoid collisions.

If we consider Swift's experiment in terms of the general questions raised above on page 99, we may note the following points:

1. Better methods of throwing and catching important. There are certainly better if not best methods of throwing and catching. For example, to avoid collisions, it is better to give the balls a circular motion, so that they go up on one line and come down on another. Moreover, of the possible circular motions it is better to use one from right to left instead of one in which the balls are thrown up at arm's length and take such a circular course as to fall close to the body. Hence a subject who hit upon the right-to-left or left-to-right circular movement early in the experiment made rapid progress. Likewise, in catching, it is better to let the ball fall into the palm held horizontally in front of the body than to catch it "with the hand high in the air and the palm forward and almost perpendicular," as was done by one of the subjects, with the result that he made relatively little improvement during the month that he used this second method.

2. Verbal directions help in this case. - The possibility of profiting by verbal instructions, supplemented perhaps

by imitation, seems perfectly evident in a case like this. Hence Swift makes the following statement:

We see in this the value of suggesting good ways of doing things while the learning is still in its early stages. If the learner goes on [without instruction], he will finally develop a plan of his own, but only after a good deal of wandering, and even then it may not be the best. (10: 182)

3. Attention sometimes on movement, sometimes on result. In some cases it would seem that the learner's attention might with advantage be centered on the movement, and in other cases on the objective result. For example, in the case of the poor method of catching described above, the substitution of the better method would seem to involve centering the attention on the movement or position of the hand. This does not mean that the subject would look at his hand, but its position would be kept clearly in mind in modifying the method. On the other hand, in the adoption of the right-toleft circular method of throwing, attention would almost certainly be centered on the objective result, namely, the path followed by the ball.

4. Trial and accidental success used in finer adjustments. After full allowance is made for the possibilities of verbal instruction and imitation in assisting the learner to use the best methods, there remains a very large rôle to be played by the method of trial and accidental success. This becomes evident in the case of tossing and catching balls when attention is called to the large part played by the fingers and wrist in skillful juggling. The finer adjustments made by these parts do not come clearly to mind at any stage in the learning, yet they are most important in catching and throwing the ball skillfully.

5. Separate training on elements not valuable here.There can be little question concerning the desirability or undesirability of giving separate training on each of the

elements which constitute the complex act in this case. By omitting catching for the time being, it would be possible to give separate attention to throwing, and vice versa; but it is doubtful whether anyone would advocate this. To be sure, efficiency in the act as a whole depends to a large extent upon efficiency in each part of the act. Hence an individual might throw well but his total efficiency remain low because he caught poorly. But since his catching is always to follow upon his throwing, it would seem best to strive to improve his catching in connection with the throwing instead of separately. Here the sequence is so intimate, or so certain, that separate mastery of one of the elements would seem to be uneconomical. There may be cases, however, which we shall discuss later, in which separate attention to the elements of a complex movement may seem to be justified.

Other investigations illustrating trial and error. — In order to bring out more clearly the part played by the method of trial and accidental success in adult learning we shall study briefly two other examples from experimental psychology. The first is an experiment conducted by J. H. Bair on learning to move the ears, and the second is, W. F. Dearborn's description of the acquisition of skill in mirror writing.

Learning to move the ears a good example.- Learning to move the ears is for most adults an almost entirely novel feat. It is learned by securing separate control of certain movements of the ears which may occur when certain general movements of the head muscles are made, including raising the brows. That is, the movement which it is desired to control is first made as a part of a general diffuse movement of the neighboring muscles, and gradually by a process of selection and elimination comes to be made by itself. The same process of selection from a more diffuse movement occurs in learning to alternate the ears, in learning to raise the brows without moving (or innervating) the ears, and in learning to raise one brow independently of the other.

Having idea of movement did not eliminate trial and error. - Bair experimented to determine whether passively experiencing the movement would enable one to make it. He stimulated the appropriate ear muscles of the subject with electricity, thus making these contract and move the ear. He found that, while this enabled the subject to identify the ear movement when he succeeded in making it by trial and accidental success, it did not enable him to produce the movement without going through the process of trial and accidental success. Subjects who had experienced the electrically stimulated movement, however, took less time in learning to move the ears voluntarily than did a group of subjects who had not had the muscle stimulated electrically.

Concerning the direction of attention during the process. of selection and elimination (for example, in learning to move one ear without moving the other) Bair says that a positive effort to inhibit the movement of one ear is not effective, but if the subject concentrates attention on the one to be moved and forgets the other, he gradually gains the separate control. Hence he concludes:

The more closely the attention can be directed to a movement to be made, and the more nearly the part of the movement desired not to be made can for the time being be forgotten, the more likely is the desired movement to be accomplished. (4: 487)

This point is brought out more clearly in the following quotation from Freeman:

The elimination of useless movements, or the selection of appropriate ones, is one of the fundamental processes in motor learning. A practical question which may be raised concerning it is whether the result can best be reached by emphasizing the movements which are to be selected or those which are to be eliminated. In general it is much better to fix attention on the movements which are to be made, and allow the superfluous movements to drop out of themselves. It is a familiar fact that the bicycle rider avoids

the ditch best by keeping his attention on the path. The nervous energy is automatically withdrawn from the channels leading to the muscles not concerned, when the nervous channels to the appropriate muscles become more open. Directions should be positive, then, rather than negative. The pupil should be shown what to do rather than what not to do. The only exception to

this rule appears when the pupil has fallen into bad habits which need to be broken up. Then it may be necessary to call attention to the thing to be avoided. (16: 25)

Trial and error, not reasoning, predominant in mirror writing. Dearborn's experiments on the acquisition by adults of skill in mirror writing illustrate clearly the part played by trial and accidental success but do not throw much light on the relative efficiency of methods of instruction.