terug naar homepage
The philosophy of consciousness
What is consciousness?
The problem of other minds
Biological criteria for consciousness
Differences in experiencing consciousness
The problem of our own mind
Theories explaining consciousness
elemental property dualism. 10
Reductive materialism (The identity theory)
Arguments for theories of consciousness
The argument from religion
The argument from introspection
The argument from ireducibility
The argument from development
The History of research of the brain
The history of Artificial intelligence
Modern neural computation networks
New developments in brain research 34
Creating pictures using sound waves
Creating pictures using magnetic waves
The PET scan technique
Exploring the dynamic brain with imaging
The discovery of synchrony
The prospects of the mind/brain discussion
There have been many spellbounding cases in the annals of neurology. One of them was the case of Phineas Gage. Gage was the foreman of a railway gang laying track in Vermont. In 1848, when he was tamping a charge of dynamite at the bottom of a hole, a spark ignated the dynamite, sending the tamping rod up the hole and through the left frontal area of Gage's head. A few minutes later Gage was conscious and speaking. After convalescing from an ensuing infection he put on his boots and went to town appearantly completely recovered. However in the words of his friends "He was no longer Gage." his personality had changed dramatically.
This true story is a remarkable example of how fascinating the connection between the brain and man's conscious state in fact is. The accident showed that due too the impact of an external influence the poor man's behaviour was dramatically altered. The question was raised: How can this change be explained? It indirectly indicated that personality, or perhaps even consciousness, is coupled in some mysterious way to the brain. How are these mental and the materialistic properties linked to each other? Once again this inevitable philosophical question came to the fore.
In fact it was already much earlier that the first notes about this dilemma were written. Especially during the last three centuries a close bond was founded between philosophy and brain research, which was probably the result of the rapid developments in the field of neurology.
In this paper the overlap between neurobiology and the philosophy of consciousness is investigated. The main questions throughout the paper are: What has been the influence of brain research on the philosophy of consciousness in the recent history? What fields of research will most likely influence the discussion in the near future and in what way? Finally the question is posed: Where does the research of the brain eventually lead to with regards to our understanding of consciousness?
In the first chapter a short overview will be given from the history of research of the brain that had taken place during the last three hunderd years. Chapter 2 focusses on the most recent developments in brain research (like brain scanning techniques) that might have interesting consequences for the mind/brain discussion. Chapter 3 gives a short introduction of the philosophy of consciousness, while in chapter 4 an overview will be given from the current theories how consciousness can be connected to material (for example brains). Both monistic and dualistic ideas will be discussed and in chapter 5 some of the arguments for and against each stream of theories will be summed. In chapter 6 the developements of Artificial Intelligence will be investigated, and finally in chapter 7 the prospects of the mind/brain discussion will be discussed.
.c2.The History of research of the brain:
In this chapter an overview will be given from the research of the brain over the last three centuries. Descartes representation of cerebral function was without any doubt the most complete at the end of the 17th century (ref 2). Although he did not execute any experiments himself, his ideas will still be discussed in this chapter because they formed the first real hypothesis of how brains function.
One remarkable point in Descartes' ideas was that he believed in a hydraulic model: Not electricity but "animal spirit", which was viewed as a liquid, was responsible for sensory information transmission. (It was also the same liquid that was responsible for enervation and muscle activation). In fact Descartes did not really distinguish between cervical fluid and other body fluids. He thought that the sensory stimulation induced the heart to pump animal spirit in the nervous system. The brain then selectively permitted the fluid to pass by opening and closing specific pores and this was the level where the information was processed. The result of all this could be a flux of animal spirit (guided by the nerves) to the muscles resulting in specific muscle contraction. The whole concept as described was placed by Decartes in a very mechanical perspective. The body was viewed as a machine in which cause (sensory input) and response (movement) could be followed and predicted logically.
The pineal gland however had a very special place in the framework. It was regarded as the centre of imagination and common sense and in addition the key position in the signal transduction from sensory input to motory output. An image of the object was formed on the pineal gland, which itself then leaned toward the side bearing the image. In so leaning, the pineal gland pushed on certain tubes and closed them, while in the mean time other tubes were released and became permeable. The animal spirit then flowed out to the appropriate muscle and caused them to contract. In addition Descartes believed that there was a great difference between body and soul. He quoted: "The soul of reason cannot in any way originate from the power of matter". It is lodged in the body (pineal gland) but not merely like a "pilot in a boat". In contrast to these dualistic viewpoint towards humans, according to Descartes animals did not have a soul. "If there were machines that would have the organs and the features of a monkey we would have no means of recognition that they were not of the same nature as these animals." In fact here Descartes already hit the basic philosophical problem of the other minds consciousness. He based his statement on the idea that animals do not have language and therefore no consciousness.
At the end of the 17th century a more refined separation between (motor) function and location in the brain were established; The higher parts of the brain were responsible for the "higher" functions like complex motor function or voluntary movement. The inferior parts of the brain descending into the spinal cord in contrast contained lower functions like involuntary movements. By the time well known involuntary movements was reflexes like in the knee. In fact the identification of reflexes was bound to have an incredible impact in the ideas about the function of the brain and resulted in a complete new way of thinking. In the first reflex experiments with frogs it was established that stimulation of the nerve as well as touching the foot both resulted in a muscle contraction. It was also known that the spinal cord had an important function in the reflex as Mashall demonstrated in 1850. So it was obvious that some lower functions were directly regulated by reflexes. The question then was raised how far could reflex function be extrapolated to higher regions in the brain? The controversy between Pfluger and Lotze demonstrates the difference in opinions (ref 16). Pfluger supported the notion that the most complex behaviour was only a summation of simple reflexes. Lotze on the other hand supported the thesis that a simple reflex was only an arbitrary unit of behaviour which always remained under the control of a higher "centre". It is interesting to see that both views are integrated in the modern conception of brain function which combines the reflex in the totality of the nervous system. This debate also induced the new idea of reflectionism. Herbert Spencer (who's ideas were mostly in line with those of Pfluger) posed the double thesis: The reflex is always a psychic act, and the psyche is an assemblage of reflexes. His ideas pointed in a very mechanistic direction. Not only could the reflex be extrapolated to the highest level of the brain (the whole brain could be understood in terms of reflexes) he also claimed that psyche was the result of all reflexes. The consequence of this is that psyche (or consciousness) according to Spencer was not located in one particular brain centre.
Until the end of the 19th century the concept of a higher and a lower brain separation were kept in that simple form. New findings from several researchers however quickly specified ideas especially about the higher brain parts where they laid more emphasis on the importance of the cortex. In 1778 Sabourant already noted that if one could follow the nerve fibres from a paralysed muscle to their origin in the brain one might determine were the disorder in the brain was located. Almost a century later the ideas of Sabourant formed the foundation for the works of several scientists who all focused on the function of the cortex. In 1870 Fritsch and Hitzig discovered that certain parts of the cerebral cortex were excitable. They placed two fine platinum electrodes on the brain, several millimetres separated from each other and applied current from a battery. In this experiment the intensity of the stimulation was very weak. When the electrodes were placed on the anterior part of the brain the passage of current produced a contraction of a limited group of muscles in the half of the body opposite the stimulated side. Stimulation of a neighbouring zone resulted in a contraction of another muscle group. In addition to these findings, two other types of experiments were executed that also stressed the importance of the cortex within motor function. First of all, experiments with animal species like apes (Fritsch and Hitzig) and dogs (Ferrier) were done. In the brains of the animals certain regions of the cortex were distroyed, either surgically or by an injection of water, and the consequences were observed. In general the lesion resulted in paralysis of a muscle group that, under normal circumstances, would contract after stimulation of the same undamaged cortical spot. Finally more (human) evidence was found for cortical motor function by Jackson (who worked as a physician in an English hospital). Jackson observed a patient who was unable to speak and in addition suffered from epileptic attacks at very irregular times. Jackson's thesis was that the patients symptoms could be explained by the cortical damage of respectively the speech and motor area. After the man had died, autopsy indeed showed two tumours exactly in the locations predicted by Jackson. At the same time of the motor area discovery also the sensory domain on the cortex was discovered. Munk noted that destruction of the sensory sphere resulted in paralysis by the suppression of primary impressions. Although these findings still were mostly based on movement, once again it must not be forgotten that movement was regarded as the primary reflection of consciousness. These results therefore did not actually "solve" any philosophical problem, but instead moved them from the whole brain in the direction of the cortex.
Theories about localisation of consciousness were also heavily influenced by findings indicating that complex behaviour was the result of co-operation between several regions in the higher brain. Instead of looking for one single structure that should contain consciousness, people now focused their attention on the links between different parts of the brain in the search for a location.
Goldstein already noted that pathologic lesion of the cortex, in addition to producing localised difficulties, simultaneously influenced domains of function like perception and language. These domain of function however were never completely abolished. The motoric disorder named apraxia was demonstrated to be a good example that more than one centre was responsible for a complex function. Liepmann described a patient who had a disorder in his right arm movements. Although the arm was not paralysed, the patient had only limited control over his arm: On the one hand, all activities of daily life placed in their normal context (like using a spoon during a meal) were easily accomplished. However, when the patient was asked to perform a gesture taken out of the normal context (make a fist, show how to use a brush), the patient failed completely. Apparently all motor functions were still present but on some higher level the communication was definitely disturbed. Liepmann placed the disconnection between visual sensory and audio memory on the one hand (which constructs the intention) and the motor memory on the other (which would execute the movement). Indeed autopsy demonstrated a lesion in the left hemisphere (which acts on the right half of the body) on the left parietal lob (explaining the disconnection between visual memory and motor centres). In addition also a lesion in the corpus callosum, which connects the two brain halves, was found.
