One of the principal claims of this site is that mental images exist within a continuum of causal modalities: from instances that may be described as clearly due to physical causes to those that seem to have only mental causes. The diagram below suggests this continuum, indicating several types of images.
Click on the image for more information.
The properties of mental images change according to their type and their position on the causal continuum. One of the principal changes we notice is that the more we move in the direction of mental causation and higher cognitive functions, the more involved images seem to be with language, and the less important their visual properties become.
For example, when an after-image is formed, no conceptual or linguistic processes appear to be involved in experiencing that particular kind of mental object. In other words, experiencing an after-image is not experiencing a linguistic event. Furthermore, this is an event that seems clearly rooted in physical causality. If we were to trace the causal chain involved in the formation of the after-image, we would find that it clearly involves physical causes at the beginning of the chain (hard-wired, neurophysiological changes in rods and cones in the retina). Finally, experiencing after-images is a form of sensation and these experiences would not normally (or, at least, pre-analytically) be included under the sorts of experiences that we would call "acts of thought" or "acts of cognition," since these acts are generally understood to involve some sort of conscious control or effort.
The situation changes with a low-level cognitive task, such as "determine the letter of the alphabet that results when a capital N is rotated 90 degrees clockwise." The processes involved in performing such a task are clearly "cognitive" (rather than merely neurophysiological) -- so we have moved into the realm of "mental causation." And, although there is a linguistic dimension (minimally, in understanding letters of the alphabet, and what is meant by "90 degree rotation") this cognitive task seems impossible to perform without an image. Clearly, the task is under conscious control and involves effort. At the same time, there is still a physical element that is a necessary part of the causal chain, since (as experimental evidence shows) the production of an image of a letter of the alphabet involves activation of cells in the visual cortex. Thus, the entire process involves mental processes and causes, but ones that also require the use of a specific image that has a non-linguistic, presentational quality, which can be associated with physical causes.
Now consider this question: "If B is taller than A is taller than C, is B taller than C?" The task could be solved without an image, using linguistic processing (logical relations) alone, but an image makes the task easier. One can "read off" the answer to the problem using the image, without much mental effort. Now we are on a different point in the continuum of causes that can be associated with mental images. Here, the connection with linguistic processing, or purely mental causes, seems stronger, while the connection to the image is much weaker than in the rotation task.
Finally, there are cognitive tasks (or acts) processed by the imagination, as symbolized by Da Vinci's drawing of a flying machine. How much of the task involves the act of drawing (or imagining) the image? Should we think of the act of image manipulation as integral to the act of the creative imagination? Perhaps. But the specific image, while it is integral to any one instance, might not play a role in another, similar, act. For example, Da Vinci could be imagining flying machines and draw another machine showing a man using mechanical wings (as he actually did in another drawing). At this end of the causal spectrum, it seems that the cognitive components of the act -- those involving mental causation (willful, conscious attention, and purposeful thought) -- are essential to the act. The act also involves high-order abstract concepts, such as thoughts about the properties of air, which require the engagement of high-order cognitive functions. It also appears that the task involves more "mental effort" than the other tasks we have considered, and the image has, as we have indicated, a secondary, or provisional status, within that task. Furthermore, imagining a new sort of flying machine involves linguistic components that would be essential to the task. In addition to the identifier "this image is of a machine that might fly" the entire task would involve sophisticated reasoning (linguistically processed) about the function and operation of the components of the imagined machine -- and these linguistic processes could be considered essential to the entire task of imaginative image generation.
We have only scratched the surface the considerations that may be brought to bear on these examples. For example, our diagram indicates the two-way nature of causal connections throughout the spectrum of mental and physical events. Some experiments indicate that merely thinking of a color can induce physical changes in the retina, so it may not be true that in every case of "seeing an after-image" only physical causes are involved. For more discussion, particularly in relation to the problem of epiphenominalism, see Advanced Philosophical Psychology.
Back to Mental Images: Introduction
Revised: 10:58 2/19/2005