INTRODUCTION
“Incurable optimist that I am, I find this recent invasion by physicists into the domains of cognitive neuroscience to be a cloud with a silver lining: for the first time in my professional life, an interloping discipline beats out philosophy for the prize of combining arrogance with ignorance about the field being invaded. Neuroscientists and psychologists who used to stare glassy-eyed and uncomprehending at philosophers arguing about the fine points of supervenience and intensionality-with –an-s now have to contend in a similar spirit with the arcane of quantum entanganglement and Bose-Einstein condensates. It is tempting to suppose that as it has become harder and harder to make progress in physics, some physicists have sought greener pastures where they can speculate with even less fear of experimental recalcitrance of clear contradiction.” (Dennett ‘Sweet Dreams: Philosophical Objections to a Science of Consciousness’ p. 10)
In this blog-post I will evaluate physicist Edward Witten’s claim that consciousness will remain a permanent mystery that science will never reduce to basic physics. In the first half of the blog-post I will evaluate Witten’s argument (such as it is), and criticise his view as lacking in sufficient justification. I will discuss why people seem so impressed by Witten’s claim despite the lack of evidence presented by him. In the second half of the blog-post I will attempt to present a philosophical justification for why Witten may hold the views on the mind that he does and show that a possible argument for his position is constructible but that Witten hasn’t presented this argument himself.
Part 1: An evaluation of Witten’s “argument” for the non-reducibility of consciousness to physics.
A little over two years ago Edward Witten declared in an interview that consciousness would more than likely forever remain a mystery which science couldn’t explain. Witten’s entire argument went as follows: Future Science will make great progress in giving functional explanations of how the brain is implemented in causing consciousness. However, it is unlikely that future science will ever get a handle in explaining why consciousness has the form that it does. Witten says he finds it inconceivable that future science will ever explain the hard problem of consciousness. He went on to speculate that he believed that science is more likely to explain the nature of big bang than it is to solve the hard-problem of consciousness.
On the face of it Witten hasn’t said anything particularly controversial. Most theorists working on consciousness are aware of the so-called hard problem of consciousness and there are various competing approaches to deal with the problem. What is surprising though is that people found Witten’s speculations particularly interesting. Given that Witten didn’t provide a single piece of evidence to support his claim, but simply asserted that he personally didn’t think there would be any progress in a field he doesn’t work in, one would imagine that his speculations would have evinced a shrug of the shoulders. But no his speculations made news. Scientific American writer John Horgan wrote a piece about how the most brilliant physicist the world has seen since Einstein has come out as a mysterian about consciousness. Horgan noted that Witten was adopting a position similar to philosopher of mind Colin McGinn and linguist and cognitive scientist Noam Chomsky. There was an air of triumph in Horgan’s article he seemed delighted to have a famous scientist on board with the mysterian position. Why Horgan was so happy is unclear. Witten didn’t present a brilliant argument to support his position, nor did he present any empirical data to support his position. All Witten did in the short interview was assert that he personally found it difficult to conceive of a solution to the hard problem of consciousness. Yet some people take this bland assertion by Witten as a vindication of the fact that consciousness will permanently remain a mystery to science.
Since there seems to be no logical reason why one would be impressed with Witten making an assertion, for which he provided no evidence, one is left to speculate why Horgan and others were so excited about Witten’s claim. One obvious reason why a person would be so impressed by Witten’s non-argument is that he is a physicist. Physics is our most basic science, and theoretically we would like to reduce all other claims to the basic facts of physics. When an expert in physics claims that such a reduction is impossible one should take notice. Such an expert in physics will know more about the fundamental facts that govern our cosmos than philosophers like Daniel Dennett or Patricia Churchland will know. So perhaps an expert like Witten given his superior knowledge of physics; should be taken seriously, when he makes claims about the reduction of something to basic physics. Perhaps. Though given that Witten didn’t present a single piece of evidence to support his position; I would argue scepticism would be a more appropriate response, rather than the howls of delight that greeted Witten’s assertion.
At the very least if we are to argue that Witten’s claim has any validity he should be interviewed to try to understand precisely how he proposes to support his claim. Furthermore when discussing the reduction of x to y; it should be acknowledged that while Witten is legitimately considered an expert on x, there is no reason whatsoever to believe that Witten knows anything about y. In other words, while everyone agrees that Witten is an expert on physics he has never demonstrated any competence or understanding of the various sciences related to the mind; neuroscience, cognitive science, behavioural science, anthropology, philosophy of mind, phenomenology etc. This should be a worry for the people who want to treat Witten as the ultimate authority on the nature of the mind and its relation to physics. I would imagine that those who aren’t worried will fall into two camps (1) Those who have documented evidence on Witten’s expertise on the nature of the mind and consciousness, (2) Methodological Dualists who think that while we need to rely on the standard methods of science to understand all other aspects of the physical world, but that when it comes to humans above the neck casual reflection is sufficient achieve understanding .
