(no subject)
“…And this computer was so amazingly intelligent that even before its data banks had been connected up, it had started from first principles with I Think Therefore I Am, and it had got as far as deducing the existence of rice pudding and income tax before anybody managed to switch it off.”
- Douglas Adams, The Hitch Hiker’s Guide to the Galaxy.
Here follows my essay for my Philosophy of Mind class, which I submitted today. I don't know how well it will go - I like it, but I worry it doesn't answer the question well enough. Ho hum. Should garner at least a pass.
No one model yet advanced manages to completely explain the mind; most have valid points (even behaviourism, which concedes that the only logically tenable position is agnostic solipsism and that we can never truly observe the minds of others), and all have objections against them. Dualism, in a form derived from but not identical to that of Descartes, seems to offer the same objection against most others, in the form of its most fundamental tenet: cogito ergo sum. This objection rests in large part on the fact that if we accept the cogito we must also accept free will, as I will demonstrate after first clarifying the position of dualism itself.
Cartesian dualism, reduced to simplest terms, comes from the sceptical analyses in the first few Meditations. In the beginning, it is possible to doubt almost everything, and we want to be certain of everything we know, or rather to discard as ‘knowledge’ anything we are not certain of. We find, of course, that the one thing we cannot doubt is that we are thinking, because to doubt is to think something. We conclude that the sole and essential property we can be certain we possess is that of thinking; we are merely a res cogitans. Later in the Meditations, we regain much of our original certainty of the world we perceive (with a few caveats). Descartes does introduce one unnecessary complexity into his otherwise elegant model: he fails to prove the existence of God, which he believes is necessary to the model; however, an agnostic variant may be constructed based rather on the immutability of logic itself. Any attempt to question the validity of logic in these matters must necessarily be recursive or selfcontradictory and therefore absurd, and so the model stands, slightly simplified.
An objection sometimes raised against the dualist model is based on its supposed nonsimplicity, at least in relation to whatever model the objector is advancing in its place. In relation to identity theory, this reflects a misunderstanding: dualism posits one hypothetical entity in addition to those observed, and that entity is the means by which the mind and the body interact. Identity theory hypothesises that mental states can be inferred directly from the physical states of the brain; this is a more easily tested (and hence more easily falsified) hypothesis, which has as yet had little evidence in its favour. The most that can be inferred from brain activity is a very vague, fuzzy indication of what the subject is thinking about, and it is not clear that more precise measurements of brain activity will ever allow for a much clearer picture; indeed, all we can really tell from brain measurements are inputs and outputs; we have very little information on actual conscious thought. Identity theory posits a whole set of physical rules that are not basic and rather emergent from the laws of physics, by which the states and processes of the brain translate to the states and processes of the mind (which are identical to them in all but name, hence the word translate rather than correspond); but it gives us no more hint of how we might find these, or what they actually are, than dualism gives of the connexion between the distinct mind and brain.
In relation to behaviourism, the simplicity objection holds no water because if behaviourism is held up to the same test, we find that it is simpler than dualism only because it omits many entities which are observed in the real world, and conflates the inputs and outputs of the mind/brain, which are necessarily distinct, as is shown in the pain objection: the sensation of pain and the behaviour exhibited by a being in pain are considered to be identical by the behaviourist theory, but they are clearly distinct.
There are two types of events in the physical universe that everyone can observe: ones that proceed logically as the effects of causes, and ones that are completely unpredictable and random. Most of the latter sort are emergent events from small cause-and-effect events, but at the quantum level there are unpredictable, random events caused by nothing but probability. However, every person can observe also his own thoughts, his own mind. It seems to make some sort of sense to him: it certainly doesn’t seem random, and it also seems to be under his own control. Indeed, if it were not under his control, what could possibly be meant by the word him (or her, or to concede to political correctness, them)? He can also observe the behaviour of others, and it seems similar to his own in most cases. From this he infers that the others also have thoughts and minds like his. This is not a completely logically tenable inference, but it is a scientifically valid one and one which one needs to make if one is to make any headway in dealing with people.
