I''ll begin this essay with a story. In 2011, having completed a Personal Growth course through Youthline, the free New Zealand youth phone counselling service, I graduated to a course intended to teach us wannabe counsellors phone counselling skills. At the first session a number of cards were given out to the group members, all of us sitting around on beanbags, each with an abstract multicoloured pattern or image on it, and we were each asked to choose the one we most identified with. I was drawn to two cards: the first was angular, sort of mathematical, and the second more free-wheeling, free-spirited. As the reader may guess, this game was New Agey pseudo-pyschological bullshit. I knew that the first card signified a personality with a kind of rational analytical way of looking at the world while the second indicated a personality more intuitive and emotional; I simply sensed this. Of course the students were not supposed to consciously realise which personality type each various card represented. At the time, I was torn between these two different ways of viewing myself but I was strongly drawn towards the first card. At the last moment, almost as though some force was propelling me, I selected the second card. It might be that I intuited that this was the card the others in the group would want me to select; perhaps the spell I was under somehow led me to the second card After I chose it, the two group facilitators beamed approvingly at me and told me that this was the card that they themselves would pick. It was almost as though some kind of supernatural force had influenced my choice. I will return to the significance of this moment later in the essay.
The purpose of this essay is, first, to critique the recent work of philosophy and science, Determined: Life Without Free Will by Robert Sapolsky, secondly to discuss quantum physics again, and thirdly to approach the idea of free will from a different perspective than that taken by Sapolsky. My aim in discussing quantum physics is to correct and clarify some ideas I presented in a recent post "Quantum Physics for Dummies and a New Idea". I will try to state my view of probability more clearly. Although it seems that in this essay I am attempting to cover quite different topics, there is significant overlap because some of Sapolsky's key arguments concern quantum physics.
A general comment first. I saw part of an interview with Robert Sapolsky by Lawrence Krauss and Sapolsky himself comes across as very nice man, although he has an evident tendency towards depression and a strong leaning towards biological fatalism. Sapolsky's intention in Determined is to disprove the existence of free will. Before I discuss Sapolksy's argument, I should remind reader that there is in fact a fairly simple well established argument against free will that runs as follows: A particular action by a person is the result of a deterministic chain of effects and causes that extends back to a time and place outside the person's 'control' or is initiated by a chance event or both. In no case can the person be said to be exercising free will because it seems intuitively obvious that a person cannot be held responsible for events that occurred in her environment, before she was born, or that occurred at random. If the reader wants to learn more about this argument, I recommend an old Youtube video "Compatibilism Debunked" by Alex O'Conner or his quite recent debate with Ben Shapiro. Alex's argument is informed by an essay, "The Powers of Rational Beings" by Peter van Inwagen but I suspect that those of us who have always been skeptical of the idea of free will have often arrived at this argument independently although few of us have articulated it in a philosophical essay or Youtube clip.
Although the argument above seems to me to inform Sapolsky's approach, his stated method is somewhat different. Sapolsky is attempting to show that science has disproved the idea of free will. Sapolsky is a professor of neuroscience but he is not attempting to show that neuroscience alone has disproved free will; rather he is saying that all the scientific disciplines taken together collectively show that free will does not exist. "Crucially, all these disciplines collectively negate free will because they are all interlinked, constituting the same ultimate body of knowledge." (p.9) Thus genetics, neurobiology, endocrinology, psychological theories concerning child development and later behaviours, sociology, and cultural anthropology, together with all the sciences I haven't listed, collectively leave no room for free will. "There's not a single crack of daylight to shoehorn in free will."(p.9) Implicit in his argument is the idea that this body of knowledge contains no holes, no puzzles, something that Rupert Sheldrake has, I believe, called "scientific triumphalism". Sapolsky sometimes, particularly in the section concerning genetics, seems to be trying to overwhelm the reader with jargon as if to say, "Don't try to say the science is wrong or incomplete because that will just signal to the world you are stupid; we scientists understand everything."
Although I have argued consistently in this blog that free will does not exist, I am a contrarian and so intend in this essay to show that Sapolsky's arguments are flawed. I shall argue that the science is indeed incomplete. There is a particular aspect of Sapolsky's argument in particular that I think can be criticised: although the explicit aim of his book is to disprove the idea of free will, implicitly he is also seeking to show that any religious or mystical view is also incorrect, that the supernatural does not exist. In the world we live in, religious faith and a belief in free will go hand in hand; at the same time, it is common for atheists to deny that free will exists. However it is possible to believe in the supernatural and deny that free will exists (just as it is possible to be a materialist atheist and embrace the notion of free will as Daniel Dennett for instance does). It is possible to believe this without subscribing to Calvinism. One aim of this essay is to show that it may be possible to "shoehorn" in the supernatural. It may even be possible to somehow salvage the notion of "free will", something I will attempt at the end of the essay.
Even though I intend to criticise Sapolsky, I still think people should read his book. He is engaging and often funny and (usually) has his heart in the right place. The absolute best thing about the book is that he kills Evolutionary Psychology, the idea that our behaviour is genetically determined. (I don't intend to discuss this part of the book in this essay but refer the reader to Determined itself.) However I believe he makes mistakes. If you do read it, just remember that he is engaging in motivated reasoning and may be cherry picking studies that support his position.
