CHAPTER 1

Everyone thinks that they know what is meant by emotion. After all, we feel what we call emotion all the time in our daily lives. However, once one tries to define what we feel and give those feelings names, one finds oneself in an ever-widening intellectual quicksand. Definitions of emotion use terms such as physical, psychological, psychical and mental, and assume easy distinctions between emotion, cognition and volition. One has first to question what is meant by these terms, and one cannot do this without being involved in the highly controversial issue, the Mind-Brain debate, with all its philosophical and ideological implications. Recent research into emotion is at, or near, the centre of this debate and cannot, therefore, be ignored.

If one consults the Shorter Oxford English Dictionary (1973), one finds that mind is "the seat of consciousness, thoughts, volitions, and feelings", or "memory", and that brain is "taken as the seat of sensation, the organ of thought, memory, or imagination". Mental is "concerned with the phenomena of mind", and physical is "of or pertaining to material nature; pertaining to or connected with matter; material; opp(osite) to psychical, mental, spiritual". Emotion is "a mental feeling or affection (e.g. of pain, desire, hope, etc.) as dist(inct) from cognitions or volitions". Although these definitions may seem quite satisfactory for everyday use, they beg a lot of questions if one starts to analyse them more carefully. For psychologists, who have spent a lot of energy arguing whether emotion is a mental or physical phenomenon, these definitions from the S.O.E.D. are clearly inadequate.

If one uses terms like spiritual, used above as an antonym of physical, in certain scientific contexts, one risks being treated with derision or simply ignored. Many people, who are conscious of this, yet believe the term to have a real meaning, are careful to point out that they use it only in the context of religious experience, claiming that one cannot mix science and religion. Yet those who so easily deride the believers, do not hesitate to use the mental/physical distinction freely, not only in everyday conversation, where language, often unconsciously, reflects more traditional ways of thought, but also in academic contexts where care with one's terminology is so essential. The careless use of the word mind in the latter context can lay one open to classification as a 'monist' or 'dualist', with all the implications this involves.

The problem of distinguishing between mental and physical, and related concepts, is by no means new, and it reappears whenever human behaviour is under discussion. From the earliest times, some people, the dualists, have argued that mind and body are two separate entities, whereas others, the monists, prefer the idea that they are different aspects of the same phenomenon. Although most philosophical and ideological theories from the early Greeks to the present day fit roughly into one category or the other, the single or dual principles in each case involve different concepts and, with dualism, different relations between these concepts. For example, Plato made the distinction between body and soul, and regarded the latter as the pilot of the former, a belief that was continued in Judaeo-Christian thought. He also saw the brain as the seat of the soul.

However, the terms dualism and monism were only coined in the 18th century, and are usually associated with Descartes’ dualism of mind and matter. Johnson-Laird (1987: 250), quotes Descartes as saying that "the body is an automaton, but the mind has free will and therefore lies outside the realm of scientific explanation". According to Descartes' theory, however, the non-material mind functioned through the pineal gland in the brain, and controlled the natural animal instincts of the automaton-like body. This posed a problem about the nature of the pineal gland, or brain, as a sort of interface between mind and matter, which set in motion a chain of thought that led to the Brain-Mind identity theory held by materialists today.

The scientific method of subjecting everything to empirical testing developed in the following centuries, and materialism was given increased impetus by Darwin’s theory of evolution. Belief in a soul or mind, and the notion that it distinguished man from the animals, was therefore undermined by the impossibility of physically verifying its existence, and by the logic of evolutionary theory. If one adds to this the growing view in the 19th and 20th centuries of the individual as a product of the social environment, the refutation of dualism of mind and matter by Ryle (1949) seemed to many to be the ultimate conclusion to be drawn.

Ryle’s negation of the "ghost in the machine", a soul or mind controlling the automaton of the body, was seen as a philosophical vindication of the scientific approach to psychology by the Behaviourists. As he said (Ryle 1949: 308): "The general trend of this book will undoubtedly, and harmlessly, be stigmatized as "behaviourist"". The brain was a physical organ, part of the body, and the subject of physically-based analysis. Any notion of Mind, as neuroscientists such as Bunge (1977) pointed out, is thus "unexplainable by science". These opinions, of course, conflict strongly with a long tradition in philosophy and religion which would consider such a materialistic conclusion as simplistic. However, before turning to the controversial problem of the nature of mind, let us take a brief look at the research on the brain which has fuelled this debate.

Whatever the arguments about the existence of mind, one point on which modern monists and dualists are in agreement is the existence of the human brain. Both will marvel at its complexity and agree that it is the physical organ which is either identical with or somehow related to mind. People have been trying to describe its structure and how it functions for centuries, yet the more we understand about the brain, the more complex it appears.

One of the more interesting results of the theory of evolution has been the research that has been carried out to discover how the human brain evolved. The brains of various species have been examined both from the point of view of physical shape and size in relation to body mass, and from that of the development of brain functions to suit different physical, environmental and social needs. A comparative study of this kind shows that the more intelligent an animal, the more complex the brain structure, that the development of intelligence follows a certain evolutionary pattern, and that the brains of apes are the most similar to human brains. Many are the theories that have been proposed for the why and how of the development of the superiority of the human brain over that of other species, and long have been the discussions over the exact quality of this superiority. Some theorists suggest that the evolution of the brain is still in progress: Carl Sagan (1977)[1] expresses the hope that, aided by super-computers, our brain power will now evolve ever more quickly.

A considerable amount is already known about the anatomy of the brain. Although it is not my purpose to discuss either this anatomy, or how the different areas of the brain, including those which appear to be involved with language, function or develop, as far as the Mind-Brain Debate is concerned, it is important to make certain points clear.

One of the positive results of Behaviourism was the stimulus it gave to research into the brain. A theory which saw the brain as a sort of 'black box' which processed the stimuli received from sensory information and then produced a response was naturally interested in working out the mechanics involved. Some of these experiments have been done on animal brains and while they tell us a lot about the animal brains, they are also considered important for the information that can be deduced from them about the human brain, a fact which in itself presumes a certain type of evolutionary process. Experiments on normal human brains can be conducted using technology like electrode systems and brain scanners, as well as drugs. Scientists using the latest techniques, like NMR, or nuclear-magnetic-resonance imaging, and EPI, or echo-planar imaging, in conjunction with high-powered computers, even claim that they are able to record the brain's activity accurately enough to show which areas process different aspects of conceptualisation and language[2]. Much of the more traditional research has, however, been based on the observation of people suffering from different forms of brain damage.