I doubt that anyone will present evidence of Witten’s years of engagement with the sciences of the mind; though I would love if they did, I suspect the reason Witten worshipers don’t think he needs any training on the mind is because they are implicit methodological dualists. It strains credulity, given the stunning amount we have learned in the sciences of the mind over the last hundred and fifty years, that anyone could hold such an absurd anti-scientific world view. Nonetheless, a substantial amount of philosophers and physicists believe that when it comes to human consciousness, casual reflection on the contents of experience is sufficient to justify wide ranging metaphysical conclusions.
Science and Introspection: A Brief Interlude
As stated above it is unclear whether Witten has any knowledge of the relevant sciences of the mind; or whether he is just using his intuitive understanding of the consciousness to demonstrate that it cannot be reduced to physics. If Witten, or his followers are merely relying on casual introspection to support their claims, caution is warranted; scientific experimentation has demonstrated that consciousness may not be all that it pre-theoretically seems to be.
In Nick Chater’s 2018 book ‘The Mind is Flat’, he argues that current experimental data demonstrates that our mind doesn’t have the structure that casual introspective inspection implies. There is a certain sense in which Chater’s book reads like an updated version of Dennett’s 1991 ‘Consciousness Explained’. In fact Chater begins his book with a quote from Dennett’s ‘Consciousness Explained’:
“When we claim to be just using our powers of inner observation, we are always actually engaging in a sort of impromptu theorizing-and we are remarkable gullible theorizers, precisely because there is so little to ‘observe’ and so much to pontificate about without fear of contradiction” (Dennett ‘Consciousness Explained p. 68)
Chater argues in a similar manner to Dennett. He thinks that our so called introspective observation is a kind of impromptu theorising. Chater even uses the same example as Dennett to illustrate his point. It seems upon casual inspection that our entire visual field is in colour. An experiment which anyone can do is to place a playing card on front of you and to then move it to the periphery of one’s visual field. A typical person who is asked would guess that the card will be seen in colour whether it is at the centre or at the periphery of one’s visual field. However, as anyone who has done the experiment would attest when the card is at the periphery of one’s visual field it actually appears to be black and white. The result of this simple test, Dennett and Chater argue, shows that sometimes our supposed direct introspective experiences are actually just implicit theories we are using.
Chater presents a list of experimental data to show that introspection isn’t as reliable as we may uncritically assume:
(1) Impossible Objects: These are objects created by the Swedish artist Oscar Reutervard. These works of art give the appearance of being three dimensional objects. However when we analyse the objects we discover that the objects would be impossible to build as three dimensional. Chater draws three conclusions from these impossible objects (A) We aren’t inspecting some inner object that the mind has constructed as a mirror of reality, the mind is better thought of as an inference machine than as mirror of reality (B) We aren’t inspecting these impossible objects as a coherent whole, rather the mind is inspecting different parts of the object at different times in a piecemeal way. (C) Our misplaced confidence in introspection. We assume we are inspecting an actual three dimensional object, but we are in fact doing no such thing.
(2) Our assumption is that our visual field is uniform that we see colour at the periphery of our visual field. However, it is well known experimentally that our peripheral vision is both blind and blurry (‘The Mind is Flat’ p. 42). So again what we think we are experiencing through casual introspection, is now revealed to be false through detailed experimental studies. As Chater notes when discussing simple illusions: “Our visual experience can depend, rather dramatically, on where we are looking-and we are certainly unable to ‘load up’ this entire image into our minds- even though it is actually very simple and repetitive.” (ibid p. 43).
