Here is a simple experiment you can do at home. Hold your hand a foot in front of your face with the thumb turned towards you, so that you are seeing your hand in profile. Have a look at it: it will look like a normal solid object in front of you. Now shift your gaze to fixate on an object about ten feet behind your hand, keeping your hand in the same position. You will find that your hand doubles in front of you (you may have to shift your hand slightly to get this effect): you see two hands not one. If you now close one eye you will find that the second hand disappears, only to reappear if you open both eyes. The reason for this effect is binocular disparity: each eye receives separate images of the hand, which are usually synthesized by the brain, but in this setup the two images are not synthesized—focusing on the distant object inhibits binocular synthesis. Why this is so is hard to explain—we don’t normally experience a doubling of nearer objects as we fixate on more distant objects—but it is a robust phenomenon. It works not only with hands, of course—books will produce the same doubling effect.
From this simple experiment we can draw two conclusions, by no means trivial. First, there exists a perceptual unconscious: whenever we see anything there are two images in the mind derived from the two eyes. Unlike the image that results from binocular synthesis, these images are unconscious: you are seeing those two hands unconsciously whenever you see your one hand consciously. Every (binocular) visual perception involves processing a pair of visual representations corresponding to the image on each retina. The experiment you just performed merely brings what is normally unconscious to consciousness. Second, it is possible to do psychology based purely on introspective evidence, and indeed it is hard to see how the doubling of the image could be detected without the use of introspection. Of course, the effect can be checked inter-subjectively, but for each subject the crucial information is derived from introspection. There is nothing methodologically wrong with that.
There is another effect that can be obtained from the experimental setup described. If you position your hand correctly, while fixated on a distant object, you will find that your hand, or a part of it, disappears from view: it becomes transparent or simply invisible. This is a disconcerting phenomenon, as your hand appears to melt away, losing solidity and opacity. It is not easy to describe exactly what happens: it is not as if your hand is suddenly removed from your visual field; rather, it is seen as (partly) invisible. But you cannot see an invisible object! The hand is registering visually but it has been rendered transparent. Clearly the hand is still in front of your eyes and is having its usual impact on your retinas, sending visual signals to the brain; but the result is not a seen hand but an unseen hand, or a seen un-hand. You stop seeing what you are “seeing”. Again, this strange perceptual state results from binocular disparity: your two eyes are getting a full view of the distant object, because the two different images capture the full reality of the stimulus, without a gap—but there is an interposed hand there that is also seen. Presumably a visual system could just accept that the hand is not blocking your view of the distant object, which it clearly is not, and accordingly produce an image of both near hand and distant unblocked object. But our visual system does not do that—instead it does away with the image of the hand. It decides not to see the hand. Why? Because it operates under the assumption that if a solid opaque object is between your eyes and some distant object, occluding that object, then it must be the case that some portion of the distant object is not seen. If there is an occluding proximal object, then some of the distal object must be occluded. But if that distal object is not in fact occluded, because of binocular disparity, then the only conclusion the brain will accept is that there is no occluding proximal object—thus it “disappears” the occluding object. It tells you there is nothing solid and opaque there in order to explain why it is that you see the uninterrupted whole of the distant object.
The brain thereby comes to a false conclusion—a visual illusion—but it does so for intelligible reasons. It would rather believe that there is an invisible or transparent hand in front of you than that a visible hand can be interposed between you and a distant object the view of which is not blocked by that hand. It is perfectly possible to see around the hand from the angles afforded by the two eyes, so that the two retinal images of the distant object can be synthesized into a single continuous visual image; but the brain prefers to believe that the reason the distant object can be seen in its entirety is that the interposed object has been rendered invisible. It is as if the brain does not understand its own binocular disparity system! If you look at the distant object with one eye while holding your hand in front of you, your hand remains stubbornly visible and opaque, with no tendency to melt or clarify. The brain accepts the fact that its view has been blocked. But when two eyes are involved we have blocking without a break in the distant object—and this the brain finds unacceptable. Each eye fills in what the other lacks because of the interposed hand, thus giving an uninterrupted view of the distant object; so the brain simply removes the hand from the visual field. It resolves the paradox of the blocked-but-seen object by rendering the blocking object invisible.
Of course, your brain knows quite well that solid objects don’t just disappear because your eyes focus on different things, but it is prepared to draw that conclusion in the circumstances described. And it is under no illusion that your hand has literally disappeared or turned transparent, since it is right there in front of your eyes sending in the usual signals; the brain only renders it invisible because it is well aware that it is there. One can only wonder at the frantic unconscious reasoning that the brain goes through order to decide to make the hand invisible despite its obvious visibility.
It would be possible in theory to attach a blocking object to the head in such a way that the brain always renders it invisible: it would have to be positioned in just the right place between the eyes so that binocular disparity could do the requisite filling in. In this setup the brain would constantly countermand the evidence of its own eyes and treat the object as if it wasn’t there, or it might be taken to exist in a transparent ghostly form. The object would instantly leap into visibility if the subject were to close one eye, but if monocular vision were impossible for the subject the interposed object may never be detected—though it is detected by the subject’s brain, only to be “disappeared”. The object would be unconsciously seen but consciously unseen. The brain sees it but it sends out instructions to the conscious subject not to see it—the subject may never suspect what his brain knows very well.
The brain constructs a unitary visual world from the disparate data supplied by the retinas of each eye. Normally this works smoothly enough and the perceiver doesn’t notice anything anomalous—the dual basis has no phenomenological counterpart. But in odd cases, even quite simple ones to arrange, the scaffolding of vision becomes apparent; then we experience perceptual anomalies. The “invisible hand” illusion is one of these, and it shows the complex nature of the underlying unconscious visual processes. The brain not only makes the world visible to us; it can also make it invisible, as the occasion demands. It can deny the evidence of the senses.