An error of perception is often called an "illusion", though this term is commonly reserved for errors that are large and curious. When one who is being awakened by a bell perceives it as a tom-tom, that is an illusion. An {451} illusion consists in responding to a sensory stimulus by perceiving something that is not really there. The stimulus is there, but not the fact which it is taken to indicate. Illusion is false perception.

The study of illusions is of value, not only as showing how far a given kind of perception can be trusted, but also as throwing light on the process of perception. When a process goes wrong, it sometimes reveals its inner mechanism more clearly than when everything is running smoothly. Errors of any kind are meat to the psychologist.

Illusions may be classified under several headings according to the factors that are operative in causing the deception.

1. Illusions due to peculiarities of the sense organs.

Here the stimulus is distorted by the sense organ and so may easily be taken as the sign of an unreal fact.

Separate the points of a pair of compasses by about three-quarters of an inch, and draw them across the mouth, one point above it and the other below; you will get the illusion of the points separating as they approach the middle of the mouth (where the sensory nerve supply is greatest), and coming together again as they are drawn to the cheek at the other side.

Under this same general head belong also after-images and contrast colors, and also double vision whenever for any reason the two eyes are not accurately converged upon an object. The fact that a vertical line appears longer than an equal horizontal is supposed to depend upon some peculiarity of the retina. Aside from the use of this class of illusions in the detailed study of the different senses, the chief thing to learn from them is they so seldom are full-fledged illusions, because they are ignored or allowed for, and not taken as the signs of facts. An after-image would constitute a genuine illusion if it were taken for some real {452} thing out there; but as a matter of fact, though after-images occur very frequently--slight ones practically every time the eyes are turned--they are ignored to such an extent that the student of psychology, when he reads about them, often thinks them to be something unusual and lying outside of his own experience. The same is true of double images. This all goes to show how strong is the tendency to disregard mere sensation in the interest of getting objective facts.

2. Illusions due to preoccupation or mental set.

When an insane person hears the creaking of a rocking-chair as the voice of some one calling him bad names, it is because he is preoccupied with suspicion. We might almost call this an hallucination, [Footnote: See p. 375.] since he is projecting his own auditory images and taking them for real sensations; it is, at any rate, an extreme instance of illusion. In a milder form, similar illusions are often momentarily present in a perfectly normal person, as when he is searching for a lost object and thinks he sees it whenever anything remotely similar to the desired object meets his eyes; or as when the mother, with the baby upstairs very much on her mind, imagines she hears him crying when the cat yowls or the next-door neighbors start their phonograph. The ghost-seeing and burglar-hearing illusions belong here as well. The mental set facilitates responses that are congruous with itself.

3. Illusions of the response-by-analogy type.

This is probably the commonest source of everyday illusions, and the same principle, as we have seen, is operative in a host of correct perceptions. Perceiving the obliquely presented rectangle as a rectangle is an example of correct perception of this type. Perceiving the buzzing of a fly as an aeroplane is the same sort of response only that it happens to be incorrect. If the present stimulus has something in {453} common with the stimulus which has in the past aroused a certain perception, we may make the same response now as we did before--especially, of course, if the present mantel set favors this response.

Fig. 67.--The Ladd-Franklin illusion of monocular perspective. Close one eye, and hold the book so that the other eye is at the common center from which the lines radiate; this center is about 5 inches from the figure. Hold the book horizontally, and just a little below the eye.

A good instance of this type is the "proofreader's illusion", so called, perhaps, because the professional proofreader is less subject to it than any one else. The one most subjcet to it is the author of a book, for whom it is almost impossible to find every misspelled word and other typographical error in reading the proof. Almost every book comes out with a few such errors, in spite of having been scanned repeatedly by several people. A couple of misprints have purposely been left in the last few lines for the reader's benefit. If the word as printed has enough resemblance to the right word, it arouses the same percept and enables the reader to get the sense and pass on satisfied. {454} Before we began to pore over books and pictures, the lines that we saw usually were the outlines of solid objects, and now it requires only a bare diagram of lines to arouse in us the perception of a solid object seen in perspective. An outline drawing, like those of the cube and staircase used to illustrate ambiguous perspective, is more readily seen as a solid object than as a flat figure.

Fig. 68.--Aristotle's illusion.

Another illusion of this general type dates away back to Aristotle. Cross two fingers, perhaps best the second and third, and touch a marble with the crossed part of both fingers, and it seems to be two marbles; or, you can use the side of your pencil as the stimulus. In the customary position of the fingers, the stimuli thus received would mean two objects.

A much more modern illusion of the same general type is afforded by the moving pictures. The pictures do not actually show an object in motion; they simply show the object in a series of motionless positions, caught by instantaneous photography. The projector shows the series of snap-shots in rapid succession, and conceals them by a shutter while they are shifted, so as to avoid the blur that would occur if the picture were itself moved before the eyes. But the series of snap-shots has so much in common with the visual stimulus got from an actually present moving object that we make the same perceptive response. {455} The same illusion in a rudimentary form can be produced by holding the forefinger upright three or four inches in front of the nose, and looking at it while winking first the one eye and then the other. Looked at with the right eye alone it appears to be more to one side and looked at with the left eye alone it appears to be more to the other side; and when the one eye is closed and the other simultaneously opened, the finger seems actually to move from one position to the other.

Fig. 69.--The pan illusion. The two pan-shaped outlines are practically identical, but it is hard to compare the corresponding sides--hard to isolate from the total figure just the elements that you need to compare.

4. Illusions due to imperfect isolation of the fact to be perceived.

Here belong, probably, most of the illusions produced in the psychological laboratory by odd combinations of lines, etc. A figure is so drawn as to make it difficult to isolate the fact to be observed, and when the observer attempts to perceive it, he falls into error. He thinks he is perceiving one fact, when he is perceiving another. The best example is the MÜller-Lyer figure, in which two equal lines are embellished with extra lines at their ends; you are supposed to perceive the lengths of the two main lines, but you are very apt to take the whole figure in the rough and perceive the distances between its chief parts. You do not succeed in isolating the precise fact you wish to observe.{456}