To begin, I want to say something about the term 'consciousness'. Sapolsky, at the beginning of a section concerning whether or not free will needs to be conscious, says "I don't understand what consciousness is, can't define it" (p. 31). Sapolsky's bete noire, Daniel Dennett, has argued that free will exists but consciousness doesn't. People who either can't define the term 'consciousness' or argue that it doesn't exist at all have not really considered the meaning of the word. Presumably you, dear reader, are reading this on your computer or smart device – you are conscious of this essay, conscious of your computer or smart device, you are conscious of the furniture in the room. You are not conscious of the people in the next apartment over unless they start banging on the wall. More esoterically, you are presently conscious of some part of your mind and unconscious of other parts – you are not currently bringing into conscious awareness memories from your childhood for instance. Such memories can be described as subconscious because sometimes you bring them into the conscious part of your mind but at other times they sink back into the unconscious part. There may be constituents of your mind that you never bring into conscious awareness at all, such as the part of your mind or brain that regulates digestion – such constituents can proper be described as being part of the unconscious. The term 'consciousness' simply describes the activity of being conscious of things in either the outer or inner world. This seems self-evident to me. Those who struggle to understand it or deny that it exists must presumably be working with some loose definition of 'consciousness' that is wrong or which I cannot understand.
Sapolsky's discussion of consciousness comes in the context of an early chapter about experiments first carried out by Benjamin Libet in 1983. Subjects were asked to press a button when they felt like it and then report on when they decided to press it. During the experiment, subjects were hooked up to an EEG machine. What the experimenters found is that the brains of subjects decided to press the button about three hundred milliseconds before the subjects were consciously aware of having made the decision. This finding is of course supposed to be profound. However I would like to point out that the notion that our decisions are at least influenced and perhaps fully determined by unconscious mechanisms goes back at least as far as Sigmund Freud. Yes, some of Freud's ideas, such as the Oedipal complex, seem ridiculous now – but he should be given credit for pretty much inventing the concepts of the conscious and unconscious mind. Freud was of course completely ignorant of neuroscience but the neuroscientists who thought they'd made a major breakthrough in 1983 should at least have been a little aware of Freud. Moreover we cannot be certain that this highly artificial experiment proves that all our decisions are made unconsciously first. (In mentioning all this, I am not critiquing Sapolsky because Sapolsky himself says, when discussing scientific and philosophical debates arising from these experiments, "these Libetian debates [...] are irrelevant to thinking about free will." Sapolsky's purpose is to show that the chain of cause and effect that leads to an action, regardless of when that chain becomes conscious, never includes a juncture which we can describe as 'freely willed'.)
My first criticism of Determined concerns Sapolsky's uncritical acceptance of psychiatric labelling. Sapolsky argues very consistently through the second half of the book that psychiatric labels let people off the hook with respect to responsibility. We now, he argues, recognise that 'schizophrenia' is a 'brain disease' and so when, as very occasionally happens, someone diagnosed schizophrenic does something antisocial, we can say that the person is 'not guilty by reason of insanity'. Similarly if we recognise that Autism Spectrum Disorder is simply indicative of neurodiversity in the population, that bipolar disorder, ADHD, dyslexia, and other psychiatric conditions are neurological diseases and so nothing the people afflicted with them can do anything about, we can forgive people their eccentricities. Furthermore people with such conditions can forgive themselves, tell themselves "it's not my fault, it's my brain's fault." Sapolsky goes so far as to suggest that many people who are obese are not responsible for their obesity because they carry genes that affect their metabolism or how food tastes to them. The problem here is that Sapolsky simply assumes, without ever expressing it directly or explicitly defending this view, that such conditions are congenital, incurable, and lifelong. For instance, he cites research that has supposedly found that people destined for a diagnosis of 'schizophrenia' were often bedwetters as children. (Someone should tell Sarah Silverman this.) This could lead us to look for clusters of symptoms in children and diagnose them as 'pre-schizophrenic' – something that I think could constitute a self-fulfilling prophecy. Sapolsky's attitude to labels evinces his biological fatalism.
There are two ways to argue against Sapolsky here, the first to invoke a kind of Rupert-Sheldrake-inspired-magic and the second more materialist, more neurosciency. I'll describe the first way briefly. I suspect that labelling a person with a psychiatric disorder actually somehow supernaturally affects the person, is effectively equivalent to laying a curse on that person, and that if we really want to help the 'mentally ill', we need to totally change society's and psychiatry's whole way of viewing and understanding 'mental illness'. We need to recognise that it can be temporary, that it can arise in a person's life as a result of proximal circumstances and later pass away. A major reason why people usually don't recover from 'psychiatric disorders' is simply that everyone, including the afflicted sufferers, assume that they cannot.
I may have readers who find this idea a bit woo-woo and so I will sketch out another more rational seeming rejoinder. One idea in neuroscience that Sapolsky touches on but does not explore thoroughly is neuroplasticity. Consider first that if a person regularly works out at a gym and has the right type of endocrine system, he or she will gain muscle mass. The bones in the racket-holding arm of a tennis player will actually become stronger through continued use. Similarly, although a few decades ago people thought brain damage was irreparable, we now know the human brain can change and even fix itself. A person can have a stroke that impairs her ability to use language and through workaround solutions the brain of the person can reacquire this linguistic faculty. A famous example of a New Zealander who had a stroke, lost his ability to verbally communicate, but seems to have fully recovered is poet and academic CK Stead. Presuming you remember some ideas in this essay and that materialism is true, I have literally caused a physical change in your brain. I'll give one more fascinating example of neuroplasticity. Before GPS, London cabbies had to memorise London's whole topography – this was called "The Knowledge." It was found that a part of these cabbies' brains had become larger than those who had a more quotidian feel for London's layout.
