First Philosophy or Scientism?

The right way to proceed, according to Cummins, is to "pick a theoretical framework and ask what explanatory role mental representation plays in that framework and what the representation relation must be if that explanatory role is to be well grounded." In other words, one takes a theory such as orthodox computationalism and then one asks: what must the nature of mental representation be if this theory is to be both true and explanatory?

So the question of mental representation is not a question of 'first philosophy,' i.e., a question to be settled independently of, and prior to, empirical research, but a question in the philosophy of science exactly analogous to the following question in the philosophy of physics: what is the nature of space given that General Relativity is true and explanatory? The properties of physical space are for physics to determine. Philosophy's role is correspondingly modest, that of a handmaiden. To coin a phrase: philosophia ancilla scientiae, philosophy is the handmaiden of science, similarly as it was the handmaiden of theology in the Medieval period. Philosophy becomes the philosophy of science. And according to Quine, who is quoted by Cummins on the fontispiece of his book, "Philosophy of science is philosophy enough."

It ought to be the same, Cummins thinks, with psychology. Whether or not there are mental representations becomes the question whether or not the best theories in cognitive science and psychology posit mental representations, and the nature of the representation relation is to be read off from whichever theory is taken to be true and most explanatory.

But there is a problem with this view. If we want to know about physical space and the nature of matter, we turn to the physicist. And if we want to know about the brain, we turn to the neuroscientist. But if we want to know about the relation of mind and brain, we cannot base ourselves solely on empirical science.

Here is one consideration. The extant empirical theories imply the existence of mental representations. But surely their existence is not obvious. Looking at a photograph of a mountain, I become aware of the mountain and some of its features via the picture. Here it makes sense to say of the picture that it is a representation of the mountain. But when I look directly at a mountain, there is no phenomenological evidence of any epistemic intermediary or representation: I see the mountain itself. I don't see sense-data or representations or any kind of epistemic deputy.

Perhaps I will be told that I am nonetheless aware of an internal image but not aware of being aware of it. Compare the case of walking into a room and seeing a depiction of Hillary Clinton so realistic and convincing that I take it to be Hillary herself. In such a case I am aware of an image, but not aware of being aware of an image. Why couldn't it be the same with outer perception? One crucial difference is that I can come to be aware of the Hillary-image as an image. But I cannot come to be aware of any supposed internal image that mediates my outer perception. This is particularly obvious if the internal image is a brain state. I cannot, while gazing at the mountain, 'focus inwardly' and become aware of the brain state that is supposedly mediating my perception of the mountain. Given this fact, I suggest that it is unintelligible to say that there is an internal image that mediates outer perception. The word 'image is' being misused. An image that I cannot become aware of as an image is in no intelligible sense an image of something.

No doubt the brain and its states are part of the causal basis of perception. And there is no doubt that an organism's brain is in (some) of the states it is in because of what is happening in its environment, e.g., bright light is being reflected from a snow field into the organism's eyes. But to say that some of the brain states represent the environment makes no sense assuming that 'represent' has the sense it has when we attend to the phenomenology of representative consciousness. To speak of material representations literally in the head is to read back into a third person conception of the world notions that make sense only from a first-person point of view.

You may agree with what I just said or not. But the discussion we will have about these matters surely does not belong to any empirical science. It belongs to first philosophy. The question of whether there are mental representations at all, for example, cannot arise within disciplines that presupposes their existence. And the relation between a first- and third-person view of the world cannot be treated within an exclusively third-person point of view. Finally, there is the point that the claim that science alone can clarify these questions is itself unscientific and so an instance of (negative) first philosophy.