Descartes’ position is that the mind is an object (or in his old-school words: “a substance”), and like any object (this chair, this book, the dog, the Earth) it has features, and it undergoes changes. What distinguishes the mind from all other objects (at least, all material objects) is that none of the features of the mind are features that material objects have (in virtue of which they are material): it has no extension, shape, mass, size, color, surfaces, etc. One of the mind’s intrinsic features: being immortal. Some of the mind’s features that have a temporal duration (its occurent features): desiring, doubting, feeling a pain, hoping, worrying, dreaming, etc. Now, does the mind have a spatial location, even though it has no material features? This is a tough question. I don’t see any a priori reason why anything that is completely invisible (and therefore non-material) couldn’t have a spatial location. We just wouldn’t be able to tell, by way of (at least simple methods of) public observation, that it had one. But the fact that we couldn’t tell doesn’t imply that it couldn’t have one. (If you don’t like the example of something being completely invisible and therefore having no material features, then consider other examples: gravity, the weak nuclear force, the strong nuclear force, and electromagnetic fields do not have material features, and nevertheless have spatial locations.) I think Descartes might agree with me here. He seems to think that a mind doesn’t NEED to have a spatial location but some of them sometimes DO: for instance, I think he would say that my mind is presently located in the space of my body. (For some reason, Jaegwon Kim is committed to the position that Descartes' conception of mind requires that the mind never has a spatial location. Can't figure out why). What would explain the present spatial location of my mind is God. At any rate, the traditional problem of mind-body causal interaction arises from the conception of the mind as having no material features, and of the body as having only material features. One traditional objection to there being a causal interaction goes like this. Causation between two objects requires that the objects share some of the same features. Since the mind and body share none of the same features, there is no causal interaction between the mind and body. This assumption about causation begs the question against Descartes, who maintains that there is mind-body causal interaction. And there doesn’t seem to be any independent reason for believing that two causally connected objects need to share some of the same features. Another traditional objection to interaction is that causation between two objects requires that the objects share some material features (for example: they need to have physical contact). Since two objects share some material features only if none is a mind, there is no causal interaction between mind and body. This assumption, again, begs the question against Descartes. And there doesn’t seem to be any independent reason for believing that two causally connected objects need to share some material features.

Louis Loeb suggests that Descartes could respond to these traditional objections by adopting a Humean position on causation: deny that mind-body causation is anything more than a constant conjunction between events. The reasoning seems to go like this: If mind and body events are causally connected to one another just in virtue of there being a constant conjunction between them, then the features that they have (material or non-material) are not necessary to there being a causal connection between them. Jaegwon Kim’s reply to Loeb is ingenious. His argument goes like this.


1st stage of the argument

Mind-body union requires that these body parts are mine, and these body parts are mine only if they’re not anyone else’s; for example, this arm is mine only if it’s not anyone else’s. There is mind-body union. So, this arm is mine, and not anyone else’s.

2nd stage of the argument

This arm is mine only if there is a causal connection between my mind and my arm; for example, that I will that this arm goes up causes this arm to go up; and this arm getting bruised causes me to have a feeling of pain. Now, this causal connection is either just a constant conjunction between events, or not.

Suppose that it’s just a constant conjunction between events. But then there might be a situation where anytime that I will that this arm goes up, you also will that some other arm goes up, and, afterwards, this arm goes up, and that other arm goes up. But then this arm would be as much yours as mine, because this arm would be causally connected to your mind as much as it is causally connected to mine. But this arm is mine, and not anyone else’s. Therefore, the causal connection between my mind and my arm is not just a constant conjunction between events. Therefore, by extension of reasoning, the causal connection between mind and body is not just a constant conjunction between events.


The one thing I think that is wrong with Kim's argument is his (seemingly contingent) claim that this arm is mine: it needs to be stated as a necessary claim (that this arm is necessarily mine) in order for the argument to work. And I don’t know how he can argue for that necessary claim. This is how I would revise the 2nd stage of his argument:


This arm is mine only if there is a causal connection between my mind and my arm; for example, that I will that this arm goes up causes this arm to go up; and this arm getting bruised causes me to have a feeling of pain. Now, this causal connection is either just a constant conjunction between events, or not. If it’s just a constant conjunction, then that I will that this arm goes up causes this arm to go up IF I will that this arm goes up, and afterward this arm goes up. But that’s not a sufficient condition. Because if, unbeknownst to me, you were to put an iron bracelet on this arm, attach a magnet to the ceiling, and if, when I will that this arm goes up, before I am able to act on my will, you activate the magnet, and this arm goes up, THEN I would not be causing this arm to go up. The reason is: I would not be experiencing causing the arm to go up in virtue of acting on my will, because I would not be acting on my will; in fact, I would report to you that I did not cause my arm to go up! Therefore, the causal connection between my mind and my arm is not just a constant conjunction between events. Therefore, by extension of reasoning, the causal connection between mind and body is not just a constant conjunction between events.