The crux of this argument lies in that a person observes that his own thoughts are decidedly nonrandom, and (with the weight of a century of psychological inquiry behind him) that those of others are essentially unpredictable. He also knows that every person thinks this way: that his own thoughts are nonrandom and those of others are unpredictable. It appears we have a third type of event: nonrandom, but unpredictable. This seems to be the perfect definition of free will.
The concept of thought is tied very closely to that of free will, because if something has no free will, it is difficult to see how said something can be said to be thinking. After all, the mind is a rather special thing, and it appeals to our egos as well as our intuitions to think that a specific mind (as opposed, for now, to any generic mind) could not be accurately simulated. It may be realised through different means, if we supplement our brains with electronics or manage to transplant our minds into machines; but it could not be completely accurately simulated. If there were not already an identity theory of mind, we might be inclined to term this the Identity Principle; but the two are not related. The principle in question is that any given mind is a single, necessarily unique entity, impossible to be replicated. This is why our intuitive response to Searle's Chinese Room objection to functionalism is that it is not thinking in any meaningful sense of the word. The Chinese Room is not only demonstrably finite: it is built on a rigid set of rules, and in addition those rules and its internal states are both completely knowable (and thus replicable) and unchangeable. The cogito denies functionalism as it applies to the Chinese Room and analogous models, because the cogito demands free will and such models clearly lack it. However, the cogito does not completely discredit functionalism in all its forms, as we shall see later.
The hardest problem this implication poses is that very few people would reject the cogito; it seems to suffer no counterarguments. Smart’s proposal of identity theory is in essence a long and sophisticated argument against dualism from personal incredulity, a logical fallacy he fails even to keep up for the entire time, and drops when he tries to object to the steps from the conceivability of dualism to its possibility and from that possibility to actuality. In fact, it can be argued that if the mind and brain were actually identical then dualism would be inconceivable. Yet Smart does not mention the cogito, despite the fact that (or possibly because) it undermines his own proposal. The strongest form of identity theory draws a necessary identity between the mind and the brain; but even if such identity was not necessary, it would still imply a hard deterministic worldview, which it seems Smart does indeed hold. The only alternative, drawn from quantum physics, is a mind/brain whose states are not determined by absolute rules but by probabilities: this we might term an absurdist relative of determinism, but it is as tenable as hard determinism. However, both are only tenable at all until we consider the implications of the free will objection. If we consider hard or absurdist determinism, we have no free will, and we are in essence machines, not qualitatively different from computers or any other machine except for the materials we are made of. It seems impossible to draw a boundary at what level of complexity defines thought: any such boundary would be to a large degree arbitrary. We are in no doubt that simple machines and the computers we have today are not thinking: what makes us different from them? Identity theory claims that nothing does, and the implication from that is that we’re not thinking either, because the cogito tells us that the definition of mind cannot be an arbitrary one; yet it also tells us that the one thing it is logically impossible to believe is that one is not thinking.
We must consider now what is necessary for anything to be called a mind, for we do not want to be chauvinists, and claim that only things with human brains have minds, for there are things with human brains that definitely do not have minds and there are beings with differently-designed brains that are suspected of possessing sentience perhaps lesser but of the same qualitative sort as ours. However, it seems that while our perceptions are limited by our biology - we can distinguish a finite number of colours, we cannot perceive sounds above a certain frequency, and so on - the possibilities of the mind, the res cogitans itself, must necessarily be infinite, because it is a truly free thing. This allows for two models: a form of functionalism in which the workings of the mind, its processes and states, are nonetheless impossible to describe or define anything like completely; or dualism. Both have their problems.
The major objections to dualism itself seem mostly to expose gaps in the theory, where it needs improvement, and do not discount the theory itself as a whole. The most common one is the causality objection: how can the immaterial mind interact with the material brain and body? It is possible that the answer to this is merely unknown rather than unknowable; dualism does not in its proper form presume to explain how the connexion might be realised.
Functionalism may provide the simpler alternative we are looking for. Each brain, like each mind, is unique (and let us leave the possible paradox of a set of entities that are all infinite and yet each unique for another day); the mind has infinite capability, but only a tiny, finite subset of those limitless possibilities is realised by any given mind, and while identity theory is necessarily deterministic and thus cannot even allow for infinite possibilities, a slightly unorthodox form of functionalism may allow for it.
