Clever Hans Part 6

I was not satisfied with convincing myself subjectively of the facts indicated, but sought to fix them objectively, by means of a graphic method. For this purpose I used the device mentioned by Prof. R. Sommer for the a.n.a.lysis of expressive movements.[18] The purpose for which Prof. Sommer's apparatus had been constructed, was to record the involuntary tremor and movement of the hand. These movements, of course, take place in the three dimensions of s.p.a.ce. By means of three levers it is possible to record the movements upon the flat surface of a smoked paper fastened to the revolving drum of the kymograph, the movements in each direction being recorded by a separate lever, in such a way that the three curves thus made represent the a.n.a.lysis of a single movement into its three dimensional components. By making slight changes, which tended to complicate the experiment somewhat, I adapted the apparatus to the measurement of movements of the head. The method of experimentation was the following. The subject whose movements were to be registered, was placed in the device in such a way that his trunk and head were bent slightly forward, the latter a little more than the former. This, it will be remembered, was the usual position of the questioner when working with the horse. Three levers were attached to his head in such a way that every movement backward or forward would act upon the first lever, every movement to the right or left would move the second, and every movement of the head upward or downward would be recorded by the third. With regard to the sensitivity of the machine, micrometric determination showed that when the subject was properly installed, movements through so small a distance as 1/10 millimeter could be accurately ascertained. The subject was carefully instructed to remain as quiet as possible, but without constraint. Voluntary movements were thus obviated. But the question arose: were not the involuntary movements thus suffering a loss?--And it was upon them that we were experimenting. The question cannot be put aside summarily, but experience taught us that the movements in question, nevertheless, did appear quite effectually, if one could have the right kind of subjects at one's command. We need hardly mention that besides the two persons immediately concerned--I, myself, attended to the apparatus--there was no one else present, and that the subject was not allowed to see the curves produced on the kymograph. Besides the registration of the head-movements, I also undertook to register the respiratory-movements of the subject. This was done by means of the so-called pneumograph, attached to which was a lever recording the thoracic expansion and contraction. This was for the purpose of ascertaining the relationship, which might eventually be found to exist, between the release of psychic tension, on the one hand, and respiration, on the other.

The subject was now told to think of some number, which, of course, was unknown to me. At a given moment I was to tap upon one of a series of keys arranged like those of a piano, with the middle finger of my right hand--corresponding to the right forefoot of the horse. The questioner observed my key, I, his head,--just what had happened in the experiments with Hans,--and as soon as I perceived the involuntary closing signal I reacted upon it by releasing, suddenly, another key upon the same keyboard, which I had in the meantime been pressing down with my second finger, thus marking what with Hans had been called the backstep. Each key was connected with a separate electro-magnet, and these in turn with markers, in such a manner that pressure upon the keys closed two electric circuits and, releasing the keys, opened them, and both the closing and the opening were recorded upon the smoked paper by means of the markers. And, finally, in order to ascertain the time relations of all these processes, a time-marker indicated the time in fifth-seconds upon the revolving kymograph record. The time-curve was recorded just below the other curves.

Of the curves[P] thus obtained under the most equable conditions possible, we publish seven which show the great general uniformity of the tests made upon the horse with those made in the laboratory. The role of questioner was undertaken at different times by Mr. Schillings and the students of philosophy, Messrs. von Allesch, Chaym and K. Zoege von Manteuffel. To all of them I am greatly indebted for their unselfish services in these laborious tests. The experiments with von Allesch and Chaym, who were among the most suitable of my subjects, were conducted absolutely without knowledge on their part of the nature of the phenomena which I was observing. Neither of them knew anything about the expressive movements in which they were unconsciously indulging, and furthermore, since they kept their heads bowed during the entire course of these experiments, they did not perceive what it was that I was observing. It is interesting to note that Chaym on the occasion of his only visit to the horse, immediately received a number of correct responses. Without a doubt von Allesch would have met with equal success. The other two subjects (von M. and Sch.) went through this series of tests, possessing some knowledge of the nature of the movements involved. Conditions were such that they (and especially Mr.