(3) Our intuition that we can see words all over the page when we read it is also an illusion. Chater discussed experimental data which uses ‘eye gaze tracking devices’ to demonstrate that in computer screen tests which consist of mostly blanked out words, interspersed with windows of words, if the computer maps the windows of texts to the eye gaze movements the person will not be aware that the screen consists of mostly blanked out words. Chater draws the following conclusion from the study: “The results suggest that the eye and the brain picks up little outside a very narrow ‘window’. Indeed, we can go a bit further the evidence suggests that we can only read one word at a time…Looking at a crowd, it turns out that you can only recognize one person at a time; looking at a colourful scene, you can only report colours or details of things that you are looking at directly. This does not imply, of course, that you pick up no information at all about objects you are not attending to- just that this information is extremely sparse. (ibid p. 46)
Other pieces of evidence which can be used to demonstrated that our consciousness experience isn’t as detailed as we like to think is the phenomenon of change-blindness. Change-blindness has been studied many different times experimentally. In his 2005 book ‘Sweet Dreams: Philosophical Objections to a Science of Consciousness’ discussed the phenomenon of change-blindness and the difficulties that it raises for defenders of qualia.
In his classes Dennett often used a demonstration from an experimental study by Rensink et al ‘To See or Not to See: The Need for Attention to Perceive Changes in Scenes’ (1997), as a way of tapping his students intuitions about the notion of qualia. In the demonstration two almost identical photographs are shown for almost 250 milliseconds each, separated by a 290 millisecond mask. This is repeated in alternation till the subject notices a difference between picture A and picture B. It usually takes between 20 and 30 seconds till people notice the differences between the pictures ( ‘Sweet Dreams’ p. 82).
When Dennett presents this test to his students, after they have finished it and indicated that they noticed the changes between the pictures, he asks them a question:
“Now before you noticed the panel changing colour, were your colour qualia for that region changing? We know that the cones in your retinas in the regions where the light from the panel fell were responding differently every quarter of a second, and we can be sure that these differences in transducer output were creating differences farther up the pathways of colour vision in your cortex. But were your qualia changing back and forth- white/brown/white brown-in time with the colour changes on the screen? Since one of the defining properties of qualia is their subjectivity, their “first-person accessibility,” presumably nobody knows – or could know- the answer to this question better than you. So what is your answer? Were the qualia changing or not? (ibid p. 83)
Dennett’s question is an important one to ask. It is sometimes assumed that we have direct access to our subjective experiences, that we know them in an intimate way that isn’t accessible to third person science. In the philosophical literature, these direct experiences are known as; qualia. But as Dennett notes the phenomenon of change blindness casts serious doubts on whether we understand qualia as directly as we think we do.
There are three possible answers to Dennett’s above question. (1) Yes my colour qualia were changing even though I wasn’t aware of that fact. (2) No my colour qualia wasn’t changing despite the changes occurring in my brain as I viewed the images (3) I don’t know (various reasons I don’t know could be): (A) I now am aware that I never really understood the concept of qualia all along, (B) I do understand the concept of qualia but I have no first person access to my qualia in this case, and third person science cannot get access to this qualia either.
Obviously any of the above answers will cause serious difficulties for the concept of qualia. The experiment shows that we don’t just have unproblematic access to our subjective states. None of the experimental data discussed by Dennett or Chater will be news to students of consciousness. Philosophers, psychologists, neuroscientists etc have debated the significance of these experimental data and what they demonstrate about the nature of consciousness for years. I don’t bring up this data because I think it is conclusive against introspection as a tool; in fact I don’t think the experiments do rule out careful introspection as a tool. I bring up this data because there is no evidence whatsoever that Witten claims about the nature of consciousness and its relation to physics deal with any of this experimental work in the science of the mind.
Consciousness seems to be unique to those who are interested in understanding how the mind works. No serious student of language argues that by virtue of being a language using creature we can casually reflect on the nature of language and make pronouncements on syntactic structure. Imagine the response if some physicist, who the popular press labelled the new Newton, pronounced that he believed by casual reflection on his own language using abilities that, it is more likely that we would understand the big bang than that we would ever reduce syntactic structure to basic physics.
I am certain that any responsible linguist, would adduce a series of facts that have been discovered re-the empirical research into syntax. She would then ask whether the physicist demonstrated any understanding of the relevant facts; whether the physicist’s intuitive understanding of syntax matched the precise models that have been created by scientific investigation. With this done it would then be a matter of demonstrating to the physicist that her conception of syntax wasn’t sufficiently precise enough to understand the nature of the subject matter, scientific research was needed, not casual reflection on one’s own experience as a native speaker of English . However, if the physicist was open minded and wanted to learn what science has told us about the nature of syntax, the physicist could become a productive collaborator on the project.
However, in the case of Witten and consciousness things are more difficult to ascertain, as we don’t know why he thinks that consciousness cannot be reduced to physics. All we have to go on are Witten’s bland assertions, and the howls of delight by mysterians to have an authority of a physicist on their team. Until Witten, or his supporters, manage to flesh out his views on consciousness, and whether they are sensitive to the relevant scientific facts it is very difficult to judge the truth value of his claims.