The findings of apraxia together with the knowledge of motor and sensory areas strongly encouraged connectionists in their ideas that behaviour could in principal be fully understood in terms of physiology. Later developments in the anatomy and physiology of the cerebral cortex during the 1950's allowed the identification of distinct cortical projection areas which were directly connected to the sensory organs. Here however, in contrast to the concept of connectionism, the idea of convergence was introduced. Connectionism implies that a product of the nervous system (a movement for example) is constructed from elements which are progressively connected with each other. In the view of convergence the information of the projection areas then was transferred to associative areas (which in fact occupy the largest part of the cortex) in a way that one area receives information from several sensory modalities at once. So in this concept the primary information of each source is spread out over a number of associative areas and there integrated with other primary information. The consequences of convergence for the mind/brain problem was that most people who believed in a mechanistic explanation had to admit that, although perhaps still possible, the solution had to be much more complicated than connectionists claimed initially.
It is interesting to note that also a sery of experiments was executed based on philosophical thoughts about "will". Schoppenhauer had claimed: "Will and mind are not synonymous, but almost. the will is the root of the mind" (ref 29). In the same era Biran stated: "If the individual did not will, or were not determined to begin to move himself, he would know nothing . . . he would not suspect any sort of existence, he would not even have the idea of his own self. I act, therefore I am" (ref 25). He was convinced that willed movement was based on will. In addition, voluntary contractility should be accompanied by a subjective experience of "individual feelings of effort". Maine de Biran had the intuition that the feeling of effort was in fact the feeling of resistance to action on the part of the will. Without something that resists there is no effort. Several patients with paralysis (which causes a maximal resistance) were studied. In experiments they were asked to move their paralysed limb and subsequently they were questioned about their "feeling of effort". In this way it was tried to examine the subjective (feeling of effort) in an objective approach. The observations pointed in the following direction: When the paralysis was total, no real sensation of effort was accompanied with a contraction attempt. However in less severe cases of paralysis each effort of the will was accompanied by the sensation having the limb held down by an enormous weight. Analogous sensations were reported by the rare volunteers for experiments of total paralysis due to curarisation. (Our current knowledge tells us that curare blocks the neural transmission at the neuromuscular junction.)
One of the most important consequences of this type of experiments was that it raised the question about the origin of the feeling of effort and free will. These new findings discouraged the idea of behaviourism (which claims that sensory input is the foundation of any motor activity), for in that case free will was in fact the result of this input. But what if no external stimuli are responsible? A sensible assumption was that somewhere in the body, neuronal stimuli might be generated internally without the help of external stimuli. This idea not only encouraged the search for spontaneous neural activity, it also resulted the discovery of the "sensory" receptors by Sherrington. (The sensory receptors are now known to send information to the brain about the internal state of muscles and organs.)
One of the first obvious explanations for spontaneous neural activity came from Lorentz who claimed: "When an instinctive behaviour is stopped for a long time, the threshold for stimuli which trigger is lowered. The lowering of the threshold can in certain cases approach zero, that is, the instinctive movement in question can take off without any external stimulus."(ref 11).
Von Holsts identification of motor centre oscillators from 1930 and onwards was an important development for the spontaneous nervous activity. Von Holst came with the idea that certain neural elements localised in the motor centres oscillated according to some permanent rhythmic activity. Neurones did not merely serve to transmit information from place to place, they fabricated information themselves. Nowadays there is proof for the existence of neuronal networks that are able to produce excitations as in the oscillators (for example the central pattern generators which are involved in rhythmic movements). Also cells have been described that are involved in the biological clock, a function that is individually capable of producing action potentials. It is interesting to note the remarkable comparison between the mechanism by which the clock cells produce excitations, and hypothetical mechanism described by Lorentz.
The last discovery that will be discussed here is electroencephalography that was first described by Hans Berger. In his publication he demonstrated the spontaneous neural activity in numerous parts in the brain. His work also changed views considerably about the dynamic properties of the neural network. In 1929, Berger himself thought that the measured electrical activity was best described as the reflection of "psychic energy". The brain, he thought, was the site of metabolic activity like no other tissue. This activity liberated a chemical energy which was subsequently transformed into heat electricity and psychic energy. It was not until 1965 that the potentials in neurones could be measured and so the theory of psychic energy was abandoned.
This overview of the research of the brain is far from complete. It does however demonstrate a number of interesting relations between biological research and philosophical theories about the mind/brain problem.
Firstly, it demonstrates that some philosophical questions are not solved by new findings, but instead are emphrased in terms of new concepts. Take for example the mind/brain problem: It already existed centuries ago, it is still not solved now and it will probably never be solved at all. The emphasis of the discussion on the other hand indeed did change over the years. The arguments that are used in the current discussion are mostly derived from (thought) experiments with computers (see also chapter 6). A hundred years ago this was unthinkable, for the simple reason that computers did not exist. Another example is given by reflexes and consciousness in relation to the discovery of the cortex: At first people were arguing about whether it is possible or not to explain most human behaviour with reflexes. After the discovery of the cortex the same discussion changed and centered around the question whether it is possible to explain function of the cortex by reflexes.
The second interesting observation that can be made throughout history is that people's opinion sometimes dramatically changed after a new insight in the function of the brain was gained. For example when it was found that brains also generate impulses without receiving sensory input, reflexionism suffered a tremendous blow in its support.
Thirdly, it is shown that as a result of intense research, the support of materialistic theories has increased. The knowledge of how the brain functions had been expanded enormously over the last decades. As a result, also people's ideas about the mind/brain problem could be specified. At the time of Descartes, people suspected that the brain had something to do with mental activity. The finding of reflexes showed the mechanism by which some of the responses were achieved by the neural system. Later when the cortex was discovered, people started to realise the existence of projection areas and their function. Neuroscience taught us how to explain someones behaviour by showing the biological underlying mechanism, the activity in certain brain structures. The result of this for the mind/brain problem is that it is almost impossible nowadays to support a theory that denies the power of the materialistic property completely. At least most neuroscientists either support a materialistic theory, or alternatively a dualistic theory that takes a lot of the materialistic foundation into account (property dualism).
Further reading and references (nr):
2, 8, 11, 14, 15, 16, 22, 25, 29
.c2.New developments in brain research
In chapter 1 a short overview was given from research of the brain over the last three centuries. Like most branches of science, also neurology rapidly extended of its knowledge during the last few decades.
For a long time brains could only be studied with an indirect approach. The physiology of dead brains was intensively investigated, patients with certain deseases were used for experiments etc. These approaches provided a hugh amount of information about the brain. With this information it was possible to construct a model of how the brain was likely to function. Non of these early experiments however was able to give direct information how the living brain works in a normal healthy person.
Recent developments have brought some progress in exploration of this gap in the knowledge of the brain. At present there are several methods available that are able to measure the activity in the brain of a living human being. With most methods it is possible to obtain a 3-dimensional picture of the working brain. Each of these methods has its strengths and weaknesses, for example the costs of the equipment, the resolution, and the labour that is necessary to prosess the data. In the next paragraphs several techniques and developments will be discussed.
.c4.Creating pictures using sound waves:
Since the discovery of ultrasound, several applications in medical science have been developed. An important one is the non invasive technique that creates images of internal structures without opening the skin barrier. The technique works by directing a pulse of acoustic waves at some body structure and then measuring the strenght and speed of the waves that bounce of the body and return as echoes. Assuming an average speed for sound waves through body tissue (about 1540 meters per second) the ultrasound device assembles the various echo times into an image. Organs and connective tissue tend to reflect sound waves quite differently, so that the outlines can be recognised easily. Ultrasound is also capable of producing pictures of moving blood, which is very useful in identifying active organs or (brain) subregions. The resolution of this method is 0.5 to 1 millimeter.
The technique has a big advantage in comparison to other investigation methods, ultrasound is widely believed to pose virtually no threat of side effects. Despite this advantage there is however one limitation: Sound waves penetrate very poorly through bone, and this fact might appear to make ultrasound imaging in adults impossible. Only in new born babies, in whom the bones of the skull are quite thin and have not fused together, it is still possible to investigate the living brain. This restriction however makes it at least difficult to do research at some brain functions, because babies cannot be asked to perform a certain task as can be done in adults.
.c4.Creating pictures using magnetic waves:
At present there are three imaging techniques that are based on magnetic waves, namely MRI, NMR and MSI.
MRI, which stands for Magnetic Resonance Imaging, is the best known technique of all three. During an MRI measurement the patient is positioned inside a large magnet. The magnetic field acting on the patient is just powerful enough to bring the magnetic poles of the hydrogen nuclei in the body into realign under the influence of the magnetic field. When a single strong pulse of radio waves is fired, the nuclei are knocked into disarray but then realign under the influence of the magnetic field. As they do so, the various types of cells emit a distinct radio signal on their own, which a computer transduces into a visual image. MRI is primarily able to detect hydrogen molecules, which are abundantly present in water. Therefore, tissue that contains high amounts of water is distinguised easily on the MRI image from other tissues; Fluid filled spaces appear quite dark whereas fatty tissues show up as bright areas.
Similar to MRI and using much of the same equipment, is Magnetic Resonance Spectroscopy (MRS). Here the magnet and radio waves are tuned for other atoms than hydrogen. Phosphorous atoms, for example, form part of phosphate molecules which are involved in energy metabolism. With MRS it is therefore possible to distinguise metabolic active areas in the brain from relatively inactive areas.
