Can Computers See?

Can Computers See?


Gregory Koukl

If they can, does this refute Christianity?

divider

I had an interesting question asked of me last week. It had to do with what philosopher's call the mind/body problem. I answered that question with an illustration and I have gotten a response in the mail to the question that was raised. I want to spend some time responding to this because it really helps us to work through this issue. It helps to make a case, I think, for the existence of the soul, which is very important.

Now the mind/body problem relates to a question: Are we simply a physical body and that's it, or are we a physical body that houses a soul and the soul is what animates the body--it gives the body ultimate life, gives it personhood, gives it identity, the soul is who you are?

The mind/body problem is an important one even for those who are not philosophers because this is not simply an academic issue. What is at stake is the existence of the soul. Some will argue that all that exists is the physical world. Non-physical things like souls simply do not exist. A physicalist would give this assessment of a human being. He would say that all so-called mind activity can be exhaustively explained in physical terms. It can be reduced to physical activity in the brain--C-fibers firing and chemical reactions. I heard one talk show host say not long ago that, after all, an emotion is just a chemical reaction. Well, this plays into this notion that we are only our bodies. Since everything can be explained in physical terms--in terms of the physical activity of the brain--there is no need to posit this eerie, mystical, unscientific, and religious thing called the soul.

Why is this important? It is important because there is something else going on here. What is going on is the argument that if there is no soul then there cannot possibly be any life after death. Think about it. What is life after death? It is when the body dies and the soul goes on. After the physical body dies there is something that remains and continues living. But if the physical body is what you are, then when it dies, you die. And if someone can demonstrate scientifically, without a doubt, that human beings can be exhaustively quantified in physical terms--that there is no mind separate from the brain, that there is no soul separate from the human body, that the soul is just a word we used to use to describe something we thought was there but now have learned that it is not there at all, that there is only the brain--if that can be demonstrated to be true, the game is up for Christians. There is nothing else to discuss, nothing to preach, because why talk about salvation for eternity if once we die we're gone? That is why there is tremendous philosophical currency riding on this particular question. That is why it is very important.

I used an illustration last week to disprove the notion that we are just our bodies and that all soulish activity can actually be explained in strictly physicalistic terms.

By the way, part of this discussion entails the notion that colors are merely physical wave lengths of light. The argument that colors are wave lengths that cause a chemical reaction in your eye that is physical is meant to argue against the notion that colors can exist or be apprehended in some kind of non- physical realm. My illustration that I gave was to make the point that colors are not wave lengths because it is possible to see color when there is no light. If you can see color when there is no light, then colors can't be light. They must be something different than light. They may be associated with light, but they must be different than light.

Here is the illustration I gave. Close your eyes and picture your mother. You are a child. She's in the kitchen washing dishes. Have you got the picture? What color dress is she wearing? What color is there? Some say red. Some say blue. Some say green or white. The point is, there is a color in your imagination. You can conjure up a color with your mind. Note I didn't say brain because your brain isn't conjuring up the color. Your mind is. How do I know it's not the brain? Because you cannot crack your brain open and ever find your mother in there wearing a certain color dress. What you are seeing is not something physical. If is was physical, it would occupy a physical location. You can't find it physically. She's not in there. Nor is the color in there. Not in your brain. It is in your mind. Therefore, it is not physical. It cannot be equated with brain activity. It can't be identified as the same thing as brain activity, and this becomes as argument for the existence of the soul.

I received a letter from a Christian who questioned my illustration. Greg from Redondo Beach offered the counter-example that a computer can supposedly see color, therefore this argument for the mind is compromised.

I'm very glad that he is thinking about this and is responding to this because he is raising questions he might run into. I'm glad to be able to sharpen my point and make the case for the existence of the soul.

I'd like to read the paragraph in which he responds to my illustration. He says this: "I caught a little of your show last weekend and am troubled about the mind/brain question. I'm not sure of the example "If you crack your brain open, you won't find the color blue," is valid since a computer can be programmed [Here's the heart of it] to detect blue with a sensor or camera and verbalize the word blue. The computer is matter which contains in it a representation of the color blue. [He suggests here that a thought is just a complex representation that would be on the physicalist order of things.] So it contains a representation of the color blue, and when you smash the computer there is likewise no blue. But in this case, mere matter and programming, without needing a soul to do it, contain blue. My hunch is there is still a kernel which may require a separate soul, and I can't imagine how any machine or physical brain could achieve self-motivated, independent, free action that was not programmed in by a Designer." Now that's the free moral agency question and I will come to that later in the show. But the earlier challenge is still a good one.

He is saying, You know I don't buy your illustration, Koukl, and here are the reasons. You may still have a point on other grounds and those other grounds are the grounds of free moral agency. But I think that your illustration about color in the mind is flawed because it does seem that computer can recognize color and have a kind of representation of the color blue, even though when you smash a computer you won't find the color blue in there. So that seems to compromise your argument. If we can invent a machine, a computer that can recognize blue and identify it with the name blue, then why do I argue that a soul is necessary to do that in humans? It is an excellent question.

First, let me make a quick observation. To say that it is not necessary to assert the soul's part in this doesn't establish that only physical processes are involved. In other words, just because you could show that a machine could see blue in that way doesn't mean that the brain is seeing blue. It doesn't obviate or make unnecessary the existence of the soul. It's just a way of arguing that it ain't necessarily so. So just because we can show that my point doesn't hold and that a machine can see a color, it doesn't mean that we don't have a soul. It just seems to weaken my argument for the necessity of a soul. Even if a machine could do that, it doesn't mean that in the case of humans a machine is in fact doing that. That is a separate issue. But I am actually going to argue the harder case. I am going to argue that it is not even possible for a machine--and here I mean even biological machines like brains--to do what the soul can do.

Our specific argument is going to focus on the notion of being able to see a color. So our question is, Can a machine see color?

My argument in brief goes something like this: Physical things have physical characteristics. The elements of consciousness--thinking, intending, believing, having sensations and feelings (and that's what we are talking about here when you can see the quality of a color)--these are things that don't have physical properties, therefore they are not physical or material. They are non-physical, they are immaterial. The mind is not the brain. You see my argument?

Let's see how this objection applies to my argument.

Our question here is this: Can a machine--whether a computer or a brain--do the same thing that a soul does? I hold that the machine can't see the color blue. Greg from Redondo suggests that maybe it can. So let me make this distinction. It comes from the words that were used here--talking about a computer that can be programmed to detect blue. I must make a very important distinction here between detecting or measuring and seeing . Seeing is a way of measuring. I could have a visual impression of an object lying along side a ruler. I am seeing the object along side the ruler and my sight allows me to measure that ruler at 12 inches, but I could still measure that ruler at 12 inches even if I was blind. I could touch the ends of the object and the ends of the ruler and see that they are the same. I could put them next to each other and correlate the size of the object with the size of the ruler and determine that the object is one foot long. So I can measure it or detect it without seeing.

But my argument goes beyond that with regards to the soul. All I have to do to demonstrate that there is more to man than his brain is demonstrate that activities that we go through cannot be accounted for by a mere physical process. This accounting or inability to account for something is not a limitation of science--but that it can't be accounted for even in principle. Now that point is important because if something cannot be explained by physical processes even in principle, and we can demonstrate that to be the case, then we don't have to worry about somewhere down the line people discovering something in science that refutes our argument as if it is founded just in the contingencies of scientific discovery.

My point is that this has nothing to do with science. I am trying to make the point and prove the point that it doesn't matter how much scientific sophistication we have, we will never be able to do certain kinds of things because they are the kinds of things that do not respond to physical assessment principally because they are not physical things. There is no science that we will ever discover, for example, that will help us to know what is morally right and wrong. Science doesn't measure that kind of thing. Science measures physical things. A moral is not a physical thing, therefore science cannot measure morality. Rather, science is the subject of morality. In other words, morality impinges upon science, not the other way around.

I'm talking, though, about whether a computer can actually see color. The argument has been, A computer can measure color. That seems to be the case. But can it see color? That is what is at stake here. That's why I made the very important distinction between measuring and seeing. My argument is that a machine can measure, but it can't measure in one of the ways that you and I measure. It can't measure by seeing, that is, by having a visual impression. Why not? Machines can't have visual impressions. Minds have visual impressions. That is a characteristic of consciousness, not of machinery. Therefore, if minds have visual impressions and machines can't have visual impressions then the mind and machines--even the machine of the brain--are not the same thing.

Here is my response to the computer illustration--the computer that can be programmed to detect blue with a sensor and then alert us with the sound blue. Notice I didn't say the word blue as was used in the letter. A word is a physical symbol that stands for something else that is not physical. That non-physical thing is the word's meaning. The computer can't use words like that. It doesn't consciously use symbols. It merely makes a noise and the noise means something to us. So the computer can measure a wave length that we see visually as blue. Yes, it can do that. It can alert us with the sound that we know as the word blue, but is this evidence that the computer sees blue? The answer is no. The computer doesn't see color. It can only measure a wave length, and it can't even do that if light is not available and there is no light in my brain when I imagine my mother washing the dishes wearing a colored dress.

Greg writes in his letter: "The computer contains in it a representation of the color blue." Well, you know he is absolutely right about that. It contains a representation of the color blue, but I think he has conceded more than he realizes with this remark. What is the representation? If all of what blue is is captured in the wiring of the computer--or in the wiring of my brain by analogy through some kind of an electro-chemical reaction--then it doesn't make any sense to say that it represents something else. If the physical description is what blue is, then there is nothing to represent. Of course, that is what a physicalist would argue--blue is the chemical reaction in your brain, not that it represents chemical reactions in your brain.

I'm going to prove this very simply to you in case the change hasn't fallen into the meter yet. I realized this can be kind of an obscure concepts.

Picture the color blue. Now, describe the quality of your impression in physical terms. Tell me what blue looks like. You can't do it. It is not possible to tell me what blue looks like, unless of course you compare it to another color that I am familiar with. In other words, you can only describe a sensory impression by using other sensory impressions. You can't describe it in purely physical terms in any way that captures what you are experiencing.

If I asked you to describe blue to me in physical terms, you would say that it has the wave length of .902 milacrons (I just made that up). Does that help me at all to know what it is that you see right now in your mind when you see the color blue? It doesn't help one single bit. And the reason is that the blue is not physical in that sense. It can apply to physical things, but the visual impression--the thing that you see--is not physical. That's why a computer can't see blue.

The point that I am making is that there is a difference between measuring a color and the color itself. If we say that the measurement represents the color, as Greg did, then it is clear that the color is something different from the measurement--in this case a measurement of wave length. And that's why we say the wave length represents the color. The color is what we see. The wave length is simply the wave of light that produces this visual sensation of blue.

Greg also said that the computer contained the color blue, but what does it mean to contain blue? If the programming is the blue, then how doe the programming contain the blue? It doesn't. The blue is something else. Blue is what we see. Our minds don't simply detect a wave length. We see color.

In fact, a blind person could be equipped with a scanner attached to his brain that would enable him to detect colors much like a computer detects color. It would just communicate the wave length of color through the means of maybe a sound or something like that. Different colored wave lengths might have different pitches of sound corresponding to them. And since sound has a continuous gradient scale just like the colors in a rainbow, one could get an exact correlation of certain sound wave lengths to certain color wave lengths. So a blind person, knowing the sound correlations, could identify from sound cues any particular color that was before him. But would it make any sense in that circumstance to say that the blind person was actually seeing the color? It wouldn't at all. Why? Because he's blind. He sees nothing. That is precisely my point. A computer is the same way--even a sophisticated computer like the brain.