More about the content of perceptual experience

 

It’s really amazing how many different kinds of information we can get using our familiar five senses. Take visual experience. When you look at charts in a newspaper you can see how the economies of distant foreign countries are doing, or how much sugar has cost over the past ten years. When you look at your computer you can see how much of some program it has downloaded, or what percentage of the different colors of light some favorite shade consists of or which of your friends are at their computers. If you were a soldier with special vision-scopes you could learn even more exotic things using your sight like how heat or cold things in your environs were or which objects were in range of your guns. I think there are even computer people (the ones who design user interfaces) whose whole job consists of coming up with ways to pour more and more different kinds of information into the familiar old senses of sight and hearing in a convenient and intuitive way. 

Now, not only is there a sense in which you can ‘see’ all these things e.g. that there is a warm object within shooting range, that sugar has been getting cheaper, that your friend is online, but (I claim) there’s no firm distinction between the sense in which one can see how tables and chairs in one’s environment (or the facing surfaces thereof) and the sense in which one can see all these exotica. Actually, the idea of a firm distinction between basic information conveyed by the senses and further information which we don’t see but infer on the basis of what we do see is so unfashionable that I probably don’t need to do too much arguing against it. I’ll just give one example from the borderline territory. Take looking at two objects which are the same size, but one is one is close and the other is far away. Do we see that the two objects are the same size, or do we see that they look different sizes and correlate this knowledge with our perceptions of distance and infer that they are the same size? To the extent that there is a difference between ‘seeing that P’ and ‘inferring on the basis of what you see that P’ I’m inclined to think that it’s a matter of a particular perceivers habits and occurant thoughts. When you are first trained to spot poison ivy you might look at it and directly believe that the plant in front of you has a certain number of leaves, shape etc and then go through a checklist and infer that it was poisen ivy. Later, after a bit of (hopefully not to itchy) experience you might look at the plant and directly see that there was some poison ivy over there.

I’m telling you all these platitudes in order to explain my motivation for thinking something which seems to be much more controversial about the content of experience. This is that, if you just consider what it feels like to have a perceptual experience, this feeling on its own doesn’t have any kind of content. What gives a perceptual experience content is not a matter of how it feels to have that experience but rather of what role experiences of that type play in the cognitive economy of the perceiver and their interactions with the environment.

The phrase ‘cognitive economy’ in the above is admittedly a bit slimy, so let me put things more concretely before I go on to say why I think the platitudes above support my controversial thesis. You have a certain experience when you see red things, an experience which we might say represents the presence of red objects in one’s environment. Now my claim is that what gives this experience is content (that there’s red stuff around) isn’t how it feels to you to have the experience but rather how that experience is usually caused in you and what kinds of consequences it usually has. The experience is, for example, typically caused by red things, and it typically results in the storing of beliefs that there are red things around, and in your going back to the area where you had that experience if you are ordered to pick up something red. We can see that this is true by thinking about what you would say if we fiddled with your brain or put color lenses on you which switch around the expereinces of red and green (as in the classic ‘inverted spectrum’ problem). At first you might say that you ‘saw that stuff looked green’ and then inferred that it must really be red, but eventually (just like with the poison ivy) you would directly look at sunsets and ripe tomatoes and conclude that they were red without any indetermediate inference. I –and, I think, most normal English speakers- would call that seeing that they were red. Thus, by considering what we would say if the phenomenal aspects of an experience were changed while its cognitive role remained the same, I think we can tell that what makes your experience an experience of seeing that there is something red (i.e. what gives it the content that it has) is not how it feels but its cognitive role. So, if you just think about phenomenal feels these don’t have any content. Only once we plug these feels into a particular cognitive economy where they are reliably caused by certain things and lead to certain kinds of information storage and behavior modifications can we speak of them as inheriting content from their role within a particular person.

This controversial thesis that phenomenal feels on their own don’t have content/truth conditions/veridicality conditions has a little to do with something more familiar, namely disjunctivism about content. The disjunctivist would agree that an experience’s content depends on something more then how it feels to have it. But they think that the ‘something more’ is how the particular experience in question is caused (by  a horse or by a cow-in-the-night?), while I think the ‘something more’ is how that kind of experience is usually caused and what you tend to do with it. The dialectic here is even a little bit more complicated since the disjunctivist theory of exactly how the content of an experience depends on external factors about how it was caused uses a distinction between veridical and non-veridical ways of having a certain experience-feeling which my view says is incoherent. Specifically, the disjunctivist wants to say that the class of phenomenally indistinguishable experiences which can be caused either by a horse or a cow at night has certain veridicality conditions associated with it, and that these veridicality conditions determine the content of each of the experiences belonging to the class. For example, the condition might be that for one of these class of experiences to be veridical it has to be caused by a horse. Then the fact that one of these experiences caused by a horse is veridical and another, caused by a cow, is not will help determine different contents for these two experiences: in one case that there is a horse in front of one and in the other case that there seems to be a horse in front of one. Thus, inspite of my superficial agreement with the disjunctivist that the content of an experience depends on more then what it feels like to have that experience, our two views are deeply incompatible. For while I think that specific phenomenal feels have nothing associated with them, not content nor truth conditions nor veridicality conditions, the disjunctivist thinks that they do have veridicality conditions which play a role in determining the content of an overall experience.

Now, why do I think that the platitudes above motivate this view that phenomenal feels (or, if you like, classes of phenomenally indistinguishable experiences) don’t have anything like content/truth-conditions/veridicality conditions on their own? The first point, remember, was that our senses can give us information about a wide variety of different things. Now a slightly strong, but I think still plausible, point is that phenomenally very similar experiences can give us radically different information depending on the set up surrounding them. The same blotches could tell a person with heat-sensitive goggles that the enemy is approaching, or someone looking at a colored chart that the geographical distribution of a plague was receding or someone looking at a close up picture in an online newspaper that their abstract artwork had been ripped off etc. The second relatively uncontroversial point above was that the distinction between what you can see and what you can infer by seeing is not a firm line, but more a matter of habit and training (like what happens with the poison ivy). Finally, if we add a third claim that the content of an experience is what you can see by having that experience then we can put these pieces together as follows. The same (or indistinguishable) phenomenal experience can let one infer lots of different stuff depending on what context one has that experience in. And the difference between inferring something based on e.g. a visual experience and seeing that thing is merely a matter of familiarity and habit. So not only could one infer lots of different stuff from having the same phenomenal feel depending on context, but having that phenomenal feel could let one see lots of different stuff. And if the content of an experience is what having that experience could let one see, then there won’t be any interesting facts about the content of phenomenal feels. For there’s almost no limit to the different things which (in a suitable context and suitable training) that experience could let one see.

And, by the way, I think the same thing applies to veridicality conditions. For, I gather that the idea of veridical and non-veridical experience goes something like this. You have an experience as-if-of-a-horse, that experience will be veridical if it is caused by a horse, and no veridical if it is caused by a cow in the dark. The veridicality condition associated with that class of experiences (that there is a horse in front of one) is exactly what having one of those as-if-of-a-horse experiences could let you see, namely, that there’s a horse. So intuitively speaking the veridicality conditions associated with a class of phenomenally indistinguishable experiences should be the conditions which having one of those experiences can allow you to see obtain. But if, as argued above, there are no interesting limits to what you can see by having a certain phenomenal feeling/having one of a class of phenomenally indistinguishable experiences, then there will be no such veridicality conditions.

In this little essay I have only described my positive motivations for thinking that phenomenal experiences don’t have truth or veridicality conditions on their own, but only inherit these from being embedded in a perceiver’s cognitive economy in a certain way. I haven’t addressed any arguments against this view, which say that phenomenal experiences must have already have content on their own. This is unfortunate because there are lots of interesting arguments against my view, and lots of people who disagree with me would agree that there are prima facie reasons to think that phenomenal feels don’t have contents but think that the above mentioned arguments ultimately carry the day.

So, the next task is going to be to consider some of these arguments that phenomenal feels must have content: that we need to invoke such content to make sense of the dependence of experience on expectation, optical illusions, the evolutionary history of perception or the way that experiences can justify beliefs.

2.

In the section above I have claimed that through the results of training, special goggles, spectrum inversion and the like the same experience can acquire a radically different role in a person’s cognitive economy and a radically different content. And I have cited this as positive reason to think that what gives an experience its content is not how it feels to have that experience but the role which that kind of experience plays in the conginitive economy of the perceiver. Now, a person might be inspired by psychological and phenomenological research to make the following objection to this claim. What a person’s experience feels like depends on a lot more then the patterns of light which hit their eyes. In particular it’s well known that a person’s previous expectiations, and familiarity with certain kinds of objects can make a subtle but noticeable difference to what they experience. So you might just deny that in the cases which I have mentioned people really have the *same* experience. Perhaps the very same changes which train a person so that e.g. they can immediately see poison ivy also subtly change their phenomenal experience when they look at plants, so that the perilous leaves visually ‘stand out’ more or the like. And maybe this change in phenomenal feel is what gives the experience a different content.

So, what I’m going to do in this section is contrast two theories about how a certain kind of experience can acquire different content and argue that the theory on which phenomenal feel doesn’t determine content should be preferred. Let’s start by just describing the difference between the two accounts. Both accounts agree that roughly the same type of experience e.g. the kind of color experience you now get from fires and ripe tomatoes could acquire a different content if you wore spectrum-inverting lenses for a while and got used to it. But they disagree about how this would happen. On my account, as you adjusted to the fact that red things now looked to you the way that green things used to, and got into the habit of using your new perceptions to determine whether tomatoes were ripe or traffic lights allowed you to go, the kind of phenomenal experience which you now got from green things would acquire a new role in your cognitive economy, and in virtue of that it would count as having a different content. On the alternative account the same kinds of changes would occur to how you used that experience, but these changes to your background expectations would cause your experience to subtly change and it would be in virtue of _this_ subtle difference in the experience would determine that it had a new content. As I have indicated in the little chart below, both views agree that the processes which give some broad experience-type a new cognitive role can cause a subtly difference in how it feels to have that experience. But the accounts disagree as to whether an experience-type acquires a different content in virtue of the subtle change in phenomenal feel or of the new cognitive roll.

 

Background changes =>new cognitive role = new content

                                  => subtly different experience

vs.

Background changes => new cognitive role

                                  => subtly different experience = new content

 

Common causes so its difficult to see just what makes the difference

Some cases where there doesn’t seem to be an experiential change

            -can say that there *is* a change it’s just very subtle

            - can say that when there isn’t a change the facts in question never become parts of experience’s content

            Can also consider what we would say about alternative possible worlds where the psychological law connecting the two is broken.

            Easer to explain talk about experiences in terms of cognitive role account than to explain anything on the experience account. Also its easer to explain learning, etc.