- Douglas Adams, The Hitch Hiker’s Guide to the Galaxy.
Here follows my essay for my Philosophy of Mind class, which I submitted today. I don't know how well it will go - I like it, but I worry it doesn't answer the question well enough. Ho hum. Should garner at least a pass.
No one model yet advanced manages to completely explain the mind; most have valid points (even behaviourism, which concedes that the only logically tenable position is agnostic solipsism and that we can never truly observe the minds of others), and all have objections against them. Dualism, in a form derived from but not identical to that of Descartes, seems to offer the same objection against most others, in the form of its most fundamental tenet: cogito ergo sum. This objection rests in large part on the fact that if we accept the cogito we must also accept free will, as I will demonstrate after first clarifying the position of dualism itself.
Cartesian dualism, reduced to simplest terms, comes from the sceptical analyses in the first few Meditations. In the beginning, it is possible to doubt almost everything, and we want to be certain of everything we know, or rather to discard as ‘knowledge’ anything we are not certain of. We find, of course, that the one thing we cannot doubt is that we are thinking, because to doubt is to think something. We conclude that the sole and essential property we can be certain we possess is that of thinking; we are merely a res cogitans. Later in the Meditations, we regain much of our original certainty of the world we perceive (with a few caveats). Descartes does introduce one unnecessary complexity into his otherwise elegant model: he fails to prove the existence of God, which he believes is necessary to the model; however, an agnostic variant may be constructed based rather on the immutability of logic itself. Any attempt to question the validity of logic in these matters must necessarily be recursive or selfcontradictory and therefore absurd, and so the model stands, slightly simplified.
An objection sometimes raised against the dualist model is based on its supposed nonsimplicity, at least in relation to whatever model the objector is advancing in its place. In relation to identity theory, this reflects a misunderstanding: dualism posits one hypothetical entity in addition to those observed, and that entity is the means by which the mind and the body interact. Identity theory hypothesises that mental states can be inferred directly from the physical states of the brain; this is a more easily tested (and hence more easily falsified) hypothesis, which has as yet had little evidence in its favour. The most that can be inferred from brain activity is a very vague, fuzzy indication of what the subject is thinking about, and it is not clear that more precise measurements of brain activity will ever allow for a much clearer picture; indeed, all we can really tell from brain measurements are inputs and outputs; we have very little information on actual conscious thought. Identity theory posits a whole set of physical rules that are not basic and rather emergent from the laws of physics, by which the states and processes of the brain translate to the states and processes of the mind (which are identical to them in all but name, hence the word translate rather than correspond); but it gives us no more hint of how we might find these, or what they actually are, than dualism gives of the connexion between the distinct mind and brain.
In relation to behaviourism, the simplicity objection holds no water because if behaviourism is held up to the same test, we find that it is simpler than dualism only because it omits many entities which are observed in the real world, and conflates the inputs and outputs of the mind/brain, which are necessarily distinct, as is shown in the pain objection: the sensation of pain and the behaviour exhibited by a being in pain are considered to be identical by the behaviourist theory, but they are clearly distinct.
There are two types of events in the physical universe that everyone can observe: ones that proceed logically as the effects of causes, and ones that are completely unpredictable and random. Most of the latter sort are emergent events from small cause-and-effect events, but at the quantum level there are unpredictable, random events caused by nothing but probability. However, every person can observe also his own thoughts, his own mind. It seems to make some sort of sense to him: it certainly doesn’t seem random, and it also seems to be under his own control. Indeed, if it were not under his control, what could possibly be meant by the word him (or her, or to concede to political correctness, them)? He can also observe the behaviour of others, and it seems similar to his own in most cases. From this he infers that the others also have thoughts and minds like his. This is not a completely logically tenable inference, but it is a scientifically valid one and one which one needs to make if one is to make any headway in dealing with people.
The crux of this argument lies in that a person observes that his own thoughts are decidedly nonrandom, and (with the weight of a century of psychological inquiry behind him) that those of others are essentially unpredictable. He also knows that every person thinks this way: that his own thoughts are nonrandom and those of others are unpredictable. It appears we have a third type of event: nonrandom, but unpredictable. This seems to be the perfect definition of free will.