It is ironic that mysterians feel so desperately insecure about their philosophical stance that they feel the need to appeal to the authority of people like Witten, because the mysterian position is a perfectly sensible one. Rather than engaging in appeals to authority, I would suggest that a more sensible approach to adopt would be to simply outline your own reason for thinking that mysterianism is more likely to be true than its competitors.

Part 2: A Proposed Explication of why Witten is a Mysterian
There are various different difficulties with trying to explicate consciousness interms of underlying brain states. Over the last fifty or so years as we have learned more and more about behavioural distinctions between different types of consciousness, and have developed good psychological models to explain various different types of conscious and unconscious states it has become much more tractable to study the neural correlates of consciousness.
However, it has been noted by a number of theorists, that none of these studies demonstrate why the particular conscious states have the qualitative feel that they do. There are a number of different ways of explicating this intuition which are now pretty standard. It is worth though briefly mentioning through famous intuition pumps which are designed to make the difficulties of reducing consciousness experience to the physical more salient.
(1) What is it like to be a bat?
The philosopher Thomas Nagel’s famous 1974 paper ‘What is it like to be a bat?’, was a thought experiment which was designed to illustrate the illusive nature of subjective experience from a third person perspective. Nagel noted that a scientist who is studying the bat, has a variety of different avenues available to him in his study. The scientist can study the bat from the point of view of behavioural tests, naturalistic observations, neuroscientific studies etc. However, Nagel argued that the Bat has a subjective world of his own which will be missing from any third person study done by the scientist. The Bat as he flies about will have a world of experience which will have a specific quality that will forever elude any understanding from the point of view of third person science.
Nagel believes that by reflecting on the subjective world of the bat it should be obvious that it’s direct first person experience is not the kind of thing which we could understand by appeal to third person science. In his ‘Consciousness Explained’ Dennett noted that Nagel’s choice of a Bat as his example was inspired. A bat as a mammal is enough like us to make us confident that it is conscious creature. If Nagel had chosen a fly or an earthworm as his exemplar a lot of people would have been less confident that there was any reason to believe that the creature in question was conscious. While if Nagel had chose a fellow primate who is much closer to us biologically a lot of people wouldn’t have been as confident in Nagel’s intuition that we couldn’t understand its mental states using third person science.
Nagel’s thought experiment has influenced a lot of scientists and philosophers that our subjective states and their precise nature are something that we cannot understand using third person science. And his thought experiment does have some initial plausibility. It does seem plausible to assume that Bat’s have subjective states which can only be truly understood through direct experience, and that third person science doesn’t tell us about these states or about how they are realised.
This of course leads us to the question of whether when Witten is arguing that the mind will never be explained interms of physics; he is pumping a similar intuition to Thomas Nagel? Witten does seem to be alluding to the so called Hard-Problem of consciousness when he asserts his mysterian position. Like Nagel, Witten argues that, while science will eventually give us a good functional understanding of the workings of the neural correlates of consciousness; but the actual subjective states we have; and why they have the quality they have, is something that won’t be accounted for by third person science.
In terms of the conclusions they adopt Nagel and Witten seem to be reading from the same page. Whether they reached their conclusions via similar arguments is a different question. Nagel’s thought experiment involved reflections on the conscious experience of a Bat, and his argument is meant to go through for all conscious creatures. Witten never mentioned whether he was referring to the consciousness of all mammals, or just human consciousness, whether he thought all animals or even insects were conscious. Whether there were different kinds of consciousness, and if so whether all of these different types of conscious states were equally un-amenable to scientific reduction. When Witten argues that consciousness isn’t suitable for scientific reduction it isn’t clear what the extension of consciousness is for him, or whether he is appealing to the type of intuitions that moved Nagel.
(2) The Problem of Mary:
While the relation of the intuitions that Nagel was pumping to the intuitions that Witten is pumping are difficult to demarcate; another intuition pump may be closer to the concerns that move Witten. I am thinking of Frank Jackson’s famous thought experiment about Mary a colour blind neuroscientist who in the thought experiment knows all the scientific facts that there are to know about the physics and psychology of colour vision. One day Mary has her colour vision fixed and she begins to perceive colour for the first time. The question then asked is whether Mary will learn anything new about colour vision when she first sees the colour red.
