Seeing What One Is Prompted to

The power ofverbal labels to influence what we see can be seen in the following example. Look at the figure below:

Skyline

Now, look at the same figure rotated:

Letter

What do you see now? In this case, once you "get it" and see the letter, it is difficult to see it as any other object.

Words have a pronounced effect on what we see, which may either facilitate perception or make it difficult to see things outside the linguistic box.

6.3 Illusions that require the observer to fill in the missing pieces. In the Poggen-dorff illusion (left), the diagonal line at lower left, when extended by the viewer's imagination, does not seem to line up with the upper diagonal line. Yet, if you measure them, they fit perfectly. At right, the broken line shows where people imagine the line should extend. In the modified Muller-Lyer illusion (right), the distance between the apexes of the left and center "wings" and that between the center and right "wings" are equal, even though the space on the left looks much longer than the one on the right.

6.3 Illusions that require the observer to fill in the missing pieces. In the Poggen-dorff illusion (left), the diagonal line at lower left, when extended by the viewer's imagination, does not seem to line up with the upper diagonal line. Yet, if you measure them, they fit perfectly. At right, the broken line shows where people imagine the line should extend. In the modified Muller-Lyer illusion (right), the distance between the apexes of the left and center "wings" and that between the center and right "wings" are equal, even though the space on the left looks much longer than the one on the right.

perceived as being different in size if they are perceived as located at different distances. Artists from the earliest cave painters on have practiced this law, but it was during the Renaissance that artists systematically used contextual cues to create the illusion of depth and three-dimensionality.

Some visual illusions call upon the observer to supply missing material, as in the case of the Poggendorff illusion and the modified Muller-Lyer illusion shown in figure 6.3. Both of these illusions have been part of the standard psychological repertoire for 150 years—the Poggendorff illusion was first published in 1860— but the effect was known to artists before that time. (See later discussion of Rubens.) The illusions of "misalignment" and "unequal gaps" are compelling, even though much of the illusion is created by the interior lines being imagined by the observer rather than sensed. Such is the power of our mind to see things as they ought to be, rather than as they are. This illusion is a bit more fundamental than the illusion created in the mind of hockey fans, which is based on culturally learned biases—a more "cerebral" illusion.

The propensity to "see" things that do not exist, apart from those truly psychotic hallucinogenic episodes, is illustrated in the illusions shown in figure 6.4. These types of figures are called illusionary contours, as illusory lines and outlines of forms are perceived. Examine figure 6.4. Do you "see" the horizontal and vertical lines in A? These "lines" are clearly perceived by most viewers. On closer inspec-

"lllr

6.4 Phantom figures: two imaginary lines, a triangle, and a column. In A we "see" a horizontal or vertical line; in B a white triangle seems to float above the surface of the page and is whiter than the background; in C there are two illusions: there appears to be a white column surrounded by semicircles, and the diagonal line seems to be misaligned.

tion, we understand that the illusory lines are defined by the parallel white lines. In B you have the sensation that you are looking at a white triangle (commonly called a Kanizsa triangle) which is described by just having three corners physically present. In C, a compound illusion is created. The illusion of a white column is created by the surrounding semicircles. That illusion is so strong that it further creates a Poggendorff illusion with the diagonal line: here an illusion creates another illusion.

In spite of the strong subjective impression, somehow we know that these visual objects are illusionary. Some of the salient features of visual illusions are: they appear to be figure rather than ground, they appear to hover over the background, and they are more saturated than the background (in figure 6.4B and C the illusion appears to be whiter than the background). There are some additional compelling features of these illusionary contours. In the case of the Kanizsa triangle, ifyou stare at the imaginary triangle for a few seconds and then cover the little "Pac-Man" circles, the illusion of the triangle still exists. The same thing happens for the column in 6.4C.

Here we offer two explanations for the natural tendency to see things that do not physically exist, both based on a model of limited processing capacity. We're not crazy when we see illusions—quite the opposite. Seeing illusions is not only normal, but necessary for survival.

• When we see an illusion, it may be a very clever means to see things based on only fragmentary cues—psychologists call this redintegration. When only part of a visual field is present, there is a tendency to make an object whole or to form a good Gestalt and, by doing so, to "see" a basic object easily understood. When we "see" illusions, we disambiguate features that otherwise seem to be a jumble of unconnected visual noise.