The mechanisms which transmit sensory information from various parts of the body to the area of the brain responsible for receiving the information have been studied in detail. For example, research into visual perception has demonstrated how neurons conduct the image received on the retina to the visual cortex, and research into other senses, such as hearing and touch, has led to similar data[3]. Certain types of cells which typically appear in the different areas have been identified. Neuroscientists, such as Hundert (1987), believe they are able to identify the areas of the brain which receive, process and, presumably, understand different sensory information. However, it is also recognised that the pathways of the brain cannot be described as simple one-way stimulus-response systems. Information supplied by the senses would appear to enter the brain at certain known points. However, the processing and eventual comprehension of the information, which in its turn prompts an appropriate response, would seem to involve complex systems of inter-communication between several parts of the brain. Researchers have managed to analyse the way some of these complex systems work, and, although few will deny the difficulties involved, there is a certain optimism that, with time, a fairly complete account of how the different parts of brain work and interrelate can be given. What would appear to be lacking is some generally acceptable theory on what actually coordinates all these functions.

One of the most complex areas of brain research is related to the nature of memory. It would seem that there are different types of memory for different functions ranging from a short term memory, which is useful for immediate recollection, to the long-term memory which enables us to remember things from our early childhood, and which is somehow stored in the brain. Experiments have shown that memories of the distant past can be stimulated by the application of electrodes to the brain[4]. This has encouraged the view that the brain resembles a computer in which information can be stored and retrieved by mechanical means, and has lead people to speculate on just how this physical mechanism works. One of the most difficult problems to solve would seem to be not so much how we form memory, but how and why we select and reject the material to be memorised.

The big difference between animal and human brains is the human brain’s capacity for language. The parts of the brain associated with language, and in particular those known as Broca’s and Wernicke's areas, have little or no counterpart in the brains of other species, although some scientists claim that traces of them can be found in the brains of the higher primates. Yet this does not mean that other species do not communicate, and one can point to a certain evolution of this type of communication into language.

While the Behaviourists were working in their laboratories elaborating on their S-R (stimulus-response) and conditioning theories, others observed behaviour in the world about them. It is significant that so much of this research is centred on the problem of intra- and inter-species communication. In the 30s, 40s and 50s, Konrad Lorenz and Niko Tinenbergen, who became known as ethologists, studied the behaviour of birds and animals, and were able to show that animal behaviour and communication were, to a large extent, genetically imprinted. Konrad Lorenz (1952) discovered that greylag goslings will accept the first moving object they meet after hatching as their 'mother', whereas ordinary farm ducklings are fussier - they expect the 'mother' to be nearer the ground and to quack! In much of the animal world the maternal instinct is genetically imprinted, as anyone who has witnessed even domestic animals such as cats, bitches or cows giving birth, will know. Man and the higher primates, however, have to acquire the knowledge of looking after their young from their social groups.

Bird and animal signals and sounds appear to be universal within the species and even develop in individuals kept in isolation from the rest of their species. The dance 'language' of the bees studied by Frisch (1967) showed that this type of communication can be highly complex. Animal communication seems to evolve in the interests of survival of the species, the genetic imprinting probably mutating to keep pace with changing needs. Psychologists such as Maynard Smith (1976), and Krebs (1987) have shown how animal signals evolve and become ritualised from more primitive behaviour, so that the present meaning of the signal may not be immediately apparent to the untrained observer.

The behaviour of the higher primates is of particular interest in any comparison of human and animal intelligence. There has been some well-publicised research into the behaviour of chimpanzees and gorillas in the wild by people such as Jane Goodall and Dian Hussey. Nearly as well-known are the experiments to train chimpanzees to use American Sign Language, or other symbol systems, for communication with the human experimenters. The emphasis of research is on how these animals can, or could be made to, reason and communicate. Sceptics (and behaviourists?) tend to view attempts at intra-species communication as clever conditioning by the experimenters. Although, despite the genetic imprinting described above, no-one doubts that behaviour can also be controlled and modified by conditioning, in the present intellectual climate which, according to Blakemore and Greenfield(1987: vii), allows that "even the humble laboratory rat [has] proved too resourceful, too thoughtful, too rich in insight to be treated as if it were a mindless machine", perhaps attitudes are changing.

The research into animal communication had eventually to point to the possibility of something similar for human beings. However, because of the traditional belief that animals and humans are essentially different, Chomsky’s proposal that certain language structures are part of our genetic heritage produced strong reactions in research both inside and outside linguistics, as the implications were important not only to philosophy and linguistics, but also to the natural sciences. Neurobiologists felt justified in treating language as another function of the brain, and Harris (1987: 507) quotes the biologist J.Z. Young (1978) as defining grammar as "the set of brain programs by which sentences are generated". However, the philosopher, Peter Hacker (1987: 487), describes the analogy used by Young as a "conceptual confusion". Harris (1987: 516), a linguist, goes on to describe this definition as a grossly simplistic view of what is known as either syntax or semantics, and describes the biologists' claim as so much "neuro-nonsense". What started as the theory of a philosopher /linguist has led to biological work based on his theory, but is contested by others in his own discipline.

It is significant, though, that neurobiologists feel the need to face the challenge presented by language. Recent research by neurologists such as Damasio and Damasio (1992) has shown that phonemes and syntax are processed in a specific part of the brain, that conceptual systems can be allocated to other parts, and that there are a third set of structures which mediate between the two. This research has been based on tests with people who previously had normal language faculties but who, because of brain lesions, now suffer from damage to parts of these three systems. Since lesions to the specific areas of the brain produce specific language disabilities, one can only conclude that a certain language capacity is innate. The extent to which we can postulate the innateness of whole syntactic or semantic structures, in the Chomskyian sense, is still a controversial question, but there seems little doubt that some language function is innate which, as a result of the developmental process, in the Piagetian sense, develops during childhood in a definable manner.

Emotion, too, seems to have its space in the brain, particularly in the limbic system. It is here, says Sagan (1977: 64), that we find the pituitary gland which dominates the human endocrine system, and "the mood-altering qualities of endocrine imbalances give us the important hint about the connection of the limbic system with states of mind". One can also say the reverse, that states of mind (which can be emotional) cause endocrine imbalances which provoke psychosomatic illness; for example, gastric ulcers. Another part of the limbic system, the amygdala, is particularly associated with fear and aggression. If stimulated electrically in animals it can "rouse them to unbelievable states of fear or frenzy" and, if extirpated from "ferocious animals" they "become docile"(ibid: 64). Drugs which are used to control or stimulate mood usually act on the limbic system.

