Materialism and free will

I have written before about the apparent contradiction between free will and materialism (the idea that the universe is exclusively comprised of particles that obey physical laws). The problem is easy enough to state: if every particle in the universe behaves in a manner governed by a combination of random chance and predictable laws, how can a physical entity like the brain respond to stimuli in a way that is neither random nor determined?

Joshua Gold of the University of Pennsylvania and Michael Shadlen of the University of Washington recently summarized some experiments on monkeys that illuminate this issue. They found that they could use a computer to predict how monkeys will respond to visual stimuli, suggesting that such mental functions are automatic.

Of course, there is a big difference between parts of mental life like maintaining a steady heartbeat and tracking a moving object visually and those like making ethical decisions. That said, I continue to be unable to see what mechanism could exist between the former and the latter, and which could square our intuitive belief in free will with what we know about the functioning of the universe. That being said, we do not have any reason to act as though free will does not exist. The reason for that is simple: if free will doesn’t exist, we don’t have any influence over what we believe or how we act, while if it does exist we certainly want to behave appropriately. As such, if we do have any scope to choose, we should choose to believe in free will.

Author: Milan

In the spring of 2005, I graduated from the University of British Columbia with a degree in International Relations and a general focus in the area of environmental politics. In the fall of 2005, I began reading for an M.Phil in IR at Wadham College, Oxford. Outside school, I am very interested in photography, writing, and the outdoors. I am writing this blog to keep in touch with friends and family around the world, provide a more personal view of graduate student life in Oxford, and pass on some lessons I've learned here.

73 thoughts on “Materialism and free will”

  1. “if free will doesn’t exist, we don’t have any influence over what we believe or how we act, while if it does exist we certainly want to behave appropriately.”

    In my view, it’s worse than this. If free will doesn’t exist, then, given the current meanings of the terms, we don’t “act” or “believe” at all. We might “behave” and “influence”, however.

  2. It is exciting to think about what could happen, if the level of computing power humanity has continues to increase exponentially. If it does, it seems plausible that one of two interesting situations would arise:

    1) True artificial intelligence – the emergence of non-biological machines that perceive themselves as self-aware, as far as we can tell. The whole ‘ghost in the machine’ idea brought to life

    2) The final disproof of free will – as in the thought experiment, we become able to model every particle and law within a human brain and predict how it will respond to any input, with some uncertainty from random chance

    The first situation would extend the mystery of materialism and free will. If it can seem to arise from some combination of circuits and software, where does the physical explanation lie?

    The second would settle the question, though it is really odd to think what it would mean sociologically and psychologically. After all, the experiment would prove that our reaction to the results of the experiment would be highly predictable.

  3. While many smart people are working on it, it is not obvious that any amount of computing will lead to mental events. Sure, we could simulate mental events, or at least their superficial properties, but there is no reason to think that simulating their superficial properties is the same as simulating what actual (unbearably complex) events are going on underneath them

    To think about what thought “is”, it’s useful to think about how we model other natural phenomena that exceed our ability to model “from the ground up”. Weather and climate, for instance – we can say some meaningful things about “what is going on”, but we say these in spite of not knowing the internal mechanisms. One reason for this in the case of weather is the internal mechanism is not small – it’s actually an unbelievably complicated many-level set of relations between many different kinds of objects.

    Thought is also many-levelled and unbelievably complex. Thought, for instance, includes perception – which means it includes one’s entire body (whatever kind of body one happens to have). Thought also includes the object-thought, however, which means it includes the situation in which one is thinking. In the same way, the climate includes the sea, the sun, the land, cities, compounds being added or taken away from the atmosphere etc…

    So, like climate, we will probably be able to make better and better models which predict how thought will act or react in different situations, in response to different changes in input etc… But this is not the same as reproducing the “inner mechanism”, who’s complexity might forever exceed our ability to grasp.

  4. It’s true that we may never be able to tell whether a machine has an inner mental life or not – just as we cannot tell if our fellow human beings are philosophical zombies.

    That said, just having a computer that can pass the Turing Test would be impressive.

  5. As for modeling, it is possible that there are some phenomena that are so sensitive to minute differences in initial conditions that modeling is practically impossible, even with vast computing power.

    Bunimovich Stadium is a simple demonstration of how small variations in inputs can generate vastly different outputs, even in a relatively simple system characterized by easily comprehensible rules.

    Of course, complexity of the Bunimovich Stadium type is not indicative of free will or awareness.

  6. Thought, for instance, includes perception – which means it includes one’s entire body (whatever kind of body one happens to have).

    Having a body isn’t really a problem for the electronic-brain-in-a-jar thought experiment. The sensations associated with having a body are just one set of inputs you would feed into the simulated brain.

  7. I think what would be impressive is not a computer passing a turing test, but an experiment in which a human being might be controlled like a robot. Such experiments are already possible with rats (you can move the rate around like a remote controlled car). The crucial thing we can learn from this experiment is whether or not the subject of the experimented experiences the other-controlled movements as his or her own will.

    I’ve heard about existing experiments which have already shown that when you decide something, some sort of testing equipment can tell that you have decided a half second before you yourself are conscious of the decision.

    These kinds of experiments, much more than AI modelling, will/do throw into question assumptions we make about action and responsibility.

    On the other hand, Kant already knew this – if you treat the human as an object of scientific inquiry, you will inevitably learn that it is no different than any other object of enquiry into law-governed objects.

    The point is, you (Milan) already have the right attitude toward’s freewill – it’s an assumption that we are better off making than not making. The reasons why we are better off making this assumption run right down to the basic assumptions behind any society governed by norms – that people can be held responsible for things they do. This also relates to the problem of the continuity of identity across time.

  8. The rats and insects, as I understand it, are ‘steered’ by exploiting the reflexes associated with their whiskers – like when you hit a human’s knee with a hammer.