Recently a third application has been developed which is called Magnetic Source Imaging (MSI). It is based on the knowledge that every electrical discharge (for example a firing neuron in the brain) is accompanied by a very weak magnetic field. The extremely sensitive detectors that are located all around the head of the patient, are able to locates the source of this magnetic field precisely.
Although imaging using magnetic resonance is not hindered by the skull (as is the case using ultrasound), there are some other limits using these techniques. The equipment that is needed is (at least at this moment) very expensive. Furthermore, the highest resolution that can be obtained using these techniques lays around 2 millimeters, which is quite low, knowing the fine structure of the brain. Despite these limitations, magnetic resonance is capable of providing a good picture of the living brain; With MRI, the internal structure can be examined, while using NMR and SMI also local neuronal activity can be measured.
.c4.The PET scan technique:
PET scan is a technique that stands for Positron Emission Tomography. A PET scan measures the distribution in the body of a "labeled" substance that the patient has received shortly before the scan. Adding a radioactive label to compounds as glucose permits researchers to monitor roughly the metabolic activity at particular locations in the brain. The cells take up the radioactively tagged glucose, the glucose is metabolized, and the transiently radioactive atoms remain inside the cell, giving off positively charged particles. These "positrons" quickly collide with nearby electrons, then give rise to gamma radiation, which can be detected outside the body and mapped by a computer.
Like NMR measurements, PET scan is able to distinguise metabolically active areas from relatively inactive ones. The advantage of this techinque compared to the magnetic resonance ones, is that no expensive super magnets are required. On the other hand, the use of radio active materials on the long run are known to increase the chance of severe deseases like cancer.
.c4.Exploring the dynamic brain with imaging:
The imaging techniques provide the chance to have a closer look at how the brain functions in a living human being. Especially NMR, SMI and PET offer a large number of opportunities in this new field of research.
Recently it was investigated which parts of the brain are involved in speech, using the PET scan technique. Speech is known to be one of the most complex behaviours that human beings perform. As will be shown in chapter 3, some people believe that the capacity of speech is a requirement for, and is directly linked to consciousness. There are numerous reasons to be interested in this research.
To investigate which parts of the brain specifically are involved in speech, the following approach was used. A testing person was put into a position where the brain activity could be measured with the PET scan. First, a background activity of the brain was established. This is to say, the neural activity in the testing person that is present when he or she is just sitting and doing nothing. This background activity should not have anything to do with speech. Then, the person is asked to read four words. The first word is constructed of symbols having the caracteristic angles and shapes of letters. The next word is constructed of real letters grouped into units about the lenght of average words, but unpronounceable. Then a word is shown with strings of letters that look like normal words and that could be pronounced but are not words in the English language. Finally a word is shown that has a normal meaning in English. The sequense could look like this:
The goal of this project is to dismantle how language is read and recognised. The first combination of signs is taken to distinguish the purely visual aspect of reading. Then gradually some linguistic properties are introduced. The results show that an area known as the medial extrastriate cortex came notably into action as the task reached the third and fourth letter combination (see figure 1).
Further experiments were done with sound and words. Here the testing person was asked to listen to words, and to tell whether they rhymed. Both this data together with experiments in the near future are likely to increase tremendously the neurologic understanding of the working brain. With imaging it becomes possible to investigate how different aspects involved in language (like reading and hearing words, recognition of them, memory function) can be found in specific brain regions. Furthermore, the cooperation between these subregions can be studied, which is at the moment an essential point in the understanding of language.
The research is not limited to images of learning and motor function. Because it permits close observation of how the brain is using energy, for whatever function, PET actually allows researchers to map what may be called the anatomy of emotion (ref 1).
Patients who suffered from a panic disorder were studied during a panic attack that was deliberately induced by the infusion of sodium lactate. In a second experiment the feeling of anxiety was induced by telling the testing person that they were about to receive an electrical discharge. It was revealed that in both experiments the same specific region in the brain demonstrated an increased activity.
Recent PET data suggests that a number of interesting features like anxiety, attention and mental activity are centered in the right hemisphere of the brain. Further studies using PET imaging will most likely provide more detailed information about the activity of this area.
.c3.The discovery of synchrony:
Until recently the activity of large numbers of cells played an important role in neurophysiological research. Measurements were performed via encephalogram and in field potentials on cell groups only. Since the technical improvements allowed to measure individual cell potentials, researchers have done so exstensively. The message conveyed by the neuron was thought to be defined entirely by the amplitude of the response and its provances. As a consequence, in single unit studies, time received relatively little attention as a demension for coding.
Recently, however, a dramatic change in attitude and interest occured, as at the moment many think that temporal relations between the responses of distributed neurons are as important a code as relations between respons amplitude.
To understand how this time related coding works, first the principal of neural oscillation must be understood. With the old measuring equipment, oscillating neural activity was found at different wavelenghts. With field potential measurements, oscillations in the delta range (0.5 to 4 Hz) was found during sleep and physiological states like coma. Oscillation in the theta range (6-7 Hz) was found during states of attentive arousal. Oscillation in the 10 Hz range, also known as alpha activity, was found during drowsiness or during states of relaxation.
New measuring techniques however were able to bring oscillations into light in the ranges 15-30 Hz and 30-60 Hz (the beta and gamma ranges respectively). The amplitudes of these high frequency oscillations were usually small, thus indicating that the group of neurons engaged is small or dissipated. According to researchers in the new area, findings indicate that neurons in cortical networks have the tendency to engage in syngronous activity in different distinct frequency bands. Hereby, the probability of occurance of synchronous activity in a particular frequency range depends on the central state of the brain, on the presence of sensory signals, and on the occurance of motor acts. In other words, oscillation activity in the beta and gamma range is believed to code for specific information in certain mental activities (like perception and motor function).
This hypothesis could provide a fruitful approach to another problem in the understanding of information transfer, which commonly is referred to as the binding problem. The incoming information from sensory organs like the eye is believed to elicit responses in a large number of spatially distributed neurons, each of which encodes only a partial aspect of the object (for example colour, contour etc). The binding problem deals with the question how these distributed activities are reintegrated in order to generate unambiguous representations of objects in the brain. One proposal has been that there are areas in the brain where all these distributed activities reconverse onto neurons that respond in a highly selective manner only to those constellations of features that characterize a particular natural object. One of the important objections that have been raised argues that there would probably not be enough cells in the brain if all distinguishable objects including their many different views would each have to be presented by a speciallized neuron.
In chapter 1 and 2 both the history of brain research together with the newest promising techniques related to the study of consciousness have been discussed. Besides brain research, also a lot of developments have taken place in the field of philosophy of consciousness. Especially the last centuries demonstrate an interesting parallel evolution of brain research and philosophy, with some remarkable interactions between the two. In chapter 4, a number of philosophical theories to explain consciousness will be discussed. In that chapter, also attention will be paid to the influence of brain research on the development of the theories. First however, a short introduction will be given to the subject of consciousness in chapter 3.
Further reading and references (nr):
1, 17, 31
.c2.The philosophy of consciousness:
Before we are able to say anything about consciousness first two very important issues must be recognised, namely (1) what is consciousness and (2) what problems occur in investigating it?
.c3.What is consciousness?:
The first question, "what is consciousness?", sounds easy enough but in fact to find a proper definition of consciousness is a severe problem. It might remind us what a philosopher once said about time: "If one speaks about time I know exactly the meaning of the word, what it is, how to apply it and so forth. However, if one would ask me to give a definition of time then I am afraid this is beyond my powers." And perhaps the same counts for consciousness. Everybody knows what it is as a result of daily life experience, it is difficult however to give exactly one strict definition. In fact, as will be shown later, during history several types of definitions have been applied by scientists and philosophers. The definition of the term consciousness as used in this theses is perhaps best reflected by how J.C. Eccles defined it: "Consciousness is the conscious experience which each of us has privately for him or her self" (ref 3). So it is the primary reality that each human being experiences, and what makes us say in daily life person X has consciousness and that rock over there has not. With this example it is illustrated how most people instinctively feel about consciousness.
To conclude, there are many possible difinitions thinkable to describe consciousness. In this paper we will however focus on how consciousness, by the definition of J. Eccles, can be explained. If in the text the word consciousness is given then it refers to Eccles' definition, unless stated otherwise.
.c3.Studying consciousness: The problem of other minds:
The second important issue in the philosophy of consciousness that must be recognised relates to problems that occur in investigating it. How do we exactly know that someone or something possesses consciousness? Perhaps this is the item that makes the whole subject so interesting. The striking answer is: We donít. There is no possible way we can check that someone else has consciousness, because for that, you would have to be that person yourself. This dilemma is called "the problem of other minds."
So why do we presume that other people have the same conscious states as we have? The answer to this question will probably be that we "human beings" are all very similar to each other in a lot of ways; We are all build of the same flesh, blood and bones. Furthermore we have the same blueprint, our bodies are arranged roughly in the same way when compared to each other. Not only do we all have brains, a perception system composed of eyes ears etc., even the internal structures of these features show a remarkable resemblance. This is what make us decide that, very likely, other people possess the same conscious capabilities as we do ourselves. Letís say this decision is based on "the argument of analogy".
Although this assumption sounds very plausible there are still a few severe objections against this argument. Firstly, the generalisation as mentioned is based on only one single case. We know from ourselves that we have consciousness. In addition the behaviour of other people together with their resemblance in physical structure makes us suppose that for other people the same should count. However this assumption is quite dangerous. This is best illustrated by the next analogy. Imagine that you hate the sensation of pain. It feels unpleasant and therefore you try to avoid it. All the people in your environment say they hate pain as well. Can we now make the generalisation that everyone hates to get hurt? In fact we can not. There certainly are some people who do enjoy pain, although this is definitely a minority of the whole population. In general we can say, the more cases one uses in a generalisation the higher is the predicting value. In the case of consciousness the extrapolation is grounded only on one case. Therefore the argument of analogy is heavily undermined.