Schillings) could not be prevented from obtaining some knowledge of the essentials, at least. However, it would be wrong to suppose that for this reason the results were more favorable, owing, mayhap, to voluntary efforts on the part of the subject. The contrary was true. The two subjects who had no knowledge of the character of the reactions upon which my responses depended, retained their normal habits, unchanged, throughout the series,--whereas the last-named two, afraid lest their knowledge vitiate the result, lost more and more of their power of concentration and within a short time were in a condition of tense inhibition, which is all the more conceivable, since they had had no psychological training whatever.[Q]

[Footnote P: For registering the curves a Hering kymograph was used, with a loop 2-1/2 metres long. The kymograph rested on felt. With the aid of the Marey model a pneumographic record was taken now of the thoracic, now of the abdominal, breathing, never both simultaneously, since this was extrinsic to my purpose, and it would have made the whole experiment too complex. The time was recorded by means of the Jacquet chronograph. For purposes of making more exact measurements the acoustic current interrupter of Bernstein was used, attuned to 100 vibrations per second. But this necessitated such rapid revolution of the drum of the kymograph that the curves were not compact enough for purposes of demonstration. The levers were all fitted with micrometer adjustments. They wrote tangentially and, except the one registering the breathing curve, all points lay in one vertical line. The error of deflection and that due to the rondure of the writing-surface were both very slight on account of the comparative length of the levers and the small extent of the excursions, and for that reason synchronous points lie practically in one perpendicular. Only the breathing curve has been moved somewhat to the left, 7.5 millimeters in figures 6 and 7, 2 millimeters in figure 8, 4.5 millimeters in figure 9. (When the breathing was very profound, as occasionally happened, the error of deflection would, of course, have to be taken into account.) The curves here used as ill.u.s.trations have been reproduced in the exact size of the originals by the zinco-graphic method, though somewhat compressed vertically in order to economize s.p.a.ce.]

[Footnote Q: My own expressive movements, on the other hand, are as p.r.o.nounced as ever. I still find the attempt to suppress them as difficult now as when I was working with the horse (page 57). I could not, of course, procure a curve of these movements of my own.]

Their movements, which at first were quite profuse, decreased more and more, so that in the case of von Manteuffel the percentage of my successful responses sank from 73% correct responses in 90 tests to 20% in a total of 20 tests,--and in the case of Schillings from 75-100% to 23% in a series of 35 tests. The curves obtained with von Manteuffel as subject, which I am here publishing (figures 8 and 15), are, however, true to his normal habits. The same is true of the two first curves of Schillings (figures 10 and 11), whereas the third (figure 12) shows distinctly the traces of the state of inhibition into which he fell, and represents the same condition as when Mr. Schillings, while preoccupied, tried to work with Hans. All the finer details of the phenomena in question, were likewise unknown to these two subjects.

For purposes of a clearer understanding of the various curves, figure 5 is inserted to give the general scheme of their arrangement.

[Ill.u.s.tration: FIG. 5.]

All curves are to be read like script from left to right. The first is the breathing curve of the questioner, the second, third and fourth curves represent his head movements,--all translated through the workings of the levers into up-and-down movements. The objective direction of these head movements is indicated by the arrows. It will be noted that (because the lever in question was one with two arms, and therefore reverses all movements made) each lowering of the head is indicated by a rise in the fourth curve, and each raising of the head is recorded by a sinking in the same curve. The records of the head movements forward and backward and to the left and right (curves 2 and 3) are two and one-half times the size of the actual movements; while the curve of the movements up and down (curve 4)--which is of especial interest to us--is five times its actual size. The fifth and sixth curves, which record my own responses, represent the taps of the horse,--the fifth indicating the number of taps and the sixth the back-step, which was Hans's reaction when he noted the head-jerk of the questioner. The seventh, the lowest line, indicates the time in fifth-seconds. Since the rate at which the drum revolved was not uniform for all the tests, the fifth-second marks do not appear the same distance apart in all the records, but are farther apart the greater the rapidity with which the drum revolved. For the experiment itself this is quite immaterial. Figures 6 to 9 correspond in detail with the diagram just described. Figures 10 to 12 differ only in that the breathing and back-step curves (the first and sixth in the diagram) are lacking. In these there is no response on my part to the head-jerk of the subject, but tapping was continued _ad libitum_ (in the case of the ill.u.s.trations here given I tapped to 5). When these latter curves were taken the ordering and the technique of the experiments had not yet been perfected. When this was finally done, Mr. Schillings, who acted as subject in those tests, had to be eliminated from the ranks of appropriate subjects on account of the increasing inhibitions, which gradually developed as described on page 120.

a.n.a.lysis of such curves is rather difficult, and those of different subjects cannot be directly compared. It is necessary to make a study of the normal curve of each subject taken when his affective state could be described as "indifferent". The influences of the purely physiological processes, such as pulse[R] and respiration, must also be determined.

