Hesperus is Bosphorus

A group blog by philosophers in and from Turkey

Seeming to see: a doxastic theory of perceptual sensation

with 9 comments

How does blue show up in the experience of blue, if it’s not a blue quale? Perhaps in the same way blue is present to you when you’re thinking of something blue.

I am going to imagine that every perceiver omits the step that leads from (1) the neurological effects of sensory stimulation through (2) the presence of a quale to (3) belief about what one is perceiving. In the absence of (2) I am going to propose that to have a sense impression is to acquire a belief. But let me be more precise.

I’m going to propose a doxastic theory of sensory states. (doxa: belief, in Greek.) On this theory, seeming to see something blue is becoming disposed to believe that one sees something blue. It’s analyzed in terms of inclinations to believe things about what one is directly seeing.

(1)                    S seems to see something F iff S is inclined to believe that he/she directly          sees that something F is there.

The theory interprets ‘It seems to me that …’ as ‘I am inclined to believe that …’. This expresses a familiar propositional attitude. And one ordinary use of ‘it seems to me’ is indeed to speak about what one is inclined to believe: “It seems to me that all politicians are corrupt.”

I will call your attention to three primary features of the doxastic theory. First, the blue which seeming to see involves is real, genuine blue—the blue of the sky, the blue of my jeans, and not an ineffable blue indescribable in any public language. Secondly, absolutely nothing is blue when I only seem to see something blue. Thirdly, the blue is internal to the object of my experience.

My experience is an attitude (it seems to me that) described in terms of a propositional object (I see something blue). The blue is an aspect of the proposition, a concept internal to it.

The doxastic theory agrees with qualia theory that sensations are private events: (1) only I can have my sensations (as only I can have my thoughts), and (2) I may well have privileged epistemic access to my sensations (something to investigate later).

One sign of the success of a doxastic theory within a functionalist picture is that it solves, or dissolves, the eternal problem of the inverted spectrum.        Still, questions naturally remain. Is the problem of the inverted spectrum a by-product of bad philosophy as Wittgenstein might think, or does it have deeper roots so that it can trouble even non-philosophers? It’s a good question and I don’t know the answer.

We may see the reasonableness of (1) from the failure of two simpler proposals.

(1´)                  S seems to see something F iff S is inclined to believe that something F is there.

(1′) fails to distinguish perceptual and non-perceptual ways of acquiring belief. You may become inclined to believe that something black is there because you hallucinate a cat or because someone tells you something black is there. Only in the first case is there a visual sense impression.

(1´´)                 S seems to see something F iff S is inclined to believe she/he sees that something F is there

As Berkeleysays, you can see either directly or via inference and prior belief. For example, you look at the mountain top and directly gain the belief that something is white: you directly see that it is white. At the same time you indirectly gain the belief that something is cold, via your inference that the white stuff is snow and your prior belief that snow is white. You have a sense impression of white, but you don’t have a sense impression of cold.

In order to complete this presentation of a doxastic theory I’ll need to sketch the relation between seeming to see and actual seeing—both seeing that something is the case and seeing an object. Here’s my analysis of propositional seeing.

(2)                   S sees that p iff the fact that p causes, in the proper direct manner, S to believe that p

(2) is merely the structure or schema of an analysis, because it does not clarify the manner in which the perceived fact must cause the belief. This manner of causation will have to satisfy several conditions:

(2a)      It distinguishes perceptual ways of getting belief from rumors, theorizing, and other non-perceptual ways.

(2b)     It is such as to justify S’s belief to such an extent that the belief qualifies as knowledge. (Remember that seeing that p is a species of acquiring knowledge that p.)

(2c)      It requires proper use of the eyes, which is necessary if perception is to qualify as seeing rather than some other sense modality.

Finally, I analyze the seeing of objects, in terms of propositional seeing.

(3)                    S sees x iff for some property F, S sees that x is F

In the light of (2), (1) turns out to analyze seeming to see as an inclination to believe things about causation and one’s own beliefs. Having a visual sensation will require both a kind of self-consciousness and a capacity for second-order belief: belief about one’s own beliefs.

So seeming to see is deeply cognitive and conceptual. One who seems to perceive is not sensing sense-data or having experiences with qualia. There are no representations that are really F. We may say, though, that seeming to perceive is an act of representing—to be precise, representing that one directly sees that something is F.

These analyses make seeing that p the fundamental concept of perception. It is both conceptually and temporally prior to seeming to see and to the seeing of objects.

I consider these analyses to be elements in a theory of direct perception, both because there are no intermediate objects of perception or awareness, such as qualia, and because the theory makes perceiving facts about real objects prior to having perceptual sensations.