.c4.Biological criteria for consciousness:
The extrapolation from one human being to another is a simple one in the sense that there is a lot of resemblance between organisms belonging to the same species. But what if this resemblance is not that optimal? What if for example we try to establish the conscious state of an animal like a cat or a dog? Although also they do have brains, a perceptual system etc., do they have in addition a certain level of consciousness? Their brains are smaller and so in a sense less developed than those of humans. Where do we have to draw the line between conscious and unconscious when we follow that line from a human being down to a single cell?
This is definitively a difficult question that cannot be answered with certainty. Because we are human beings we will never get in the position of for example a dog or an amoeba or any lower animal species. What we can do however, is try to identify certain requirements for higher conscious states (although it must be said that also these are subject to personal opinion). Once when a number of consciousness related features has been identified one can try to find the place on the evolutionary line where consciousness might have appeared. To illustrate what is meant here, a few theories of possible requirements for consciousness will be discussed.
The first and most obvious requirement that could be made is the presence of a neuronal network that can function as a transmitter and modulator of information. Yet it is interesting to see that already with this criterium a number of life forms can be excluded from having conscious states. In fact neuronal networks are only found in the kingdom of animals and for example not in that of plants. So according to this criterium all plant life does not have any consciousness at all. Furthermore when the requirements for a network are considered one can easily see that a number of co-operating cells is required. Therefore also single cellular animal life is by this definition deprived of consciousness. According to existing hierarchical classification the first animal that would fall into the conscious group would be something like polyps that posses a very primitive undifferentiated neural system.
Apart from the requirement of the neural network itself there are a few interesting structures or requirements within the neural system that might be quoted as requirements for consciousness. Take for example the requirement of sensory input. When we look at the way we use the word "consciousness" in our daily language, it reveals the connection to awareness. If one wants to be aware of something in the outside world, not only a computational unit is needed, but also an input of exterior information. The more information is given to- and processed by the network (the brain), the higher the awareness of the outer world can be. One of the criteria that some people use for a conscious being is the capability to obtain information from the outside world. The better and more complicated the internal representation, the higher will be the awareness and the consciousness of the outer world.
Another interesting theory about a requirement has been formulated namely that of memory. According to this criteria, to be conscious means more than just receiving and processing information from the outside world, consciousness also involves a certain part which can be called self consciousness. For this however, memory is needed. The argument runs as follows: To have a certain state of self consciousness one must be able to realise that the person that one is today, is the same person that inhabited ones body 10 years ago. Self consciousness is only experienced when this continuity in time is recognised and for this a memory is necessary.
The last theory of requirement for consciousness is derived from a linguistic perspective and it claims that, to posses (self)consciousness, the organism must have developed the capability to use some sophisticated form of communication like language. If we take for example the sentence "I am ...", to be able to consider oneself as "I" one needs the linguistic capabilities to understand the word, or to be more precise, to connect the meaning of that word to a word. The same counts for the word "am". Without the linguistic basis one is never able to think about oneself as a conscious being. This idea of language coupled to the mind is also reflected in an other theory that claims that consciousness is coupled to a centre in the brain that has a major function within speech.
.c4.Differences in experiencing consciousness:
All these biological theories about the possible requirements can give some idea about the likeliness that other organisms have consciousness. However, it poses also the second central question in the problem of other minds, how do other humans or animals experience their form of consciousness.
We can never be sure of the other organisms experience, just as we can not be absolutely sure that other organisms have consciousness in the first place. For example, is the experience of the colour red the same for different people? Imagine that in one single person the experience of red and green have been substituted. So this person will have the experience of green when he is looking at a red object, although he will refer to it as a red object (for the reason of that everybody always had done so, he defined that experience as being red). Although you might think that you have the same experience as that person when you are looking at and talking about a "red" object, in fact the experience of red can vary dramatically. There is no way ever to find out that the experience in the other person is different from yours.
Yet another problem occurs with respect to the form of conscious experience. Although in the thought experiment one can claim on the ground of analogy that the experiences of other people are likely to resemble, a great problem remains with regards to other animal species who do have a completely different morphologic structure.
A famous illustration of this problem is given with the thought experiment once posed by Nagel in which he invites the reader to imagine what it is like to be a bat (ref 26). It is not so difficult to imagine what it is like to be a bat with a human perceptual system. For that, one just needs to put on a pair of spider web wings and find a way to stick oneself upside down to the ceiling. The real problem is: What is it like for a bat to be a bat? Here it becomes clear that the main difficulty arises due the completely different perceptual systems of humans and bats. Bats are in fact almost blind but instead they have evolved a highly advanced echolocation system. Via echolocation the bat is able to fly in narrow caves without bumping into walls, catching prey which is sometimes moving in very irregular patterns, the bat uses echolocation in a lot of situations where we would use eyesight. Now what we would like to know is the subjective state how is it like for a bat to be a bat, preferably from a objective perspective. It seems quite impossible to achieve this. What we might accomplish however, is to get some idea of the subjective state by looking in an objective way at the behaviour and morphology of the animal. If we understand that the animal uses echolocation as a navigation system, that knowledge might give us some clues from which position or situation we might get an idea of what it is like to be a bat. For example, it is probably much easier for a blind person to get in the perspective of a bat than it is for a deaf person.
.c3.Studying consciousness: The problem of our own mind:
As we have seen there are a lot of problems in investigating other peoples' consciousness. There are in addition also a few problems in the study of self-consciousness. It is clear that extrospection is always mediated by sensory organs and, as a result, it is possible that errors are introduced during the process of acquiring information. The traditional view claims that observations about one self via introspection are not mediated by sensory organs, and therefore not subjected to introduction of incorrect information. In other words, if introspection tells us that we feel or desire something, how can we possibly be wrong about it?
Recently a few objections to this infallibility have been posed. Firstly, from an evolutionary point of view, brains have been selected on their external perceptive possibilities. An organism that is able to process visual information and is able to respond correctly to the environment, has advantages in comparison to organisms that can not. At some stage introspection must have become possible as well, but because the selection of the brain never really focused on it, the argument runs that therefore it is unlikely that introspection should give us more reliable information than extrospection. Secondly there is an objective argument derived from the network theory. This theory tells that the brain can best be regarded as a furious active theoriser. The brain has (in contrast to most analogue computers) a parallel information processing structure. This internal structure involves a certain part of speculation which enables it to proces information highly efficiently. However the price to be paid is that under some conditions the speculations might result in a misinterpretation of the given information and errors can be introduced. (For example, when a person looks at a picture of a car with a tree standing in front of it, the person will probably postulate that the car is one solid object. In contrast of the observation, the picture could just as well be a reflection of a front and a back part of a car, with a tree just covering the missing part!) So because the whole brain is constructed according to the network principle, mistakes can be introduced everywhere in the brain in all observations. Therefore also introspection is not completely reliable.
In this chapter some problems with respect to the study of consciousness have been brought under attention. In the next chapter we shall investigate what philosophical standpoints exist in explaining how mental states are linked to material states.
Further reading and references (nr):
3, 24, 26.
When someone asked me about the consciousness of animals I always have to think about my cat. I suspect that she has some awareness of her environment, she even might be conscious of certain people in her life, but she is absolutely very conscious about one thing, and that is her dinner!
.c2.Theories explaining consciousness:
Plato thought that reality was not to be found in the physical world (world of sensibilia) but rather in the non-physical world of intelligibilia, a world that housed mathematical entities and other non-physical objects of thought. Only by turning away from the sensible world and by contemplating the non-physical objects of intellection could real understanding be achieved. Interaction with the physical world might yield opinions of mere belief, but it could never yield knowledge of reality. Pure contemplation was the proper occupation of the rational soul, itself a non-physical substance capable of existence independently of the physical body. Since the theory of Plato postulates two kinds of substances, mind stuff and material stuff, it was known as dualism.
Both dualistic and monistic (mostly materialistic) theories have a long history. Philosophers have tried to explain with these theories the concept of life, and in a later stage a part of this discussion, namely the problem of consiousness. In chapter 2 part of the history has been discussed. In the early days the differences between the two were very profound. For example, vitalists (which is very near to dualism) believed that a vital substance was necessary to make a living organism "alive". The early claims of mechanism were that all science could be derived from mechanics, and as a result that living creatures could be treated as machines in understanding them. Although people sometimes still claim to be either dualist or monist, nowadays it wil be hard to find any scientist that will support either of these views. The differences between the the two have become smaller as time went by.
The aim of this chapter is to get a good overview over all types of basic theories that have been evolved to explain consciousness.
The first main stream that can be recognised is called dualism. Although the first forms of dualism are generally accepted to be superseded, some newer adapted forms are still widely supported. Especially in religious groups dualism is quite popular for reasons that will become clear when the central thesis is understood. Dualism can be distinguished from other ideas in its central claim that conscious intelligence resides in something non-physical, something forever beyond the scope of sciences like physics, neurophysiology and computer sciences. During the centuries different forms of dualism have been developed and supported as will be discussed below.
.c4.Substance dualism:; The distinguishing claim of this view is that each mind is a distinct non-physical thing whose identity is independent of any physical body to which it might be temporarily "attached".