And even so, an interpretation of the curve becomes possible only when a large ma.s.s of material is at hand, and when the introspections of the subject are taken into consideration. The following remarks, therefore, are not based solely upon the ill.u.s.trations given, but upon the ma.s.s total of my results.

[Footnote R: Slight head movements accompanying the pulse-beat were until recently regarded as the symptom of certain diseases of the vascular system (the so-called symptom of Nusset), but H. Frenkel has now shown them to exist also in normal individuals.[19] I myself discovered such movements (lateral as well as sagittal) more or less p.r.o.nounced in all the curves obtained from my subjects. The most striking case was that of a young physician whose circulatory system was perfectly healthy. In most instances I was able to note these oscillatory movements directly and to count them without much difficulty. For purposes of control the radial pulse was always determined at the same time. The observation of the phenomenon appears to be especially easy in the case of somewhat full-blooded individuals.]

In beginning our a.n.a.lysis, let us take first the breathing curve. Our results here were quite in accord with the view taken by Zoneff and Meumann,[20] who believe that in the respiration is to be found a good index of the affective tone of the subject's mental state. In the greater number of cases it was possible to conclude as to the degree of concentration of attention,--and when this was very great, it was even possible to get a clue as to the number thought of. Since the high degree of tension, under which a subject labored during a test, would be accompanied by strong affective coloring, we cannot regard as normal any of the curves here reproduced (with the exception of the two high points in figure 9). Although breathing was always deep and regular before and after a test, during the test it was less deep and irregular. Very often it was suspended altogether (figures 7, 8 and 9). In ordinary life we often notice that highly concentrated attention is usually accompanied by non-voluntary inhibition of movements in the musculature which, for the moment, is not directly involved; the man lost in thought slackens his pace and finally stands still, the intent listener or looker-on holds his breath.

Of the three curves registering the movements of the head, we find that nothing peculiarly characteristic is revealed by the two upper ones, giving the movements up and down, and to the right and left, respectively. They are the ordinary tremor-like movements and indicate nothing beyond the fact that the subject is unable to hold his head absolutely quiet for even one second. It is the third line that is of interest to us, for it is here that the oft-mentioned head-jerk (which indicates arrival--in the counting--at the number expected) registers itself. The moment of the head-jerk corresponds, almost without exception, with the moment of the first deep inhalation,--just as one would be led to expect from common experience. But we are not to regard the head-jerk as a result of the inhalation, for it also occurs when the subject complies with the request that he hold his breath during the test. The actual height of the jerks recorded in figures 6 to 12 was 1/4 to 1-1/2 millimeters and the average height obtained from the forty curves of these four subjects was 1 millimeter. There is great individual variation: the greatest height that was obtained from the records was 2-3/10 millimeters, the lowest 1/10 millimeter. The variations within the records of the several individuals are comparatively slight and are evidently dependent, in the main, upon the degree of concentration of attention. Thus in the case of von Allesch, where in 75 tests the average height of the jerk is 1 millimeter, the mean variation is 4/10 millimeter. If, in order to obtain some idea of the size of Mr. von Osten's movements,[S] we compared the values gained in the laboratory with those which would probably obtain in his case, we would say that his head movements were more minute than almost any of those of which we obtained records. At the most they could not have been more than 1/5 millimeter (when measured in terms of the distance through which the brim of his broad hat moved, they would appear to be about 1-1/2 times as large. See page 49.) The movements of Mr. Schillings, on the other hand, were certainly four or five times as great as those of Mr. von Osten, and occasionally even greater than that. When we turn to consider the time-interval elapsing between the subject's final head-jerk and my reaction (as recorded in the sixth curve), we find that the reaction-time averages 3/10 seconds, a value which agrees very favorably with that estimated for the horse (page 56). Thus it appears that man and beast have the same reaction-time--though we must bear in mind that I worked under some difficulty, since I had to care for the apparatus.

[Footnote S: In a special series of experiments a subject was instructed to execute rapid head movements as minute and as evenly as possible. These were registered objectively and at the same time I made judgments concerning them. The results showed that my judgments were most exact in the case of the most minute jerks. The thing that made it especially easy to judge the movements of Mr. von Osten under normal conditions, (page 220), was their extraordinary evenness, such as I have not met with in any other individual.]

Let us now turn to a discussion of the several figures.

Figure 6 (von Allesch) gives a typical view of the great, and at the same time economic concentration of attention characteristic of the subject. Respiration (first curve) is not so profound as usual, yet is changed very little. The head-jerk (fourth curve) is of medium height.