• In addition, identifying figures as being distinct from background was of fundamental importance in adapting to terrestrial objects. In order to "see" objects, strong line and edge detectors evolved in the eye and brain. Furthermore, the criteria for detecting lines or forms, such as a triangle, a circle, and even a face, were not exact. Thus, poorly defined lines, incomplete objects, and even illusionary lines and forms were admitted as valid images and processed as if they were physical forms.

In the next section, we will see that some illusions are so convincing that they stimulate brain regions in a similar way to actual stimuli.

Neurological Activity Associated with Visual Illusions

Tracing the pathways from visual input to the eye and to the brain has been made possible through traditional psychophysical methods, as well as recent developments in neuroimaging techniques. Not so many years ago, optical illusions of the type just discussed were considered witchery; many worked their way into the magician's bag of tricks and are still used today. Artists employed optical illusions for centuries. Now illusions are the subject of serious neurological studies.

Investigations of the neurological substrates of illusions make fascinating science as well as raising some interesting theoretical issues. Subjective experience tells us we "see" a line, triangle, or column, and these experiential phenomena have been corroborated by laboratory studies. The deeper question is, Is seeing an illusion neurologically similar to seeing the actual object it represents?

The question is important. If illusions and real objects share the same neurological space, this gives support to the idea that the brain, not the eye, contributes to perceptual distortion. Are the illusionary lines, triangle, and column simple shell games played out in the physical world that are misapprehended by the eye, or are illusions actually represented by the brain in a way similar to physical stimuli? We will now look at the neurological studies of illusions.

The Ponzo illusion, which is so prominent in art and everyday life, is neuro-logically curious. Return to figure 6.2. No doubt the top line in the right-hand drawing appears longer (as well as more distant). Suppose you present the illusion in color, with red lines and a green background. The illusion continues much as it does in black and white, but suppose the illusion is presented in color with the lines iso-luminant (equal in brightness) with the background. Surprisingly, the illusion diminishes. It seems that there is a segregation of function within the visual pathways involved in the Ponzo illusion. Specialized cells are turned on only by isoluminant stimuli, as in the case of lines and backgrounds of equal brightness. These cells appear to be depth-insensitive. (See Livingstone and Hubel 1988 for further details, including perceived movement, color, and depth.) As we shall see shortly, artists have effectively used the Ponzo illusion (or a variation of it) in their work to create an illusion of depth, intuitively using figure-ground renditions composed of contrasting luminance. It may be that we evolved excellent cues for depth perception that work best with well-illuminated features and background, as might be found in nature.

Perhaps even more startling neurological support for visual anomalies is found in studies of contour illusions and the Kanizsa illusion. Meticulous work has been done in the Netherlands (see Peterhans and von der Heydt 1989, 1991; von der Heydt and Peterhans 1989) with monkeys, whose visual cortex is similar to that ofhumans. Using tiny threadlike electrical probes, the von der Heydt group demonstrated that cortical cells respond to contour illusions much as they do to actual visual signals. If shown a moving vertical white line on a black background, single cells in the visual area called V2 (an area associated with color, form, and depth perception) respond. The striking finding, however, is that the same cells also respond to an illusory contour (like the one shown in figure 6.4^) presented in the same orientation. The effect has been replicated many times with different stimuli. Such findings have led some to conclude that there are contour cells that integrate output from earlier structures (the primary visual cortex V1).

Cortical activity has also been demonstrated with the Kanizsa triangles (see figure 6.4.B) by Peterhans and von der Heydt (1991). The cells that "fire" to a real triangle in the V2 area also react to a Kanizsa triangle. From these experiments, it appears that at least some visual illusions have a basis in cortical physiology. In light of these studies, terms such as "visual illusions" and "optical illusions" now seem somewhat misleading, as the illusionary effect seems to be cortical and, in some instances, specifically V2-related. While some think "visual" and "optical" to be more peripheral, the term "V2 illusions" will probably not replace the other terms.

And, in fact, we do not know all there is to know about the interactions between higher-order processing centers and visual illusions. The careful work just reported has isolated parts of the brain that suggest a correspondence between visual signals and visual illusions. Much more is involved in the psychology of art, but these observations are important steps in our understanding of this hugely complicated and fascinating topic. We now turn our attention to a broader class of illusions devised by artists and architects to make the world appear as they imagine it.

Was this article helpful?

0 0

Post a comment