Substance dualism can most likely best be illustrated by the ideas of the first important dualistic philosopher, namely Descartes. His ideas are sometimes also quoted as Cartesian dualism. As we will see in chapter 5, Descartes clearly distinguished on one hand the physical world which was extended in space. Every instance of it was measurable, had precise properties like height, weight and an accurate position in space. In Descartesí view this world was very mechanistic (his brain model is a good example of that) and could be understood and predicted completely by laws of physics. On the other hand, according to Descartes, one feature could not be understood in terms of substance, and that was the conscious reason of man. When one purely introspectively observes oneself there was a conscious element left that could not be fit into the mechanical world. Descartes also strongly correlated consciousness with thinking, which is something with no mass whatever, no shape whatever and no position anywhere in space. This "mind-stuff" therefore could be best explained by an additional dimension which contained all non-physical thinking substance.
The feature in which cartesian dualism is quite extreme is the claims that consciousness can not be localised in any particular spatial position under any circumstance. In a newer form of substance dualism, which will be referred to here as popular dualism, is less radical in this point. This theory describes a person as a "ghost in the machine" where the machine is the human body (or even the brain) and the ghost is the spiritual substance that is quite unlike physical matter in itís internal constitution but with a spatial localisation that does correspond to matter. So unlike Cartesian dualism, the mind is localised within the body in the head.
The advantage of the latter theory is that it is more successful in explaining the central problem that arises with substance dualism, namely the gap between the physical and the mental substance. If both dimensions exist (as is claimed in this theory) and together explain the phenomenon of conscious beings, then how are they coupled to each other? Both Cartesian and popular dualism fail to bridge this gap completely, however, the latter is capable of suggesting that as a result of spatial localisation and direct contact with the brain in some form or another the two can be connected. For example, matter can be considered as very condensed energy (Einsteinís relation E=MC2). Perhaps mind-stuff is a well behaved manifestation of energy as well, but in a different form.
.c4.Property dualism:; The basic idea of the theories under this heading is that while there is no substance to be dealt with here beyond the physical brain, the brain has a set of properties possessed by no other physical object. The properties in question are all characteristics of conscious intelligence like having a pain, feeling A, thinking B, desiring that C and so forth.
Like substance dualism, also with property dualism different positions can be recognised. The oldest one of them is called epiphenomenalism (epi- = above-). This view claims that mental phenomena are not a part of the physical phenomena, but merely emerge when a certain level of complexity is reached in the growing brain. In addition, epiphenomenalism claims that these mental phenomena do not have any causal effects on the physical world in return. The sensory input that comes from the outside world is being processed and results in a mechanistic way in our actions and responses towards the events in the outside world. Somewhere along the line of processing, our consciousness (in terms of feeling A and desiring B) arises, but this has no causal effect in return on the brains or our responses. Epiphenomenalism thus claims that free will and our experience of voluntary responding do not exist and are nothing more than an illusion. Free will must be understood in a completly different way. It is our experience of having several different possibilities to respond in a particular situation. The imaginative force of every distinct possibility and the preference of one of them might in fact have a completely understandable physiological fundament that suffices all physical laws. So free will is not the fact that we have free choice, but instead the experience that we think we have.
It is however exactly the point of free will that was hard to believe and probably as a result the next adjusted theory was formulated : Interactionist property dualism differs only in one essential point from the previous theory in that it claims that mental properties indeed do have causal effects on the brain, and thereby on behaviour. With this solution the problem of free will is easily avoided.
The last dualist theories are already coming close to materialistic theories. In both cases consciousness is an emerging property requiring a physical structure that is solid, alive and has a complex internal organisation. However, dualism makes the further claim that mental properties are irreducible, they are novel properties that are beyond prediction and explanation in physical science.
One at least puzzling thing here is that it might be hard to claim physical irreducibility on the one hand and evolutionary emergence on the other both at the same time. The first suggests for consciousness an independence of matter, while the latter tells us the opposite. This dilemma resulted in the formulation of the last theory that will be discussed here, which is called .c4.elemental property dualism.
The idea of elemental property dualism is that the emergent property on matter is not strictly necessary. Mental states must be regarded as an extra fundamental property that has been here from the universes inception, just like other properties like length, mass, electric charge and so forth. The problem with this view is that, unlike electromagnetic properties, the postulated mental properties emerge only in very large physical system. One could argue that it is hard to believe that a basic property needs a complete organism to be manifested.
The second group of views with regard to explaining the mind/body problem can be summarised under the name materialism. This stream claims, in contrast to dualism, that in principle physical matter is enough to explain the existence of conscious states. Several slightly different interpretations of materialism have evolved over the years which will all be discussed below.
.c4.Reductive materialism (The identity theory):; This materialistic theory, like all other theories of its kind, claims that nothing more than matter is necessary to explain consciousness simply because conscious states are physical states. A good analogy to explain what is meant here is be the phenomenon of light. The first descriptions of light were based on human macroscopic observations, light as we know it in daily life. The phenomenon of light nowadays has been captured in an theoretic reduction that explains macroscopic things on a microscopic basis; The conception of understanding light is in terms of electromagnetic waves. Both understandings together tell us that light is electromagnetic waves. Reductive materialism now suggests that the same claim can be made for consciousness and matter, namely conscious states are physical states.
What makes it particularly reductive materialism is the assumption that each type of mental state is numerically identical with some type of physical state. So there is a one-to-one match up between mental states and physical states. This is the point that raises some objections which have been adressed in an adapted version of reductive materialism.
.c4.Functionalism:; Like the previous theory, functionalism claims that mental states are equal to physical states. Furthermore it ascribes specific mental states (like for example pain) to a physical system when the system (1) is complex enough (like brains) to entail consciousness, and (2) has the right functional input that we know to result in that particular mental state in our selves. More specifically, the essential feature of any type of mental state is the set of causal relations it bears to (1) the environment, (2) other types of mental states and (3) bodily behaviour. Pain for example characteristically results from some sort of physical damage, it causes other mental states aiming at relief as well as particular bodily behaviour like wincing or nursing the traumatised area.
The theory of functionalism differs in one particular point from the previous one in that it abandoned the one-to-one match up between mental and physical states. It argues that although a particular mental state might be the result of one physical state, there might very well be other physical states that would result in the same mental state as well. For example, the structure of the brain is capable of achieving a particular mental state. but, functionalism argues, an advanced (futuristic) computer which is constituted in a completely different way from completely different matter (metals instead of carbons) might have exactly the same properties. To summarise, functionalism claims that one instance of a given type of mental states is numerically identical with some specific physical state in some physical system or another.
In fact this example illustrates not only the theory of functionalism, but also the current direction in which the largest group of supporters can be found, the computer world and Artificial Intelligence. This subject will be discussed in a separate chapter later on.
.c4.Eliminative materialism:; The last materialistic approach that will now be described is strongly influenced by ideas coming from the psychology of language. The theory of eliminative materialism like behaviourism also claims that no one-to-one match up will be found, but on entirely different ground. It argues that our common sense psychological framework is a false and radically misleading conception of the nature of cognitive activity. All older frameworks will on the long run be eliminated by the matured neuroscience.
To understand fully the background of eliminative materialism, two additional theories must be explained; the theoretical network thesis and the theory of folk psychology. The first thesis tells us something about where linguistic expressions and in this particular case theoretical terms get their meaning. It claims the meaning of a term is usually not expressed in one single explicit definition, but rather on the network of principles it is derived from. For example, one can describe the theoretical term "electron" as being the unit of electricity, however this tells us little if we do not understand the conception "electricity". Both terms can therefore best be described within the framework of the electromagnetic theory that does not only postulate the existence of electric charge, but also electric force fields, magnetic force fields, etc. Electromagnetic theory on itís turn is integrated within the molecule theory, and so forth.
The thesis of folk psychology tries to give an answer to the question: How are we able to predict other peoples behaviour (in terms of feelings, pains) in daily life? Here we are not talking about a small group of researchers or philosophers who try to understand mental states, but merely the way that people (who are not particularly occupied with this question) apply their intuition in every day life. The theory of folk psychology assumes that the intuition that we nowadays posses is founded on numerous individual findings of people, experienced over the millennia, that resulted in a large number of reductions which in the end had a very reliable predicting value. We can uncover literally hundreds and hundreds of these common sense generalisations in our daily life, such as the following: "Persons in pain tend to want to relief that pain", or "persons who are angry tend to frown". All assumptions together form the framework of folk psychology.
Now returning to eliminative materialism, the conception of folk psychology is an old efficient framework that on the long run will be replaced by the even better framework of futuristic neuroscience. The one-to-one match up between neuroscience and mental experiences is therefore false because these mental states can only be described in linguistic terms which are in turn derived from and integrated within the network of folk psychology.
In this chapter an overview was given from a number of central ideas about how to bridge the gap between mind and matter. It should not be forgotten that the meaning of this overview was not to quote all possible theories evolved yet because for that purpose infinitely more space should be needed. In this chapter it was tried to provide a kind of "philosophical map" on which most of the currently evolved theories can be classified. As we have seen usually new theories are specified or adapted versions of older theories and therefore this overview might also be very useful in understanding the position of a new theory about mind and matter. In the next chapter the philosophical evidence for and against the theories of this chapter will be evaluated. Chapter 5 and 6 will focus on the current discussion that is going on due the rapid developments in the fields of neuroscience and computer technique.
Further reading and references (nr):
6, 8, 9, 18, 27, 28, 32, 33.
.c2.Arguments for theories of consciousness:
In this chapter we will have a closer look at some of the arguments that played an important role in the discussion between dualism and materialism. As we shall see, most of the arguments pleading for dualistic theories will often plead against materialistic ones and visa versa.
.c3.The argument from religion:
The first argument that will be discussed is the argument of religion and is mostly used to support dualistic theories. Often people do not choose position in the mind/matter discussion solely on grounds of findings, also peoples believes tend to have great influences on peoples viewpoints. Religious people often argue that life is not without meaning and that some sort of higher force or power is existing where people are only a small inferiour part of. Here the emphasis is laid on the fact that this seems to be quite impossible without having consciousness in other terms than matter only.
Against this reasoning can be argued that it does not prove anything. A belief, on which this argument is founded in the first place, is not a finding, but an opinion in itís origin. Opinions or beliefs within the scope of religion can vary dramatically. Sometimes different opinions are even contradicting. To use these beliefs as proof for a dualistic theory of consciousness would therefore be without much force.
.c3.The argument from introspection:
A more universal consideration pleading for dualism is the argument from introspection which tries to appeal to the common experience. We experience consciousness in terms of emotions, fluxes of thoughts, desires etc. It seems that these mental states and properties as revealed to us by introspection could hardly be more different from physical states than thinkable. If we take for example the experience of pain. If one would have to describe how it feels from the point of introspection if one's finger is hit with a blow with a hamer, the description will probably be something like: very unpleasant, painful etc. These types of descriptions seem to be completely different from explanations in terms of electrical pulses, molecular bindings and enzymatic prosesses. The argument from introspection claims that we experience the pain, and not the molecular or physical processes that might be connected to it in some way. The verdict of introspection therefore seems strongly on the side of some forms of dualism, on the side of property dualism at a minimum. At the same time introspection is pleading against materialism. It claims that by allowing the reduction of mental states to physical states, materialism denies the mental values as they are experienced by us. To illustrate what is meant here, the next thought experiment might be helpful. Imagine that in the far future neuroscience has extended it's knowledge that it is aware of, and understands everything that is going on in the brain. Now imagine at that point in time lives a scientist who is colourblind and cannot "experience" the colour red. Although the scientist exactly knows how the "sensation of red" is accomplished via a certain phisiological process, according to dualists the researcher still doesnot know what the experience of red really is.
Although this argument seems to have much more force than the argument from religion, there are still a few things that can be said against it. Introspection assumes that our faculty of inner observation or introspection reveals things as they really are in their innermost nature. This assumption is suspect because we already know that our other forms of observation, sight, hearing, touch and so on do no such things. The red surface of an apple doesnot look like a matrix of molecules reflecting photons of certain critical wavelenghts, but that is what it is.The sound of a flute doesnot sound like a sinusoidal compression wave train in the atmosphere, but that is what it is. If oneís pains and hopes and beliefs do not introspectively seem like electrochemical states in a neural network, that may be only because our faculty of introspection, like our other senses is not sufficiently penetrating to reveal these things. The argument is therefore without force unless we can somehow argue that our faculty of introspection is quite different from all other forms of observation and this is probably the fundamental point in which dualism and materialism differ in opinions.
.c3.The argument from ireducibility:
Another cluster of important considerations with respect to the dualism/materialism debate can be collected under the argument from irreducibility. Here one points to a variety of mental phenomena where it seems clear that no purely physical explanation could possibly account for what is going on.
In fact this argument brings us back to a completely different discussion namely between reductionism and holism, which is perhaps as complex as the mind/brain debate itself. In short, reductionism claims that features at a higher level can be reduced to features of a lower level without loss of content. For example, organisms can be reduced to individual organs where each organ is a part of the organism. If one understands the function of each individual organ, one would understand the whole organism as well. Organs can be reduced to cells, cells can be reduced to molecules, molecules to atoms, quarks etc. Holism agrees that for complex organisations sometimes building subunits can be distinguished, however in contrast to reductionism it claims that no reduction can be made without loss of part of the original feature. In other words, the complexity of the whole organism is greater than the sum of the value of all the individual building units.
When we extrapolate the reductionism/holism debate to the dualism/materialism discussion it will be clear that reductionism is on the side of materialism, and holism on the side of dualism. If the argument of irreducibility claims that no possible physical explanation could account for what is going on, materialism can reply that this is due to our current restricted knowledge of neuroscience. Perhaps now in this century we are not able to to explain all mental phenomena by purely physical processes, but as time goes by perhaps in the future we will. For example, Descartes was highly fascinated by man's mathmatical capabilities, and he strongly correlated them to consciousness. As we all know now, the average personal computer at present posesses the calculating capabilities that in many ways outrun our own capabilities completely. Perhaps in the far future we are able to understand and copy things of the human brain far better than we now can imagine.
So reductionism seems to be an important issue in the debate of dualism and materialism. There are however different ways in which reductionism can be applied. It can be used as the ultimate means to reduce any complex structure into something of a lower level. Reductionism can also be used as a method only. Popper demonstrated that it is indeed possible to be a dualist on one side, and apply reductionism as a method on the other (ref 4).
Firstly he demonstrated that scientists have to be reductionists in the sense that nothing is a greater succes in science than a successful reduction (such as Newtonís reduction of Keplerís and Galileoís laws to his theory of gravity, and his correction of them). A successful reduction is perhaps the most successful form conceivable of all scientific explanations, since it achieves (what Meyerson stressed:) an identification of the unknown with the known.
Secondly Popper suggested that, whatever the philosophical attitude of scientists towards holism, they have to welcome reductonism as a method. Holists are in fact critical reductionists in a sense that they refer to the incompleteness of most reductions. About this issue can be said that hardly any major reduction in science has ever been completely successful: there is almost always an unresolved redidue left by even the most successful attempt of reduction. In addition Popper claims unsuccessful reductions can on the other hand can be very useful, because also from them can be learnt a great deal.
In the article Popper illustrates these theses by two examples of reduction, one derived from mathematics (the reduction of rational fractions in ordered pairs of natual numbers) and a second derived from physics dealing with the reduction of all matter to basic physical properties and extentions.
.c3.The argument from development:
Further support for materialism is usually obtained from the process in which consciousness arises, like during embrional developement and in evolution. If mental states cannot be reduced to physical states then how can one explain the development from a fertilised egg cell (which posesses little or no consciousness at all) to a human being that has the complete range of mental states as we know them? Furthermore, the evolution theory implies that higher animal forms (with higher states of consciousness) have evolved from lower life forms like bacteria (or even a self-replicating molecule). Somewhere along the line from protists to mammals consciousness must have arisen. Materialsts suggest that, if the process of evolution from lower to higher lifeforms can be understood in physical terms (which is claimed by reductionists) then why shouldnot the same be one for the appearance of consciousness?
The next argument in the discussion between dualism and materialism is urged by materialists and it appeals for the greater simplicity of their view. It is a principle of rational methodology that, if all else is equal, the simpler of two competing hypothesises should be preferred. This principle is somethomes calles "Ockham's Razor" after William of Ockham, the first philosopher who came up with this thesis: "Do not multiply entities beyond what is strictly necessary to explain the phenomena". The materialist postulates only one kind of substance (physical matter) and one class of properties (physical properties), whereas the dualist postulates two classes of substances and two classes of properties. To no explanatory advantage charges the materialist. This is not yet a decisive point against dualism since neither a dualistic nor a materialistic approach can explain all the phenomena to be explained. However, it seems clear that at this point physical matter is capable of explaining much more than spiritual matter and this brings us to another point against dualism: the relative explanatory incompetence. It runs as follows: Consider very briefly the explanatory resources already available to the neuroscience. We know that the brain exsists and we begin to have an good idea of its organisation. We know much of it's microstructure, how neurons are organised, how they are connected to sensory organs and muscles. At present we can explain a number of mental disorders (for example apraxia coused by a tumour) purely on neurological, and thus physical grounds. All together, the framework of neuroscience is capable of explaining and predicting quite something.
Compare now what neurosciene can tell us about the brain, and what can be done with this knowledge, with what dualism can tell us about spiritual substance, and what can be done with those assumptions. Dualists cannot tell us what the mind stuff is made of, the elements that make it up, the laws that govern their behaviour or what so ever. In fact dualism can do none of these things because no detailed theory of mind stuff has ever been formulated. Compared to the rich resources and explanatory successes of current materialism it is an empty space waiting for a theory of mind to be put in it.
So the argument of explanatory impotence tells us that although the theory might be valid, there are not as much indications supporting it as there are supporting neuroscience and materialism.
Further reading and references (nr):
4, 8, 9, 15, 21, 23.
Over the last decades computers have evolved rapidly. Some fifty years ago it would have been unthinkable what people nowadays can do with computers. Not only did they appear to be fast calculators of complex mathematical problems, they also showed to be very useful in lots of other fields. The modern world is almost unimaginable without the role of computers. In this chapter the influence of the computer on the mind/brain problem will be discussed.
.c3.The history of Artificial intelligence:
The great power of computers lays in the field that is very similar to the process that we call "thinking". The first computers were in fact no more than calculators. Making calculations however is very often seen as thinking; when a pupil on the secondary school manages to solve a "difficult" equation, the teacher response could very well be "Good thinking John!". This very same equation could be solved by some appropriately programmed computers in seconds.
The question can be raised: "Can computers think?", or more generally, "Can machines think?". The answer of this question depends largely on ones definition of "thinking". If we define "thinking" as "the capability to solve mathematical problems" then perhaps the answer is yes, computers can think. It seems obvious however that human thinking includes much more than solving mathematical equations. Many of these skills are, at least at this moment, not executable by computers.
Also the question "Do computers have consciousness?" depends on ones definition of the word consciousness. In the beginning of this paper, consciousness has been defined as "The experience that everyone has privately for him or herself". As is shown earlier, this is not the only definition possible. A large group of researchers who are working in the field of Artificial Intelligence have a different functional approach. Everyone or everything that manifests intelligent- or thinking behaviour, per definition possesses consciousness.
Allen Turing was the first person who put together a test procedure that was made to discriminate between thinking and non-thinking computers. The Turing test can best be referred to as an imitation game. In this game, a human being was replaced by a computer, in such a way that it was not visible for the outer world. It is for example possible to communicate with either a human being or a computer via a modem; the only thing that one can see is the text that is sent by either of them. Via the computer screen one could ask all kinds of questions to the human/computer and evaluate the replies. A computer passes the Turing test when it can not be distinguished from a human being. So if a computer is able to act like it can think, it can think, according to the Turing test. The same counts for intelligence: if the computer behaves intelligently, it is by definition intelligent.
Several programs have been made that are capable of communicating like humans do. The first of these programs was called ELIZA and it was based on repeating part of the sentence that was given to the computer. A conversation with ELIZA could be as follows:
ELIZA -How do you feel today?
Human -I feel fine.
ELIZA -Why do you feel fine today?
Human -Because I have my day off.
ELIZA -You have your day off?
Human -Yes, it is holiday and I am going to Spain next week.
ELIZA -Oh, I see.
This conversation shows that the computer is able to respond quite resonably. The computer however does not have a clue what the conversation is about. After a few minutes most people probably would be bored and stop communicating. In the newer program SHRDLU the makers tried to increase the level of understanding. It is however very difficult to do this. The problem is that if one wants to use language correctly, one needs to have a lot of knowledge about the surrounding world. One needs a framework in which the meaning of all words are connected to each other, and coupled to some phenomenon of the outside world in the right context.
A few remarks should be made about the Turing test. First of all, the Turing test was invented in the 1950's which is, on the scale of time regarding to computer developments, a long time ago. By that time only very primitive computers were known. At this moment computers are able to make much better simulations then in the 50's.
One of the main objections against the Turing test has been that it is only capable of imitating a very limited part of human beings. It might give answers to questions like a human would do, but it is still not capable of for example picking up an item from the table, driving a car, etc. One could argue that if a computer is programmed well, it might just be able to execute these behaviours. At this moment computers are not capable of doing that, but in the future they might be. In fact at this moment most people agree that probably in the future this will be possible.
The emphasis in the discussion seems to have shifted from the Turing test to a new question. The main point in the discussion now seems to be: Is consciousness based on computational procedures?
The conversation program ELIZA for example is capable of imitating a human being behind a modem. ELIZA therefore might pass the Turing test. Within the program however, there does not seem to be any level of understanding what the conversation is about. For this reason some sceptics argue that the Turing test does not give the answer whether a computer has consciousness or not. A computer that passes the Turing test is a successful imitation of a conscious being, but therefore it does not necessarily posses any state of consciousness itself. For that, at least a computational structure is required that is able to understand the meaning of the input before it responds with the output.
.c3.Modern neural computation networks:;
Neural computing networks are a part within the computer science that is occupied with the development of new computational structures. The idea is to take the principles from information processing like it happens in the brain, and apply them in new computers.
There are a number of important differences between information processing in computers and in the brain. One of the most obvious and important is that brains work parallel and a computer does not.
The information processing in a computer works as follows: The complete processing from input to output is devided in a number of steps. The input will first be processed by the first step. The outcome of the first step then will be processed by the second step. After this one the next step follows, and this goes on until the complete processing is carried out. Every step is executed very quickly, however the computer does only one step at the time.
Brains on the contrary are capable of doing more than one processing step at the time. Brains are capable of working parallel and this is the result of its different architecture in comparison to the computer. As is discussed in the previous chapter, the basis unit of the brain is a neuron which is capable of processing information. It receives information from preceding cells via nerve fibres called dendrites, the information is processed, and it sends this information via a single long fibre (the axon) to the next cells. Different cells are capable of processing information at the same time. Furthermore, the cells are arranged in such a way that individual cells receive information from a number of preceding cells, and also send their information to a number of following cells. This way of processing information is called parallel processing. At certain places in the brain (at the cortex) a number of layers can be discriminated. Each layer consists of neurons that receive information from a higher layer only, and strictly send it to a lower layer. This structure is schematically depicted in figure 2.
At this moment the human brain is superior to a digital computer in the performance of many tasks. A good example is the processing of visual information: a one-year-old baby is much better and faster in recognising objects, faces, and so on than even the most advanced AI system running on the fastest computer. There are many features that would be desirable to take over in artificial systems: Computation in the brain is robust and fault tolerant. Nerve cells die every day without affecting its performance significantly; It is flexible. It can easily adjust to a new environment by "learning" and does not have to be reprogrammed; It can deal with information that is fuzzy, probabilistic, noisy or inconsistent; it is small, compact and uses very little power.
An important step in the development of neural networks was the construction of the McCulloch-Pitts neuron. This neuron is an small processing unit that essentially works the same as a biological neuron. The only difference is that it is a very simplified version of its biological counterpart. In figure 3 a schematic diagram of the McCulloch-Pitts neuron is given.
McCulloch and Pitts proposed a simple model of a neuron as a binary threshold unit. Specifically, the model neuron computes a weighted sum of its input from other units, and outputs are one or a zero according to whether this sum is above or below a certain threshold. (In a biological system this threshold model is also assumed, and only when the total input exceeds the threshold, an action potential is supposed to be generated.)
Most of the current research runs on (silicon) neural networks that are like brains constructed in multi-layered connections. It is interesting to note that most networks are not exactly programmed, but instead they are given numerous examples from which they "learn" how to make a calculation. The research partly focuses on how these learning processes can be achieved by altering the threshold of neurons as a result of their input.
It is worth remarking that the typical cycle time of neurons is a few milliseconds, which is about a million times slower than their silicon counterparts, which are called semi-conductor gates. Nevertheless, the brain can do very fast processing for tasks like vision, motor control, and decisions on the basis of incomplete and noisy data. This is obviously possible only because billions of neurons operate simultaneously. The capacity of a system that operates parallel with switching times of current semiconductor devices therefore would be far beyond our imagination.
Almost everything in the field of neural computation has been done by simulating the networks on serial computers. Neural network (VLSI) chips however have been created. The main problem with making neural network chips is that one needs a lot of connections, often some fraction of the square of the number of units. The space taken up by the connections is usually the limiting factor for the size of the network. The neural chips made so far contain of the order of 100 units, which is too few for most practical applications.
Potential alternatives to integrated circuit chips include optical computers. The field is very young but electro-optical and optical associative memories have already been proposed or built.
It should not be forgotten that the main purpose of most Artificial Intelligence research is to create better and quicker computers that might have many useful applications. An optical recognition system for example that can deal with noisy data would be very helpful in many industrial processes. These new computers are built according to the hierarchy known from the biological brain. It must be stressed that most of this work is not occupied with the question whether it is possible or not to put consciousness into a computer. Some of the developments however do give some interesting information with regards to this question.
As is shown earlier, the main point of discussion nowadays is whether it is possible to ascribe consciousness to a computational system. It is interesting to see that, as a result of the AI work, some computational systems in computers seem to have come much closer to that of the brain than ever before. In chapter 1 some arguable criteria for consciousness in biological neural networks have been given.
The first criteria that will be discussed here is that of memory. At the moment both brains and computers posses some sort of memory. It is interesting to see that in a neural parallel computer network the structure of the memory resembles biological memory much more than traditional computers. One could argue that, using the argument of analogy, it is more likely that this new type of computer has a conscious state.
The same counts for the criteria of input of information; both traditional computers and brains are (or in case of computers: can be) connected to sensory devices. The brain for example is connected with the eye, while computers can be linked to electronical variants of the eye. In a parallel computer network the information is processed roughly in the same way as it takes place in the brain. It is theoretically even possible to create the network in such a way that an internal representation of the external world is present within the network. Also here, according to the argument of analogy, it might be more likely that these computers posses consciousness.
As time goes by, AI will probably be capable of reproducing more and more of the human brain in computer networks. As these developments go on, the analogy between computers and brains will increase. There will however always remain two fundamental differences between brains and computers:
The first difference is that computer network will likely always be a simplification of the biological counterpart. AI is capable of imitating the biological currents, but these are so complex, and include most likely so many unimportant details that they will probably never be introduced in any AI network.
The second difference between computers and brains is even more fundamental and refers to the basis substance the networks are made from: Brains are made of carbons and computer networks are made of silicones. AI might be able to transfer all basic principles of the brain to a network, but there will always remain a difference between the building stones the networks are made of. People who believe that some computers do have consciousness, in fact believe that these network principles are enough to entail consciousness.
The last interesting point of discussion is about the simulation of parallel networks on serial computers. At the moment most of the work with parallel networks is done this way. Some researchers share the opinion that, because a serial computer can simulate a computer with possible conscious states, the same conscious states should be ascribed to the serial computer as well. Opponents of this argument have objections against the simulation step. They claim that there is a difference between representation and reality: The representation of something with consciousness does not automatically ascribe consciousness to that simulation itself. For example, the simulation of a cow does not give milk either. One could argue however that in some cases the representation is in fact the same as the perceived reality. For example, one could make a mathematical model to provide proof for a certain problem. The representation of proof is in this case proof itself.
Further reading and references (nr):
5, 7, 10, 12, 13, 19, 20, 30, 34
.c2.The prospects of the mind/brain discussion:
In the previous chapters several aspects of brain research, that is related to the mind/brain discussion, have been discussed. What will be the influence of the new developments in neurobiology on the current mind/brain discussion?