It occurs just at the proper moment,--the subject had thought of 2, and had directed his attention economically. This attention was of the kind described as type I on page 93. The lowering of the head, (recorded in the figure by a rise in the curve), immediately following upon the head-jerk upward, is irrelevant.

In figure 7 (Chaym) we have a record of a different nature. Respiration was inhibited throughout the test,--(the small waves are due to the pulsating of the heart); immediately after the test deep breathing takes place. Tension steadily increased till 3, the number expected, was reached. The head, accordingly, gradually sank a little forward. The head-jerk ensued during an interval beginning just before the reaching of the goal and ended immediately after. The movement was predominantly backward, its upward direction being only through a distance of 1/4 millimeter. (This subject was not so strongly motor as the preceding one.) The reaction followed promptly as seen in curve 6. It was the decided raising of the head which follows the head-jerk, that prevented the usual back-step with the left foot, when the subject was working with Hans.

[Ill.u.s.tration: FIG. 6.]

[Ill.u.s.tration: FIG. 7.]

Figure 8 (von Manteuffel) is typical of strong and at the same time economical concentration. Respiration, normally deep and very regular, is for a time completely inhibited. Tension rises steadily and the head gradually inclines forward. In the interval between the number before the final one and the final one the subject makes a sudden bend forward and immediately upon reaching the final number gives a violent jerk of the head, upward. The attention here would be characterized as being of type III, described on page 94. (Owing to lack of s.p.a.ce it is impossible to give an example of type II, which is only to be found in the case of very large numbers.)

Figure 9 (von Allesch) is expressive of great, but--according to the subject's introspection--not economical concentration. Respiration, which before and after the test was quite regular, during the test itself shows a pause. (The tiny waves are due to the heart-beat.) The subject had thought of 5, and this number is accompanied by a decided head-jerk. But we note that even before the final jerk a number of less p.r.o.nounced jerks occur--the result of poorly regulated psychic tension.

[Ill.u.s.tration: FIG. 8.]

[Ill.u.s.tration: FIG. 9.]

Figure 10 (Schillings) depicts a very high degree of uneconomical concentration. There was sudden concentration at the beginning of the test, and a steady increase throughout its course. Accordingly Mr.

Schillings bent forward at the start, and inclined still farther forward at the second--and just before the third--tap. But at 3 there is a sudden upward jerk. The number thought of had been 4, tension therefore had exploded, as it were, too soon.

[Ill.u.s.tration: FIG. 10.]

Figure 11 (again of Schillings) gives indications, on the other hand, of a medium and economic concentration of attention, which is more normal in character. The number thought of was 4.

[Ill.u.s.tration: FIG. 11.]

Figure 12 (Schillings again) is indicative of a low degree of psychic tension. With the very first tap the head begins to rise and continues to do so throughout the test. A true final jerk does not occur, we note rather in all three curves registering the head movements, slight time-marking movements, especially in the second curve. In the third curve they are at first minute, but increase steadily in size until the fourth tap, after which they suddenly disappear. The subject had, as a matter of fact, thought of the number 4, but it is hardly probable that Hans would have reacted properly upon these stimuli.

[Ill.u.s.tration: FIG. 12.]

Mr. Schillings had thought of the same number in all three tests given in figures 10, 11 and 12. The probabilities are that if he had been working with the horse at the time, in the first case Hans would have reacted with three taps with the right foot and a final tap with the left, as a result of the questioner's bending forward again after the premature head-jerk at 3. In the second instance the horse would probably have given four taps with the right foot, and in the third, the chances are that he would have continued to tap beyond the 4.

These curves give, on the whole, a fair idea of the intensity and of the course of attention of the various subjects.

Let us now consider a number of records which ill.u.s.trate the expressive movements involved in the process of thinking of such concepts as "up", "down", etc. Their arrangement is identical with the scheme given in figure 5, with the exception that the tapping curves (the sixth and seventh) do not appear. The subject was asked to think of any of the words "up", "down", "right", "left", "yes", "no", etc. He was to begin to conceive them vividly when the command "Now!" was given. This moment is recorded in figures 13 to 15 on the fifth curve. What has been said on page 123 with regard to respiration, holds also in these instances: only the first rise recorded in figure 14 can be regarded as normal.

The magnitude of these movements varies between 1/2 and 3 millimeters.

The records of the subject whose movements were most extensive, show an average of 1-7/10 millimeter (based on 50 tests), with a mean variation of 6/10 millimeter. Lack of s.p.a.ce precludes the reproduction of more than three records.