The conclusion typically drawn is that despite knowing all of the physical facts about colour vision, Mary would indeed learn something new when she first saw red. She would learn what red is like from the first person point of view. The intuition being pumped here is that no amount of third person science would be sufficient to generate direct first person experience.
A typical objection to the Mary thought experiment is that we have absolutely no idea of what it would be like to know ALL the scientific facts about colour vision. Given the obvious fact that we truly do not have any conception of what it would mean to know all the facts about colour vision; any extrapolations we make about what Mary would learn are entirely baseless.
A defender of Witten could argue that given that he is one of the more knowledgeable physicists alive today; his intuitions about having all physical knowledge of X would be more robust than your average philosopher. So we should take his intuitions about the Mary case more seriously than the intuitions of the average philosopher. Such a critic would have a point. However, once we acknowledge two points Witten’s intuitions on Mary seem just as un reliable as anyone else’s intuitions. (1) Witten may know a lot about physics but this doesn’t really give him much insight into what it would be like to know everything about the physics (unless we add the implausible premise that Witten is close to a complete theory of everything), (2) While Witten may know a lot about physics, there is no evidence he knows anything whatsoever about perceptual psychology. So there is really no compelling reason to think that Witten is any more capable of understanding how someone with ALL knowledge of colour vision would react upon receiving their sight than anyone else.
Again though it is unclear given the lack of detail in Witten’s argument for mysterianism whether his intuitions on the topic are derived from the intuitions pumped by Nagel or Jackson. The only theorist Witten does mention in his interview was the mathematician and physicist Roger Penrose. Penrose in his 1989 ‘Shadows of the Mind’ famously argued that the mind cannot be a computational device. He noted that Godel’s Incompleteness Theorem proved the incompleteness of mathematics by showing that in any mathematical proof we will be intuitively able to recognise certain truths that cannot be formally proven within the system. If someone argued that the brain was a computational device then they wouldn’t be able to explain how we grasp certain truths which cannot be computationally explicated. Therefore, Penrose argued, since the computational theory of mind is our best working model of the brain, and it can be proven mathematically to be incomplete, we must draw the conclusion that we don’t have a physical explanation of our mental capacities.
Again with Penrose’s argument we have no idea whether it is guiding Witten’s intuitions. All Witten did say re- Penrose was that he didn’t think that Penrose’s supposed solution to the problem would work.
At this stage even though we cannot say for sure what intuitions Witten was relying on re the non-reducibility of consciousness to physics, we at have at least briefly discussed a few problems with reductionism which may have influenced Witten. We will now proceed to discuss how these difficulties may have led to Witten being a mysterian about consciousness.
(3) Mysterianism as a Philosophical position:
A philosopher who is arguing that there are problems with reducing psychological predicates to physical predicates doesn’t automatically have to become a mysterian. As we briefly discussed above Roger Penrose thinks that understanding consciousness in terms of basic physics is at present not tractable. But he argues that if we make suitable modifications to our understanding of basic physics then this may solve the problem. So Penrose, despite having difficulties with current attempts to reduce the mind to the brain, doesn’t adopt a materialist approach to the mind body problem.
Furthermore, other scientists who don’t think that the mind can be reduced to physics aren’t directly led to a mysterian solution. There are various different approaches adopted by scientists and philosophers on this topic. There is the neutral monism of Ernst Mach, John Dewey, Bertrand Russell, and William James. There is Eliminative Materialism held by the Churchland’s (who are eliminativists about propositional attitudes but not consciousness), Keith Frankish seems to be an eliminativist about consciousness and to a degree Dan Dennett is too (Dennett is not an eliminativist about propositional attitudes but seems to be about consciousness). Anomalous Monism argued for by Donald Davidson. Panpsychism has defenders in people like Galen Strawson, and David Chalmers amongst others. There many other competing views; epiphenomenalism, property dualism, substance dualism etc.
I do not here have space to evaluate these various competing positions in understanding the relation of the mind to the body. I just bring them up to demonstrate that many, philosophers and scientists see great difficulties in reducing the mind to the body; however they are not led by these difficulties to a mysterian position. So even if Witten did accept some of the difficulties put forth by Nagel, Jackson, and Penrose with reducing the mental to the physical; it should be noted that mysterianism doesn’t automatically follow from these difficulties.
