Lecture Notes for January 13 and January 20th

Hellar (2000) poses some questions that confront one when they begin to study touch, representation, and blindness. These questions are valuable because they indicate areas where philosophical and psychological frameworks make an impact on the study of pictures and the blind.

First question: Is information in the stimulus impoverished or rich?

Three different approaches on this issue are those of Helmholtz, Kant, and Gibson.

The Helmholtzian approach is an empiricist approach that emphasizes that perception depends on learning and that this learning is largely inferential in nature.

The basic picture of perception is that the only innate knowledge is sensation and some basic inferential abilities such as associations or simple logical rules like modus ponens.

These sensations are the basic stimulus and they are very impoverished in that they lack any inherent meaning or order. Instead that order and meaning are built up through cognitive processes of learning and logic and this is how perception of the world arises.

These cognitive processes remain active even after learning has taken place because the sensory data of the sensations is always ambiguous, variable, and incomplete. Learning and logic are always needed in order to overcome these difficulties Æ unconscious inference.

The working of these higher inferential processes can be seen in perceptual illusion. They reflect that normal processes that are compensating for the impoverished stimulus e.g. shape constancy mechanisms and depth perception mechanism at work to create size illusions as in the Ames room illusion.

Helmholzian approach is the dominant approach in psychology. Another important approach in connection with spatial perception is Immanuel Kant.

For Kant, many of the features that we perceive are not actual aspects of reality but rather reflect an innate structure of pattern that the mind actually imposes on reality. Kant was critical of the idea that we could empirically learn about space. For Kant experience presupposed spatial organization in order to be meaningful experience from which we could learn. Without spatial organization we would not have perceptual experience but just meaningless chaos from which nothing could be learned. For Kant we do not perceive 3 dimensions we perceive in 3 dimensions, i.e., it is the form of our perceptual experience and without that form imposed on our perceptual there would be no perceptual information at all.

Kant could point to two important facts about spatial perception. We perceive/conceive a lot spatial relations as necessary, e.g. the interior angles of a triangle add to 180 degrees. We know that this must be the case for all triangles even though we cannot establish such a conclusion empirically. The second point is that one cannot logically derive a third dimension from two dimensional sensory data.

This Kantian approach figures in discussions about the possibility of amodal perception of spatial information, which we will discuss later.
The 3rd approach differs significantly from both the Helmholtzian and Kantian approach. It is an approach that goes back to Aristotle but is more recently associated with Gibson.

Aristotle argues against Plato who distrusted the senses as not providing knowledge and being inferior to intellectual thought. Aristotle argued that our senses could be trusted and that perceptual errors were due to interference in the medium or malfunction in the organ or disagreement between people. He argues that how you correct perceptual error tells you what the actual cause of the illusion is. Typically we clear the medium, correct organ malfunction, or we get intersubjective agreement in order to determine reality. That tells us the sources of the errors. As long as the conditions for correction were met our perception and reality were identical in form, hence in-formation. We merely had to pick up this information from the world.

Aristotle’s theory is very commonsensical and comports well with our everyday experience. It lasted for 1000years. However significant events called it into question. The Copernican revolution lead to important consequences for psychology. If Copernicus is correct then perception that is free from interference, malfunction, and gets intersubjective agreement is still an illusion, e.g. the sun rising in the east and setting in the west.

So our senses cannot be trusted. They do not provide us with veridical information. Instead the information is ambiguous, variable, and incomplete. Hence the Helmholtzian approach. The senses are always potentially misleading and must be constantly corrected by inferential processes.

However, the Aristotelian approach as been resurrected by Gibson in his approach of ecological realism. In this approach the stimulus is rich with information. This information contains invariant higher order patterns that specify features of the world. So the solutions to most of the problems confronting perception can be found be closely examining the nature of the stimulus rather than looking inside the person.

We discussed the looming reflect and how the information for it involves higher order invariance that is complete, clear and therefore specific to indicate an object is looming.

What about Gibson and perceptual illusions? They are due to artificial experimental situations in which the complex interactions between variables have been removed through the experimental control of variables. This situation is not ecologically valid, i.e. the results will not generalize to the real world. The complex interaction between variable that carries higher order patterns does not exist in the experimental situation, and that is what is most relevant in the real world. Hence the illusions are the result of such artificial impoverishment of the stimulus. So the study of illusions or sensations (also artificially impoverished perception) cannot increase our understanding of perception.

So instead of perception being the pick up or simple sensations followed by large amounts of inferential processing, we have that perception is the sophisticated pick up of complex patterns that contain a lot of information that is directly about the world.

All of this leads to another question form Hellar.

What are the role of illusion in touch and the source of this illusion? Although we won’t be directly addressing this we can wonder if pictures, especially tactile/haptic pictures are illusory in nature.

Another question that follows form our first question about the nature of the stimulus is whether touch is normally active or passive. Is processing similar in active and passive touch, or static holding? The ecological approach is that touch is normally active and that passive touch is artificial and impoverished. It causes one to focus on sensations rather than high order patterns.

In contrast a more Helmholzian approach would not predict as great a difference between active and passive touch. Perception runs off of sensations, which are basically the same in both conditions. In fact, perhaps better in passive touch. It is difficult to say what Kant might conclude. Perhaps that spatial relations could equally be imposed on active and passive perception. Where Kant and Helmholtz would differ would be in their predictions about the role of experience and learning, and therefore about the perceptually abilities of the congenitally blind.
This takes us to another of Heller’s questions. Is there intermodal equivalence or non-equivalence between touch and vision? Is it possible to get the same information through different senses.

Gibson made reference to an older argument by Locke, which actually goes back to Galileo and the period of the scientific revolution. Locke argued that we can trust information that is intermodal. Primary qualities such as number or shape are really in the world since such information can be sensed through more than one sense, and such intermodal coherence and consistently gives us reason to believe that these are qualities of the object, i.e., objective properties of the world. In contrast to such mathematical properties are secondary qualities that exist only in one sense and therefore could not be trusted to be in the object, but were merely subject-ive in nature. So reality was perceived through higher order intermodal patterns. According to Gibson this argument is basically correct and there is such intermodal equivalence, which is a valuable source of information about the world.

We can ask, is there real intermodal (intersensory equivalence) and to what degree? Does this equivalence point to an even higher order amodal sense of space like Kant advocated. Does one sense dominate over others, e.g. vision over touch in the train moving illusion?

Another of Heller’s question concerns the role of such higher level cognition and representation. In particular what is the role of mental imagery in touch perception by the blind. One intuitive idea is that when a blind person is perceiving tactile/haptic pictures that what they are doing is forming an internal visual mental image that one internally "looks at". Much class discussion about this point is where the second lecture ended.