Towards the end of Determined, Sapolsky briefly discusses homosexuality and because this used to be an obsession of mine and because it is related to the present discussion, it may be useful to say something about it. In 1991, Simon Le Vay discovered that a region in the brains of gay men was smaller than the corresponding region in the brains of straight men, in fact comparable in size to the corresponding region in straight women. Sapolsky does not say this but I have learnt from other sources that this region is in the visual cortex and I would wager is involved when heterosexual men surveil the world for attractive females and/or is involved in sexual fantasy. Sapolsky, in a gratuitous display of virtue signalling, quotes Lady Gaga: "God makes no mistakes, I'm on the right track, baby, I was born this way." However, when we consider neuroplasticity, an alternative explanation to the idea that gay men are born gay can be considered. It might be that Le Vay and other gay men make a 'decision' to be gay in early adolescence and, as a result of being gay for many years, this part of their brains atrophies or never grows at all.
If this idea of neuroplasticity is at all true, it suggests that people diagnosed with psychiatric disorders can recover. Certainly I have recovered from the psychoses I experienced years ago and hopefully this essay is halfway coherent and so demonstrates this. Readers may say that this apparent cogency is evidence that the drugs I am legally compelled to take are working and so I need to say something briefly about the dosage I am currently on. At the beginning of last year I was taking 300mgs of Olanzapine a fortnight; I am now only taking 210mgs a month. It has gradually been reduced over the last two years and I have had two injections at the new dosage level. I have not experienced any recurrence of psychotic symptoms, feel more clear headed, can concentrate better, am reading more, and am generally feeling happier. My aim when I see my psychiatrist next is to persuade him to reduce my dosage to 150mgs with the goal of getting off medication entirely within six months. This though is still a very difficult ask. The point here is that if a person diagnosed (or perhaps misdiagnosed) with schizophrenia can fully recover and go off drugs, the same may be true for all other psychiatric disorders. (Of course, schizophrenia is presumed to be a neurotransmitter imbalance rather than a result of brain damage, but the argument I have made can be generalised to cover this.) It may seem that Sapolsky's prescription is more compassionate, more humane, than mine because it seems that I am suggesting that people with psychiatric disorders are in a sense responsible for them – but this is not the case. I think that people diagnosed with psychiatric disorders can take steps towards recovery but that they require help. Friends, family, and society generally are collectively responsible for helping people recover from 'mental illness'.
I turn now to my second criticism of Determined. Sapolsky cites research claiming that rain forest dwellers tend to invent polytheistic religions while desert dwellers tend to produce monotheistic religions, research that suggests that there is a causal link between type of religion and climate. I need to give a little admittedly amateurish history lesson to show the mistake here. From the little I know, the first monotheistic people were the Jews. But the people from whom the Jews descended were polytheists. Incredible fact alert: Yahweh was originally a thunder god, part of a pantheon, and had a wife. The ancient people who became the Jews came to practice something known as monolatry which means they tended to worship one god while acknowledging the existence of other gods; over time the cult of Yahweh took over and the rest of the pantheon, including the wife, slipped away into oblivion. We arrived at the Judaism gradually built up over centuries and documented in the Old Testament. What about the other monotheistic religions? Christianity, of course, sprang out of Judaism, and Islam, through some process about which I am not entirely clear because I haven't looked closely at the history of Islam, sprang out of the older two Abrahamic religions in the early seventh century. There are a few other monotheistic religions such as the Druze faith, which developed out of Islam in the eleventh century in modern day Lebanon, and the Baha'i faith which developed in the nineteenth century in Iran and took inspiration from all the other major world religions. Although I have not looked at all the monotheistic religions (there are about ten) it seems likely that the fact that monotheistic religions have all, I think, arisen in the Middle East is a historical coincidence rather than evidence that arid climates somehow deterministically cultivate a belief that there is only one God. Sapolsky is making the classic error undergraduates are warned against: he is mistaking correlation for causation. In order to know if desert climates somehow naturally engender monotheistic religions we would need to run an experiment in which we play out world history from somewhat different initial conditions on a hundred Earths and see if a correlation between deserts and monotheism arises that is statistically significant. Otherwise we are entitled to doubt Sapolsky's claim.
Sapolsky's assertion that there is a causal link between desert cultures and monotheism is not a key part of his overarching argument but the reason I have focussed on it is that it may by indicative of other errors in the book, other moments where he has presented speculation as fact, that I and others may not pick on.
Even if the monotheism claim is fairly trivial, there is another strand in his argument, a strand that resurfaces repeatedly throughout Sapolsky's book, that really does deserve to be challenged. This strand concerns experiments involving 'priming'. To define this term, I shall simply quote Wikipedia: "Priming is the idea that exposure to one stimulus may influence a response to a subsequent stimulus, without conscious guidance or intention." For instance, early on Sapolsky says, "Ask someone to name their favourite detergent, and if you have unconsciously cued them earlier with the word ocean, they become more likely to answer, 'Tide'" (p. 16). Sapolsky mentions a study that shows "that disgusting smells make subjects less accepting of gay marriage" (p. 47) If a religious person has to unscramble the letters "lehl" or "neehav" they will, apparently, become more prosocial. (p.258) Although these examples fall purely within psychology, sometimes the examples Sapolsky gives also involve endocrinology or EEG scans, neuroscience. Sapolsky's intention is clear: we tend to associate free will with "conscious guidance or intention" and if the behaviours of a group of people can by influenced (on average slightly more than would occur by chance alone) by prior stimuli, this weakens the belief that our decisions are rational, are the result of "free will".
The problem here is with the whole concept of 'priming'. In Freud's day, psychology was very different than it is today. Freud and his followers were therapists who would listen to their patients, attempt to interpret their dreams, and sought to devise elaborate and sometimes silly theories concerning individual pathologies and general features of the human conscious and unconscious mind. Today, though, psychology is trying very hard to be a proper science – and that involves experiments. 'Priming', a term coined by Segal and Cofer in 1960, is a useful concept for psychologists today because it enables researchers to take a number of generic individuals, expose them to a stimulus and measure the results, all in the lab. These days academic psychologists all over the world are performing 'priming' experiments hoping to get novel results, get published in reputable journals, get lots of citations, and bring in all the consequent funding. Of course, this creates perverse incentives. In the post "Threading the Needle", I briefly discussed 'the replication crisis' (sometimes referred to as the reproducibility crisis). Good science does not just involve publishing surprising findings, it also involves repeating experiments carried out by others to ensure that the original studies were sound, weren't artefacts or accidents. Unfortunately repeating others' experiments is not particularly sexy and so it was not until the last five or ten years that it was realised that a great deal of science, including apparent findings that had become canonical, were not replicable. This is especially true of psychology. In fact, in 2022 a giant within the field of psychology, Daniel Kahneman, said in a lecture that behavioural priming research was "effectively dead." In Determined, Sapolsky briefly mentions the replication crisis and says that he is only considering "findings whose brand conclusions have been independently replicated." (p. 47n.) I would like to suggest that the crisis is so bad that we should regard all research into priming as suspect.
The concept of priming seems to me reminiscent of an earlier fad within psychology, behaviourism, an approach which has not entirely gone away but which is much less fashionable today than it once was. Readers may have heard of Pavlov's dog who was conditioned to salivate at the sound of a bell. I believe animals like dogs and cats and octopuses have inner lives that we know little or nothing about because they can't tell us about them – I have often thought that when Pavlov's dog heard the bell, she probably thought to herself, "Goody, that means someone is going to feed me soon" and salivated accordingly. This way of thinking about living things, that they have minds as well as bodies, is precisely what behaviourists rejected because the inner life is unobservable from the outside. This brings me to another idea concerning priming. In a recent Youtube video, Rupert Sheldrake discusses the reproducibility crisis and then segues into a very intriguing idea, that scientists can psychically affect the results of their experiments. He argues that chemists, for example, may be able to telekinetically affect the rate at which chemical reactions occur. I suspect this may go a little too far but, if we grant for a moment that paranormal phenomena are real, then it raises the possibility that when priming experiments actually do produce statistically significant results, it may be because the experimenters and experiment itself psychically affect the behaviour of the subjects. What others think of a person may affect the person mentally and perhaps even behaviourally. This is the reason for the story I told at the beginning of this essay. I note by the way that Sheldrake himself believes in free will but I am not myself here defending free will but rather proposing a kind of supernatural causation.
I'll consider one last study that Salpolsky alludes to. Four social psychologists (Cohen, Nisbett, Bowdle, and Schwarz) carried out three experiments in 1996 which explicitly sought to prove that a "culture of honour" exists among White males from the Southern US that can affect their "cognitions, emotions, behaviours, and physiological reactions" in certain circumstances. Just before the experiment was supposed to take place, each subject was bumped into by a confederate in the corridor outside the lab and called an "asshole". Subsequently it was found that subjects from the South were "more likely to think their masculine reputation was threatened", had higher cortisol and testosterone levels, and were more likely to engage in aggressive and dominant behaviour. Subjects from the North were pretty much unaffected. The problem with these experiments is that although the researchers assert baldly that these reactions show the effects of this "culture of honour", there are a number of other possible explanations. Perhaps people in the South are just nicer than people in the North and so Northerners, having been exposed to such conduct growing up, found it easier to shrug off. Perhaps the Southerners were already finding it difficult living in Michigan (where the experiments took place) and so were more predisposed to take offence. Perhaps they sensed that they had been chosen to participate in the experiments simply because they were from the South and were consequently more vigilant, more paranoid. The issue here is what Philosophers of Science call "the underdetermination of theory by data" – we can't be sure that these reactions show the pervasive influence of a Southern "culture of honour". Sapolksy employs this example to argue that culture affects behaviour, is transmitted intergenerationally, and can persist over centuries. However it seems here that Sapolsky is again presenting speculation as fact.
The aim of Determined is to convince readers that free will does not exist and so Sapolsky must consider counterarguments, something he does in the middle of the book. He groups these counter arguments under three general headings: chaos theory, complexity, and quantum physics. As promised at the beginning of the essay, it is the third that I shall now turn to.
In the first of the two chapters on quantum physics, Sapolsky sketches out a rough picture of what quantum physics is and of its implications. He isn't a physicist and admits this – but we don't need to be quantum physicists to see its relevance to a discussion of determinism. Put very simply, quantum physics postulates that at a subatomic level events occur randomly. Quantum physics necessarily involves indeterminacy. In order to discuss "randomness", we need some sense of the meaning of the term, and so Sapolsky illustrates the idea by talking about it would mean for a particle to move randomly. This passage is too long to quote here but involves the idea that a randomly moving particle will always most likely be found at the point where it was last observed and has a kind of fractal trajectory in that the more we zoom in the path it has taken the more zigzags we find.
My criticism here is that Sapolsky is presenting one conception of randomness as though it is definitive. It is not the only conception. I myself tend to think of randomness as involving an event with a number of possible outcomes – for instance, when a die is rolled, we take the view that there are six possible results, each of which is equally probable. If a very large number of factors are involved, if there are many events all with the same probability distribution and all independently contributing to a value, we get a normal distribution – this is why people's heights fall on a bell curve. (The wave function associated with a simple quantum harmonic oscillator in its ground state, incidentally, is also described by a normal distribution.) Sapolsky's view of randomness is of absolute randomness. This is like saying that we can roll a die and get a banana. My view is that there can be weighted or constrained randomness. This will become relevant later.
The issue for those us who are fascinated with quantum physics is that it lends itself to what Sapolsky calls "mystical gibberish" (p. 215). Ordinary people usually don't understand much quantum physics but have picked up on the idea that it is seriously weird – this leads people to slap the word quantum on everything as though religious or New Age beliefs have scientific backing. Sapolsky heaps scorn on Deepak Chopra for instance for his 1989 book Quantum Healing and adduces a journal on "quantum psychotherapy" which contains an article called "Quantum Logic of the Unconscious and Schizophrenia" as more evidence of mystical gibberish. What Sapolsky misses is that these supernatural associations must have come from somewhere and presumably originated with people who actually did understand quantum physics. Two defenders of the free will hypothesis that Sapolsky discusses, two who seek to find it in the quantum realm, are Stuart Hameroff and Roger Penrose – Penrose won the Nobel Prize in Physics in 2020 and so should not be dismissed as a quack. It is such quantum defenders of free will that Sapolsky seeks to criticise in the second chapter on quantum physics.
Years ago, when wondering about quantum mechanics, I thought that perhaps consciousness caused wave function collapse. Hameroff and Penrose argue, surprisingly, that wave function collapses cause consciousness. The idea is that subatomic apparently random events occurring in microtubules (organelles found in eukaryote cells like neurones) collectively generate free will. In arguing against this proposal, Sapolsky enlists the support of a number of neuroscientists and philosophers who also deem it stupid. Sapolsky argues that a brain is so much vastly larger than an electron wandering about randomly in a single organelle in a single brain cell that there is no way a single quantum event involving the electron could result in a human action. He says, "This is the bubbling-up problem in going from quantum indeterminacy at the subatomic level up to brains producing behaviour – you'd need to have a staggeringly large number of such random events occurring at the same time, place, and direction. Instead, most experts conclude that the more likely scenario is that any quantum event gets lost in the noise of a staggering number of other quantum events occurring at different times and directions." (p. 220). Sapolsky quotes a philosopher called Hobbs: "The law of large numbers, combined with the sheer number of quantum events occurring in any macro-level object, assure us that the effects of random quantum-level fluctuations are entirely predictable at the macro level, much the way that the profits of casinos are predictable, even though based on millions of 'purely chance' events." (p. 221). To pummel his point home, Sapolsky gets into the nitty gritty of neurobiology, pointing out that for a single neurone to activate its neighbour there are enormous number of neurotransmitter molecules and other processes involved, and, even if it does, it's still only affecting one neurone out of 100 billion.
Here is the question though. Is the human brain like a casino or more like the weather? Or, to put it another way, does the law of large numbers apply to the human brain? Famously, if a butterfly flaps its wings in the Amazon it can lead to a tornado in Kansas. This is the lesson of chaos theory. I am not engaging in idle hair-splitting here – Sapolsky devotes two chapters to chaos theory earlier in the book. Chaotic systems are in essence unpredictable: a tiny difference in initial conditions between two systems, a difference so small that it is effectively immeasurable, can lead to enormous differences later on. A famous example in physics of chaos at work is the three-body problem. There is no general solution to this problem. Even if we could task the most powerful computer in the world with finding the solution it would in principle be unable to do so. Sapolksy's reason for bringing in chaos theory is that some free-will proponents have argued that the brain is a chaotic system, therefore fundamentally unpredictable, and that this unpredictability is equivalent to indeterminacy. Sapolsky replies that unpredictability is not the same thing as indeterminacy: just because we don't know the cause of a behaviour doesn't mean it was causeless, 'freely willed'. He gives the example of behavioural variant frontotemporal dementia. In 1922, no one knew the cause of it. But today we can predict with high accuracy that an individual will develop it in later life if he was born with a mutation in a gene called MAPT. The mistake Sapolsky makes here is that this example has nothing to do with chaos theory: a chaotic system, such as the system that causes the tornado in Kansa or the three body problem, in principle can only be very approximately solved if solved at all. More pertinently to my present argument is that in the chapters on chaos theory, Sapolsky seems to concede that the brain might be a chaotic system and then in the chapters on quantum physics conveniently forgets this, citing only experts who think the brain is a kind of stable or linear system. If however there is both subatomic randomness in the brain and chaotic amplification of these random factors by the brain, it seems quite reasonable to believe that indeterminate quantum events could lead to indeterminate behaviour.