Figure 13 (von Allesch) shows the movement accompanying the thought of "up", a slight raise of the head, recorded in the fourth curve. (The thought of "down" is accompanied by a corresponding downward movement.)

[Ill.u.s.tration: FIG. 13.]

Figures 14 (von Allesch) and 15 (von Manteuffel) ill.u.s.trate the nod which is a.s.sociated with the thought of "yes" in the case of two subjects. It is essentially the same in both: the head is lowered and then raised. The first of the two subjects is more decidedly motor, and his movements therefore were somewhat the more extensive. In the case of the second subject the nod proper is followed by another which is somewhat less extensive.

[Ill.u.s.tration: FIG. 14.]

[Ill.u.s.tration: FIG. 15.]

A number of other experiments were carried out which corresponded with the color-selecting tests made upon Hans. (Page 78.) Five sheets of white paper, 1/2 meter long and 1/4 meter wide, were arranged in a series upon the floor, 1/4 meter apart. A dot marked the middle of each.

The experimenter stood at a distance of 7-1/2 meters and directly opposite the middle sheet. At about 1/2 meter to the right or left of him stood the subject who took the part of the "horse". The problem of the experimenter was to indicate to the "horse" a certain one of the five sheets, but without the use of word or gesture. I at first undertook the role of "horse", whereas the others consecutively played the part of questioner. All of them looked fixedly at the sheet which they had in mind. Besides, it usually happened that they would turn at least their heads, and often their bodies, more or less in the direction of the particular sheet--and this without purpose or knowledge on their part, but purely as a result of concentration upon the sheet they wished me to point out. One of the experimenters remarked, quite casually, that he had noted that I always made a better judgment, the more intently he thought of the sheet. Others often admitted that, when I had made an error, they had not imagined the sheet vividly, or had been debating whether or not to decide to think of the neighboring sheet--the one I had designated. This indecision could be noticed by the direction of the eyes. But the following table shows how uniform, on the whole, was the behavior of the various persons when under the guidance of the same impulse. The number of tests was 200 in each case. All errors were of the same character. Neighboring sheets were mistaken for each other, and the errors were never of more than one position to either side.

Their number can easily be obtained by subtracting the percentage of correct inferences from the total, 100%.

Experimenter: v. A. B. C. Mrs. v. H. K. Miss v. L.

Correct inferences: 88% 88% 77% 81% 77% 82%

It will be seen that the number of correct interpretations is quite high and in none of the cases does it deviate far from the mean average of 82%.

I based my judgment as to the direction of the subject's eyes, upon an imaginary line perpendicular to the center of the cornea. (This perpendicular does not always coincide with the subject's line of vision, which was the thing I was after, but this cannot be directly obtained. This, of course, was what made the judgment a rather difficult matter.) My judgment as to the direction of the head I based largely on the direction of the nose, (to express it more accurately: upon the direction of the median plane.) I purposely noted only the position of the experimenter and not the movement which led up to it. When I tried to do the latter, the results were not always satisfactory, because the head and eyes of the person would frequently, in the process of adjustment, move beyond the goal and thus lead me into error. An attempt was made to make each judgment as independent as possible of the preceding one. But usually, after a few tests, an unintentional a.s.sociation became established between certain att.i.tudes and the different places in the series of papers. Often all that was necessary was to observe the experimenter in order to know which of the places he had in mind, it was not necessary to look at the papers at all. Every change in the position of the person would, of course, make the a.s.sociation thus established, useless.

Later, the subjects and I changed roles, I took the part of the experimenter and they the part of the "horse". The number of tests in each case was 200 as before. Here, too, errors were, with but one exception, never more than of one place to either side. Whether the error was one place to the right or one place to the left appeared to depend upon the position of the person making the judgment, i. e., it depended on whether he stood at my right or at my left. The following results were obtained:

Subject ("horse"): v. A. B. C. Mrs. v. H. K. Miss v. L.

Correct inferences: 76% 79% 75% 81% 77% 74%

A certain agreement can be seen in these results. The average of correct inferences is somewhat lower than that which was obtained by me (page 135), 77% as over against 82%. This is probably due to the fact that the subjects had had so little practice compared with me.

With one of these subjects, Mr. Koffka, a student of philosophy, I carried these tests somewhat further, varying them partly by increasing the number of sheets of paper, partly by decreasing the distance between them. The increase in the number of sheets made only a slight difference in the results. With 200 tests in each case I obtained the following results: