In Defense of Indirect Realism

Indirect realism is an easy position to argue for. Simply consider the actual mechanics of perception (say visual perception). First there is the fact that objects never directly present themselves to our eyes, only photons reach us, and many things could have generated those photons, not just the object we think that we see. Even the information generated by photons striking the eye is not directly incorporated into consciousness. Our unconscious preprocesses this information, which we can deduce from the existence of optical illusions and the fact that concepts are present as part of visual experience (when you see a tree you don’t just have a certain visual sensation, you see a tree). Thus it is only information about the external world, transformed several times, that becomes part of consciousness, and what better word for it could there be than sense data?

But wait! Didn’t I give a defense of direct realism earlier? Which one is right? Well they both are. Yes both direct and indirect realism are correct, what seems like real differences between them are simply different ways of talking about the same phenomena. When you ask an indirect realist what experience is about he or she is likely to say that it really about the sense data that is part of consciousness. On the other hand when you ask a direct realist what experience is about they will say it is really about objects in the external world that cause our experiences. What we have here is not a real disagreement, but a disagreement as to what about-ness in perceptual experience is. In theory neutral terms an indirect realist uses “about” to designate the part of consciousness that is responsible for the perceptual experience, and the direct realist uses “about” to designate the causes of the experience.

So, given that direct realism and indirect realism are compatible with each other, and that their apparent incompatibility is a result of using the same word in different ways, where does that leave us? One concern that must be addressed is whether indirect realism naturally leads to idealism (or solipsism), because if it does, and if direct realism is not an alternative position but a complementary position, then it seems we might be forced to be idealists. The argument goes as follows: if our experiences of two people, or the experiences of a single person at different times, aren’t really about the same thing, (in indirect realist terms) then what is to prevent us from concluding that the entire world we know is simply a construction by our minds, and that there is nothing real behind it all?

What we need then is an argument against idealism that doesn’t rely on direct realism. That argument is as follows: We notice a constancy in our sensory experience; when we look at things from different angles, when we examine them with different senses, and when we observe them at different times, in all of these cases our senses give basically consistent information about the object that we are supposedly perceiving. The most likely explanation is that there is something causing this consistency, something that is not part of our conscious minds. We simply call that something “physical reality” (although it really could be anything, even a computer simulation). Additionally our experiences of “physical reality” seem to correlate with the reported experiences of other people, and once again this implies it is likely that “physical reality” is the same for all people. From this we can conclude that the most likely explanation (by far) for our experience is that there is something external to us that is responsible for the consistency of experience, something that we call reality, for lack of a better word.

Thus, with the threat of collapsing into idealism put aside for the moment, direct and indirect realism can coexist without problems. Of course the disagreement as to how exactly we should use the word “about” in philosophy hasn’t been resolved, but “about-ness” is complicated in its own way, more than the direct/indirect realism debate reveals, and thus its exact nature will have to be left for another day.

On Dreaming

Many philosophers (for example Descartes) have wondered how we can know that what we think of as our waking existence is not a dream. This is slightly different from the possibility of being a brain in a vat or being deceived by an evil demon, since dreams are products of our own minds. Believing that one is living in a dream is roughly the same as solipsism (at least some kinds of solipsism), and the other possibilities mentioned require something to exist outside of the thinker.

A simpler question to ask is “should we act as though life were a dream?” A little game theory reveals that we shouldn’t. We start be considering the four possibilities and their outcomes. If our lives are a dream and we act as though we are living in a dream everything is fine. Likewise if our lives are a dream and we act as though they aren’t a dream everything is fine as well. Of course if our lives aren’t a dream, and we act as though they were the outcome is disastrous. Finally if our lives aren’t a dream and we act as through they aren’t then everything is fine. We ask ourselves then how we should act given that we don’t know if we are in a dream or not. Well given that one course of action can lead to disaster and the other can’t we should pick the sure thing, and thus should act as though life isn’t a dream.

This doesn’t answer the fundamental question however, I admit. A traditional way to answer that question of how we can know that we aren’t in a dream is to rely on some capacity (such as pain, clear thought, … whatever) that is available to us when we are awake but not when we are asleep. Although we might believe that we are awake when we are asleep we never have that missing capacity, and thus having that capacity is sure evidence that we are awake. Let us say then that our waking capacities are {A, B, C} and that our capacities while asleep are {A, B}. The missing C then is our irrefutable evidence that we are awake.

What this answer misses however is the possibility that our whole lives (that we remember at least) are the product of some dreaming state that is unlike the normal dreams that we have at night because it does provide us with C. There is also the possibility that if we ever became fully awake there would be some additional capacity, D, which we don’t have now that would serve as evidence of the state of being “really” awake. Although the traditional response does rule out the possibility that our lives are not an ordinary dream it does not exclude the possibility of “super-dreams” such as this.

I argue that we can be sure we aren’t living in a super dream because of our limitations. I don’t mean our physical capabilities, but our mental capabilities, specifically those mental tasks that we can’t but that other people can. For example consider proving an obscure theorem in topology. I probably can’t do it, even after hours of effort, even if a great deal depends on me being able to prove it. However if I open up a topology textbook I can probably find a proof of the theorem and be able to understand that proof. If my life is a dream where did that proof come from? Clearly it wasn’t part of my conscious mind. Likewise the idea that somehow I unconsciously knew the proof all along seems unlikely, since I didn’t have any “proof intuitions” helping me out.

The only conclusion that is consistent with the evidence is that there is something outside our minds, something that is able to create things such as proofs and the other mental accomplishments that we ourselves are not able to do. Yes it is true that we might be deceived in a dream into believing that we had seen a proof when we had not, but if we have the actual proof in front of us there is no denying it exists, and thus no reason to believe that we are in a “super-dream”. Of course this doesn’t rule out the possibility that our entire experience is an illusion, but it does rule out certain kinds of solipsism, which is an important first step.

Are Rocks Conscious?

This post is divided into two parts. In part 1 I will discuss how various philosophical positions on the mind and consciousness would answer this seemingly ridiculous question (and perhaps their answers to it should cause one to doubt their validity). Part 2 is more serious, and focuses on showing why a description, for example a functional description, cannot be given to an arbitrary system, thus showing that rocks cannot be conscious under certain views of what consciousness is.

Part 1:

Dualism
Dualism is the position that the mental is a completely different thing from the physical. In some theories the mental is considered to be a separate substance, subject to its own laws, while in other theories the mental is separate from the physical because the mental properties are completely independent of the physical properties. In either case the physical and the mental are only tenuously connected at best. Because of this most dualists would be forced to concede that it would be possible for rocks to be conscious. After all, the mental substance or mental properties are not dependant on physical properties, so there is no reason to say that the rock could not be conscious. It’s hard to say if it is even unlikely for the rock to be conscious, because there is no test under dualism that could be performed to determine what is and what is not conscious. (If you believe that such a test might exist you are no longer a dualist.)

Idealism
Idealism is the view that everything is part of the mind. Idealists hold that the world does not exist independently of us because it is defined completely by our perceptions of it. Although idealism seems preposterous there are a handful of intelligent people who defend it, and thus it would be rash to dismiss it out of hand. Unfortunately under idealism it is nearly impossible to have certain knowledge about anything, since none of our perceptions have access to the underlying reality. Even under systems of idealism which hold that other minds exist independently of ourselves it is hard to identify what aspects of the world represent interaction with those minds. Thus under idealism it is possible that the rock may be one of the other minds, but you would never be able to find out if it was, so it is somewhat of an empty question.

Behaviorism
Behaviorism is a theory in which consciousness is identified with certain types of behavior, especially linguistic behavior about beliefs and thoughts. Under behaviorism we could say that rocks are not conscious, because they do not display any kind of behavior.

Identity theory
Identity theory is a position which holds that consciousness is identical with certain physical properties of the brain. Under identity theory a certain pattern of neural activity is the same thing as a certain thought. There are also less strict versions of identity theory in which certain classes of neural activity can be identified with a thought. At first you might think that this rules out the possibility of conscious rocks, since there is nothing in a rock that could count as the correct type of neural activity. However identity theory doesn’t rule out other kinds of consciousness, such as alien consciousness, that are identified with completely different physical features of the alien’s “brain”. Thus it is possible under identity theory that there exists a kind of rock-consciousness that can be identified with various physical properties of rocks.

Biological naturalism
Biological naturalism is the view that consciousness is a product both of our biology and the process of evolution. Biological naturalism can be rather difficult to understand so I will skip over the details. It is enough to say that rocks cannot be conscious under biological naturalism since the rocks present now did not come about as a result of an evolutionary process from earlier rocks, and thus it is not possible for them to be conscious.

Epiphenomenalism
Epiphenomenalism is the view that the mind is somehow caused by the operation of the brain. Unfortunately epiphenomenalism doesn’t say exactly how the brain causes the mind to exist, and thus it is possible that rocks somehow also cause minds to exist.

Functionalism
Functionalism is the view that consciousness can be identified with some operations of the brain. However functionalism holds that it is the nature of the operations (i.e. how they processes information) that is to be identified with consciousness (in contrast to identity theory which identified consciousness with some specific properties of the brain). In theory anything that performed the same operations would be conscious, weather it be human brain or computer. Since a rocks does not process information we can conclude that there is no functional description associated with them, and thus that rocks are not conscious.

Part 2:

There is one problem with my argument that functionalism shows rocks to lack consciousness, and that is the assumption that there is no computational description that can be given to a rock. In fact any theory that relies on descriptions needs have criterion determining when it is justified to apply a given description and when it is not. The problem is that without such a criterion it may seem that anything can be given any description. For example consider the a small section of the brain shown at several time intervals, where . is inert matter and – is an electrical charge.

...   ...   ...
-..   .-.   ..-
...   ...   ...

We might describe this situation as the movement of a signal from one part of the brain to the other. On the other hand consider a rock, where each . is an atom.

...   ...   ...
...   ...   ...
...   ...   ...

Since it doesn’t look like there is any activity me might think that there is no signal propagation. But consider the case where we suddenly decided to label one atom in each time interval as special.

...   ...   ...
#..   .#.   ..#
...   ...   ...

Now it seems that we do have a description of the rock that indicates signal propagation. We might extend this kind of description to the whole rock, at each instant distinguishing some atoms from the rest, so that when viewed as a whole we could give the rock a computational description that indicates it has consciousness.

To make the claim that the computational description of our brain is valid and the one of the rock is not we have to make some distinctions. First we can separate out the computational description from our physical descriptions, weather it be the charged/unchanged description of the brain or the special/non-special atoms description of the rock. We can say that given either of these descriptions the computational description is warranted, meaning that we now need to show why the charged/unchanged description is different from the special/non-special description. The key difference between these two is where the information that makes up each description comes from. In the charged/uncharged description the information comes from a rule and the physical properties of the system (the rule having to do with the ratio of electrons to protons). In contrast the description we are trying to apply to the rocks gets its information completely from the rules; the rules tell us what kind of description we need to justify a computational description on top of it, and so we label the atoms appropriately. We can then reject the second kind of description because it is independent of the physical basis, which is contrary to our intuition that a valid description must have a basis on what it is describing.

There is also a slightly more interesting, and difficult, case to ponder over, which may cast some doubt as to our criterion for what is a valid description. Consider a computational description of water, where the movement and interaction of water molecules is the basis for such a description (meeting our criteria). Usually the computational description is meaningless, but occasionally the bucket may be shaken in just the right way that for a few moments the computational description is meaningful, and possibly even indicative of consciousness. Is this really plausible? I think it might be, even though it is so improbable that it would never happen, but I will leave a more detailed analysis of this for another time.

Why Materialism is Preferable to Idealism

David Hoffman, author of Visual Intelligence (amazon), is preparing to present a paper detailing a new version of idealism, and I have had the privilege of reading an advance draft of that paper. I will not link to that draft here, out of politeness, and because I do not have permission, but it is sufficient to state that I cannot refute the formulation of idealism presented within it, and may it be possible to refute it at all. However Hoffman’s paper does not refute materialism either, and thus we are left with two theories, conscious realism (Hoffman’s idealism) and materialism, both of which purport to explain the world, and both of which cannot be right. To decide between them we need to rely on other techniques, specifically Occam’s razor and falsifiability.

First though let me give you a brief overview of conscious realism in comparison to materialism. I have included a couple of quotes below from Hoffman’s draft. I must stress that since drafts change it would be inappropriate to rely on these quotes for anything professional.

The argument for conscious realism follows from basically a single premise, that the content of our experiences is created by consciousness and not by “the world”. I wouldn’t dispute this point (although some philosophers would), since materialists such as myself would argue that “the world” is simply a meaningless arrangement of physical elements. From this he concludes that: “Something does exist whether or not you look at the moon, and that something triggers your visual system to construct a moon icon. But that something that exists independent of you is not the moon. The moon is an icon of your MUI, and therefore depends on your perception for its existence.” So far both idealism and materialism are in agreement. “The something that exists independent of your perceptions is always, according to conscious realism, systems of conscious agents. Consciousness is fundamental in the universe, not a fitfully emerging latecomer contorting the senseless face of matter.” This part of the conclusion is not warranted by the premises, because the trigger of the moon icon in my perception could equally likely be the meaningless physical description of the world as it could be interactions between conscious agents. However even though it is not warranted it is not necessarily wrong and thus our need other tools to decide between materialism and conscious realism.

Let us consider how the theories in question might answer a question such as the following: “Why do we all perceive the moon in basically the same way?”. Materialism says that there is a single real physical reality, which affects us all in the same way, thus resulting in a similar concept of moon being triggered in us all (although not the same concept, for example some people see the face and some the rabbit, we must assume the trigger is modified by the physical make-up of the brain). Hoffman would have to answer this by asserting that all the conscious agents are interacting, constantly, to ensure that they are all triggered in a similar fashion, and that this interaction is preserved throughout time, and that the interaction would be modified if something happened on the moon to change its appearance, maintaining the illusion of a single physical reality. You can see then why we might prefer materialism because of its simplicity, for in Hoffman’s view there must be interaction between conscious agents concerning every aspect of “the world” that they agree on, and that these interactions must be altered appropriately if there are changes in “the world” in a way that makes the world seem consistent and independent of us. Thus Occam’s razor would lead us to materialism. Additionally materialism has the ability to be falsified while conscious realism does not. (i.e. we would conclude that materialism was false if different concepts were triggered by the same material substance when there were no differences between the minds in order to explain why they were triggered differently, such as past experience.) For these reasons I am compelled to accept materialism as a working hypothesis over conscious realism.