Mysterianism typically starts out with the commonsense position that humans being animals will have limits to the things they can understand. Chomsky described the position as follows:
“As for the matter of cognitive reach, if humans are part of the natural world, not supernatural beings, then human intelligence has it scope and limits, determined by initial design. We can thus anticipate that certain questions will not fall within their cognitive reach, just as rats are unable to run mazes with numerical properties, lacking the appropriate concepts. Such questions, we might call ‘mysteries-for-humans’, just as some questions pose ‘mysteries-for-rats’. Among these mysteries may be questions we raise, and others we do not know how to formulate properly or at all.” (Chomsky ‘New Horizons in the study of Language and Mind’ p.107)

This position is perfectly sensible; as a human animal whose brain can only operate within certain fixed parameters it is perfectly sensible to suggest that there may be some problems which our brain is incapable of solving. Chomsky has suggested that ‘consciousness’ and ‘Freewill’ may fall with the realm of mysteries for those with human type intelligence.
Chomsky though is careful to note that he is only suggesting that ‘consciousness’ may be a permanent mystery to human intelligence. He isn’t claiming to have proof that consciousness falls within the mystery category. It is one thing to argue that humans being animals with a fixed cognitive structure may have cognitive limitations which makes some questions unanswerable for us. It is another thing to draw an actual line which demonstrates precisely where this line falls and to place certain subject matters on one side or the other of the problems/mysteries divide.
Colin McGinn argued that we have good reason to think that consciousness falls within the mystery category. As we saw above we have many different competing candidates for dealing with consciousness and at present we don’t have any way of deciding between these positions. McGinn thinks that the reason we cannot decide between these various positions, is because consciousness is beyond our capacity to understand. He argues that while we have made great progress in the last 2000 years in understanding cognitive structures in the brain, neural correlates of consciousness, etc we have made very little progress in understanding the hard problem of consciousness. He thinks that the best explanation of this difficulty in solving the hard problem of consciousness is that it is a mystery for creatures of our cognitive structures.
Personally I don’t agree with McGinn’s position that the hard problem of consciousness is a mystery. I don’t think that the disputes between the various different competing camps is as tractable as Mcginn thinks. Nonetheless I think that Chomsky is certainly correct in raising the possibility that consciousness may be a permanent mystery to humans.
As I said at the outset it is difficult to know why Witten is pessimistic about the possibility of reducing psychology to physics as he doesn’t present an argument. I have thus far sketched a possible justification for Witten. (1) People like Nagel, Jackson, and Penrose have presented good arguments that indicate that it may not be possible to explain subjective states in terms of third person science. (2) Chomsky has shown that humans being animals may have cognitive limitations which may mean that certain subjects will remain permanent mysteries for them, (3) Given the seemingly intractable situation with competing theories of how to reduce the mind to the body it is plausible that consciousness falls within the realm of mysteries.
Whether Witten is actually arguing in this manner is difficult to know as he is so sketchy of his reasoning on the topic. Anyway my reconstruction of Witten’s possible reasoning processes is the only sensible justification I can give of his vague comments.

“This principle, as the reader will remember, states that, if a certain event calls out a certain response, and if another event is experienced just before it, or at the same moment, in time that other event will tend to call out the response which, originally, only the first event would call out. This applies to both muscles and to glands; it is because it applies to glands that words are capable of causing emotions. Moreover, we cannot set limits to the length of the chain of associations that may be established. If you hold an infant’s limbs, you call out a rage reaction; this appears to be an ‘unlearned reaction’. If, you and no one else, repeatedly hold an infant’s limbs, the mere sight of you will call out a rage reaction after a time. When the infant learns to talk your name may have the same effect. If, later, he learns you are an optician, he may come to hate all opticians; this may lead him to hate Spinoza because he made spectacles, and thence he may come to hate metaphysicians and Jews. For doing so he will no doubt have the most admirable reasons, which will seem to him to be his real ones; he will never suspect the process of conditioning by which he arrived at his enthusiasm for the Ku Klux Klan.” (Bertrand Russell ‘An Outline of Philosophy’ pp. 88-89)
In the above incredible passage Russell presents a toy example of how through a process of classical conditioning a child as he grows up into an adult could end up developing an irrational hatred of a particular group. On this picture a child would begin with an instinctive anger about being having his limbs held. Only one person is guilty of holding the child’s limbs, and that person is a visiting scientist. Through classical conditioning the child eventually comes to associate the scientist with being held. The mere sight of the scientist is enough to get the child to feel rage. In this respect the child is like Pavlov’s dog who salivates every time he hears a bell ring. Russell’s extrapolations are thus far consistent with known scientific data . Russell’s example is similar to the studies done by both Pavlov and Watson in classical conditioning. In Pavlov’s famous experiment the dog has an unconditioned reaction to food being presented (he salivates), and through conditioning; a bell being rung just before the food is presented, the dog salivates at the sight of the bell. The dog salivating to the bell is called a conditioned response. Russell’s toy example is similar in that the rage is an unconditioned response to being held and the conditioned response is the rage elicited at the sight of the scientist. Russell’s toy example stands up to critical scrutiny so far. Though to be more complete in his analysis, he should have discussed the fact that such conditioning processes wouldn’t necessarily continue on indefinitely, once the scientist had discontinued with the practice of holding the child’s limbs. Russell’s analysis implies that this classical conditioning will necessarily continue on to adulthood resulting in poorly understood rage. However in Pavlov’s original experiments it was noted that after a period of time where the unconditioned stimulus (the food) is presented without the conditioned stimulus (the bell), the dog ceases to salivate at the sight of the bell. Russell though is just uncritically assuming that the conditioned stimulus will continue throughout the child’s life.
The evidence for extinction occurring in classical conditioning is overwhelming. There have hundreds of studies done on extinction in behavioural science in the years since Russell wrote his ‘An Outline of Philosophy’. Russell cannot be held to account for not predicting the results of experiments which were not yet performed. Nonetheless, he is guilty of uncritically assuming that the effects of the conditioning will continue throughout the child’s life. This assumption was unjustified; even Pavlov writing years before Russell had noted the effects of extinction on conditioning.
A possible reason for Russell assuming that the effects of the conditioning would remain throughout the child’s life, were the existence of phobias. Some people spend their entire lives afraid of certain creatures because of experiences in their childhood. Russell greatly admired the work of J.B. Watson and discussed his work in ‘An Outline of Philosophy’ in great detail. Watson’s experiment with Little Albert was viewed at the time as a way of inducing a phobia in the child using classical conditioning. The child Little Albert is introduced to a white rat which he plays with for a while. Watson then makes a loud noise every time the child went to touch the rat. After a while the child is presented with the white rat and immediately becomes upset at the sight of the white rat despite no loud noise being made. It is then shown that the child generalises his fear of the rat to other furry creatures such as dogs, teddy bears etc. Russell may have extrapolated from the demonstrative fact that some people have irrational phobias in adulthood, and Watson inducing a phobia in a child through classical conditioning to the conclusion that adult phobias are the result of unintended classical conditioning of children in childhood. Such a conclusion obviously doesn’t follow from facts Watson presented. There wasn’t any evidence that Watson’s classical conditioning would last into adulthood.
If Russell had asserted that it was possible that adult phobias were the result of classical conditioning then his assertion would have been a reasonable hypothesis worth trying to test. However, Russell instead just assumed that classical conditioning in childhood would last into adulthood. It is possible that Psychoanalysis, which when Russell was writing his ‘Outline of Philosophy’ was world famous, implicitly influenced Russell’s thinking. Freud claimed to have demonstrated that childhood trauma can unconsciously influence adult behaviour and thought processes. Freud had justified his claim with his clinical work with patients who he claimed he cured by bringing their unconscious traumas to consciousness. Freud even argued that a lot of our explanations of our behaviour could be shown to be merely rationalizations that do not correspond with the real causes of our behaviour. As we saw above Russell made a similar point to Freud about the adult giving a rationalization to explain his racist views.
Given the esteem that Freud’s views on the importance of childhood experiences in determining adult behaviour, was held in when Russell wrote his ‘An Outline of Philosophy’, it is highly probable that Russell was influenced by Freud’s views. This would make sense of Russell’s uncritical assumption that classical conditioning of rage to a particular person would last into childhood. Russell may have reasoned as follows: (1) We know from Freud’s clinical work that trauma from childhood can last into adulthood, (2) Watson induced a phobia into Little Albert using classical conditioning, (3) Given Freud’s discoveries it is reasonable to assume that the classical conditioning used by Watson could last into adulthood, (4) Watson’s experiments are an excellent tool to explain adults with irrational phobias, this is indirect evidence that the conditioning lasts into adulthood. Obviously, if this was Russell’s reasoning process then it was non-demonstrative. Nonetheless it could be considered a reasonable inference to the best explanation. Russell’s imagined reasoning process would have been reasonable at the time. But since then the amply demonstrated notion of extinction in classical conditioning and the many methodological flaws with Freud’s psychoanalysis casts serious doubt as to whether the classical conditioning Russell is appealing to would actually continue to affect the adult’s behaviour. On the other hand, the existence of spontaneous recovery of conditioned responses after they have undergone extinction could be used as a point in Russell’s favour.