Then again perhaps Sapolsky and the other quantum skeptics are indeed right, that a single quantum fluctuation gets lost in the noise. Let us grant this for a moment. There is another argument I would like to make in favour of 'indeterminism'. In order to set it out, I need to digress and discuss Neo-Darwinism or what is sometimes called the Modern Synthesis.
Modern evolutionary biologists have a very simple story to explain the evolution of organisms. We begin with a population of a single species inhabiting a particular environment, occupying a particular environmental niche. Every now and again a random DNA mutation occurs: usually these mutations are detrimental to the organism but sometimes a mutation is very slightly beneficial to the individual born with it. This individual, if also blessed with a little good luck, will have more offspring than its compatriots. Over time the number of individuals who possess this beneficial mutation will increase until they become the majority. Eventually it will be possible to say that the whole population, the whole species, has changed. This is how a species evolves. Because this mass change has resulted originally from an interaction between a single DNA mutation in a single individual and the organism's environment, it seems reasonable to describe evolution as a chaotic process (although it may be helpful to bring in another notion from chaos theory, 'strange attractors').
I once wholeheartedly subscribed to this doctrine. But then in 2013 I realised that there was a significant problem with it and in the years since decided that this simple little story just didn't make any sense at all. It would take me too far afield to go into detail about why I came to this conclusion in this essay. Suffice it to say that my rejection of the Neo-Darwinist story did not lead me to reject evolution altogether but rather to realise that some other process must be involved. In early 2018, a solution came to me, in fact via a voice. When a mutation occurs, it must occur to a large number of individuals in a population at the same time. It is almost as if some kind of collective spirit or idea that embraces the whole population simultaneously affects the whole population .
The relation of this insight to the brain should be obvious. As quoted above, Sapolksy says, "you'd need to have a staggeringly large number of such random events occurring at the same time, place, and direction." But perhaps something like this is exactly what happens. Perhaps our behaviours result at least partially from a large number of quantum fluctuations that are coordinated. To reiterate, I am not here defending free will. What I am suggesting is that our behaviours may not be wholly biologically determined but may also be supernaturally or spiritually influenced. Quantum physics provides a bridge between the materialist universe and its supernatural counterpart. This influence results either from the chaotic amplification of quantum events by the brain or from the coordination of a number of quantum events within the brain that ultimately can be traced back to causal factors in a sense outside the brain. At this point I want to bring in another key plank in Sapolsky's case against free will. Sapolsky argues that if our decisions are undetermined, the mind would throw up so many possible options, perhaps an infinite number, that it would take a lifetime to choose between them, to select an appropriate behaviour for a particular circumstance. The problem with this argument is, as mentioned above, that Sapolsky views apparent randomness as absolute randomness. However it could be that biology only picks out a discreet number of possible alternatives. It could also be, and this is my preferred postulate, that each variable associated with an action lies on a normal distribution. Yes, extreme values are possible but are very unlikely – 95 percent of the possible values for a variable lie within two standard deviations of the mean if the variable can be described as normally distributed. Thus Sapolsky's argument here can be countered if we suppose that randomness is constrained or weighted rather than absolute. Although, of course, as I've just said, I am not here defending free will but rather arguing that our actions are underdetermined by biology.
Sapolsky has another response to defenders of quantum free will. He says that if our behaviours are at root random, we cannot be held responsible for them. My perspective on this idea should be deducible from what I will have to say about free will towards the end of the essay.
Whew. This essay has swollen into a monster. It probably would have been more sensible for me to talk only about Determined in this post. However in a recent post, "Quantum Physics for Dummies and a New Idea" I made a couple of errors that I feel I should correct. I also want to develop and clarify some ideas I proposed in that post. So at this point I want to move away from Sapolsky and discuss quantum physics more directly.
In that recent post, I discussed wave-particle duality. Although I did not say this directly, I implied that electrons and photons can both be described by the Schrodinger equation or Dirac equation. If any physicists read this blog, they will have of course spotted the obvious error. In particle physics there are two types of particle: fermions and bosons. The Dirac equation applies to the former and the Klein-Gordon equation to the later. Electrons are fermions and photons are bosons. When writing that previous post, I actually did realise that the Schrodinger equation does not apply to photons but wasn't clear about this. However, with respect to the example I discussed, diffraction, it doesn't matter which equation is used because the width of the bands that we conceive of as 'appearing' on the screen depend only on the width of the aperture, the wavelength of the particle, and the distance to the screen. Because we presume that both photons and moving electrons have characteristic wavelengths, the fact that each are described by different equations is irrelevant to this example.
The purpose of the recent post was to describe as clearly as possible the Measurement Problem. I have thought of another way to approach this puzzle. Let us suppose that we can make two different types of measurement: primary measurements and secondary measurements. Some experimenters want to perform a diffraction experiment involving electrons. Before they begin, they measure the width of the aperture and the distance to the screen and make a hopefully quite accurate estimate of the velocity of the electrons they are going to fire through the aperture. These can be described as primary measurements and are used to arrive at an initial approximation of the wave function. The experimenters also measure the location on the screen each electron arrives at: each of these can be described as a secondary measurement. My claim is that the experimenters formulate a model of the wave function of a particular electron based on the primary measurements but then must amend this model by taking into account the secondary measurement. The wave function is probabilistic and probability estimates must be revised to take into account new information; furthermore it is not just the present and future wave function that must be revised but also the particle's past history. With each additional measurement, the wave function changes, although, because of the Heisenberg Uncertainty Principle, no number of measurements will enable us to arrive at an absolutely accurate description of the particle's past history, present, and future. (There is a snag involved with this picture that I shall come back to in a moment.) Although I proposed, provisionally, the idea that we have two different kinds of measurement, this distinction should in fact be discarded and we should actually treat all measurements similarly.
In the recent post I described the three most prominent interpretations of quantum mechanics: the Copenhagen interpretation, the Many Worlds interpretation, and De Broglie-Bohr Pilot Wave Theory. The essay was a bit muddled because I wasn't sure which if any of these interpretations was compatible with my own theory. I realise now that none of them are. This is because all three treat primary and secondary measurements as quite distinct. The Copenhagen interpretation supposes that primary measurements determine a wave function that is real and objective; secondary measurements have no effect on the wave function. All they do is confirm, in aggregate (that is, after many electrons are fired through), the probability distribution given by applying the Born rule to this objective function. Many Worlds also presumes that the wave function is objective and, once determined by primary measurements, unchangeable. As I said in the earlier essay, according to this interpretation 'wave function collapse' never occurs. Rather every possible secondary measurement occurs, each in its own universe. I have thought in the past that my own theory might be compatible with Pilot Wave Theory but now suspect it is not. This is because Pilot Wave theory also assumes that the wave function is objective and cannot be affected by secondary measurements. In supposing that every measurement changes the wave function it is possible that I am making an enormous error somewhere but I don't know what this error could be.
What is the snag? In the previous essay I imagined a scientist, call her Jane, who has a conception of the wave function in her own mind based on primary measurements and then 'updates' her conception based on an observation of where the electron actually lands on the screen. The problem here is that although it is fairly easy for Jane to work out a wave function for the electron involved in a diffraction experiment based on primary measurements, it may well be fiendishly difficult for her to calculate a wave function that incorporates her knowledge of where the electron has actually made contact with the screen. So the idea that she 'updates' her conception based on the secondary measurement only makes sense if Jane can instantly solve a fiendishly difficult and perhaps impossible mathematical problem in her mind very quickly, something I think unlikely.
This brings me to the fundamental mystery. What is probability? In the past I have said that any probability estimate is 'subjective' in the sense that it is made by a person based on incomplete information. I can illustrate this with a scenario I have deployed several times before in this blog. Three people, Brenda, Don, and Amy, are asked to estimate the probability that the top card of a shuffled deck is the Queen of Hearts. Brenda knows nothing about the deck and so, applying the Principle of Indifference and thus assuming that each combination of cards is equally probable, estimates the probability as being 1/52. Don has more information. He knows that the top thirteen cards are all hearts and so estimates the probability as being 1/13. Amy, however, knows for sure that the top card is the Queen of Hearts and so estimates the probability as 100 percent. I view the universe as deterministic although, as I said above, I suspect that there is both natural and supernatural causation. If the universe is deterministic then if a person somehow knew everything, possessed all the information, she would estimate all probabilities as either being certainties or impossibilities. It is only because people lack some relevant information that probability estimates come into play. Furthermore one person can possess different information than another. This is why I have said that all probability estimates are subjective. However I am not saying that such estimates are completely arbitrary, that a person can just make any estimate he or she wants. In pure mathematics, the study of probability is a big deal: there are empirically well established axioms like the Principle of Indifference, the Law of Large Numbers, and the Central Limit Theorem (although it is possible that people sometimes apply these rules in situations where they shouldn't). The problem is that the philosophical foundations of probability theory do not seem to be well established: this is why the Monty Hall problem confused so many people in the 1970's and why the Sleeping Beauty problem confuses so many people today. Somehow we need to regard a probability estimate as subjective and objective at the same time. This is something I can't speculate on any further on in this essay.
The questions about probability I have raised in the above paragraphs are, of course, not wholly original with me although I believe I came up with them at least somewhat independently. Bayesian probability is well established. However, even though I am acquainted with very little scholarly research concerning quantum physics, I sense strongly that even though quantum mechanics is necessarily probabilistic, quantum physicists themselves have rarely if ever attempted to interpret quantum mechanics from a Bayesian perspective. Quantum mechanics is strange enough as it is: if we all decided that the wave function squared actually describes our knowledge about where and when a particle is rather than objective reality, it could cause us all to descend into the abyss of German transcendental idealism. Physicists themselves exhibit a reflexive antipathy towards interpretations of quantum physics that involve consciousness. This explains the popularity of 'objective collapse theories' and of the idea of 'decoherence' which many physicists wrongly believe solves the measurement problem. (I considered talking about decoherence in this essay but it is long enough as it is.)
And now, finally, I shall try to salvage some notion of 'free will'.