    I don’t think either reflex action tells us much that is important about free will.

    That said, the ideal way to test the brain-in-a-jar would be to simulate a specific person, expose the real person and the simulation to the same inputs, and see if they react identically.

  9. Commands and rewards

    Electrodes were implanted into areas of the rat brain responsible for sensing reward as well as those that process signals from their whiskers.

    The commands and rewards were transmitted by radio from a laptop computer to a backpack receiver strapped to each rat.

    The scientists were able to make the rats run, turn, jump and climb where they wanted from distances of up to 500 metres (1640 feet) away.

    The ratbots negotiated an obstacle course which involved climbing a vertical ladder, running along a narrow ledge, hopping down a flight of steps, squeezing through a hoop and descending a steep ramp.

    The scientists say, “Our rats were easily guided through pipes and across elevated runways and ledges, and could be instructed to climb, or jump from, any surface that offered sufficient purchase.

    “We were also able to guide rats in systematically exploring large, collapsed piles of concrete rubble and to direct them through environments that they would normally avoid, such as brightly lit, open arenas.”

    A “turn left” signal was interpreted by the rat’s brain as a “touch” on its left whiskers. If the rat correctly followed the cue and turned left, its reward-centre was stimulated, filling the rat with a feeling of well-being.

    Dr Talwar says, “This is an animal with 200 million years of evolution behind it. Rats have native intelligence which is a lot better than artificial intelligence.”

  10. So, it goes beyond reflexes, to incorporate reward centre stimulation as well.

    Still, I don’t know how much this tells us about free will.

  11. It doesn’t tell us much at all about freewill. However, if we could run similar experiments on humans, and we could interview the humans, we might learn something.

    As for “brain in a jar”, I don’t understand. Brains in jars can’t think – to think you’re going to need some kind of body. Even if that is some very limited form of perception, there is going to need to be more than the “jar”, there will need to be a “world”. Without world, there is no content – and thought always has content.

    Also, no amount of brain in a jar tests will teach us anything about free will, – see Chinese Room Example.
    http://en.wikipedia.org/wiki/Chinese_room

  12. A brain in a jar is like a processor not plugged into a motherboard. Yes, you need to connect it to something to make it work. That said, if you have been able to build one processor that perfectly replicates the behaviour of another – when plugged into the same slot – you can claim with good authority that you understand the first processor well enough to simulate it accurately.

  13. The processor-processor analogy is partial, because neither is an object we expect to possess free will.

    That said, if we can build a processor that behaves just like a brain, and we continue to accept that processors don’t have free will, then it seems we have proven that brains don’t have it either (or at least that our choices are entirely predictable).

  14. I think a brain in a jar is more like an atmosphere without a planet. A processor is a reductively grasped object – we can idealize where each part is from the ground up. But a brain is more like an atmosphere – we can say some things about it inductively, but we can’t grasp the inner mechanisms that make it respond in its environment in a particular way.

  15. What if those mechanisms are actually very simple: a set of physical laws obeyed by each particle?

    If the laws are simple, then all things made of matter would be equally easy to simulate: whether they are cold lumps of rock or the minds of sentient beings.

  16. I think you have forgotten one the most basic ruless about modelling – you can’t build a model of the world (or anything indeterminably complex), because you’d need something as big as the world (or as complex, which if you are going down to the level of the particle, amounts to the same thing). And even if you could, it wouldn’t teach you anything. Models are useful only to the extent that they simplify the natural phenomena.

    If the “natural phenomena” of thinking is complex enough, it might not be simplifiable by models. Or – it might. For instance, Chomskian grammar tries to hypostatize about what kind of structures might be behind language such that humans have the ability to learn language without being taught (explicitly at least) any rules. This is an example of trying to determine certain properties some genetic mechanism has without needing to know the reductionistic account of how these properties emerge.

    Language might be an instance where models help us simplify and understand an underlying mechanism which remains essentially beyond the reach of human ability to comprehend. Notice though, that the chaos at the bottom doesn’t mean concepts can never be appropriately applied.

  17. you can’t build a model of the world (or anything indeterminably complex), because you’d need something as big as the world (or as complex, which if you are going down to the level of the particle, amounts to the same thing)

    This isn’t necessarily true. If there are underlying dynamics that determine how all matter and energy interact, it is possible that combining our understanding of the mathematics of their function with raw computing power would allow us to simulate something larger than the machine we are using to do the simulating.

    It comes down to how elegant the mathematics at the base of reality are.

  18. It is a tiny bit like the difference between vector and raster graphics.

    A bitmap is just a checkerboard of pixels of various colours. Performing an operation on one (such as flipping it) requires a very predictable amount of memory, time, and computing power in combination.

    A vector graphic is a set of equations, describing something like a character in a TrueType font. It can be manipulated in more perfect and elegant ways than a bitmap of the character.

    If the universe is, at the very base, like vector graphics, it could be possible to partially simulate on computers we could actually build.

  19. Good point about elegance, and vector/scalar graphics. However, even simple relations between particles can become unbearably complex as the numbers of particles increase. This can be in principle overcome by computing power only if the increase in complexity is linear – if the complexity of relations between matter and energy increases exponentially as the number of entities involved in the relation increase, we probably have good reason to believe that thought is un-model-able without simplification.

  20. Alternatively, there could be mathematical shortcuts that let us solve exponentially more difficult problems using only linearly increasing computing power.

    This could be akin to specialized algorithms for rendering graphics, or interpolating data from a Bayer array.

  21. To speculate wildly, maybe in the distant future we will be able to build custom universes, full of computing power, in order to simulate things in whichever universes we care to live in.

    In any event, determinism that is hard to prove is quite distinct from free will.

  22. “determinism that is hard to prove is quite distinct from free will.”