Written by shvoss

April 12, 2012 at 2:51 pm

Posted in Uncategorized

9 Responses

Subscribe to comments with RSS.

  1. HI Stephen,

    Interesting post. As you know, I’m also into defending a direct realist theory of perception, except I think that there is space, and perhaps a need, for sensations in such a theory. In particular – I think that in our direct perception of colours sensations play something like a reference fixing role. So I think that our colour CONCEPTS can be thought of as something like definite descriptions – for example “the cause of this sensation” but with such descriptions being used referentially, not attributively. Such descriptions can be thought of as fixing the reference of our colour concepts, but don’t contain any information about the nature of colours. According to such an account sensations (which I think, in your language would be qualia), play an essential role in the formation of our colour concepts (and hence in our beliefs about colour) but are not the immediate objects of perception.

    Anyway – here’s my question. I assume on your account that beliefs about colours contain colour concepts. I wonder how these colour concepts work on your account. How is their reference fixed? How are they individuated? How, on your account, could we learn such concepts?

    Lucas Thorpe

    April 16, 2012 at 12:52 pm

    • On your view, Lucas, evidently I understand red as that property of things that causes “this sort” of quale / sensation in my mind. Your view is that I understand red only in causal terms and it remains open which property plays this causal role.

      This view strikes me as entirely consistent with Locke’s. Locke holds that I conceive red first of all as a kind of idea in the mind. Then I formulate the concept of a “quality” as a power in a thing to cause an idea. This allows me to formulate a second notion of red, namely as the power in a thing to cause ideas that are red in the primary sense of the word.

      On such a view I must, logically, have the notion of a red quale / idea before I have the notion of a red thing, since I’ve defined the latter in terms of the former. I must then be more directly aware of red as a quale than of red as a property of things.

      That suggests two immediate critical thoughts to me. First, that such a theory can’t be called direct realism. Second, that any theory that holds that there are intrinsic qualities of our sensory experiences of which we are directly aware is subject to a raft of paralyzing objections. Only one of those objections is that it makes my knowledge of red objects inferential and derivative from my knowledge of red qualia.

      Your turn!


      April 16, 2012 at 2:31 pm

      • Hi Stephen, this will have to be quick. But:
        (1) I don’t want to say that redness (as an objective quality) is to be identified with “a power in a thing to cause an idea”. On my view (which is influenced by Thomas Reid) we use the causal power of the quality to fix the reference – but I don’t want to identify the quality with this causal power. I think we do not know what ‘redness’ is. Perception does not entail that we know the essence of the thing or quality perceived, but merely that we are able to distinguish one thing from another. This is a topic for scientific research. Given what I know of colour science, I suspect that (objective) colours are extremely disjunctive qualities.

        (2) I don’t think the qualia (or sensation) is an idea. The qualia is not what we perceive – but how we perceive.

        Lucas Thorpe

        April 16, 2012 at 2:40 pm

        • Hi once more!

          I don’t think that these clarifications remove any of the sting from my objections. Here’s why.

          1. Instead of saying a color is a power to cause a quale, you’d like to sayit’s a quality that possesses the power to cause a quale. Either way, our understanding of the quale is logically prior to our understanding of the color. Indirect realism.

          2. You’d like to say the quale is not a Lockean idea but a way we perceive — sometimes we perceive “redly” for example. Either way, given the definition of quality in terms of quale, we must be capable of realizing that we’re perceiving redly as a precondition of being capable of realizing that we’re perceiving something red.

          3. I believe your formulation lets us formulate the central question: What is it to perceive in a “red” way, “redly”? And which comes first, redly or red?

          4. Whether one is a direct realist will depend on which one believes is conceptually prior — to perceive in a red way or to perceive that a thing is red.

          5. With Sellars, I hold that the quality red is prior in that our primitive perceptual capacity is to recognize that some things are red and some things aren’t. Then, noticing that sometimes things look different than they are, we arrive at a notion of a way of perceiving — which is often but not always found in perceiving that things are red and which is defined in terms of perceiving that things are red.

          6. Back in my original post, I defined “S has a red sensation” in terms of the more basic concept of red. You could say that this was my stab at defining “S is perceiving redly”.


          April 18, 2012 at 7:22 pm

  2. Hi Stephen,
    Let me just comment on one of your points. You say [and I’m numbering your sentences]:
    ” (1) Instead of saying a color is a power to cause a quale, you’d like to sayit’s a quality that possesses the power to cause a quale. (2) Either way, our understanding of the quale is logically prior to our understanding of the color. (3) Indirect realism.”