What I think is ingenious about Kim’s argument is the connection he makes between there being a mind-body union and there being mind-body causation. I think this connection can be used to strengthen John Searle’s position that we experience causation whenever we perform intentional actions and whenever we have perceptions.

And therefore your version of AI is going to what?

Doesn't trying to create AI sort of fly in the face of cartesian dualism?

I don’t see any reason why AI has to be inconsistent with property dualism, at least not a limited version of property dualism that just takes the phenomenal character of our mental states as not having any material features. What am I talking about here? Well, the phenomenal character of a mental state is the way it feels like to be in that state when you’re in it. It’s the character of the experience that you have when you’re in the state, so it’s OBSERVABLE ONLY FOR A FIRST-PERSON PERSPECTIVE. Let me give some examples. The phenomenal character of the state that I’m in when I’m desiring that I eat a cheeseburger is the way it feels like to me when I’m desiring that I eat it. The way it feels like to me is not observable from a third-person perspective, although what causes me to have the feeling is. Another example: the phenomenal character of the state that I’m in when I’m seeing a red rose is the way the red rose looks to me as I’m seeing it. The phenomenal character of the state that I’m in when I burn my hand on the stove is the way it feels to me when I’m burning it. At any rate, since we are machines, biological machines—machines evolved by a process of natural selection—I don’t see any reason why we couldn’t create a machine that duplicates the causal powers that we have that cause us to have mental states with a phenomenal character. Because similar causes have similar effects, I imagine that the phenomenal character of their mental states would not have any material features, just like us.

James,
I think you could have left Descartes and dualism in the past where they belong. I don’t think your references to personal perspective are incorrect but to get involved in a discussion of dualism is of little reward. More to the point are your references regarding the effectiveness of physical sensation on an entity. When we have sensation, from any sense, it is involved in the process of evaluating our ability to survive. It is the information we apply to decide our next series of actions. Some situations require little or no action while others require immediate and severe actions. The point is this is what we reference when we discuss what we feel about some something. It is a compilation of all experiences we have had regarding some object or situation. And most important it serves to warn and motivate the animal world and us in general.

The question is or should be can a similar situation be created in a machine? Not only can it but, depending on how you see these kinds of events, it is! To understand what I am referring to we must take a broader perspective. One that looks to the action / reaction aspect of input devices in general. One that looks at animal emotional systems as a series of input / output devices that have specific value. Just as a fire alarm system has an association with the health of a building. Feelings are the devices that describe our limits. We sense several homeostatic conditions. Conditions such as blood characteristics, temperature, taste, smooth, rough, and the list goes on. All contribute to experience. Our survival has depended on this and this is propagated by Natural Selection. By approaching this as an input / output situation we can compare our emotionality substructure to that of action / reaction devices. When you mentioned that “I don’t see any reason why we couldn’t create a machine that duplicates the causal powers that we have that cause us to have mental states with a phenomenal character” it seems to me that you are approaching this conclusion.

The confusion about the need of emotion in a machine speaks to the level that these principals are misunderstood. Emotions and feelings are the currency of homeostatic condition. Direct mechanical comparisons are devices such as temperature controls, A to D converters, current sensors, voltage sensors, vibration sensors and so on. With these devices we can begin to build machines that can duplicate the emotional systems of the animal world. The key is to know why!

Lex Ricketts

It’s useful to think of some mental states in terms of the passive and active causal roles they typically have for the physical systems that have them. For example, it’s useful to think of pain in terms of the sensory stimulations that typically cause it (tissue damage), and the behaviors (and other mental states) that it typically causes (wincing, groaning, withdrawal behavior, avoidance behavior, a state of fear, a state of distress). The passive causal role the state has makes up the INPUT; the active causal role the state has makes up the OUTPUT. This is all just a model though: the functionalist model of the mind. It certainly doesn’t establish that there is nothing more to being in a mental state than being in the right kind of causal role. The one crucial thing the model leaves out is the phenomenal character (i.e. the experience) of being in the mental state. You can instantiate the causal role typically associated with being in pain and still not have the experience of being in pain. And if you don’t have the experience, then you’re still not in pain. Now, is there an evolutionary explanation for why some mental states are typically associated with certain kinds of causal roles? Of course there is. Does this establish that there is nothing more to being in a mental state than being in the right kind of causal role? Certainly not.

It’s reasonable to question whether the experience you have, when you’re having a mental state, has any material features. Why? Because the experience you’re having is only observable from a first-person perspective. And material features are observable from a third-person perspective (that is, they are publicly observable) Since the experience you're having is not observable from a third-person perspective, it doesn't have any material features.

And therefore your version of AI is going to what? Or was that a suggestion of why?

Some interesting points. Perhaps there are some aspects to our experience that are only experienced and have no causal effects what so ever. But I am inclined to believe that every aspect of our experience is potentially useful via actions we may take. If something salient grabs your attention, your brain is more inclined to evoke some actions relating to that salient phenomenon, and more likely to generate associations with that phenomenon, associations that could be used to inform decisions about actions at some later time.