    Maybe not. Maybe “free will” is the way humans have learned to deal with other people. In the same way that we economize with respect to every other natural phenomena (I don’t need to know the reductive relations of all the particles involved to know I will get hypothermia if I spend too much time in ice water), we economize in our relations to others. We don’t need to know “perfectly” who other people are in order to be able to deal with them, work with them, converse and enjoy experiences with them.

    Maybe “Free will” is just the idea that other people, unlike objects, need to be treated with autonomy and respect if one is going to flourish with them.

  23. Maybe “Free will” is just the idea that other people, unlike objects, need to be treated with autonomy and respect if one is going to flourish with them.

    I like the idea here. It doesn’t really get at the question of whether or not free will is compatible with physical law, but it does touch on the issue that humans (and many animals) perceive the world as being made up of two very different categories – animate and inanimate, agents and non-agents, or (very loosely) free will and no free will. And it makes good sense to interact with these categories differently.

    I wonder how we would come up with a clean division between the two categories that would admit sufficiently complex robots into the people category? My feeling here is that for most people what will matter is not some proof that the robot has or does not have free will, but rather a certain threshold of complexity of behaviour.

    However, complex behaviour alone is not quite it. Many things have very complex behaviour – computers, river systems, weather patterns, etc.

    I think the second ingredient required for us to ascribe “free will” is goal-directed behaviour. Physical systems are conspicuously not goal directed. Animals are. Plants are too, but perhaps in ways which are too simple (superficially at least).

    Right now computers have high complexity of behaviour (perceived as much more complex in behaviour than plants, I would guess), but they are purely reactive. As soon as their behaviour becomes goal directed (and those goals are to some degree internally managed, not simply responses to direct commands), I think we will very quickly begin to treat machines in the same way we do other “animate” things. To begin with, probably at a level of respect equivalent to very simple animals, but as their complexity increases, I would guess they will gradually ascend towards the “person” category.

    ============
    Actually, as I wrote this, an interesting question occurs to me. Why do we ascribe “free will” to animals but not plants? (Indeed, do you ascribe free will to animals? Or just to humans?). As you think about them a bit more, the distinctions here are not entirely obvious!

  24. I think what would be impressive is not a computer passing a turing test, but an experiment in which a human being might be controlled like a robot. Such experiments are already possible with rats (you can move the rate around like a remote controlled car).

    You can do the same with humans actually, although it’s not the same as the rather spooky example with rats. It’s just a cheap hack based on disrupting the inner ear. See here:
    http://www.youtube.com/watch?v=Kf0E9llkZIU

    Apparently the sensation of being piloted involuntarily is rather unsettling.

  25. On the overall topic of free will, the perspective I like best is Milan’s pay-off matrix. Whether or not we have free will, it makes sense to believe that we do. This has stuck with me from our discussion a few years back – I’ve discussed it with people several times since.

    It is genuinely hard to see how there is room for free will in a universe governed by physical law. The obvious wriggle room is that it is also very hard to see how consciousness fits in with physical law. And consciousness is pretty much the one thing we cannot doubt we have. Perhaps in consciousness there is still enough mystery to leave the free will question open for now.

    On the topic of consciousness, I have always been fascinated by the China brain thought experiment (not to be confused with the Chinese room experiment). It leaves me with no strong convictions either way.
    Could such meta-minds exist? If (information processing == consciousness), i.e. consciousness is a necessary epiphenomenon of certain kinds of computation, then it seems that they must. I lean towards that conclusion, but it certainly raises all sorts of strange questions.

  26. Is there any way consciousness could be an illusion?

    Say, for instance, that our processes of both logical and emotional reasoning are automatic – they will always produce identical output from identical inputs.

    If so, consciousness could be like riding on a mine cart with your head in a vice. You see everything going by, and experience various sensations, but it is all pre-determined by initial conditions and automated patterns of response.

  27. “Is there any way consciousness could be an illusion?”

    Depends what you mean by illusion. Take anything – a chair, a feeling, a painting, and think about the difference between how you perceive it, and how you would describe it scientifically. So, they are all “illusions”. But at the same time, something real is manifested in them.

    So, if by “illusion” you mean “is not correctly understood by our normal ways of thinking about it”, then of course it’s an illusion.

    But, if by “illusion” you mean something like a purposeful deception, or an idea of something that primarily conceals what a thing is, it’s less obvious consciousness is that kind of illusion.

    For comparison, think of any scientific concept which isn’t contemporarily recognized as true, i.e. matter, or gravity. No one believes in matter (in the pre-Newtonian sense), or gravity (in the Newtonian sense). So, were/are these beliefs illusory? And if so, in which sense? There may not be a strict demarcation between the senses.

  28. If so, consciousness could be like riding on a mine cart with your head in a vice. You see everything going by, and experience various sensations, but it is all pre-determined by initial conditions and automated patterns of response.

    Right, that sounds like epiphenomenalism – the idea that conscious experience is a kind of side effect of physical processes, but doesn’t have any impact on them. Which is entirely possible.
    I don’t think, however, that you can say that consciousness itself is an illusion (although you can quite easily argue that the conscious experience of free will is). Either way, we still need some way to explain the conscious experience within physics, which seems to me a genuinely hard problem.

  29. So, if by “illusion” you mean “is not correctly understood by our normal ways of thinking about it”, then of course it’s an illusion.

    By ‘illusion’ I meant what Mark identified more precisely: the possibility that we are just ‘along for the ride’ when it comes to what we think and feel, and that the process is not interactive.

  30. Why do we ascribe “free will” to animals but not plants? (Indeed, do you ascribe free will to animals? Or just to humans?).

    This is something that bothered me about Douglas Hofstadter’s book I Am a Strange Loop. He repeatedly asserts that humans have a vastly different mental life from other animals, but doesn’t provide what I consider to be a satisfactory explanation for why.

