In a talk, I mentioned that it is impossible to point to oneself; one always points to another person. There is an objection to this argument, namely, that one can point to oneself in a mirror. There are ways to respond to that critique that are mainly of a biological nature, to do with the relatively late evolutionary history of self-recognition in mirrors. However, there is a more fundamental metaphysical question: who does one see when one sees a reflection in a mirror? In some ways, a mirror is a special kind of visual illusion. Here, there can be two lines of argument:(a) That visual illusions -and mirror images in particular- teach us a lot about the workings of normal perception. This is the dominant line in philosophy as well as in cognitive science.(b) The heterodox Gibsonian view, and the view consistent with my “action potential” theory of mind is the exact opposite, that illusions do not tell us much about the normal operations of the mind. If so, mirror images are no more problematic than mirages.
I could make (b) stronger by doing a little bit of ordinary language philosophy and say that the verb ‘see’ is being used in particularly slippery ways when we say
(i) I see a bird (flying in the air above me).
(ii) I can see myself (in a mirror).
Cant resist punning; I can already see a future paper called “What we say about what we see, or why we shouldn’t take reflections at face value.” Recent work in cognitive science suggests that babies learn to move their bodies and make facial gestures by imitating others. Therefore, the ‘original face’ is not my own, but my mother’s. Therefore, the mirror image, is only contingently ones own face, i.e., it could be someone else’s, it just happens to be my own. There is no logical necessity that I see myself in a mirror. Therefore, I know that I have a face and I have eyes (potentially, if not actually, in the Aristotlean sense) because I have a more primal bodily sympathy/empathy with others.

To the extent that I empathize with the person I see in the mirror, I see my face. The self-image is not the self. The same goes for self-percept, self-concept etc. A Buddhist style of argumentation might say that there is nothing more to the self than self-images, self-percepts and self-concepts. If that is correct, then the mirror-image self is as much of a self as the self from the inside out. However, if there is more to the self than the self-image, we could argue that the self is more than its images and reflections.There is an interesting see-saw battle between the reductionist who wants to get rid of the self all together and the holist, who says that there is more to the self than its images.

A heavy use of the mirror self as an intuition pump leads toward an internalization of reflection, from mirrors in the world to our own capacity to hold thoughts, with the ultimate aim of reducing the self itself to a series of images. I, on the other hand, am trying to externalise thoughts and images, saying that these are projections, in between the world and the self, but are neither one, nor the other. To conclude:

(a) We can expand the line of inquiry initiated above to perceptions as a whole, and not just reflections. The reductionist -the Buddhist, for example- argues that perceptual objects are nothing more than an agglomeration of experiences, which in turn are nothing more than images, sounds etc. In other words, if the self is nothing more than the sum of reflections, by symmetry, the objects of the self should also should be nothing more than a sum of perceptions. This is what I should have said earlier, when I said that mirrors and mirages have the same function; if mirrors can be used to understand the self, mirages can be used to understand perception. To the extent that the self and its objects are not images stitched together, we can neither reduce the self to a sum of reflections, nor can we reduce an object to a sum of views.
(b) Introduce the idea of a-thing-from-its-own-side as a refinement of what I was calling the self from the inside out. Is the self “a-thing-from-its-own-side,” which, for example, is that from which we point, rather than what we point at. Like the snake trying to eat its own tail, it might be possible for the pointer to be the pointed, but one has to be very hungry before consuming oneself.