The concept of thought is tied very closely to that of free will, because if something has no free will, it is difficult to see how said something can be said to be thinking. After all, the mind is a rather special thing, and it appeals to our egos as well as our intuitions to think that a specific mind (as opposed, for now, to any generic mind) could not be accurately simulated. It may be realised through different means, if we supplement our brains with electronics or manage to transplant our minds into machines; but it could not be completely accurately simulated. If there were not already an identity theory of mind, we might be inclined to term this the Identity Principle; but the two are not related. The principle in question is that any given mind is a single, necessarily unique entity, impossible to be replicated. This is why our intuitive response to Searle's Chinese Room objection to functionalism is that it is not thinking in any meaningful sense of the word. The Chinese Room is not only demonstrably finite: it is built on a rigid set of rules, and in addition those rules and its internal states are both completely knowable (and thus replicable) and unchangeable. The cogito denies functionalism as it applies to the Chinese Room and analogous models, because the cogito demands free will and such models clearly lack it. However, the cogito does not completely discredit functionalism in all its forms, as we shall see later.
The hardest problem this implication poses is that very few people would reject the cogito; it seems to suffer no counterarguments. Smart’s proposal of identity theory is in essence a long and sophisticated argument against dualism from personal incredulity, a logical fallacy he fails even to keep up for the entire time, and drops when he tries to object to the steps from the conceivability of dualism to its possibility and from that possibility to actuality. In fact, it can be argued that if the mind and brain were actually identical then dualism would be inconceivable. Yet Smart does not mention the cogito, despite the fact that (or possibly because) it undermines his own proposal. The strongest form of identity theory draws a necessary identity between the mind and the brain; but even if such identity was not necessary, it would still imply a hard deterministic worldview, which it seems Smart does indeed hold. The only alternative, drawn from quantum physics, is a mind/brain whose states are not determined by absolute rules but by probabilities: this we might term an absurdist relative of determinism, but it is as tenable as hard determinism. However, both are only tenable at all until we consider the implications of the free will objection. If we consider hard or absurdist determinism, we have no free will, and we are in essence machines, not qualitatively different from computers or any other machine except for the materials we are made of. It seems impossible to draw a boundary at what level of complexity defines thought: any such boundary would be to a large degree arbitrary. We are in no doubt that simple machines and the computers we have today are not thinking: what makes us different from them? Identity theory claims that nothing does, and the implication from that is that we’re not thinking either, because the cogito tells us that the definition of mind cannot be an arbitrary one; yet it also tells us that the one thing it is logically impossible to believe is that one is not thinking.
We must consider now what is necessary for anything to be called a mind, for we do not want to be chauvinists, and claim that only things with human brains have minds, for there are things with human brains that definitely do not have minds and there are beings with differently-designed brains that are suspected of possessing sentience perhaps lesser but of the same qualitative sort as ours. However, it seems that while our perceptions are limited by our biology - we can distinguish a finite number of colours, we cannot perceive sounds above a certain frequency, and so on - the possibilities of the mind, the res cogitans itself, must necessarily be infinite, because it is a truly free thing. This allows for two models: a form of functionalism in which the workings of the mind, its processes and states, are nonetheless impossible to describe or define anything like completely; or dualism. Both have their problems.
The major objections to dualism itself seem mostly to expose gaps in the theory, where it needs improvement, and do not discount the theory itself as a whole. The most common one is the causality objection: how can the immaterial mind interact with the material brain and body? It is possible that the answer to this is merely unknown rather than unknowable; dualism does not in its proper form presume to explain how the connexion might be realised.
Functionalism may provide the simpler alternative we are looking for. Each brain, like each mind, is unique (and let us leave the possible paradox of a set of entities that are all infinite and yet each unique for another day); the mind has infinite capability, but only a tiny, finite subset of those limitless possibilities is realised by any given mind, and while identity theory is necessarily deterministic and thus cannot even allow for infinite possibilities, a slightly unorthodox form of functionalism may allow for it.