Every aspect of the qualia of seeing the color red could be accounted for by considering the countless associations the brain has with that color. Just because we are not consciously aware of those associations and the ways they end up influencing our behavior does not mean they do not exist. That unconscious associations DO exist and influence our decisions/actions is something that has been experimentally verified on many occasions. So I believe that the reason it feels the way it does to experience the color red is because of the associations (either acquired during your lifetime, or inherited because the associations had some survival value to your ancestors) your nervous system has with the activations of those particular cone receptors in your retinas. And the fact that you are even able to enervate the muscles in your hands to type a message about the indescribable aspects of your subjective experience indicates to me that even those indescribable aspects of your subjective experiences do have causal effects on the material world.

It’s important to speak as precisely as possible about these matters. And, by the way, I don’t have any firm commitments on any of this stuff. I just find the subject to be infinitely fascinating.

There’s the mental state that you’re in, and there’s the experience that you have when you’re in it. These are, in principle, distinct. It’s logically possible to have a mental state without having the experience that is typically associated with it, or without having an experience at all. I suggested earlier that you need to have the experience of being in pain in order to be in the state of pain. My basis for that conclusion is inductive, not a priori:

Whenever we’ve been in a state of pain, and have been conscious of it, we’ve had the experience of being in pain.
Therefore, whenever we are in a state of pain, and are conscious of it, we will have the experience of being in pain.

This argument stands or falls depending on what it is to be conscious of being in a state of pain. Since it’s not clear what it is, I can easily use it to render my position irrefutable. You can try refuting the premise or the conclusion by giving me an example of a case where someone clearly seems to be in a state of pain, but doesn’t have the experience of being in pain, and I could accept those facts, but say that he is not conscious of being in a state of pain, and on those grounds, rule out that case as a counterexample. If we take out the vague expression, “and have been conscious of it”, from my position, then it doesn’t seem obviously clear anymore whether it is true. For example, take the case where you touch the hot frying pain, without knowing that it’s hot, and your hand reflexively pulls away, and it is only then that you experience being in pain—a burning sensation in your hand. Now, when you are touching the hot frying pain, you are not experiencing being in pain, but are you, nevertheless, in a state of pain? That seems to be a reasonable interpretation. (The reason for saying that you’re in pain is that you’re satisfying the functionalist definition of being in a state of pain: you’re instantiating the right kind of causal role that is associated with being in a state of pain: tissue damage causes you to be in a state of pain; the state of pain causes you to withdraw your hand—who needs the experience?, your state of pain saved you from further harm without it, and in doing so, it plays a role in your survival and reproductive success, and if it’s a hereditary trait, the survival and reproductive success of your species). But I could defend my position by saying that you’re not in a state of pain at all, when you are touching the hot frying pain. Or I could add the vague expression, “and have been conscious of it”, to my position, and then say that you are in a state of pain, but just not conscious of it. AT any rate, my point of going into this is to say that I’m not that committed to the position that having an experience is necessary in order to be in a mental state.


You might be thinking to yourself: If we could be in a state of pain without ever having the experience of being in pain, then why the hell do we have the experience of being in pain? Otherwise put: if we could instantiate the passive and active causal role associated with being in a state of pain without ever experiencing being in pain, then why do we have the experience of being in pain? Why is there an experience of being in pain if the experience has no active causal powers? You might think that this intuitive grounds for a reductio ad absurdum against the position that the experience of pain might not have any active causal powers, and therewith, the position that the experience of pain is not necessary to being in a state of pain. (it’s important to recognize the connection here: if the experience of pain is not necessary to being in a state of pain, then the experience of pain might not have any active causal powers. Why? Because it is the experience of pain’s having active causal powers that is what would make it a necessary condition to being in a state of pain, given that being in a state of pain is just having the right kind of active and passive causal role). The problem with the intuition: there are many examples in nature of features that a physical state has that don’t play an active causal role, at least with respect to the physical system instantiating the state, given the functional definition of the system instantiating the state. The thumping sound that the human heart makes is an example of this. It doesn’t play an active causal role in the pumping of the blood throughout the body. It’s completely epiphenomenal, in this respect. So, maybe the experience of pain, and, by extension, all experiences, are like that?!

I can talk about what your desires, hopes, and other attitudes are, based on my conversational interactions with you, and I can talk about your hair style, suntan, and nose job based on my perceptual interactions with you. Since these interactions all occur on the same level of description (the level of medium-sized objects: chairs, trees, squirrels, coffee mugs), we get the impression that they are all features that occur on the same level of description. But they’re not. Desires, hopes, and other attitudes occur at a lower level of description than hair styles, suntans, and nose jobs. They occur at a microlevel of description, the same level of description as the neural processes that determine them. The peculiarity about them is that they are privately observable. Everything else on the microlevel is publicly observable (albeit indirectly so), just like everything on the medium-sized level. It is because attitudes are observable (albeit privately), without the aid of technology, that we can talk about them with the same facility as we talk about many things on the medium-sized level, which are also observable (albeit publicly), without the aid of technology.