In the remarkable advances that have been made, and the threats that those might be felt to pose, there is one area of technology -- the area of computation and artificial intelligence -- that gets rather close to the heart of human concerns. One thing people have been very exercised about is the possibility that before long there might be machines that will not only be able to think, but able to think about us , and and that on this subject they will take a somewhat dim view.

Recently I found myself in a discussion with Professor Igor Alexander who attained considerable public attention for his claim that MAGNUS, a machine he has built, is conscious and has a measure of free will. The occasion of our discussion was a BBC radio programme -- Start the Week with Melvin Bragg. About a week before, The Daily Telegraph of London had produced an article entitled 'I'm Magnus: I think therefore I am' in which someone called Warwick Collins, a defence expert and writer, warned that thinking computers could end up wiping out the human race unless their development is policed. The relevant section of the article runs:

Mr Collins, who wrote a novel, Computer One, based on his fears that computers will one day annihilate their creators, called for the government to establish rules on the attributes that scientists can give machines. He likened the threat posed by computers with the ability to repair themselves, to that of nuclear holocaust in the 1960s. Mr Collins said that within 30-50 years a web of millions of computers could be acting as a single organism. "If they're allowed to design and manufacture other computers, as they soon will be able to, then they have become a species". Mr Collins said plutonium stores or biological weapons could be released by military computers in an attack on the entire human race. If computers no longer needed humans and saw them as a threat they could, without warning, said Mr Collins wipe us out.

Against the background of such hysteria it seems appropriate for a philosopher to turn down the volume a little, and to think reflectively about that possibility. The question 'Could the Soul be Software?' nicely combines, as it were, ancient and modern, because while talk of software picks up on a phraseology and a way of thinking that really only came into being with computers, talk of the soul seems very ancient. And for that very reason it might be thought that these two ideas do not fit together very well. What we need to do first of all, therefore, is try to analyse that question -- 'Could the Soul be Software?' -- by thinking about what might be meant by the soul and what somebody might mean by software, and see what conclusions can be drawn.

To the extent that anybody would use the term 'the soul' nowadays, they tend to equate the soul with the mind, the thinking thing. Consequently we could revise our question along in the following lines -- 'Could the Mind be Software?'. In fact we can extend the notion of software a little. Since it is correlative with that of hardware, and thus related to a general understanding of how computers work, the original question can be expressed something like this: 'Could the Mind be a Computer?'

Of course, some people think not only that the human mind could be a computer, but that it is. I am doubtful that minds or souls are to be thought of as computers, and my purpose here is to justify that doubt by showing first, that computers can't think, second that there are souls, and third that souls aren't software.

I Thinking about the nature of the soul falls broadly into three historical periods. I mean that reflection about the nature of the soul or the mind -- for the moment we can take the two to be interchangeable -- can be divided roughly into these three periods: the pre-modern, the modern and the contemporary. Though we obviously want to end up with contemporary thinking, it is instructive to say something brief about each of these.

In the earliest phase of classical antiquity, people's thinking about the nature of persons, the nature of soul, is very animistic, in this sense: it is concerned with what it is that makes things to be alive. In early antiquity, generally speaking, people were fairly promiscuous in their attitudes about where life principles could be found and so believed that many things were animated by souls. This general idea is given considerable philosophical refinement in Plato and in Aristotle. In the Patristic and Mediaeval periods, roughly the thousand years from the intellectual establishment of Christianity to shortly before the Reformation, thought on these matters was dominated by Christianized versions of the beliefs of classical antiquity. So, Plato's view of the soul as an immaterial entity in some way associated with or inhabiting the body is very much to the fore in St Augustine, and Aristotle's view, about which there is more to be said later, is to be found re-expressed in Aquinas. It is interesting to note that Islamic and Jewish theology of the twelfth and thirteenth centuries do a similar thing to Christianity - they develop a religious version of Aristotle.

Moving into the modern period we find that, while people carry forward the idea that what makes human beings the kinds of things they are is some property or substance possessed of a power of thinking, they have slightly different, and changing, theories about what it is. Nevertheless in the modern no less than the premodern period there is broad agreement on the supposition that human beings are distinguished from other things by having powers of thought and rational deliberation issuing in action.

Now when we get to the contemporary period a very large question looms in the minds of thinkers, philosophers and scientists. It is this: Whereas up to this point people were quite happy to allow that whatever it was that made human beings special and distinctive, it could not be identified with matter, that the mind or the soul went beyond the material, in the contemporary period people thought such an idea no longer tenable. In part this was because they believed in something known, roughly, as the unity of knowledge and the unity of the world. According to this belief in unity, the whole of reality must in some way be continuous. So if there are persons, or minds, or, if you like, souls, and it is these that make human beings distinctive, they must nevertheless be accommodated within and explained by the broad scientific categories that the contemporary period has adopted. It is in this period that there emerges what I shall call scientific psychology, that is to say, the attempt to try to understand thought, deliberation and action, in terms that can be accommodated within broad empirical science.

Some people were sceptical about the possibility of a scientific psychology, of course, notably religious believers who thought that since the thing that makes people people transcends the material order, scientific psychology is an impossibility. This is not to deny that aspects of human behaviour can be studied scientifically, only that what underlies human behaviour, transforms mere bodily movement into action and makes it the expression of intelligence, is not something that can itself be explained scientifically. Interestingly, in the last ten to fifteen years some scientists have themselves come to this same conclusion, and no longer believe that there can be a scientific psychology. However, instead of giving up on science, they give up on psychology. That is to say, they draw the conclusion there is no such thing as the mind as traditionally conceived. On this view, religious believers and the people of antiquity and the early modern period were right to think that the nature of the mind would, if it existed, make mind transcendent of matter. But since matter is all there is, there cannot be such a thing.

This explains the tendency in recent years towards scientific eliminativism, i. e. the view that there simply is no such thing as the mind. According to the eliminativist, the notion of the mind is a bit like the notion of a witch. There just is no such thing as a witch, There are still people who behave strangely, but we have learned better than to call them witches. Now we must think of them as, say, psychologically disturbed, though this description is not much better from the point of view of scientific eliminativism since there is no psyche to be disturbed. At any rate, what were formerly though of as witches are no more than complex lumps of matter, behaving in non-standard ways.

Thus baldly stated, scientific eliminativism can seem crazy. Isn't the thought that there are no such things as persons, because there is no such thing as thinking, itself a thought? Doesn't this position refute itself?. But scientific eliminativists are not quite as crazy as they appear. By their reckoning there are indeed sounds coming out of human heads, and people behaving in response to those sounds in ways that are patterned and regular, and so on. The mistake, they contend, is to suppose that the explanation of this range of observed behaviour must invoke such outmoded ideas as 'thought' and 'choice'.

This, in fact, is where computers and software start to enter the picture, because this kind of reductivist scientific materialism says something along the following lines: there are no such things as witches, but it does not follow that there was no behaviour which people in darker ages attributed to witchcraft. There were certainly human beings behaving in non-standard ways. The myth lay in supposing that the explanation of their behaviour required an appeal to witches or demonic possession or anything of that sort. Likewise, scientific materialism does not deny that people make noises and move in systematic ways, only that the notion of a mind or a soul is needed to explain it. All we need is the notion of the brain as a complex and intricate system which responds in systematic and regular ways to features of its environment.

How exactly would such an explanation work? Computers seem to come to our aid by providing a model of material objects that are very intricate and respond to features of their environment in complex ways. If we think of pressing on a computer keyboard as a kind of environmental input to the machine, then the machine responds by calculating something, printing something on a piece of paper, or even speaking sentences in English or some other natural language.

If we ask once more 'Could the Soul be Software?', here is one answer -- 'Yes - in a sense'. The old idea of the soul has to be discarded as a result of the movement of thought from the premodern to the contemporary period -- from the soul as the thing that thinks, to the brain as the thing that thinks, to the brain as hardware, in which the 'mind' or 'soul' is embedded, but only in something like the way of a computer program. Consequently, software, or the model of software, is what we are talking about when we talk about thinking, or choices, or decisions. That is really all there is to talk about.

II I myself think this sort of reductionism deeply mistaken, and I want not only to reveal its error, but to show what the truth of the matter is. Let us begin by focussing on the brain. What we might call 'physical personalism' holds that persons are animated human beings and that a man thinks and chooses. Brains, or souls for that matter, don't think and choose; people think and choose. This was in fact the view of the University of Aberdeen's most famous philosopher, namely Thomas Reid.

In 1751 Reid was appointed to the University of Aberdeen as a Regent, and on four occasions following his appointment he gave the graduation orations. Reid took the opportunity of these orations to lay out his philosophy and in the third he presents a consideration which is remarkably prescient with respect to this question about computers.

It is my intention at this time to set forth a few observations about the human intellect and its primary and most simple operations and to entrust them to the judgement and good faith of this learned assembly. Formerly I suspected, but now I know for certain, that the philosophy of the human intellect, even though it has been subjected to study by excellent minds in this generation and in the previous century, has yet right up to the present time been enveloped in darkness and based on hypotheses and fancies of the human mind rather than on an accurate analysis of the operations of the intellect.

Reid goes on to discuss these in various ways and he raises for us a very interesting problem. How is it possible that there should be such a thing as thinking? I can be sitting here in Scotland and thinking of St Paul's in London. How is that possible? The object is distant and yet somehow the mind can, as it were, pass through space, or indeed time -- I could think about some past event which, as I think about it, it is immediately present to me in memory. How is that possible? What Reid was concerned to deny was that the explanation could lie with any intermediaries, representations or ideas. Any theory which explains my thinking of some distant object in terms of my thinking of some internal mental object that in some way represents it, is inadequate for two reasons. First, if it were true that all we really think about are internal, mental, objects, we are not in touch with the world at all, and this seems absurd. As Reid says, it is the church I think of, not an image of the church. Second, the postulation of an intermediary does not solve the problem. To answer the question 'How do I think about things? by saying 'You think about them by thinking about representations' both leaves a gap between me and the object of thought, and worse, creates a second gap waiting to be bridged -- the gap between the representation and the thing.

It is this second issue that Reid turns to later in the third oration: 'How can there be representations of things?' He considers various possibilities, including the then dominant notion that these representations are images or pictures in the mind ; it possible to think about St Paul's churchyard in London by entertaining a mental image or picture of that churchyard. Now what makes something a picture of, say a human being? One immediately plausible answer is that it looks like a human being, but this faces a number of objections. First, which human being does it look like? Second, a simple drawing of a face looks a lot less like a human being than it looks like a series of lines. In fact, the answer to the question 'What makes this a picture of a human being?' is: we interpret it as a picture of a human being. But if this is correct, we haven't really escaped the original problem. If we explain how it is that a picture manages to be a picture of a thing by saying that we think of it in that way, we need then to explain thinking, which was what we were trying to explain in the first instance.

Reid considers an important alternative (and may well have been the first person to do so), that representation works not through images but through words. So he says:

Let us suppose that ideas represent things like symbols. In this way words and writing are known to express everything. So let the intellect therefore be instructed by ideas not in the manner of a camera obscura [not like pictures] with painted images, but like a written or printed book, teaching us many things that are external that have passed away and that will come to be. But this view does not solve the problem either, for who will interpret this book for us? If you show a book to a savage who has never heard of the use of letters, he will not know the letters are symbols, much less of what they signify. If you address someone in a foreign language perhaps your words are symbols as far as you are concerned, but they mean nothing to him.

Now I think this is a wonderfully prescient passage for at least two reasons. It anticipates a new theory -- that thought is not a matter of pictorial images but of language -- and it anticipates an objection to that view, one that is most famously associated with Wittgenstein. At one point in his reflections on these matters, Wittgenstein makes this point: if you understand Chinese, you hear a Chinese speaker, you understand what they say, But if you don't understand Chinese, you won't even know whether they're saying anything at all. To understand language you have to share the symbolic medium of the speaker. There is nothing intrinsically meaningful about any set of marks -- an English word, a Chinese character or the drawing of a cat. If I were to say that there is no more about a cat than is conveyed by the written word 'cat' or by a picture of a cat, I ignore the fact that by assigning a particular interpretation to a set of marks, we make them symbols. But we could as easily make an alternative set the same symbols.

This insight of Reid's is highly relevant to the question of whether or not computers think, because the activities of computers operate in terms of strings of marks or impulses or sounds. The question is "Does that constitute thinking?' In and of itself, it no more constitutes thinking than chalk on a blackboard constitutes writing. In order for chalk on a blackboard or ink on a page to constitute writing, it has to be associated with something else, namely a writer, that is, an intelligent thinker. Symbols and marks only mean something because thinkers invest them with meaning. Correspondingly, computer tape and impulses mean something only in so far as programmers and interpreters assign some interpretation to them. Here we encounter one good reason for scepticism about any attempt to understand human beings by analogy with computers; the analogy fails to capture one of the essential features of what it is to be a human person, namely to be a thinker.

It is also an important question whether it captures consciousness, the condition one is in when one is aware of one's environment. An interesting claim of Igor Alexander, the inventor of MAGNUS, is that a computer of such complexity is capable of consciousness, and, even more strikingly, that its consciousness can be observed. What does this ambitious claim refer to? On the left hand side of MAGNUS's screen there is an image of the thing that MAGNUS is oriented towards, and on the right hand side an image of various connection ways inside MAGNUS. If the thing that MAGNUS is oriented towards is changed, the left hand side of the screen is changed with consequent changes on the right hand side. Thus far described, however, this no more constitutes an insight into consciousness, than does the manipulation of a television camera which results in different processes and different pictures. All that is actually being observed is some aspect of the internal operations of this machine.

III Suppose that one is persuaded on these grounds that the capacity for thought and consciousness which is so central to understanding human beings, cannot either be eliminated or explained by advances in computer construction, or more generally, by any machine theory of human beings, where ought one to turn? In antiquity the power of human beings for consciousness, thinking, deliberation, action, and emotion was associated with the activity of a slightly strange entity -- the soul. Later came the attempt to accommodate this notion to scientific theory by converting it all into activities of the brain. Then, for various reasons, people came to worry about that possibility of that, and finally some just gave up on the idea of there being persons, because they were reluctant to give up on what science had shown about brains. Now if one gets to this position and thinks that it isn't satisfactory, where is one to go? Back?

It is an interesting fact about recent philosophical thinking in this area, that while the authors of essays in an anthology on the philosophy of mind twenty years ago would all have been some sort of materialist, by contrast a recent guide to current mind/body debates is divided into four sections, such is the variety of views now available. The third and fourth sections, in fact, comprise essays which in one way or another expressly reject the materialist picture. Only twenty years ago, then, orthodoxy was in one way or another materialist, whereas in the current state of things, while some materialists struggle on, other philosophers are inclined to give up materialism. There is a small, but growing band of people, who are going back to something like the ancient view of the soul.

As always in thinking about deep and difficult issues, people will tend to one extreme or the other. Suddenly the whole world becomes materialist, then materialism seems not to work, and so we all lapse back into some dualistic picture. So, let us start off with what we know. One of the things that we know about people is that we can tell a great deal about their mental states -- about how they feel, what they want, what they're thinking and so on -- simply by looking at them. A good judge of human beings is somebody who is able to tell a good deal about human beings simply by observing them. But what is being observed is not merely the motions of muscles, even less the motions of molecules, but actions, what they say and do, and what they refrain from saying and doing (You can tell a great deal about human beings by watching them when they're not moving).

Now, any account of what human beings are, has, I think, to hold fast to the fact that we can tell a great deal about them by intelligent observation. Most current philosophies of mind neglect this and do not pay sufficient attention to the fact we can tell a great deal about human beings just by looking at and listening to them. There is a well-known philosophical theory which believes precisely that, namely behaviourism, but it too veers away from the mainstream. Behaviourism holds that you can tell everything about the psychological states of human beings just by observing them. Its fault is that it loses the very interiority that is a central aspect of what it is to be a person. Not everything is on display. Interiority is an aspect of consciousness, of feelings and so on. It follows that an adequate philosophy will take account of the fact that there are such things as thoughts and actions, language, feelings and so on, and does not attempt to exclude all of that in the pursuit of a materialist account. On the other hand, it must not lapse into a position which puts everything into an interior that is entirely out of sight. This is effectively what dualism does, with the consequence that observing people is seeing the effects of their mentality or their psychology, never the thing itself. Action on this account goes on in the soul, and it may or may not have certain effects, which is why dualists are famously embarrassed by the problem of other minds. If having a mind is being possessed of an immaterial entity that (metaphorically) lurks deep inside the human body, how do I know anybody else has one, given that all I ever see is other bodies. So, we need a philosophy that avoids this difficulty and does justice to the idea that there are really are persons possessed of intelligence and so on.

Now, I think the only philosophy that does this is a position which holds that people choose and think, because then it is possible to say that you actually see the thing that does the thinking. On the pre-modern theory, you never see the thing that does the thinking since, though you see human beings, the human being isn't the soul, and it is the soul that does the thinking. On the modern view, once again you never see the thing that does the thinking. You see head and the limbs and so on, but the thing that does the thinking is the brain, and it is out of sight. Just as the first makes behaviour the effect of intelligence, so the second makes behaviour the effect of intelligence. By contrast, the view I am espousing holds that intelligence itself can be seen since it is incarnate in what you see, namely the intelligent movements of intelligent animals. Interestingly, such a position brings us back to the soul.

Consider a simple demonstration of human intelligence at work. Suppose I am drawing a square. With the edge of a pen I draw a square in space. Now what I want to say is that as I do this you see a soul in action, not the effects of intelligence but actual intelligence, intelligence publicly on display in my movements. These movements are not the effects of thought but embody thought as I move my arm. How could this be? For the answer we have to go to Aristotle and one of his mediaeval.

I said at the beginning that in antiquity people were animistic in their inclinations. They thought that the difference between a living thing and a non-living thing consisted in the fact that the living thing has something that the non-living thing lacks, a principle of life. Now a principle of life is an activating organisation. Matter is taken up in a way that is not reducible to that matter. Aristotle in his famous work the De Anima (On the Soul) identifies a vegetative soul, that is, a principle of life which a plant has, and which gives it powers of nutrition, growth and generation. But there is, he says, another kind of living thing, which is possessed of a different set of powers, powers of perception, appetite and locomotion. Still other kinds of living things have powers of memory, will and intellect. Now these different beings constitute a hierarchy because the third kind has all the properties of the second, and the second of the first, but not vice versa. A plant, for instance, is capable of nutrition, growth and generation, but in addition, a rabbit, say, is capable of locomotion, appetite and perception, while a human being is capable of nutrition, growth and generation and locomotion, appetite and perception and memory, will and intellect.

If we are to understand what it is to be a person, there is much to be said for returning to this older, Aristotlean, picture, according to which things are organised at progressively higher levels of activity. Things are the kinds of things they are in virtue of the kinds of powers they have, and activities at one level are not reducible to activities at a lower level. Just as locomotion cannot be reduced to nutrition, or perception to generation, so intellection, volition or memory cannot be reduced to perception, appetite, or locomotion. These are genuinely emergent higher level powers and capacities.

To refer to these higher activities as activities of the soul raises a further question, of course, because traditionally 'the soul' is something that might be immortal, something that might survive the body. So the question is: Does this account of the soul as a principle of the organisation of things give us any reason to think that there could be an immortal soul? Aristotle would not have thought that the vegetative soul, that by which a thing is a plant and is alive as a plant, is something that can survive the death of the plant. Once dead it is incapable of nutrition, growth and generation and so it does not exist any longer. So too with the sensitive soul. But when Aristotle came to discuss the rational soul, there are remarks in the De Anima which provide the basis of a very interesting argument. Aristotle's general principle is that a thing is as it does. A vegetative thing is what it is because it exercises vegetative powers. The nature of a thing is inferred from its capacities. This raises the question: Does anything have a power or capacity that transcends its organic nature. If so, could that power or capacity go on being exercised after the demise of the animal in which it was formerly embodied?

Intellectual activity is something that is not performed through any bodily organ. There is no organ of thought. This fact may give reason to entertain the possibility of something that survives death, something which we can call an immortal soul. I shall not pursue the issue further here. What I claim to have shown is that the soul could not be software, that there are such things as people, and that the best explanation of people is to be couched in terms of a hierarchy of activities that are not reducible downwards. The highest level of this hierarchy consists in activities that transcend material embodiment sufficiently far to make it plausible that the subject of those activities, the thinker, may indeed be something that could survive the death of its body.