In the human brain, as Eccles, in Popper and Eccles (1977: 247), observes, the limbic system has developed from "the old olfactory (smell) brain", which is of major importance to the lower mammals both as regards actual area of the brain and its function. It may be for this reason that the sensory information from our sense of smell projects directly into the limbic system before being processed by other parts of the brain[5]. However, the way the somasthetic, visual and auditory systems project sensory information is far more complex. It would appear that although the information is projected semi-directly to the limbic system, it is simultaneously processed along other pathways of the brain to the prefrontal lobes, which in their turn project to the limbic system in a way that shows that "the prefrontal and limbic systems are in reciprocal relationship and have the potentiality for continuously looping interaction" (Popper and Eccles 1977: 273). This means that our conscious perception is modified by emotion and the other appetite drives, such as hunger, thirst, sex and aggression, which depend on the limbic system. The "looping interaction" also means that there is cross-reference of information between different senses. The limbic system also, in some as yet not totally clearly understood fashion, has an important part to play in the formation of memory.

Another factor about emotion which is of interest when one is thinking of the brain and innate structures, is that, like language and psycho-motor development, it, too, seems to develop according to a certain pre-established pattern in the individual. Whether one is talking about normal development, on which to base theses of primary and secondary emotions, or abnormal development, with a view to showing how early interpersonal and social factors may affect behaviour, an individual's emotions would also seem to depend both on genetic and environmental factors, as biologists like McNaughton (1989) are at pains to demonstrate.

A well-known neurologist, António Damasio, has recently ventured into this debate. His book (1995), called Descartes' Error, uses neurological evidence to argue that Descartes´ theory of dualism was essentially flawed, and that not just the brain, but the sensory information fed in by every fibre of our bodies contributes to the way we function. He further argues that a specific area of the brain, the pre-frontal lobe, is responsible for emotion. If this area of the brain is destroyed or damaged by accident, the subject not only ceases to feel emotion, but also seems unable to collate sensory information in a way which helps him/her interact with the social environment satisfactorily. Apparently neither memory nor acquired knowledge are affected, but the individual becomes unable to make decisions, or react either sensibly or sensitively to other people. Furthermore, primates who have been subjected to excision of the pre-frontal lobe become such social misfits that they are rejected by their group in a way in which the physically handicapped are not.

If, as this type of research tends to suggest, emotion is so inextricably bound up with how the brain physically perceives and what it remembers about the world, and yet, in its turn, both affects and is affected both by the more rational brain processes, and by what are referred to as the 'subconscious' and 'unconscious', one can begin to understand why there has recently been so much interest in the relationship between the processes of the mind / brain and emotion.

Once one turns from the discipline of biology to those of psychology, linguistics and Artificial Intelligence, or AI, one finds the term mind is preferred over brain even by those who would consider themselves materialists or monists. More often than not no specific justification of this is given, but those who do feel uneasy about not making a distinction prefer to avoid mind. Chomsky talks about Language and Mind, yet neurologists would point out that when he says that our faculty for language is genetically programmed, he is presumably claiming this to be true of the physical brain, and not some immaterial mind. A dualist will understand the distinction, but a strict monist will claim that mind and brain are identical and the distinction irrelevant. Some like to think of the brain as the 'structure' and the mind as the 'function', while others prefer to make a distinction between the brain as the 'hardware' and the mind as the 'software' of the same phenomenon. If one uses this metaphor, Chomsky might be said to be referring to the 'electronic circuits' rather than the programmes - an area where the distinction between hardware and software sometimes becomes more difficult to draw.

Some writers use the word self rather than mind, either to distinguish something they regard as interacting with the brain but somehow distinct from it, or to refer to some sort of global consciousness which, while remaining part of the brain, is, in some way, the sum of all the functions of the brain. Others prefer to discuss the problem using the term consciousness. Yet a couple of decades ago or even less, to contemplate a discussion about concepts such as mind, self or consciousness in a scientific context, would have been considered old-fashioned and, at best, embarrassing. Now that the intellectual climate has changed, and mechanistic materialism has been found to be less than satisfactory, the debate has reopened, and it is not just the philosophers who are taking part, or even the representatives of the 'softer' scientific fields, such as psychologists or sociologists; distinguished mathematicians, physicists, biologists and neurologists are heavily involved as well.

When the philosopher/neurologist team, Karl Popper and John Eccles, two eminent figures in their particular fields, decided to publish their book The Self and Its Brain in 1977, in which they argue for a type of dualism they call Interactionism, they realised how controversial they were being. The very construction of their book suggests that their intention was to stir up controversy. Each presents strong arguments based on their own discipline and research in the first two sections, and the third section is devoted to transcriptions of their discussions of their views. However, the fact that the controversy has gathered momentum since then shows that others, while not necessarily agreeing with them, feel the subject to be important and relevant. In the following sections I shall attempt a brief outline of their views as well as some of the more recent controversial theories which should demonstrate the complexity of the problem. I shall not be contemplating any obviously religious theories here, although, quite obviously, any debate of this kind must reflect aspects of religious thought, either consciously or unconsciously.

Popper starts by challenging materialist reductionism, in which every level of life is the result of the one below, until one reaches the basic combinations of (particles of particles of?...) of atoms. He claims it is invalid because it is highly simplistic and does not take into account “downward causation” (the effect of higher on lower levels), and goes on to say that it implies a mechanistic determinism which no longer conforms to the findings of modern quantum physics. He argues instead in favour of emergent evolution and cosmic evolutionary stages as follows:

He then discusses the various theories down the ages that have argued for or against a distinction between mind and body. He describes himself as an agnostic who does not accept the idea of a soul, and agrees that he shares a lot of the materialists' views on material reality. His arguments in favour of Self take most theories about the evolution of the brain for granted, as well as the various theories of innate and acquired knowledge, and the importance of language in the development of human consciousness. He also points out that only the more extreme forms of materialism refuse to admit the existence of consciousness, and explores the relationship between the conscious and unconscious aspects of Mind or Self.

He argues that "the human consciousness of self transcends..... all purely biological thought" and that although "animals are conscious", "only a human being capable of speech can reflect upon himself". He continues, "I think that every organism has a programme. But I also think that only a human being can be conscious of parts of this programme, and revise them critically" (ibid: 144). He develops this theme with reference to various aspects of human consciousness, and yet does not claim to offer any grand solution, for he feels that "We know that, but we do not know how, mind and body interact" and "that this is not surprising since we have really no definite idea how physical things interact" or "how mental events interact" (ibid: 153). He predicts "that we shall not be able to build electronic computers with conscious subjective experience" because "computers are totally different from brains, whose function is not primarily to compute but to guide and balance an organism and help it stay alive. It is for this reason that the first step of nature towards an intelligent mind was the creation of life, and I think that should we artificially create an intelligent mind, we would have to follow the same path". (ibid: 208) He believes that neither philosophy nor science can explain, or explain away, human self-consciousness satisfactorily.

Eccles' contribution to the book is first to explain the principal discoveries made in relation to how the brain is constructed and how it functions. He then proceeds to examine the separation of self and brain from the neurological standpoint. He first draws attention to experiments done by Libet and Mountcastle to establish the time factor between a subject’s receiving a skin stimulus and the cortex of the brain showing suitable signs of stimulation. The crucial, and highly controversial, point of these experiments was that subjects reported experiencing the skin stimulus before cortical stimulation was reported. This would suggest that something (a self? a mind?) is conscious of what is happening without relying entirely on brain functions, a position which is obviously contrary to the monist view that our understanding of our experience can only be the result of the processing of information by the brain.

Eccles then goes on to discuss the interesting research done on 'split-brain' patients. These are people who have been subjected to commissural section, or commissurotomy, an operation which used to be performed in an attempt to alleviate acute forms of epilepsy. It meant that the corpus callosum was severed and that the two hemispheres of the brain became physically separated. Experiments seem to show that the subjects appear to be only conscious of the activities controlled by the dominant hemisphere, usually the left, which, amongst other things, is where the language centre of the brain is to be found. The right hemisphere continues to function and respond to stimuli which are received by it, in a way which Eccles describes as "having a status superior to that of the non-human primate brain" displaying "intelligent reactions" and "learning responses", but "it gives no conscious experience to the subject" (ibid: 328).

Eccles puts forward the hypothesis that these experiments show that the conscious self is not, as the monists presume, the synthesis of all neural activities, and that a case can be put for a conscious self in World 2, which is related to the brain in World 1 through a "liaison brain", which is adjacent to that part of the left hemisphere which controls ideational and linguistic competence (ibid: 327).

Eccles is not afraid to question what happens to the self-conscious mind after death nor to assume his position in favour of a "strong dualist-interactionist hypothesis" (ibid: 374). Later, in Eccles (1987: 203) he is considering the "possibility of being able to account for the mind-brain interaction in terms of the action of non-material fields, such as the probability field of quantum mechanics", but he concludes that "modern physics presents no insuperable objection to dualist-interactionism" although "we are only at the beginning of our attempts to understand the interaction [his italics] of mental events with neural events" .

The theory of Self put forward by Popper and Eccles is, obviously, not as simple as it appears here, and it should be considered in the context of other proposals to explain the phenomena referred to as Self, Consciousness and Mind. Much of contemporary discussion of these concepts has taken place since 1977, perhaps partly in reaction to Popper and Eccles' publication. There is no doubt that the fact that Eccles is a Catholic makes the scientific world suspicious that he may be trying to identify the idea of Self with the notion which is taboo in certain scientific circles, Soul.

Szengtágothai (1987) probably sums up scientific reaction to the hypothesis of an interactionist form of dualism when he says that "If understood and used metaphorically, the three-world concept [his italics] envisaged originally by Popper and elaborated by Eccles is a fascinating view of great beauty, which may be of heuristic value", but he feels that, in order to follow the idea to its logical conclusion "we should have to leave the domain of legitimate science and enter that of religion and faith". He himself, as a biologist, prefers an idea of Self which can be understood roughly as some kind of 'wholeness' which is observable in most animals, developing into "the still rudimentary self-consciousness of a non-human primate" and finally into "true human self-awareness". He believes in some kind of evolution of 'selfness', and finds emergent materialism insufficiently explanatory.

The idea of Self, or person, has two main interpretations in more scientifically based psychology, which Parfit (1987) describes as the Ego Theory and the Bundle Theory. An Ego theorist believes that there is a person, or subject, who serves to unify someone's consciousness, and that this person is a "persisting purely mental thing", the best-known form of which is the Cartesian soul, or spiritual substance. The Bundle theorist argues that one cannot impose the sort of unity proposed in a person on what is claimed is a whole life of different mental states and events which are "unified by various kinds of causal relation, such as the relations that hold between experiences and later memories of them. Each series is thus like a bundle tied up with string". A more extreme version of the Bundle theory argues against any real individualism, because each person is merely the sum of a combination of cultural, social and, possibly, genetic factors.

Not every scientist who talks about Self makes this type of distinction overtly, although one may sometimes deduce which of the interpretations is implicitly being used from the more general context. The term consciousness or the more passive awareness, is sometimes used in a scientific context in order to avoid self, but I shall not enter here into the implications of what each individual means by these terms at this stage, or what complications arise with self-consciousness and self-awareness - I include these points merely to demonstrate the complexity of the problem.

The idea of consciousness, or self-awareness, is developed by other scientists from the point of view of the individual and his relationship to others. Biologists like Barlow (1987), and experts on animal behaviour like Humphrey (1987), see consciousness as a phenomenon that has developed for evolutionary reasons to promote the type of social life of which human society is the richest and most complex. Barlow points out that studying individual nerve cells, or even groups of nerve cells and the relations between them, may provide interesting information, but that "the picture changes dramatically if you consider whole brains, for then types of behaviour occur that result from one brain interacting with another" (ibid: 363).

It is this type of reasoning that leads to what some call 'top-down' materialism, often used to counteract the type of attack made by Popper against the reductionists who favour 'bottom-up' materialism. Attempts have been made to account for higher level or macro-properties which are not causally explainable by lower level or micro-properties. These terms are taken from Searle (1987) when he argues for Mind and Brain as two aspects of the same phenomenon. He asserts that the mind is the macro property which produces consciousness, whereas the brain can be divided and reduced to its basic neurons or micro-properties. He considers his approach as an "obvious and common-sense view" and argues that "one conclusion [that] emerges clearly from even the most cursory investigation of the functioning of the brain [is that] "mental phenomena, whether conscious or unconscious, whether visual or auditory, pains, tickles, itches, thoughts and all the rest of our mental life, are caused by processes going on in the brain. " [his italics] (ibid: 220) He also claims that "in principle, we could have all of our mental life without any of the appropriate stimuli or any of the normal external behaviour" (ibid: 231). It is important that although a materialist and, from his point of view, a monist, who believes in mind-brain identity, one of the reasons for developing the above arguments is to help support his refutation of the claim by strong AI supporters that a computer "literally has a mind in exactly the same sense that you and I do".

Psychologists, in their efforts to be considered a science, still shy away from the concept of the Individual. They prefer to discuss something that is physically verifiable, like the neurophysiology of the brain or other physiological phenomena, or something which it is more respectable to study scientifically, using statistics and group testing, like social groups. The notion of an individual is so complex that it is safer to study tendencies in groups and the way they eventually affect the individual.

The scientific doubts about the nature of Self expressed above may be all very well for psychologists and scientists, but psychiatrists and psychoanalysts, who dedicate their lives to unravelling the complexity of individuals, become impatient with such theories. As Von Franz (1972: 15) says, the idea that we each have a concept of Self, or an 'ego', "is the commonest and most normal complex among human beings". Besides, most languages distinguish various forms of Self and Other, albeit in differing forms, and, at least in the Western religious and philosophical tradition, there is a complex linguistic system for referring to these phenomena.

Individual is seen as a complex concept which has different aspects. Most people who speak English or Portuguese will easily recognise body, mind and soul, or corpo, mente and alma, as different aspects of the individual, even if they have reservations about their exact interpretation. Philosophical theories on the nature of the phenomena thus expressed abound, but perhaps a brief look at how the two most influential figures in psychoanalysis in this century, Freud and Jung, have dealt with the problem is one way of giving an idea of its complexity.

The view of Freud was that the psyche of an individual could be shown to have three definable aspects, which he called the id, ego and superego. The id is the unconscious function which is formed partly by genetically programmed instinctual elements, and partly by acquired elements which, for various reasons, are repressed and not recognized by the conscious mind. The ego, although much of its activity is unconscious, is responsible for voluntary movement and thought, and functions "at the interface of the perceptual system and the internal demand system" (Badcock, 1992: 168). The superego forms part of the ego and exercises "functions of self-judgement, reality-testing and ethical and aesthetic evaluation" (ibid:173).

Furthermore, although Freud used the term subconscious in his early work, later on, he and his followers preferred to make the distinction between the conscious, the pre-conscious and the unconscious. From this point of view, as expressed by Badcock (1992), the conscious is what "occupies the attention at any given moment" (ibid:167), and the pre-conscious is "that which is not present in consciousness, but which may have access to it" (ibid: 172). According to this analysis, the unconscious refers to the contents of the mind which are not present in the conscious at any given time[6]. These definitions by Badcock, who readily admits the influence of evolutionary theory on his interpretation of Freud, fit in quite neatly with the views on the Mind or Brain expressed above.

Freud recognized that certain aspects of human behaviour were attributable to primal instincts which were inherited, and Neo-Freudians have been at pains to validate his theories in the light of modern genetic and evolutionary theory. However, the focus of his work was on the way the individual's instinctual behaviour developed dynamically under the influence of the family and the social environment.

Jung was more interested in discovering a blueprint for the human psyche which attempted to account not only for types of personality, but also for the symbols and myths found in human culture which seem to have a universal validity. At the level of the human individual, he distinguished a persona, as the personality or 'mask' which the individual presents to society, and the ego which was "a complex of ideas which constitutes the centre of my field of consciousness and appears to possess a high degree of continuity and identity". He further distinguishes between the ego, "which is only the subject of my consciousness", and the self "which is the subject of my total psyche, which also includes the unconscious" (Jung, 1971: 425).

Jung's view of the unconscious is very complex, and, he believes it is ultimately unknowable for this reason, but he insists that it is "an exclusively psychological concept, and not a philosophical concept of a metaphysical nature" (ibid: 483 [his italics]). He argues for a genetically transmitted 'collective unconscious' which includes far more than the primal instincts necessary for survival. Although he does not go so far as to argue in favour of innate ideas, he believes that tendencies to formulate certain ways of thinking are inherited[7]. This collective unconscious is dynamic and evolves both to promote and to keep pace with human cultural development. The individuation process, which is the subject of psychoanalytical analysis, is that by which the individual ego develops as a conscious entity, separate from, but dependent on the collectivity. In his opinion, although the tendency to individuality is present in all human beings, and particularly in more culturally and intellectually developed people, by no means everyone undergoes an individuation process to a degree high enough to differentiate them substantially from this collective entity.

Like Lorenz's theories on the genetic programming of animal signals, Jung's ideas, which implicitly suggested the notion of cultural evolution and, at the same time, recognized different individual capacities for development, were used to support the cause of Nazism. It is not surprising, therefore, that their theories were not acceptable to the intellectual establishment which espoused the more egalitarian cause of Behaviourism, or the analysis of Freud of basic primal instincts being strongly influenced by the individual's environment. Their theories are now more acceptable for several reasons, but perhaps the most important ones are the findings of modern genetics, and the current view of the symbiotic relationship between the individual and society.

Like most disciplines, there are various schools of thought in the area of psychiatry and psychoanalysis. Some will describe themselves as neo-Freudians, by which they mean that they maintain many of Freud's theories, albeit in a modified form, about the effect of primal instincts and early experiences. There are those who follow in Jung's footsteps and endeavour to probe the collective unconscious of Man by examining mythology, religion, dreams and various forms of literature, old and new. Others follow Adler and his disciples and give much more importance to the complex effects of social background as well as family structure on the individual throughout life. Much of this previous theory has been re-analysed in the light of neo-evolutionary theory, and yet others turn to the evidence of new drugs to show that all psychic life is the result of physical reactions anyway.

So nowadays a rampantly promiscuous man may have his behaviour explained in several different ways, depending on the point of view of his analyst. It may be seen as the result of some Freudian childhood experience, or as a manifestation of the 'Don Juan' Jungian archetype. On the other hand it may be described as the result of familial / social competition with, and influence from, his father, siblings or peers. Evolutionary theory will describe it as a sign that he possesses strong genes and is doing his best to ensure they are transmitted to the next generation, and the more physically-minded will describe it as simply a problem of chemical imbalance of hormones.

Whatever the explanation, and few sensible people would interpret the problem from the exclusivity of one point of view, there seems little doubt that the individuality and complexity of the Self is unquestioned. Psychiatry and psychoanalysis grapple every day with the phenomenon and, despite the similarity of the experiences, fantasies and dreams of many of their patients or clients, few would argue that each individual is not different and complex. Naturally, psychiatrists and psychoanalysts, whatever their differences, regard the individual and his / her emotions as their special field of interest, and perhaps have a vested interest in promoting this point of view, but most would probably view with horror the possibility that anyone might want to reproduce these phenomena in machines, or artificial intelligence.

One of the principal reasons why so many people are now engaged so heatedly in the Mind-Brain debate is because of certain claims made by cognitive scientists for AI. Science fiction is full of talking robots, and the popularity of Science Fiction films, in which they are fairly believable, in comparison with some of the weirder monsters dreamed up by Hollywood, have contributed to making the possibility of their existence acceptable, if not highly desirable, to the general consciousness. Since so much science fiction, of the more serious variety, has a tendency to reflect future fact, the hypothesis of such machines (?beings) should be taken seriously. According to strong AI enthusiasts, we shall one day be able to build computers and robots which will be equal, and probably superior, to men in intelligence, able to take decisions, think, believe and - why not? - fear, or love. Followers of the television series Star Trek will be familiar with the problems that Lieutenant Data, the computer in human shape, has been having over learning to have emotional experience and 'self-awareness'.

Now that computer science has increased the possible brain power of the computer to a level unimagined even a couple of decades ago, cognitive science, which works towards the development of AI as an imitation, or a parallel, of Biological Intelligence (= BI), and attracts people from the disciplines of psychology and linguistics as well as computer programmers, is working towards this end. The ability of the computer, once properly programmed, to out-perform most human attempts at mathematical calculations, to memorise vast quantities of information that can be recalled almost instantly, and to carry out certain mechanical but time-consuming tasks more rapidly, has encouraged this research.

Those studying the physiology of the brain may not believe that it is either possible or necessary to imitate it in order to produce a robot, but an understanding of the mechanics of the brain is expected to provide insights into how AI can simulate the functions of the brain. The mechanics by which the brain receives information from the foot, or instructs it how to move, is of interest in helping solve the problem of how to teach a robot to manoeuvre itself. The difficult problem to solve, however, is how this information can be coordinated with a perception of its surroundings so that, having lifted its (heavy steel-plated) foot, it then avoids placing it on the experimenter's big toe. This perception in the animal world is a function of the brain, and if language is the function of the human brain, by means of which we conceptualise and express our perceptions, the crucial problems for AI are to discover how the brain perceives, how it thinks and how it encodes its thoughts in language.

If a computer is to use language, language must be formally encoded in a way that is acceptable to the computer. The emphasis of the Structuralists on syntax encouraged the idea of formalisation of language, and Chomsky's view that these structures, and possibly certain semantic universals, were somehow programmed into the brain fueled the hopes of AI that semantics, too, could be formalised. The idea of deep level phrase structures and transformational rules was appealing. A lot of work had already been done on showing which verbs used which structures, and deep case theory and componential analysis of lexemes were expected to help the computer avoid sentences like The chair bit the man and make sure it produced The dog bit the man. However, The man bit the dog posed greater problems. Did one simply rule it out as too unusual? If not, how did one construct the necessary semantic premises for the computer to know when to allow it? Rules of collocation and context proved, and are still proving, difficult nuts to crack. Some experts continue to search for further refinement of analysis from the phoneme to the text; others try to work from the knowledge implicit in a certain state of affairs to the text and thence to the units of text. Many are sceptical that any mechanical device, however sophisticated, can ever imitate more than the simpler and more literal aspects of human speech. Others, like Johnson-Laird (1988), declare that although "the enterprise [of making knowledge explicit] is enormously time-consuming.. if there are any insuperable barriers to it, they have yet to be discovered".

Strong AI enthusiasts will argue that even the humble thermostat, a concrete example of a simple Turing machine, has the basic elements of intelligence. Searle (1987: 211) describes how he once asked the inventor of the term artificial intelligence, John McCarthy, who claims that "machines as simple as thermostats can be said to have beliefs", what beliefs the thermostat had. The perfectly serious answer he got was "My thermostat has three beliefs. My thermostat believes it's too hot in here, it's too cold in here, and it's just right in here".

This line of thought argues that providing a Turing machine, or computer, can be implemented in a way that a human subject is unable to distinguish between the answers provided by it and those of another human being, it is demonstrating the same 'intelligence' as that human being. Examples of this type of experiment are the computer programmes which 'interview' patients either for diagnostic or psychoanalytic reasons. The 'diagnostic' type programme asks the patient a series of questions and then computes a possible diagnosis from the data at its disposal. Claims have been made for a degree of accuracy of diagnosis surpassing that of the human (and possibly overworked) doctor. The 'psychoanalytic' programme keeps the patient talking by picking up clues from the patient's answers with which to form the next question in the interview. For example, if the patient mentions his father in response to a question like who do you consider the most important figure in your childhood?, the machine will perhaps respond with Tell me more about your father. If the interview satisfies the physical and psychological needs of the patient, certain strong AI enthusiasts might argue, what is the functional difference between the machine and a human doctor or psychoanalyst?

When A1 experts talk about the future of computers they usually restrict themselves to the term intelligence, and yet the more enthusiastic clearly include in this what others would prefer to discuss as consciousness. As we have seen, Popper did not believe it possible to produce consciousness mechanically and this theme has been taken up and expanded upon by others. One of the better known refutations of the possibility of consciousness for computers is that known as Searle's 'Chinese room'[8].

With this argument, Searle rejects the mind/brain = program/ hardware equation. He further adds that his argument has nothing to do with the difference between the 'wetware' of the brain and the 'hardware' of the computer, or whether the hardware is made of "silicon chips", "beer cans" or "some quite unknown chemical substances", and insists that "you cannot duplicate the causal powers of the brain solely in virtue of instantiating a computer program, because the computer program has to be defined purely formally" (1987: 215). He believes that "it is a typical but unstated tacit assumption behind many of the implausible contemporary doctrines concerning the mind - doctrines such as behaviourism and artificial intelligence - that it is simply impossible to accommodate a naive common-sense account of the mind with an overall scientific world-view", and that it is "the sense of desperation" caused by this assumption that leads people to propose these implausible doctrines. He then goes on to put forward his view of mind-brain identity described above.

AI experts such as Gregory (1987) defend AI by saying that Searle is assuming that the computer will remain unable to perceive the world about it and therefore be incapable of learning through experience. He asks us to "consider bringing up a baby in the Chinese room: How could it learn the meaning of the Chinese or any other symbols, in such a restricted environment?" (ibid: 240) He puts forward the case that computers can eventually be made to 'learn' in the way a child may be said to learn, and thus acquire a wider knowledge of the world similar to that of a human being - it is just a question of discovering the best way to conduct this learning process. He admits that, given that both BI, and AI have "a finite number of operating states", BI has evolved strategies for preventing itself from being overwhelmed by information, whereas AI still suffers from the danger of "combinatorial explosion" because of the difficulty in finding a way of programming the computer to select relevant, and reject irrelevant, information according to necessity. This is the reason why there is so much interest in the way human intelligence selects only what it considers essential to memorise, notice etc. He sees this problem as difficult and central to AI, but not ultimately insuperable.

Johnson-Laird (1988) says "the goal of cognitive science is to explain how the mind works. Part of the power of the discipline resides in the theory of computability". He admits that "there may be aspects of mental life that cannot be modelled in this way" and that "there may even be aspects of the mind that lie outside scientific explanation", but argues that the use of a theory of computability, whether or not it is ultimately right, introduces a measure of scientific rigour. He does not make extravagant forecasts of intelligent robots rolling off the production line in the near future, let alone of machines with 'consciousness', but he believes that, with hard work and time, considerable advances can be made in the understanding of the human mind by cognitive science.

An interesting point for linguists of a lot of the discussion involved in all this is that while the protagonists of the debate rage about the ability or otherwise of brains and computers to process syntax and semantics, they are also conscious that the 'other side', if not themselves, is trapped in its own conceptual apparatus. Thus one of Searle's (1987: 223-4) arguments in favour of a Mind = Brain solution is to compare the problem to the scientific statements "lightning is caused by an electrical discharge" or "the colour red is caused by photon emissions with a wavelength of of 600 nanometres", which, he argues, are only another way of saying "lightning just is an electrical discharge" or "red just is a photon emission of 600 nanometres". Presumably, instead of saying mind is caused by brain, we could say mind just is brain. Gregory (1987: 243), on the other hand, accuses Searle of "the unfortunate conceptual inertia to which philosophers are prone". Semantic inertia may be useful for protecting words, he says, but "language must continuously change to keep pace with new discoveries and changing concepts, or it becomes misleading". Are the objections to asserting that a thermostat has 'beliefs' based on an inability to extend such a concept to non-human entities? The problem of conceptualisation will be discussed more fully in the next chapter, so we will not not pursue this question further here.

One of the latest contributions to the Mind-Brain debate is that of Roger Penrose (1989). Like several of the people already cited, and many who have not been, he believes we simply do not know enough to be able to describe consciousness, let alone programme it into a computer. As a mathematician and a physicist, he first takes the reader on a tour of classical and modern physics and mathematics, and then proceeds to use theories from these disciplines, and others, to argue against the possibility of producing human consciousness in artificial intelligence. While agreeing that a lot of the brain's activities are algorithmic, like a Turing machine, he points out that they are usually those which can be classified as "unconscious" and can be described as "automatic", "following rules mindlessly", or "programmed". Aspects of brain activity in which consciousness is involved - "common sense", "judgement of truth", "understanding" and "artistic appraisal" - proceed "in a way that cannot be described by any algorithm" (ibid: 531). He claims that not all aspects of mathematics are algorithmic and that the very judgement of whether an algorithm is true or false is non-algorithmic and dependent on consciousness.

He believes that neither classical nor quantum physics are yet able to offer solutions to the problem of consciousness, and that further progress must be made in physics before any explanation can be attempted. Like Eccles, he refers to the type of experiment done by Libet and Mountcastle and to "the strange role of time in conscious perception" (ibid: 573) . The interesting point about these and other experiments is the time factor involved between receiving a stimulus, the brain activity related to it and the subject´s consciousness of the stimulus. The experiments are varied and complex, but the essential point that seems to emerge from them is that the time measured between stimulus and consciousness is not subjectively experienced. The other point that raises problems is that the combined intervals of time that can be measured between consciousness of receiving, plus the time for conscious processing and the time for consciously reacting would make it impossible for consciousness to enter into activities such as playing ping pong or even ordinary conversation, whatever role the subject's presuppositions and expectations may play in accelerating the process.

He discusses the putative theory of CQG ( = correct quantum gravity) and envisages that "once it has actually been found, it may then be possible to elucidate the phenomenon of consciousness", and that if this phenomenon depends upon CQG, "then consciousness itself will fit only very uncomfortably into our present conventional space-time descriptions" (ibid: 578). His view is ultimately materialist and not the dualist vision of Popper and Eccles, but he simply "cannot believe that [consciousness] is something just conjured up 'accidentally' by a complicated computation. It is the phenomenon whereby the universe's very existence is made known". (ibid: 580).

Perhaps more radical materialists will sense in some of the theories presented an element of wishful thinking of the kind put into words in a popular comic film, Creator, in which Peter O'Toole plays an eccentric scientist. This scientist expresses the feeling that when Science finally gets to the top of the Hill of Knowledge it may well find God has been sitting there waiting for it all the time! Perhaps the sceptics are right, but there are still a lot of people who would prefer to think that maybe they are not.

By this stage I can imagine readers may be wondering what all this has to do with the subject of the language of emotion, and they can be forgiven for suspecting that I may be working out some bigger ultimate question for myself. Perhaps I am. However, I was originally drawn to this debate by the fact that so many of these theoreticians come back to perceptions, feelings and emotions as a touchstone by which to judge the validity or otherwise of their theories.

Nowadays we discuss whether or not animals can be said to display elementary consciousness and emotions from the point of view of evolutionary theory, but the debate is not new. According to Clark (1987: 139-49), Spinoza argued that emotions in animals existed but were irrational and different in quality from the emotions of men. Descartes claimed that animals were machines which did not have souls, and were therefore incapable of feeling. If one is going to attribute souls only to Man, it is rather inconvenient to have to acknowledge that animals might have feelings or emotions. Yet it is significant that only those emotions which are socially disapproved of are easily attributed to animals - when Man has these emotions, he is exhorted to resist the 'bestial' side of his nature. The presumed inability of animals to feel has been basic to the way we treat them, and Thomas (1983) traces how this type of reasoning has affected the way animals have been used by human beings, whether for purposes of work, sport or food. The big quarrel between researchers using laboratory animals, or film stars wearing mink coats, and animal rights activists is a practical extension of this controversy, in an age when Man has become more conscious of his relationship to the animals.

Emotion seems to be inextricably bound up up with ideas of Self and Consciousness. Popper (1977: 127) suggests "as a wild conjecture" that "it is out of four biological functions that consciousness emerges: pain, pleasure, expectation and attention" and that "perhaps attention emerges out of primitive experiences of expectation and pleasure". As we shall see, later on, when talking of 'basic' emotions, he is not alone in thinking this.

Searle (1987: 230) maintains that one cannot explain the intentionality of mental phenomena in terms of social relationships because "the actual mental states, beliefs, desires, hopes, fears and the rest of it have causally sufficient conditions that are entirely internal to the nervous system". Here and elsewhere he includes hope, fear and anger, which are often classified as emotions, as mental phenomena.

Penrose (1989: 525) quotes as an example of consciousness "a feeling such as puzzlement, despair, or happiness". Consciousness for Penrose is something that allows us to ask things like "What is the meaning of life? Why am I here? What on earth is this Self I feel?" He also says that "it is when one sees others behaving in this strange philosophical way that one becomes "convinced [his italics] that one is dealing with individuals, other than oneself, who indeed also have minds" (ibid: 529). He maintains that there would be no problem in programming a computer to go round saying things like this, and even, presumably, programming it to give itself some plausible answers to its questions, but one cannot expect it to be conscious of what it is doing, or behave in a way we have not already pre-programmed for it.

Penrose, using the literary convention of a Prologue and an Epilogue, tells the story of the thirteen year old boy who has been allowed to attend the initiation of the new 'Ultronic' computer and is told by the Chief Designer to ask the computer a question. The question (which one is left to deduce from the context) is What does it feel like to be a computer? and the supercomputer simply replies that it cannot understand what he is talking about.

Philosophers defy AI to produce feelings and emotions in their robots and computers. What, some ask, is the point, or even morality, of doing so, even it were possible? Johnson-Laird (1988), after spending a whole book explaining the possibilities for robots, ends by talking of feelings and emotions for AI, and concludes that "the only way in which a robot could experience the same feelings as we do would be if it had the same needs and social goals and was controlled by the same internal codes as we are". But, he asks, "are your subjective experiences the same as mine?" (ibid: 384) However, he still maintains that "feelings and free will can be explained from within a computational framework" (ibid: 389-90).

Later, in Oatley and Johnson-Laird (1989: 105) we find that "Emotions function as two-fold communications that enable a repertoire of behaviours to be produced with a minimal load on the information-processing system within an organism and on the communicative system between organisms". We also find, as the subject of the article this quotation is taken from, the most complete lexicon of Emotion in English I have found so far. Ortony, Clore and Collins (1988) have also produced an analysis of the lexical field of emotion from the point of view of psychologists seeking to reduce this complex semantic area to something comprehensible to a robot. It seems that AI, challenged by the philosophers and others, has decided to respond by attempting to teach the computer how to understand emotion, or at least the language of emotion.

Many, therefore, now believe emotion is an essential factor in the establishment of human consciousness and individuality. In one of the latest books on the subject, Damasio (1994), in the very title of the book, seeks to link "Emotion, Reason and the Human Brain". Dennett (1995: 3) in his review commends Damasio's book by saying:

"There are not many factual novelties here for those who have been staying abreast of this literature, but Damasio has woven some familiar if undervalued facts together into a vision of the brain and its parts that really makes sense, biologically, psychologically and philosophically."

Sartre defended the importance of emotion half a century ago. For him, (1939: 62), emotion was not a psycho-physiological disorder that interrupts our otherwise rational existence. For him, "l'émotion n'est pas un accident, c'est un mode d'existence de la conscience, une des façons dont elle comprend (au sens heideggerien de <Verstehen>) son <Etre-dans-le-Monde>". He argues that consciousness cannot exist without emotion and asks "peut-on concevoir des consciences que ne comporteraient pas l'émotion dans leur possibilités, ou bien faut-il voir en elle une structure indispensable de la conscience?" (ibid: 15). Fifty years later it would appear that this view is becoming acceptable to conventional science.

[1] Sagan (1977) describes this in Chapter VIII 'The Future Evolution of the Brain', in which, after tracing an evolution in human intelligence related to major technological innovation down the centuries, he concludes with the statement - "The next major structural development in human intelligence is likely to be a partnership between intelligent humans and intelligent machines".

[2] See 'Looking at inner landscapes' in the Economist, 3rd July 1993.

[3] Sekuler and Blake (1985) provide a general introduction to most forms of physical perception. A more detailed account of visual perception can be found in Gordon (1989).

[4] Examples of this are quoted in Popper and Eccles (1977: 400-1), and Sagan (1977: 31-2).

[5] One wonders whether Proust, when recalling scenes of childhood with the aroma of the 'madeleines', was aware of this.

[6] Badcock (1992: 174) defines the unconscious as: A term with three related meanings. in psychoanalysis. (1) The descriptive sense: mental contents not registered in consciousness at a given time. (2) the dynamic sense: mental representations rendered unconscious by repression and contained there by defence mechanisms and resistance. (3) the topographic sense: mental phenomena which are located in systems which render them descriptively unconscious by dynamic means. [Italics in original].

[7] Jung (1991: 13) "Não se trata de idéias innatas, mas de caminhos virtuais herdados".

[8] In order to demonstrate how it is impossible for a computer, however well programmed, to understand or be conscious of what it is processing, Searle compared the position of the interior of the computer to that of himself shut up in a room with no contact with the outside world. In this room there are two baskets of Chinese symbols and a rule book in English about how to match the symbols from the two baskets. As he combines the symbols using the rule book he passes them to the outside world, which calls his efforts 'answers'. The outside world hands back more symbols, which it calls 'questions', and perhaps further rules for combining the symbols. Searle argues that, with time, he could follow the rules, or the programme, for combining the symbols to the point where his performance appears to the outside world to be that of a native speaker. However, he argues, all this can be done without his understanding a word of Chinese or having the slightest idea of what the symbols represent. Therefore, although we may programme a computer to the point where it manipulates language in a way that satisfies the outside world, it can never be said to 'understand' what it is doing.