I have some books on visual perception in which the retinal image in the human eye is described and represented (in pictures) as a "picture" or "visual representation" of the external world. This seems very misleading to me. If it is meant as a way of "illustrating" the retinal image, this should be made clear, but it isn't.

E.g., it should be made clear that, if an illustration shows a picture of a tree on the image surface of an eye or pinhole camera (and there are plenty of these out there), the green foliage area represents an area of "visible" EMR of some intensity and spectral composition incident on the surface, and the brown trunk represents an area of a different intensity and spectral composition.

This seems to be misleading, in fact it elides the amazing nature of this "thing" called the retinal image. The vision scientist J J Gibson was pretty scathing about the conception of the retinal image as a picture, writing that the "image" in the eye "is not something that can be seen."

I can't pretend to give a good definition of the retinal image, but it's a sensory stimulus, not an object that is looked at (such as a tree or a picture of a tree). I think of it as a continuously replenished pattern of incident EMR, or photons, across the photoreceptor layer (specifically at the plane of the outer segments, where photons are being absorbed), each point of which consists of EMR coming from a single direction in the visual field (i.e., a real optical image). "visible" EMR of course - but there's no perceiver perceiving it as light, or as a luminous picture.

Is all this valid?

  • $\begingroup$ Yes, the homunculus fallacy, thank you. As an amateur very interested in vision in relation to art (i.e., pictures) who has looked at quite a few books titled "Sensation and Perception" and the like, as well as books on optics, when I say "misleading," I'm saying that they slip onto that fallacy and lead us students into it. $\endgroup$
    – Tony C
    Oct 22, 2023 at 4:18
  • 1
    $\begingroup$ I would certainly agree that far too much time is spent in rudimentary psychology/neuroscience education on "flipping images" at the retina, etc; you won't find the same emphasis at a higher level of education, but it takes a lot of additional background to explain what a retinal representation actually is as well as how representations are distributed across the brain; some of the final assembly steps eventually get you to the hard problem of consciousness so it's also not quite as if it's all understood, either. $\endgroup$
    – Bryan Krause
    Oct 22, 2023 at 14:33


Browse other questions tagged or ask your own question.