    One justification he brings up, and which is interesting, is the infinitely extensible character of our language. We can go from ‘John is a man’ to ‘Jerry thinks John is a man’ to ‘Susan doesn’t know that Jerry thinks John is a man.’ It is conceivable that other animals aren’t capable of doing this, though it would take some clever experimentation to find evidence about it one way or the other.

  31. Another potentially interesting phenomenon to consider, in relation to consciousness and free will, is the success of magicians, pickpockets, and con artists in using strategies like misdirection.

    The fact that a magician can manipulate an entire audience into looking at one thing rather than another suggests not only that there are relatively high-level human responses that are fairly automatic, but also that these responses are quite similar for individuals of different ages, backgrounds, etc.

  32. One more thing – while consciousness is certainly something we should continue to try to describe, we should be wary of extreme skepticism with regards to our current conception of it. Consciousness is not like my perception of an object (about which it is quite easy to theorize my uncertainty), but the fact that “I” exist at all (in the special sense of “I” as something that can think, make declarations, etc…). All of our ideas about life and being are presumed from a position of spontaneous thought – without consciousness even the basic concept of “belief” makes no sense.

    So, if we are concerned with economy or preserving elegance in our ideas about the world, we should conceive of inquiry into the nature of consciousness as improving our understanding of something which we already, to some extent, grasp.

    I don’t think consciousness is genuinely a “hard problem”. A hard problem is: “how can we communicate with other consciousnesses”.

  33. I don’t think consciousness is genuinely a “hard problem”. A hard problem is: “how can we communicate with other consciousnesses”.

    I disagree here. Consciousness is surprising to me. It’s not at all obvious to me why it occurs. A-priori, knowing about the laws of physics but nothing else, it would seem more natural to expect that we would be philosophical zombies, with complex behavior but without “mental experience” (i.e. qualia, etc). My tentative conclusion is that zombies are for some reason impossible, but it is not at all obvious why this should be.

  34. I guess the answer to that is: “If you are one, you cannot know it because you cannot know anything. But if you have conscious experiences, you cannot be one by definition.”

  35. “A-priori, knowing about the laws of physics but nothing else, it would seem more natural to expect that we would be philosophical zombies, with complex behavior but without “mental experience” (i.e. qualia, etc).”

    This is a poorly formed thought. If you “know” about the laws of physics, you are already conscious. You would not “expect you were a phil. zombie”.

    I think the idea of a philosophical zombie is a being which is externally indistinguishable from a person, but who has no inner mental life. If that is right, it would not “feel like” anything to be a phil. zombie – if you were one, you wouldn’t know, because there would be no “you” there.

  36. “Consciousness is surprising to me. It’s not at all obvious to me why it occurs.”

    It’s not at all obvious to me why gravity occurs. Or why matter bends space. There is plenty we don’t know – what’s wrong with mental reductionism is it assumes that in the future our knowledge will be an extension of present knowledge. If you look back historically, all great gains in scientific knowledge happen when everything up to that point is either thrown out or radically re-understood.

    I see no reason why we should expect the problem of consciousness to be resolved within a paradigm where mind and matter are still considered distinct substances.

  37. “Well the brain as you know is made up of neurons. There are a hundred billion in the adult brain, and each neuron makes something like ten thousand contacts with other neurons. People have calculated that the number of permutations of brain activity exceeds the number of elementary particules in the universe.”

    http://www.youtube.com/watch?v=t0pwKzTRG5E&feature=related

  38. There is plenty we don’t know – what’s wrong with mental reductionism is it assumes that in the future our knowledge will be an extension of present knowledge.

    The question of free will can only be asked at this point, not answered.

    That said, the question is very simple: is there something in the universe that affects the outcome of events, aside from pure random chance (especially in quantum phenomena) and the laws of nature? If so, is that ‘something’ compatible with what we believe about free will?

    People have calculated that the number of permutations of brain activity exceeds the number of elementary particules in the universe.

    This is neither here nor there. Is it logically possible for some tiny gnat with hardly any neurons to have free will. It is similarly possible for brains far more complex than ours to operate purely on the basis of deterministic laws combined with random chance.

  39. Simulating the brain is hard, not least because of all the protein interactions. As PZ Myers points out – and Ray Kurzweil apparently does not understand – you cannot model the brain using DNA alone:

    Kurzweil knows nothing about how the brain works. It’s design is not encoded in the genome: what’s in the genome is a collection of molecular tools wrapped up in bits of conditional logic, the regulatory part of the genome, that makes cells responsive to interactions with a complex environment. The brain unfolds during development, by means of essential cell:cell interactions, of which we understand only a tiny fraction. The end result is a brain that is much, much more than simply the sum of the nucleotides that encode a few thousand proteins. He has to simulate all of development from his codebase in order to generate a brain simulator, and he isn’t even aware of the magnitude of that problem.

    Of course, if we could actually model the behaviour of every single particle in the brain, we would be able to model the behaviour of the proteins they make up, by definition.

  40. “Is it logically possible for some tiny gnat with hardly any neurons to have free will. It is similarly possible for brains far more complex than ours to operate purely on the basis of deterministic laws combined with random chance.”

    “is there something in the universe that affects the outcome of events, aside from pure random chance (especially in quantum phenomena) and the laws of nature? If so, is that ‘something’ compatible with what we believe about free will?”

    It’s not at all clear that there is much agreement on what we “believe about freewill”. Certainly Kant and Mill do not agree – and this divergence is still strong today (with most liberals siding with Mill, and most heirs to the German Idealist tradition siding with some version of Kantian freedom).

    The most disturbing things about these discussions actually relates to Kant – but to the theoretical rather than practical philosophy. There is all this talk of “things operating according to deterministic laws”, as if there were no humans observing them. Nothing “operates according to x” – all that actually occurs is humans use certain epistemic/linguistic structures to interpret a world which always exceeds there grasp. Everything is far more complex than anyone understands – none of our knowledge is god-type, so it does not actually make sense to say “thing operates according to deterministic laws”. All you can reflectively say is “I can adequately conceive of it as acting according to deterministic laws”, where adequacy can be cached out in terms of some kind of pragmatism or reliabalism.