For the sake of argument we will assume that Russell’s view that the classical conditioning will continue to affect the child into adulthood is correct. And move on to the next step in the above passage by Russell. He goes on to note that this form of conditioning can be extended indefinitely. Thus when the child learns to associate a name (e.g. ‘J.B Watson’) with the person who held him, the name will then elicit the same feeling of rage. Russell’s extrapolation that the name will have the same effect as the person does is again probably based on Watson’s experiment with Little Albert.
In Watson’s experiment he purported to have shown that Little Albert generalized his fear of rats to other furry animals. Russell seemed to be thinking about Albert’s purported generalization to other furry animals as a reason to assume that the child in his above example will follow a similar process. Thus the child will unconsciously move from feeling rage at the adult being present, to feeling rage when he speaks or hears the adult’s name. Russell speculates that if the child learns that the scientist who abused him was also an optician; he could hate all opticians. This could lead to him hating Spinoza who was an optician and because Spinoza was a metaphysician and a Jew, the child could generalize to a hatred of all Jews and Metaphysicians.
The process that Russell is invoking is commonly referred to as ‘stimulus generalization’ and has been well studied and shown to have effects in both classical conditioning and operant conditioning. While people have made some legitimate methodological criticisms (and ethical criticisms) of Watson’s Little Albert Experiment, there have been hundreds of better constructed experiments verifying the effect in the last ninety years. So to a degree one could argue that Russell’s toy example invoking generalization is corroborated to some degree by experimental data.
Russell’s image is of a child as a result of classical conditioning and stimulus generalization developing strange aversions which as an adult are rationalized; while the real cause is something much different. One of the problems with this emphasis is that one is at a loss to explicate precisely what features will be generalized, and why. In Watson’s Little Albert experiment the child generalized to ‘small furry things’, while in Russell’s toy example the child over a period of years generalized from the person, to the name, to a description of the person An Optician, to all Opticians, to contingent features of particular opticians.
There seems to be no reason why a person would generalize in a particular way as opposed to another way. On Russell’s view the child is implicitly using the propositional function ‘x makes me angry’ and ‘x is the scientist’. However the when the child learns that the scientist is an optician he makes the further illicit inference: x makes me angry, x is an optician therefore opticians make me angry. But there are countless different facts about the scientist that are discoverable by child. The experimentalist is a scientist, he is an optician, he is a man, he could be short, fat, thin, hunchbacked or could have any other number of features. There seems to be no reason that the child would unconsciously generalize in one way or another on Russell’s account. So while Russell’s hypothesis has some good scientific data underlying it, he hasn’t presented a model that can predict in detail they way a person will generalize unconsciously. Nonetheless he has presented a useful explanatory model that can be used to explain people holding odd beliefs that they give unconvincing reasons for why they hold them.
Interestingly the principle of generalization was proposed by psychoanalyst Matte Blanco as one of the key features of unconscious thinking. Blanco used Russell and Whitehead’s Principia-Mathematica to formalise Freud’s theory of the unconscious. One difficulty with Blanco’s work is that he didn’t have a compelling causal account of the principle of generalization and his approach at times had a Cartesian disembodied feel. Watson’s work on classical conditioning and stimulus generalization would have been a useful tool to help give a naturalistic account of the purported structure of the unconscious mind.
It is interesting to note that Russell read a paper by Blanco where the principle of generalization is discussed in detail. Russell sent Blanco a note to say that he admired Blanco’s work. It is tantalizing to speculate whether Russell put together his own use of stimulus generalization expressed in his ‘An Outline of Philosophy’ with Blanco’s similar principle. Unfortunately there isn’t a record of Russell’s thoughts on Blanco’s ‘The Unconscious as Infinite Sets’.
As noted above there are plenty of criticisms of the methodology used by psychoanalysts. So people may not be overly impressed that some aspects of psychoanalysis seem to be in agreement with Russell’s hypothesis. It is worth noting though that Relation Frame Theory and its clinical off shoot Acceptance and Commitment Therapy also makes use of implicit generalizations to explain the behaviour of people with various kinds of behavioural problems.
Neither Blanco’s work or work in Relation Frame Theory are yet predictive of what type of generalizations a person will generate in a given instance. They are rather retrospective theories that can be used to explain the verbal behaviour of the patients that they deal with. Nonetheless I think it is fair to say that at present Russell’s speculation while not proven correct; is given what contemporary data tell us, a reasonable hypothesis.