An important idea in modern philosophy is 'emergence', something Sapolsky talks about in the chapters on complexity. A traffic jam is constituted by a bunch of cars but is in a sense something else, in a sense emerges from its constituent parts. I am not a big fan of this term 'emergence' but it approaches the idea I intend to discuss. What I want to do is employ a different but closely related idea: 'levels of description'. At one level we can describe the phenomenon as 'fifty cars stuck front to back' or we can take a bigger picture view and call it a 'traffic jam'. Another example: suppose we encounter a friend, Kate, and note that she seems happy. Let us also suppose for a moment that the serotonin theory of mood is true. It seems we can say, "Kate has more serotonin circulating in her brain than usual" or we can say "Kate seems happy". These two description are both true (assuming of course that the serotonin theory is true) and both describe the same thing: Kate's mood. They are describing the same phenomenon on different levels. What I would like to suggest is that even if a person's behaviour at a particular time is determined, as Sapolsky argues, by action potentials in a neuronal network that is reacting to environmental stimuli, action potentials that are themselves completely determined not only by the stimuli but by hormones, glucose levels, upbringing, acquired culture, and genes, it is still possible to describe that behaviour, at the ordinary human level, as being the result of 'free will'. It may seem I am deploying a fallacious form of reasoning – I am simply stipulating that the meaning of 'free will' should be defined pragmatically, as being an explanation for human behaviours at the level of normal human interactions, in situations where we cannot open a hole in someone's skull and observe what the neurones and hormones are actually doing. The reason I am making this stipulation is that I believe that at the level of ordinary human actions, human behaviour is unpredictable.
Sapolsky would hate this argument. Throughout Determined he adduces examples of behaviours or phenomena that have simple causes, examples I have discussed in this essay. We can predict that a person will be diagnosed schizophrenic later in life if in childhood he wets the bed. Desert societies reliably invent monotheistic religions. Priming experiments reliably influence the choices of subjects. The culture of honour in the Southern US reliably informs Southerners' reactions to rude behaviour. A single gene almost invariably leads to a form of dementia. And so on. However when one considers one's friends, even though one tends to think of them as possessing stable characters or personalities, in conversation we can almost never predict what they will say next. If the brain is chaotic, behaviour is unpredictable. If quantum effects have supernatural origins and can influence behaviour, it becomes even more unpredictable. And if we accept both chaos theory and this quantum hypothesis, it becomes possible to believe in something like 'emergence' – we can justifiably imagine that our behaviours result from top-down causation as well as bottom-up causation.
There is another reason to defend some form of free will. More than three hundred years before the birth of Christ, Aristotle invented the 'lazy man argument'. He argued against predetermination on the grounds that if the future is fixed, there would be no point in planning for it or deliberating about it. This argument can easily be countered if we suppose that the reason people plan and deliberate is that they possess the illusion of free will. However Aristotle's argument does raise an issue for determinists. Sapolsky didn't write Determined just to earn a buck or two – he is trying to change the world. Although I object to his representation of psychiatric conditions as being lifelong, I agree with his arguments in favour of greater compassion and leniency towards criminals. Sapolsky wants to make the world a better place and thinks if we all stop believing in free will this will happen. He has a chapter called "The Ancient Gears within Us: How Does Change Happen?" In it, he spends some time discussing the neuronal circuitry of the sea slug with the goal of showing how environmental factors can cause lasting changes in the neurology of these simple organisms and by analogy much more complex organisms like people. There is a tension in the book between his biological fatalism and his lefty compassion for the victims of circumstance, his desire to create positive change. Either, despite himself, he possesses the illusion of free will, or some supernatural force he doesn't believe in is propelling him forward. I can relate to this. A couple of weeks ago I became preoccupied with issues to do with free will and Fate. It occurred to me that in 2009 I decided at some subconscious level of my mind to embark on a fourteen year project to change the world, or that something else had charged me with this task. It's not that my mission has been to persuade the world of the truth of some New Age pseudo-psychological bullshit involving quantum physics. Rather I have been trying to change people's attitudes concerning 'schizophrenia' and 'sexuality'. It seemed to me that I had finally succeeded, leaving me with the existential question, "What now?" Of course my hopeful feeling that I have changed the world depends wholly on whether people read this blog or not – but let us suppose that they do. Sapolsky is arguing that occurrences in the past push people forward; I am proposing that people can be pulled forward by the future. This pull cannot be explained away rationally if rationality involves an ordinary understanding of time and causation. This is another motivation for my pragmatic definition of 'free will'. If a person is on the right path it seems entirely permissible to describe his or her actions as arising from 'free will' even if at some deeper level they really can't.
Finally I want to defend some notion of 'free will' for one more reason. Our values and feelings should be described as real. My dislike of Vladimir Putin, the leaders of Hamas, Benjamin Netanyahu, and the whole US Republican Party are genuine emotions. My empathy for the crazy and homeless people I meet in Auckland's CBD is genuine. Emotions like 'guilt' and 'shame' are real. Sapolsky's reductionism deprives such values and feelings of the importance they need to have to guide our actions. According to Sapolsky, biological determinism is good because it lets criminals off the hook morally; I think we should show compassion and leniency towards criminals because they could always regret their actions and repent. I admit that in presenting this argument I could be committing an informal logical fallacy philosophers refer to as 'wishful thinking'. But pragmatically it is compelling.
When one writes an essay one inevitably arrives at the problem of the conclusion. Perhaps I should just summarise what I have said. Although Sapolsky claims that science collectively has disproved the existence of free will, when he adduces findings from sciences other than neuroscience, these findings are dubious. His view that psychiatric disorders are permanent is wrong; his view that desert cultures reliably produce monotheistic religions is speculative; his reliance on priming experiments is misguided. If we presume that brains chaotically amplify quantum indeterminacy or that quantum events are somehow coordinated, we can imagine the possibility of supernatural causation. Quantum mechanics should in my view be rethought as involving a Bayesian conception of probability. And I think it necessary to at least pretend that something like 'free will' exists for pragmatic reasons. In this way we finally arrive at the end of the essay. Whether I am a rational analytical person or an intuitive emotional person is a question I leave to the reader.
No comments:
Post a Comment