    This all applies a fortiori to the notion of freewill. What actually happens when someone else acts, and we conceive of them as responsible for that action? Well, you can be fairly sure that what actually happens is more complicated than you could figure out in the timespan of a thousand universes, so what matter is what form of knowledge of that will be adequate to the conditions imposed on you by real life?

  41. People have calculated that the number of permutations of brain activity exceeds the number of elementary particles in the universe.

    This is neither here nor there.

    Quite right! In fact, the number of permutations of two decks of cards comfortably exceeds the number of particles in the observable universe. For reference, the former is about 10^88 particles, and the later is 10^166 permutations.

    http://www.cs.umass.edu/~immerman/stanford/universe.html
    http://www.google.com/#hl=en&q=104!

  42. “People have calculated that the number of permutations of brain activity exceeds the number of elementary particles in the universe.

    This is neither here nor there.”

    If you want to presume that permutations of brain activity follow from an elegant set of deterministic laws, then obviously there is no freewill. But this will always remain a presumption – something not demonstrable, but which appears true given a certain angle on the evidence presented. What is difficult to grasp is that this is no less true for the deck of cards. If you began with the 10 to the 166 permutations from the deck of cards, and tried to source the elegant formula (i.e. the deck) from which these permutations arise – you may in fact discover the deck. But no where in the permutations would you find the deck, the deck would merely be a convenient posit which well explained the data.

    The point of the “trillions of permutations” point is simply that the world always exceeds your ability to grasp it in itself – the best you can do is ad hawk, (although sometimes elegant), after the fact reconstructions. The goodness of these reconstructions can be evaluated in no way other than how well they serve the purposes for which they are constructed.

    This means, if you want to conceive of human experience as merely the pre-knowable repercussions of whatever physicalist theory you prefer, you need to take seriously the fact that any account of human action which finds the origin of actions outside the action themselves throws out a priori the ground of virtually all western political thinking (Marx aside). That is a real cost, a genuine epistemic reason to question the categories from which the apparent paradox arises.

    Especially since, as I pointed out earlier, there is no real consensus on what “free will” is anyway. Maybe there is no consensus on what “deterministic” would mean either. Certainly Kantian freedom (if it is real) is as “deterministic” as any Newtonian model.

  43. Free will seems relatively easy to define in a minimal way:

    If it is possible to take a set of inputs and predict what outputs a thing will produce – within the accuracy limits established exclusively by random chance – that thing can be said to have no free will. It is a predictable entity – a machine, possibly very complex, but not a thing empowered to alter its own actions.

    Defining the full scope of free will may be hard, but it is easy to establish a minimum requirement. At the very least, a thing with free will produces outcomes that depend on more than inputs and random chance.

  44. “If it is possible to take a set of inputs and predict what outputs a thing will produce – within the accuracy limits established exclusively by random chance – that thing can be said to have no free will. It is a predictable entity – a machine, possibly very complex, but not a thing empowered to alter its own actions.”

    This is no good. Predictability is no more a disproof of freewill than the lack of predictability (chaos). The idea that a thing is “predictable” doesn’t come from the observations you make, but form fundamental assumptions you make about how to interpret those observations beforehand. If you assume that a system is either predictable or not predictable, then you can’t encounter freewill – because there is no way something that show up as freewill since everything is pre-grasped as either predictability or unpredictability.

    There is no single idea of freewill, but you can be sure that if you don’t have a good grasp on what you mean by freewill, you won’t be able to think intelligently about whether there are contradictions between that idea and other ideas you have.

  45. “At the very least, a thing with free will produces outcomes that depend on more than inputs and random chance.”

    For Kant, a free will is one which is determined according to the law of Freedom. So, it is a will determined by nothing other than inputs and law. The difference is, the law of freedom is a moral law, and it is posited to have a different kind of causation than material causation.

    Personally, I’m unimpressed with the idea of freedom is “doing what you like”. If freedom is merely following whimsy – then how is that different from a form of determinism? Well, if you look at the history of whimsy and the role of experimental psychology in marketing, you might think there isn’t a hell of a lot of difference! But if you think you are “free” when you’re buying the right kind of toothpaste, or a new camera, or a BMW – because you want it, and because you can afford it, that to me seems to have nothing to do with freedom.

    Freedom is, rather, instead of pursuing one’s whimsy – choosing to instead do what is right. So, for instance if one had the choice between a big promotion or not lying – the free choice is not to lie. We can chose to give up benefits in order to make the moral choice because we are free, because we are not merely produced as desiring agents by techniciens.

  46. “If it is possible to take a set of inputs and predict what outputs a thing will produce – within the accuracy limits established exclusively by random chance – that thing can be said to have no free will. It is a predictable entity – a machine, possibly very complex, but not a thing empowered to alter its own actions.”

    This is no good. (…)

    Tristian, I think you are dismissing this too freely (chuckle). I fail to see how building a machine that could perfectly predict an individual’s behaviour is not a disproof of free will.
    However, the interesting issue is that such a machine may be in principle impossible because of quantum randomness (as Milan allows with his reference to “within the accuracy limits established exclusively by random chance”). This is where the dragons are.
    One of the most succinct definitions I have heard for free will is that each individual is “an independent point of departure”. But how do you draw a bright line between free will and randomness? Is every quantum interaction between elementary particles not an “independent point of departure” too? What does “free will” even mean in this context?

  47. An alternative way to ask the question of whether something has free will is to imagine a circumstance like this:

    You are an entity that definitely has free will, and you encounter the entity being evaluated. On the basis of everything you know about the universe and the other entity, does it make sense to hold it responsible for its actions? It is a foolish being that blames a lightning bolt for striking it, or a big rock for rolling downhill.

    The question, then, is whether human beings are in the category of being meaningfully responsible or not. We tend to ask the question in the context of people who do things we dislike – did that mugger or rapist or murderer make a morally culpable choice, or just act in a way that was automatic in some sense or another? Of course, the question is equally applicable in relation to things we do like – does your husband or wife really deserve credit for doing romantic things on your anniversary, or are they just manifesting their characteristics as an automaton?

  48. “The question, then, is whether human beings are in the category of being meaningfully responsible or not. ”

    “does your husband or wife really deserve credit for doing romantic things on your anniversary, or are they just manifesting their characteristics as an automaton?”

    Treating these as either-or questions is to remain in a pre-Kantian way of thinking about freedom. “Either you are free, or your actions are determined”. One thing we know for sure is that your actions are determined – insofar as I consider you as an object like any other, you are exactly as predictable as any other. The predictability of things in the world is the product of an assumption that they are predictable, and it would be a strange dualism to assume that there were a certain class of things “out there” which simply refused to succumb to scientific inquiry. The dualism which allows a notion of freedom must be compatible with scientific rule based inquiry. This is why Kant posits another kind of causation which is not perceivable, but which is not disprovable. This kind of causation is lawful, and the form of it’s law is the categorical imperative, also called the “law of freedom”.

    There are other ways of thinking about freedom, sure, but none that take more seriously the conflict between freewill, science, and the universality of inquiry.

  49. “On the basis of everything you know about the universe and the other entity, does it make sense to hold it responsible for its actions? It is a foolish being that blames a lightning bolt for striking it, or a big rock for rolling downhill.”

    Is it foolish to blame a lightning bold for striking? I’d wager that for over 95% of civilized time and space, it has been normal to imbue extraordinary acts of nature with meaning, with agency. Now, I’ll be the first to say that we can now do better than imbue the natural operations of the world with a mysterious agency, but I also think it’s mistaken to dismiss most of history as simply “foolish”. There have been good evolutionary, at least group-emotionally strategic reasons behind ideas of deities causing the natural world.

    This is probably a decent way to think about human agency. Just as getting rid of the idea of god radically destabilized societies in the 20th century, the notion of ridding ourselves of agency in humans is frightening – because the entire fabric of society is based on the idea that if the kid breaks your window, it’s the kid that did it and not a random outcome of a series of theoretically predictable events.

    Now, I could situation similar claims that a priest could make as Darwin challenges the basic assumption that a theological force is behind the order of the world. I.e. “If my crops fail, I want to believe it’s God that did it and not a random series of theoretically predictable events”. Just as it’s frustrating (and socially destructive) not to be able to hold people responsible for their actions, it was (at the time considered) socially destructive not to hold God as responsible for natural actions.

  50. There is another whole set of reasons for questioning how much free will exists – namely, by pointing out how much people are manipulated by advertising, political propaganda, the education system, etc.

    We may have freedom at the level of atoms, but lose it because so many people have gotten good at controlling human thoughts and reactions.

  51. Everything is made of atoms. That is the key hypothesis. The most important hypothesis in all of biology, for example, is that everything that animals do, atoms do. In other words, there is nothing that living things do that cannot be understood from the point of view that they are made of atoms acting according to the laws of physics. This was not known from the beginning: it took some experimenting and theorizing to suggest this hypothesis, but now it is accepted, and it is the most useful theory for producing new ideas in the field of biology.”

    Feynman, Richard. Six Easy Pieces: Essentials of Physics Explained by Its Most Brilliant Teacher. p. 20 (paperback)
    Emphasis in original

  52. “If the brain were so simple we could understand it, we would be so simple we couldn’t.” – Lyall Watson

  53. “Everything is made of atoms.”

    Except atoms. And, ergo, anything made from atoms.

  54. The importance of things being made from atoms is that atoms follow predictable rules that humans can discover.

    The same is true for sub-atomic particles. The fact that atoms aren’t actually indivisible isn’t super relevant to Feynman’s point.

  55. A Real Science of Mind
    By TYLER BURGE

    In recent years popular science writing has bombarded us with titillating reports of discoveries of the brain’s psychological prowess. Such reports invade even introductory patter in biology and psychology. We are told that the brain — or some area of it sees, decides, reasons, knows, emotes, is altruistic/egotistical, or wants to make love. For example, a recent article reports a researcher’s “looking at love, quite literally, with the aid of an MRI machine.” One wonders whether lovemaking is to occur between two brains, or between a brain and a human being.

    There are three things wrong with this talk.

    First, it provides little insight into psychological phenomena. Often the discoveries amount to finding stronger activation in some area of the brain when a psychological phenomenon occurs. As if it is news that the brain is not dormant during psychological activity! The reported neuroscience is often descriptive rather than explanatory. Experiments have shown that neurobabble produces the illusion of understanding. But little of it is sufficiently detailed to aid, much less provide, psychological explanation.

    Second, brains-in-love talk conflates levels of explanation. Neurobabble piques interest in science, but obscures how science works. Individuals see, know, and want to make love. Brains don’t. Those things are psychological — not, in any evident way, neural. Brain activity is necessary for psychological phenomena, but its relation to them is complex.

    Imagine that reports of the mid-20th-century breakthroughs in biology had focused entirely on quantum mechanical interactions among elementary particles. Imagine that the reports neglected to discuss the structure or functions of DNA. Inheritance would not have been understood. The level of explanation would have been wrong. Quantum mechanics lacks a notion of function, and its relation to biology is too complex to replace biological understanding. To understand biology, one must think in biological terms.

  56. Naughty by Nature
    What should we think of people whose addled brains are driving them to nymphomania?
    By Jesse Bering
    Posted Thursday, March 3, 2011, at 7:39 AM

    If you are a materialist holding the logical belief that the human brain, with all of its buzzing neural intricacies, its pulpy, electrified, arabesque chambers and labyrinthine coves, has been carved out over countless eons by the slow-and-steady hand of natural selection, then you will grant that specific brain regions evolved because they generated behaviors that were beneficial to our ancestors. When one part of the brain is compromised—through injury, disease, or some other unfortunate event—the constellation of symptoms that result are often remarkably specific. “The brain is the physical manifestation of the personality and sense of self,” writes University of Michigan neuroscientist Shelley Batts in a 2009 issue of Behavioral Sciences and the Law, “and focal damage to brain areas can result in focal changes in behavior and personality while leaving other aspects of the self unchanged.”

    Not to get too technical, but if you’re unlucky enough to develop a lesion that interferes with the functioning of your dorsolateral prefrontal cortex, a specialized patch of neural tissue that’s intricately braided into your anterior cingulate cortex, then your working memory, strategy-formation, and planning skills are going to take a major nosedive. Something as simple as coming up with a list of groceries that you’ll need for the next few days becomes a major achievement.

    Most of us—materialist and dualist alike—have sympathy aplenty for those patients whose brain disturbances have interfered with their everyday cognitive abilities. We’re perfectly willing to accommodate their intellectual disabilities by, say, helping them create a new mnemonic strategy or giving them a pat on the back or a word of encouragement when they’re trying to remember someone’s name. Yet when chunks of gray matter that have evolved to control and inhibit, say, our sexual appetites and other Bacchanalian drives experience a similar catastrophic blowout, are we so understanding? What if those impairments lead their victims to display … oh, I don’t know, let’s call them moral disabilities? Cases of libidinal brain systems going haywire have our kind-hearted, humanistic materialism rubbing elbows—or butting heads—with our belief in free will and moral culpability.

  57. Scientists have discovered what they say are four different species of “zombie fungus” in the Brazilian rainforest, which take over the brains of their host ants, forcing them to move to a location ideally suited to the fungus before killing them.

    In a study published March 2 in the journal Plos ONE, researchers from Brazil, the United Kingdom and the United States say they began to investigate after noticing different types of fungus growing out of the bodies of carpenter ants.

    “This so-called zombie or brain-manipulating fungus alters the behaviour of the ant host, causing it to die in an exposed position, typically clinging onto and biting into the adaxial surface of shrub leaves,” the authors write.

    The fungus then grows — usually out of the ant’s head and neck region — and releases its spores.

    The fungus, Ophiocordyceps, was originally thought to be a single species, but the researchers determined that there were actually four species at work.

  58. “The importance of things being made from atoms is that atoms follow predictable rules that humans can discover.”

    Atoms don’t follow predictable rules humans can discover. I’ll give you an example: try to compute on the basis of the locations of sub atomic particles how various things should be priced.

    The moral of the story is: looking at the wrong level of resolution is a good way to never solve a problem. For instance, if you try to solve someone’s depression with drugs when the problem is they live in a pathological society (say, early 40’s Nazi germany), it’s not going to work – the person isn’t depressed because their brain is functioning improperly; they are depressed because it’s functioning properly: you shouldn’t be happy in a criminal genocidal state.

    We can’t compute social problems on a sub-atomic level. The amount of computation required, even if sub atomic particles followed cognizable laws, which they might not, very well might exceed the computational ability of machines that can be constructed within finite time. We solve computational problems that are too difficult to do on a micro level by aiming at a higher leve, where in fact we’re cleverly making simplifications so we can deal with something which is “essentially” way to complex to understand. The ability to simplify by looking at the right level of resolution is a major form of intelligence, in fact, as we get more sophisticated you might say we’re getting better at simplifying complex situations into situations that we understand. Any particular situation is near infinitely complex, so if we had to understand it perfectly we’d never understand anything.

  59. Whether we can actually compute it or not is a practical matter with limited importance. If, in theory, it is possible for a massively powerful computer to predict accurately how people will respond to particular stimuli, that seems to have important implications when it comes to the nature of free will.

    A less abstract but related idea is that people who suffer serious brain injuries may no longer be fully morally responsible for their actions.

  60. “Whether we can actually compute it or not is a practical matter with limited importance.”

    My point, which fair enough I didn’t make as clear as I should have, is that it’s quite possible that such computation does not just exceed the bounds of what we can do now, but the bounds of what is finitely possible within our universe. There is such a thing as the “most powerful computer possible”, and then we run into size limitations (it has to fit in the universe). Being is finite, which means only a finite number of computations can occur in a given amount of space and time. If the amount of computations exceeds that, then it is not just practically impossible, but theoretically impossible.

    We don’t need a fully computational model of mind to understand that people suffering from brain injuries should be less fully held responsible for their actions. We’ve held children less responsible for their actions probably forever, and we’ve only had the computational model of mind for, evolutionarily speaking, about ten seconds.

  61. That is still a practical, rather than a theoretical, limitation.

    It is possible that in the future humanity will be able to make custom universes that serve only as enormously powerful computers. The custom universe could exclude problematic rules like conservation of energy, and could consist of nothing but processing units on a galactic scale.

    That might be the level of processing required. Still, it is possible that the problem is one that can be solved, albeit not with current human capabilities.

    Obviously, humanity may never gain such capability. Indeed, it may actually be completely impossible. That said, it is a way of side-stepping the question of computing power to focus on the basic question of whether the problem can be solved with any amount of effort and resources.

  62. Could science prove that we don’t have free will? An article in Nature reports on recent experiments suggesting that our choices are not free. “We feel that we choose,” says the neuroscientist John-Dylan Haynes, “but we don’t.”

    The experiments show that, prior to the moment of conscious choice, there are correlated brain events that allow scientists to predict, with 60 to 80 percent probability, what the choice will be. Of course this might mean that the choices are partially determined by the brain events but still ultimately free. But suppose later experiments predict our choices with 100 percent probability? How could a choice be free if a scientist could predict it with certainty?

  63. It could just be that the moment of conscious choice acts to simply an internal emotional consensus which was already tending in one direction or another before the moment of conscious choice.

    It makes intuitive sense to me that between 60 and 80 percent of my decisions are not going against tendencies in my body which could be detected prior to the decision itself. But sometimes I make a decision which acts against the general feeling I have up to the time of making the decision, because I realize another set of facts which I was not considering are more salient and make me feel a different way.

    I think it would be a quite different thing to be able to predict a human’s decisions 70% than a 100% percent.

  64. Free will and the brain
    Self interest
    Where there’s a will there’s a way

    Dec 17th 2011 | from the print edition

    Who’s in Charge? Free Will and the Science of the Brain. By Michael Gazzaniga. Ecco; 260 pages; $27.99. To be published in Britain in April by Constable & Robinson; £14.99. Buy from Amazon.com, Amazon.co.uk

    ONE paradox has come under scrutiny since medical imaging of the brain became common in the 1980s. This is the apparent clash between the mechanical nature of the mind and the impression that people can will their own thoughts and actions. Michael Gazzaniga, a neuroscientist at the University of California, Santa Barbara, believes that “we are personally responsible agents and are to be held accountable for our actions, even though we live in a determined universe.”

    The idea that the mind exists separately from the body has a long history; Descartes invoked it during the 17th century. But in recent years opinion has shifted somewhat behind the notion that people are constrained by their physical embodiment. Individuals are suspected of being predisposed to eat compulsively or drink excessively, and various other characteristics are thought to have a genetic basis. Mr Gazzaniga reckons that, while such studies are useful, they give an incomplete picture of the true nature of humanity.

  65. NEURONS V FREE WILL

    The notion of free will is under attack again, this time from the advance of neuroscience. Anthony Gottlieb explains…

    From INTELLIGENT LIFE magazine, March/April 2012

    On the evening of October 10th 1769, in one of his typically curt dismissals of a philosophical problem, Dr Johnson silenced Boswell, who wanted to talk about fate and free will, by exclaiming: “Sir…we know our will is free, and there’s an end on’t.” Nearly two and a half centuries later, free will and responsibility are debated as much as ever, and the issue is taking some new twists.

    Every age finds a fresh reason to doubt the reality of human freedom. The ancient Greeks worried about Ananke, the primeval force of necessity or compulsion, and her children, the Fates, who steered human lives. Some scientifically minded Greeks, such as Leucippus in the fifth century BC, regarded the motion of atoms as controlled by Ananke, so that “everything happens…by necessity.” Medieval theologians developed a different worry: they struggled to reconcile human freedom with God’s presumed foreknowledge of all actions. And in the wake of the scientific revolution of the 17th century, philosophers grappled with the notion of a universe that was subject to invariable laws of nature. This spectre of “determinism” was a reprise of the old Greek worry about necessity, only this time with experimental and mathematical evidence to back it up.

    In the 20th century, the new science of psychology also seemed to undermine the idea of free will: Freud’s theory of unconscious drives suggested that the causes of some of our actions are not what we think they are. And then along came neuroscience, which is often thought to paint an even bleaker picture. The more we find out about the workings of the brain, the less room there seems to be in it for any kind of autonomous, rational self. Where, in the chain of events leading up to an action, could such a thing be found? Investigations of the brain show that conscious will is an “illusion”, according to the title of an influential book by a Harvard psychologist, Daniel Wegner, in 2002—a conclusion that has been echoed by many researchers since. In 2011, Sam Harris, an American writer on neuroscience and religion, wrote that free will “could not be squared with an understanding of the physical world”, and that all our behaviour “can be traced to biological events about which we have no conscious knowledge”.

    Really? There are now hopeful signs of what might be called a backlash against the brain. Hardly anybody doubts that the grey matter in our skulls underpins our thoughts and feelings, in the sense that a working brain is required for our mental life. This is not a new, or even a modern, idea: Hippocrates proclaimed as much in the fifth century BC. But there is a growing realisation among some neuroscientists that looking at flickers of activity inside our heads can be a misleading way to see how our minds work. This is because many of the distinctively human things that people do take place over time and outside their craniums. Perhaps the brain is the wrong place to look if you want to find free will. This is a theme of recent books by Michael Gazzaniga, a neuroscientist at the University of California at Santa Barbara, and Raymond Tallis, a retired British doctor and neuroscientist. As Dr Tallis puts it in his “Aping Mankind: Neuromania, Darwinitis and the Misrepresentation of Humanity”, trying to find human life in the brain is like trying to hear the rustle of a forest by listening to a seed.

  66. “All our behaviour can be traced to biological events about which we have no conscious knowledge; this has always suggested that free will is an illusion[…] The truth seems inescapable: I, as the subject of my experience, cannot know what I will next think or do until a thought or intention arises; and thoughts and intentions are caused by physical events and mental stirrings of which I am not aware.

    Many scientists and philosophers realized long ago that free will could not be squared with our growing understanding of the physical world. Nevertheless, many still deny this fact. The biologist Martin Heisenberg recently observed that some fundamental processes in the brain, like the opening and closing of ion channels and the release of synaptic vesicles, occur at random, and cannot, therefore, be determined by environmental stimuli. Thus, much of our behaviour can be considered ‘self-generated,’ and therein, he imagines, lies a basis for free will. But ‘self-generated’ in this sense means only that these events originate in the brain. The same can be said for the brain states of a chicken.

    If I were to learn that my decision to have a third cup of coffee this morning was due to a random release of neurotransmitters, how could the indeterminacy of the initiating event count as the free exercise of my will?”

    Harris, Sam. The Moral Landscape. p.103 (hardcover)

Leave a Reply

Your email address will not be published. Required fields are marked *