To answer this question, first we have to investigate what new insights these techniques actually give. Both the imaging techniques and the discovery of synchrony offer additional understanding how the brain works. The imaging techniques provide information that increases the understanding of how and where certain brain capacities function. In other words, a better localisation is found of functions like speech and reading. In addition, more detailed studies are being done on the localisation of perception (sight/hearing) and even on emotions or emotional states. The understanding of synchrony on the long run might and probably will identify certain information patterns that code for conscious experiences like "seeing an object", "feeling an emotion", "having the intention to move" etc.
To conclude, both new developments help in providing further materialistic foundation of mental processes. They offer materialistic mechanisms to locate, identify and explain certain conscious states. In chapter 1 it was shown that over the last three centuries this proces already took place and that it resulted in a general support of materialism, in a sense that it became harder to support naive dualism due to the constant materialistic findings. In fact these new techniques continue that proces. The result will be that also in the future dualistic theories will either have to take (part of) the outcome of neuroscience into account, or alternatively have to point at the limited scope of neuroscientific theories (like anti-realism does). Within the neurological framework it will become harder to deny the great explaining power and understanding of the theories.
In "The history of brain research" it was demonstrated that often a new finding had a great impact on the whole field of research. Perhaps one can speak of "trends" in research. At certain points large numbers of researchers were occupied with reflexionism, then with connectionism, and so forth. The theories that were proposed to explain behaviour were a reflection of the brain research that was being done at the same time. In this sense it is interesting to see that most current research focuses on the "dynamic" aspect of the brain. Especially the relation between brain activity and time has become more important. It is very likely that this aspect therefore will appear in the near future in materialistic explanations.
One important question remains to be considered, namely: Will neuroscience ever be able to completely understand and predict what is going on in the brain? Here I will argue that neuroscience probably will not. Science is capable of identifying the general rules and processes that govern the brain. It must not be forgotten that all scientific laws are reductions of many observations. Eventually neuroscience might reach a very high level of understanding of how the brain works. However, to predict what is going to happen, what the behaviour of someone will be on the long run and so forth, will probably not be achieved. Although generalisations can be made, the differences between individual people remain. this demands the extensive knowledge of every single molecule that is present in the brain. Even if that would be possible (perhaps with the help of imaginable super computers) one could argue that predictions will never be possible due the incompleteness of all reductions. As Popper demonstrated (ref 4) even the most successful reductions in science are never complete reductions. (Reductions don't have to be complete as long as the purpose is to use them as a method only). To conclude, neuroscience will expand the understanding of the brain dramatically, but will probably never be able to predict exactly the activity of neurons at a later timepoint. As a result, any prediction of long term human behaviour and consious states will become impossible.
The prospects of futuristic Artificial Intelligence work looks very promising. At the moment already systems are being developed that can deal with a large quantity of information in a remarkable short time. The practical applications of AI systems will probably increase significantly over the next few decades. Especially the "error tolerant" aspect as can be found in biological brains will play an important role. It is very interesting to see how successful AI is, with respect to taking neurological principles of how biological networks work, and applying them to silicon networks; in the past the multi layer principle was copied, and at the moment researchers are developing networks that include information transduction as is found in the neurological field of synchronisation. The first data looks very promising!
With respect to the mind/body problem the developments of AI will have a considerable influence. As a result of improving AI technology, computers are heading in the direction where they are increasingly capable of copying (aspects of) human behaviour in a remarkably good way. On grounds of the theories of analogy and folk psychology, I argue that this can lead to a change of people's view with regard to the questions "Can computers think?", and "Do computers have consciousness?". If a computer is capable of showing more resemblances to humans, the analogy between humans and computers becomes higher; Two objects that share at least some principles in structure are more analogue than objects that do not. Furthermore, if computers in the future start to behave like humans, it becomes easier to believe that they are conscious beings as well. A personal computer for example does not really seem to demonstrate conscious behaviour. On the contrary, a computer that is capable of understanding a (fuzzy) daily life conversation, or that seems to be frightened when you are angry at it, such a computer is easier to ascribe a conscious state to. According to the theory of folk psychology people tend to base their opinion on numerous subtle generalisations that have been evolved during history. A generalisation could be: "When someone is hit with a hammer he tends to feel pain." Another generalisation could be: "People who feel pain tend to cry." Now, if a computer demonstrates the resembling behaviour of feeling pain (crying after being hit), and it follows the generalisations that count for human beings, it becomes hard to deny that the computer doesn't have emotional states (which are conscious states). So, improvements not only bridge (part of) the gap between humans and computers concerning their analogy, they also might convince people of being conscious more easily in the future purely on folk psychological grounds. The last point is demonstrated magnificently by a fragment of the documentary "victim of the brain" (Dennet and Hofstadter) where the main character had difficulties with destroying a computer driven mouse, just because it demonstrated (human) emotional behaviour like being frightened, aggressive, or cuddly.
1. Ackerman, S.: Discovering the brain (Washington, D.C : National Academy Press, 1992).
2. Alquiné, F.: Oevres philosophique. (Paris, Garnier, 1963).
3. Ayala, F., Dobzhansky, T.: Studies in the philosophy of biology. Chapter 7, J.C. Eccles: Cerebral activity and consciousness. (Berkley, University of California Press, 1974)
4. Ayala, F., Dobzhansky, T.: Studies in the philosophy of biology. Chapter 16, K.R. Popper: Scientific reduction and the essential incompleteness of all sciense. (Berkley, University of California Press, 1974)
5. Bertram, R.: The thinking computer: Mind inside matter. (San Fransisco, Freeman, 1976).
6. Biro, J. et all: Mind, brain and function. (Norman O.K., University of Oklahoma press, 1982).
7. Bullock, T.H.: Introduction to nervous systems. (San Fransisco, Freeman, 1977).
8. Churchland, P.: Neurophilosophy. (Cambridge, Bradford Books, 1986).
9. Churchland, P.M.: Matter and consciousness. England, Bradford books, 1985).
10. Cummins, R.: What can be learnt from brainstorms? Reprinted in: D.R. Hofstadter, D. Dennet: The mind's I. (New York, Basic Books inc., 1981).
11. Delgado, J.M.: Physical control of the mind: Towards a phsychocivilised society. (New York, Harper, 1970).
12. Dennet, D.C.: Brainstorms. (Vermont, bradford books, 1978).
13. Dennet, D.C.: Where am I?. Reprinted in: D.R. Hofstadter, D. Dennet: The mind's I. (New York, Basic Books inc., 1981).
14. Descartes, R.: Discourse on method, part 5 (London, Dent, 1975).
15. Eccles, J.C.: The self and the brain. With K. Popper. (New York, Springer verslag, 1977).
16. Fearing, F.: Reflex action. Chapter 11: The Pluger Lotze controversy. (Cambridge, Mass, 1970).
17. Geopopoulus: Neuronal population coding of movement direction. Science 233:1416-1419.
18. Globus, G.G.: Consciousness and the brain. (London, Plentinum press, 1976).
19. Hertz, J. et all: Introduction to the theory of neural computation. (Redwood city, Addison Wesley publishing company, 1991).
20. Hofstadter, D.R.: The Turing test: A coffeehouse conversation. Reprinted in: D.R. Hofstadter, D. Dennet: The mind's I. (New York, Basic Books inc., 1981).
21. Hull, D.: Philosophy of biological science. (Englewood cliff NJ, Prentice Hall inc., 1974). Chapter 5.
22. Jeannerod, M.: The brain machine. (London, Harvard university press, 1985).
23. Korthals, M.: Wetenschapsleer. (Heerlen, Open Universiteit, 1987). Chapter 3.
24. Macolm,N.: Knowledge of other minds. Journal of philosophy vol LV (1958). Reprinted in: V.C. Chappell: Philosophy of mind. (Englewood Cliffs N.J., Prentice Hall, 1962).
25. Maine de Biran, F.: Influence de l'habitude sur la faculté de penser. (Paris, presses universitaires de france, 1954)
26. Nagel, T.: What is it like to be a bat?. Readings in philosophy of psychology. Reprinted in: D. Hofstadter, D. Dennet: The mind's I. (New York, Basic Books inc., 1981).
27. Plate, U.T.: Is consciousness a brain prosess? Britisch journal of psychology, vol XLVII. Reprinted in: V.C. Chappell: Philosophy of mind. (Englewood Cliffs N.J., Prentice Hall, 1962).
28. Rosenthal, D.M.: Materialism and the mind Body problem. (Englewood cliffs N.J., Prentice Hall, 1971).
29. Schoppenhauer,A.: Ueber den willen in der natur. (Frankfurt 1936).
30. Searle, J.H.: Mind, brain and programs. The behavioral and brain sciences vol 3. Cambridge university 1980. Reprinted in: D.R. Hofstadter, D. Dennet: The mind's I. (New York, Basic Books inc., 1981).
31. Singer, W.: Synchronisation of cortical activity and its putative role in information processing and learning. Annu.re.physiol. 55:349-75.
32. Smart, J.J.C.: Sensations and brainprocesses. Philosophical review, vol LXVIII, 1959. Reprinted in: V.C. Chappell: Philosophy of mind. (Englewood Cliffs N.J., Prentice Hall, 1962).
33. Stich, S.C.: From folk psychology to cognitive sciences, a case against belief. (Cambridge, Mit press Bradford, 1983).
34. Turing, A.M.: Computing machinery and intelligence. Reprinted in: D. Hofstadter, D. Dennet: The mind's I. (New York, Basic Books inc., 1981).