Clever Hans (The Horse of Mr. Von Osten) / A contribution to experimental animal and human psychology

by Oskar Pfungst

The following is a report of the account, which Mr. von Osten gave Professor Schumann and me, of the method which he had used in the instruction of the horse, and which was illustrated by actual demonstrations. I cannot testify, of course, that Mr. von Osten really did adhere to this method throughout the four years in which he tutored the horse, but I will say that I have several good reasons for believing that it was impossible for him to have trumped up this make-believe scheme afterward, merely to mislead us. Among the reasons are the following: He was always ready to give a detailed explanation of any question which we might interpose; the written statements of Major von Keller, who has known Mr. von Osten for a period of fifteen years; the testimony of General Zobel, who became acquainted with the whole process fully a year before any public exhibitions were given; the accounts given by the tenants in Mr. von Osten's house, who for years saw the process of instruction going on in the courtyard of the apartment building,—according to their account his intercourse with the horse was like that with a child at school,—he made much use of the apparatus and never did they notice anything like an habituation to respond to certain signals; and finally the appearance of the apparatus itself—some of which could not be bought at second hand—was most convincing.

The apparatus used for the work in arithmetic consisted mainly of a set of large wooden pins, a set of smaller ones (such as are to be had in toy-shops), a counting-machine, such as is commonly used in the schools, a chart upon which were pasted the numbers from 1 to 100, and finally the digits, cut large and in brass and suspended from a string. For the work in reading Mr. von Osten used the chart shown in the frontispiece of this book. Here we have the letters of the alphabet in small German script with numbers written below which serve to indicate the row, and what place in that row, the letters occupy. For tones, a small, child's organ was used with the diatonic scale C^1 to C^2, and for instruction in colors, a number of colored cloths were used.

The work in arithmetic began by placing a single wooden pin in front of Hans and then commanding him: "Raise the foot!—One!" Here we must assume that the horse had learned to respond to the command to raise the foot during the preceding period, when tapping in general had been taught. In order to get the horse to learn that he was to give only one tap, Mr. von Osten tried to control the tapping by means of holding the animal's foot, just as a teacher tries to aid a pupil in learning to write by guiding his hand. He repeated this exercise so often that finally the single tap was made. And always the right foot was insisted upon. Bread and carrots were the constant rewards.

Two of the pins were now set up and the command given: "Raise the foot!—One, two!" Mr. von Osten again aided the establishment of the proper association by using his hand as before. At the same time the two pins were pointed out, and the order was always without exception from left to right. Gradually it became unnecessary to touch the foot or to point to the pins, and instead the question was introduced: "How many are there?", in order that the horse should become accustomed to these words as an invitation to give the taps when he saw the wooden pins before him.

Then three pins were taken and the words "one, two, three" were spoken, and so on. In naming a number the preceding ones were always named along with it, in order that the normal order might thus be learned at the same time. Later the number alone, without the preceding ones, sufficed to elicit the proper number of taps. The last word of the series thus becomes characteristic of the series as a whole. It differs from all the others, and thus becomes the sign for the whole series of numbers thus named, each of which arises as a memory image at the proper place in the series and is accompanied by a tap of the foot. Thus, Mr. von Osten at any rate had accounted to himself for his success.

But Hans was not to acquire merely this relatively mechanical process of counting (hardly to be called counting), but he was to acquire also some meaning content for the number terms. For this purpose everything depended upon the concept "and". Only he who can grasp its meaning will be able to understand a number. 2 is 1 and 1, 3 is 2 and 1. Mr. von Osten had someone hold a large cloth before the horse, where the wooden pins usually were placed. He then had the cloth taken up and he would pronounce emphatically the word "and". After this had been done a number of times, he put up two of the pins and obscured them by the cloth. The cloth was again raised and the word "and" pronounced. Then Hans, as a result of his previous instruction (so Mr. von Osten thought) would give two taps at sight of the pins. The thing was repeated with three pins, then with one, and so on, and the horse would always execute the proper number of taps.

Now, five pins were set up, the three to the right being covered by the cloth. The horse tapped twice and Mr. von Osten said "two". Then the cloth was raised, Hans gave three further taps, and Mr. von Osten said "and three" with emphasis.

In this simple manner he tried to get the horse to understand that the three belongs to the two, and that both together make five. The image of the five pins as it was known from previous experience, was to be associated with the combined groups of two and three, and conversely, it was to be reproduced when these groups were presented. Later the cloth and pins were omitted and the question was asked: "How much is two and three?". The horse tapped five times. It had learned how to add. Still this could be regarded only as a mechanical process, if the horse were able to add only those numbers which had been presented together one or more times in the manner just described. And so long as we remained within the first decade, we could get twenty-five binary combinations whose sum does not exceed 10 (counting inverted orders we would have forty-five binary permutations),—all of which might have been practised separately. But as a matter of fact, Mr. von Osten did not take this course, for as he himself says, he allowed Hans to discover a great deal for himself. "Hans had to develop the multiplication table for himself."—With larger numbers and more addends, the number of combinations becomes so great that there can be no doubt they were not practised separately.

Since, after all this preliminary instruction, Hans really began to give solutions of new problems, the master believed that this was proof that he had succeeded in inculcating the inner meaning of the number concepts, and not merely an external association of memory images with certain movement responses. But he always remained within the sphere of the ideas thus developed, and adhered closely to the customary vocabulary and its usage. Every new concept, each additional word was explained anew.