3. The instructive effect is in inverse proportion to the previous comprehension of the task, and consequently is not present after a certain amount of practice.

4. The incentive effect shows itself in the rapid adoption of a scheme or plan for learning any code and in greater care to avoid errors throughout the performance.

5. The disruptive effect is in inverse proportion to the thoroughness with which the habit is established, and consequently will offset the incentive effect during the learning of a code, but will be offset by the incentive effect after the code is learned.

REFERENCES

I. YOHANSON, A. M. The influence of incentive and punishment upon reaction time. Arch. Psychol., 1922, no. 54.

2. HANSEN, C. F. Serial action as a basic measure of motor ability. Psychol. Monog., 1922, 31 (no. 140), 320-382.

3. WUNDT, W.

Outlines of psychology, 1902 (2d ed.), 67 ff.

and Woodworth, R. S. Elements of physiological psychology, 1911.

4. Ladd, G. T.,

5. JASTROW, J. The time relations of mental phenomena, 1890.

6. HENMON, V. A. C. The time of perception as a measure of differences in sensation.

Arch. Psychol., 1906, no. 8.

RESPONSE TO A SUDDEN UNEXPECTED

STIMULUS

BY EDMUND JACOBSON

Few studies have been made, either in physiology or in psychology, of the excessive response which takes place after a sudden unexpected stimulus. Yet the involuntary start is familiar to everyone. It usually occurs after a sudden noise or other strong excitation while the individual is otherwise engaged. The movement, when extensive, generally includes a straightening of the trunk due to the erector spinæ; but jerks of the limbs or head are also common, and any external muscle-group may be involved. Often the disturbance is so slight that the only external evidence is a rapid contraction of the eyelids; although the individual may report organic sensations as of 'shock.' What is the explanation of this neuromuscular excitation that follows a strong unexpected stimulus? It is evident that some of the movements are sensory adjustments to the new stimulus; e.g., turning toward the sound with wide open eyes. But the jerks of the trunk or limbs are obviously not sensory adjustments, and they stand in need of further explanation. Lehmann (1) made a study of the effects on the pulse, respiration and volume of the arm produced by a loud unexpected organ note. Féré (2) registered contractions by means of a tambour placed on the muscles of the forearm or wrist, while at other times he used a hollow ball held in the hand. He concluded that the contractions were reflexes, which varied in amplitude with the intensity of the excitation, the excitability of the subject, and the unexpectedness of the stimulus. His studies were qualitative and obviously incomplete.

In 1908 I began an experimental study of the involuntary start in order to test an hypothesis which I have since then had to abandon. It is easy to observe that the start generally occurs when the individual is said, in lay terms, to be "lost

in thought." Accordingly it seemed plausible to suppose that, upon sudden interruption of the psychophysical thought-processes, the energy from the latter might be diverted to peripheral muscles. Preliminary tests lent a certain weight to this assumption. The subject was directed to attend to the details of a coin or to some simple but interesting picture. Since the movements to be recorded were, in part, directed upward and backward, a thread was attached to the back of the head, passing up and back at an angle of 45° to the vertical, through a small steel loop, and then vertically up to a broad tin writing-point, held in position upon the drum by a second thread. This second thread was continued by a short elastic cord, one end of which was clamped a little above the upper end of the drum. The kymograph was placed in a vertical position, its lower edge six inches or more above the head of the subject. In this way upward and backward movements, or the reverse, were very crudely recorded. Without seeking to generalize, it was found, at least at times, that during marked attention the start was strong (III. in Figs. 1, 2 and 3), while with mental passivity the start might be absent altogether (o in same figures). We sought to distinguish, by advance instruction to that effect, an intermediate state of attention (I. in the same figures).

Since experimentation calls for frequent repetition of a phenomenon under like conditions, it was clear that no great progress could be made if the involuntary start should soon disappear. This unfortunate occurrence seemed probable, if, as I had previously believed, the start depended primarily upon unexpectedness. I discovered, however, that, under the conditions of the experiment, the stimulus could be repeated as often as ten to fifteen times during a period and the start would still occur, with some adaptation, to be sure, but not so much as to interfere with comparisons. The stimulus was a sudden loud noise, made by releasing one end of a strip of resilient wood (of dimensions somewhat greater than those of a meter-stick) clamped at the other end to a wooden table. The volume of sound was kept fairly constant by raising the free end to a determined height.

FIGS. 1, 2, 3. The upper line indicates movements of the head and trunk. Breaks in the lower line indicate time of the shock. At III. the subject was fully attentive; at o he was so far as possible inattentive, while at I. there was an intermediate degree of attention. The figures show the most marked start where attention was greatest, and no start during inattention.

Now upon closely watching the subjects when they were directed to be attentive, it was noted that the brows contracted, the eyes were fixed, the limbs slightly stiffened and the trunk tensely held. On the other hand, when directed to be inattentive, tension seemed to give way to relaxation. These observations gradually led me away from the hypothesis mentioned above. A simple explanation of the start seemed to be at hand. The muscles contracted with a jerk when the strong stimulus came because they were previously contracting, although in less degree, during attention. If there was no such previous contraction, as during inattention, no start took place. Evidently this simple explanation was more acceptable than my earlier complex hypothesis.

The truth of this explanation could be tested by producing, in some other way, a heightened peripheral contraction or tonus where the sudden noise might nevertheless produce a further contraction varying in amplitude with the tonus. The subject was accordingly directed to stiffen the muscles of the arms, legs, head and trunks; that is, to sit still while partially contracting his muscles. In other tests he was directed to relax his muscles as completely as possible. An intermediate condition was also directed, the muscles being held only slightly tense. These grades are respectively indicated by the numbers 3, 0, 1, in the figures. When the sound was made under these conditions, the subject while looking away gave the same involuntary start as before. Indeed, the contraction seemed in these preliminary tests very much greater with extreme muscular tenseness than with mild; while with relaxation there was no start and no shock, and the sound seemed to lose its irritating character (Figs. 4, 5 and 6).

In this way it was suggested that, in addition to the factors mentioned by other writers, the involuntary start depends upon the previous neuro-muscular tonus; and that, if there is a correlation between the interruption of thoughtprocesses or of attention and the occurrence of a start, this is because the mental activity has a motor accompaniment.

The observations begun at Harvard were continued on