    I’m not sure that this is how I would put it – but let me accept (1) and (2). Although I don’t think I’d use the word ‘understanding’ in (2). I don’t think we ‘understand’ the quale, nor do we ‘understand’ the colour. So instead of (2) I guess I would rather say something like: ‘it is the fact that our perception has a qualitative aspect that allows us to distinguish different (objective) colors” (and this a fact about us, not a logical fact about perception in general – perhaps there are beings who can perceive colors but do not have sensations).

    Anyway – let me give you (1) and (2) – I don’t see why (3) “indirect realism” follows. Like Reid I think that our CONCEPTION of something can be relative, but our PERCEPTION can be direct. In perception we immediately grasp objective differences between qualities of objects using these (relative) concepts. Why does this fact about colour concepts imply that our perception is indirect?

    Lucas Thorpe

    April 18, 2012 at 7:53 pm

    • Hi, Lucas.

      I have the feeling we are closing in on the nature of perceptual sensation. So I’ll try to open the way for more progress.

      First to clarify my present position. There are certainly sensations — cases where a person seems to perceive in a certain way. All the same, to perceive that x is F does not require one to have a sensation. For having a sensation is, I think, simply becoming inclined to believe you’re perceiving.

      What happens in perception is, first, the presence of the object to be perceived; then, lots of complex physiology; and then as a result the acquisition of the belief that there’s a red Porsche over there. Given only that the causal chain meets the proper conditions [see my discussion of (2) in the original post], the acquisition of the belief is the seeing.

      It seems to me that if by “red” I understood a quality somehow related to a sensation, or a qualitative aspect of my having a sensation, I would need a conception of that quality or aspect in order to gain whatever conception of red I manifest in coming to believe it’s a red Porsche. And that would require awareness of that quality prior to any awareness I have of the color of the Porsche. That looks like indirect realism and I believe that’s not for you.

      But in your last reply,you suggest an alternative idea — that my perception’s possession of a quale is a necessary causal condition of my having the awareness of red that I have when I see that the Porsche is red.

      If so there are two possibilities. One is that my awareness of that quale is also a necessary causal condition of my seeing that it’s red. Given what we said 2 paragraphs back, the awareness will evidently be a non-conceptual awareness, and there will be no need that it lead to a conceptual awareness of the nature of the quale. I wonder if this would be the strongest version of your view?

      The other possibility is that what’s needed for perception of the color of a car is only the existence of a quale but not my awareness of it. I am glad to agree to this. For now we are only discussing the need for a a common physiological source of my perceptual beliefs that something is red. This quale will play no role in a philosophical theory of perception since we have no consciousness of it and since it’s simply something postulated to fit the physiological part of the picture. It might turn out to be the flow of electrons from neuron 7349 to neuron 7320.


      April 20, 2012 at 10:08 am

  3. アバクロ 半袖シャツ

    アバクロ 通販

    September 12, 2013 at 4:02 am

    • Iustusmaxima has a beautiful point. Can’t see any reason to deny it!

      My wish is to show that the idea of perceiving an object can be understood in terms of the doxastic idea of perceiving that an object has a property. Can’t be done in the simple way I had proposed. Can it be done at all?

      We might suggest: S sees x iff S perceives that an object has some property or other AND the belief this involves [see my earlier analysis of ‘S perceives that p’] is a belief about x having that property.

      This raises 2 questions. First, is the suggestion itself sensible — on the right track? Second, even if it is, can we explain what it is for a belief to be a belief about a certain object?

      The answers I think are right are: Yes; and: This is pretty difficult!

      I believe that the man over there is a spy. The man over there is Quine but I have this belief because I think the man over there is Kripke and I’m sure Kripke is a spy. Is my belief about Quine or Kripke?

      The true God is just and merciful. I believe God is unjust. Is my belief about the true God — just a mistaken belief about that God — or is it a belief about an idol?

      I see a shadowy figure outside my window. I have some suspicions about who it was. They are all mistaken and go nowhere. Is my belief about anyone? — Remember, I did see someone!

      Hard questions. In the end, we might conclude: not just hard, impossible! And that would doom my new suggestion about what it is to see an object.

      iustusmaxima has done well. Where can we go from here?

      Stephen Voss


      November 17, 2013 at 2:09 pm

  4. eresting, but I would like to raise an issue about a minor part of it, namely your analysis of:

    3) S sees x iff for some property F, S sees that x is F

    I believe this gives wrong result in terms of inferential relations. What I mean by that is that if x and y are identical than from S sees x it follows that S sees y. Therefore(A) S sees x is an extensional context. But (B) S sees that x is F is an opaque context.(I cannot think any way to argue against A and B, and one can give plausible and elementary motivations for adapting them) So from that x and y are identical and that S sees that x is F it does not follow that S sees that y is F. Therefore this analysis is wrong.


    November 17, 2013 at 2:12 am

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: