Is quantum uncertainty necessary?

I just read a post (here) called “We Seem to be Zombies” by Stuart Kauffman. It repeatedly annoyed me. I try very hard not to write angry postings but really this was just more than I could stomach. I apologize to those readers who dislike angry posts and I will still try to avoid them in future.

So here is a list of annoying bits.

  1. The introduction starts with this picture: “according to some of our finest scientific minds, you are not conscious at all. You are a mechanical zombie, a calculating-machine idiot. You have no responsible free will. You are not even an ‘I’, the subject of experiences that you naively have the illusion you have. You cannot act responsibly because machines are machines, you are a machine, and like a marble rolling down the side of a bowl, mechanisms ‘happen’ they do not act. So you cannot act. You cannot be a autonomous agent blazing with life, its anguish and joy”. I would like to see the names of a handful of ‘our finest scientific minds’ who have said such things. The impression is given that this exaggerated picture is accepted science in some quarters – why not name them? My impression is that most scientists, especially those with fine minds, believe that we are living organisms not mechanical machines and that we are aware not zombies. Why this unrealistic characterization of science?

  2. After a historical look at dualism he says: “Virtually all contemporary thinkers on the subject assume that mind and brain are identical.” You would assume that he included himself in contemporary thinkers. But, no, a couple of sentences later, he reverts to a dualist mind set, mind and brain are now not identical. “if the brain is a deterministic system, like the billiard balls, the current state of the brain is entirely sufficient to determine precisely the next state of the brain. Woefully, there is NOTHING FOR MIND TO DO! Worse, there is no way for MIND TO DO AYTHING TO BRAIN ANYWAY.” The capitals are his. This does not sound like a description of a brain and mind that are identical. It sounds dualist. There is an ‘if’ at the start so it is just as possible that he is abandoning a deterministic system because he thinks it leads to dualism. Is it identity of mind/brain or causality that he is questioning? Why is this not clear?

  3. Ah, here comes the clarification. It is a confusion of mind and consciousness. “Maybe mind really is an illusion? Maybe it is an epiphenomenon?” What is the definition he is using for this ‘mind’ that can at the same time be: identical to the brain, have nothing to do while the brain is ‘doing’, and be an illusion. Come on, is he talking about consciousness or is he talking about thought (perception, cognition and action)? He is not communicating what he means to me – is he talking down to me or is he trying to confuse me with a bait and switch?

  4. Now is a history of positivism and its ilk which he shows is as much of the dead end as dualism. Then he goes on, as if positivism was alive and well, to look at computer artificial intelligence and finally the Turing machine. All this is outlined as if it is relevant to the subject of our brain/minds and not just fit for symbolic logic or digital computers. In capital letters, we learn that a Turing machine and a very particular type of artificial ‘neural’ network are absolutely defined, algorithmic and understandable in terms of entirely classical physics. So what? He says that this has led to the belief in computer science, cognitive science and neuroscience that “the mind MUST BE algorithmic” and “IF consciousness is real, consciousness will somehow emerge when sufficiently many computer chips are coupled together.” This is a funny way to abandon the spirit of positivism. For heaven’s sake, who are these people who believe that the brain works like the most mundane (or universal) computer imaginable or works like the first try at artificial intelligent networks? I am sure there are scientists who believe that the brain is probably algorithmic but how many believe it must be? The comparison of brain/mind and computer is a metaphor not some strict specification and most scientists know that. Why am I being led down the garden path to think the brain is algorithmic?

  5. So now we have the key statement: “In the world of classic physics, the only hope its proponents hold out for consciousness so that we are not zombies is that a sufficiently complex network of calculating gadgets will ’emerge’ into consciousness.” So now I am asked to accept two ideas without the help of either logic or evidence: (a) what has just been described in the terms of brain-being-computer is the sum total of what classic physics has to offer. Does biochemistry, biophysics, physiology etc. now require something more than the physics it has always used to underpin chemistry? What is it? I have to assume it is the uncertainty principle in quantum physics the only part of quantum physics that does not usually figure in chamistry; (b) mind has now become consciousness. Where did the other mind go, the one that was identical to the brain – the one that is perception, cognition and action? I insist that I will not be dealt some slight of hand that confuses thought with awareness of thought. That is like confusing the actual tree with my awareness of the tree. It will not do. And how did non-algorithmic brain/mind come to be impossible by classic physics?

  6. Next is the promise that future postings will explain synapses in terms of coupled Trans-Turing Systems. I will probably read these postings but I am not hopeful that they will be very enlightening and I fear that they will continue the unconvincing paths of this posting.

To set the record straight I will state clearly what I think is likely so. The brain is physical (as in matter and energy with no supernatural, non-physical or magical bits). The brain has biological function and its function is perception, cognition, learning, memory, action (muscular and glandular), conscious awareness and perhaps other associated processes. We have a word for this biological function, it is ‘mind’. Brain does mind in the same way as heart does circulation or stomach does digestion or lung does gas exchange. Brain is a living organ and mind is the function of that organ. Mind does not exist in the sense that the brain exists, it exists like circulation exists. Mind is not res cogitans; it is not ‘Res’ anything because it is not a thing, not an object. There is no shred of evidence that the brain does its mind function using algorithmic calculations and at least some evidence against it. Why the assumption that thought must be a fixed, stepwise procedure. There is no reason to think that the functions of the brain give predicable results in the sense of actually being able to predict them. It does not matter whether the attempt to predict uses classical physics or not – it is not in practice possible and therefore it doesn’t really matter if prediction is theoretically possible. Who cares? We can assume that consciousness is an important part of the function of the brain because it is biologically expensive. We can assume that it interacts with other aspects of the function of the brain (ie consicousness is part of mind but not a large part of it and is integrated with other parts of mind – mind being what brain does). There is nothing in this that makes us non-autonomous, machine-like things or zombies. There is nothing here that gives us freewill either, if by free we mean free from physical/material constraints. Our brain/minds make real decisions, and act on them and are then responsible for those decisions – the decisions are neither free or un-free, they are not determined or un-determined. Freewill and determinism is a dichotomy that is fictitious – neither term is true or false – but instead they are meaningless – they are not possible. They are silly answers to a silly question. It does not seem useful to bring the uncertainty of very, very small things into the discussion of processes that scan of few inches, produce electrical fields that can be measured through the skull bone, weighs many grams, and use the energy of a fifth of what we eat. It may turn out that we need quantum uncertainty in order to explain the brain/mind and if we do than so be it. If it turns out that an explanation of the brain/mind needs the uncertainty principle, that still will not give meaning to either freewill or determinism. Why?

Because if we start with the ridiculous assumption that the mind makes decision which it then imposes on the brain, we still have the problem of exactly how the mind makes decisions. Does the mind have a process? Does the mind have a mind in an endless regression, or is it random, or is there a magical way?

Postscript: Still feeling that I should not be so annoyed by Kauffman’s blog, I have set this aside and waited for his next two posts on the subject. The first read almost identically to the ‘Zombie’ blog. The next blog dealt more with the physics. I am still annoyed after a suitable wait and so I am posting this.

2 thoughts on “Is quantum uncertainty necessary?

  1. Janet,

    Despite being an ‘angry post’ I enjoyed reading it. I think you offered some cogent criticisms of Kauffman’s provocatively titled article.

    Although I followed, and found sympathy with, almost all of the post, there was a point in your concluding section where I didn’t see a connection. You wrote:

    “There is no shred of evidence that the brain does its mind function using algorithmic calculations and at least some evidence against it.”

    My understanding is that this is not the point behind the theory of universal computation. It is not a consequence of the theory that the brain processes information in a “fixed, step-wise procedure”, just that all real-world, deterministic systems are, in principle, simulable by a Turing machine. Whether in practice predictions can or cannot be made is, as you rightly say, irrelevant (the resource and time burdens of some simulations will render them prohibitively expensive) – but the fact remains that the brain could be simulated by a Turing machine.

    This point seems relevant to me when, a bit further down, you say:

    “We can assume that [consciousness] interacts with other aspects of the function of the brain (i.e. consciousness is part of mind but not a large part of it and is integrated with other parts of mind – mind being what brain does)”

    Which I understood; then you wrote:

    “There is nothing in this that makes us non-autonomous, machine-like things or zombies.”

    Let me explain why I don’t follow this.

    If a brain is simulable, then it is deterministic (in principle, if not in practice!). If the brain is deterministic, then the mind is deterministic. If the mind is determined then certain definitions of freewill are impossible. So I think that it does have a bearing on the issue after all. (As it happens, I think that freewill and determinism are compatible notions after all, but the debate hinges on the definition of ‘freewill’ in the first place.)

    I would be interested to hear your response.

    Kind regards,


    JK: Thank you for the comment. It is appreciated.
    You mention 3 ideas I put forward that you are not convinced of – in fact, that you find nearly impossible. (1) the brain is not algorithmic (2) the brain cannot be stimulated with a Turing machine (3) freewill and determinism are incompatible and both flawed.
    I will give my thoughts on (1) slowly. When you have a feedback loop with some inputs, it has a stable point in which it can sit without change and when there is a change in inputs, it snaps to the new stable point in a very short instant. If it has components with momentum, memory, springiness, capacitance, inductance or other time sensitive components , than it rattles a bit before becoming stable. There is nothing procedural (algorithmic) about this behaviour – it is not an iteration or the like any more than an op amp is. One thing we know about the brain is that it is massively parallel. There are millions of parallel feedback loops between cells in the thalamus and the cortex, and the loops overlap somewhat. Where ever an axon goes from the thalamus to the cortex another axon goes the other way. This is a mind-blowing feedback system and for any set of constraints (input) there will be one or very few stable points. There are other similar parallel sets of loops between other pairs of structures in the brain. In fact it is almost a rule of thumb for the brain that if there is a path from a to b there will be one from b to a. Further there is electrical and magnetic field feedback where the fields are formed by the neuron activity and in turn affect that activity. There is feedback resulting from chemical environments too. I see no reason to doubt that the brain can quickly and without procedure find the best-fit scenario for a given set of inputs. I believe that there may be trivial tasks that the brain uses procedures to solve, especially where the architecture of the brain has developed to solve a very particular problem. But I do not believe it is the main way that the brain operates. Besides the architecture pointing to what is in effect a very large analog system not capable of step-wise operation, there is also the question of time. Neurons are slow and they could not do what they do in the time they do it in using step wise procedures. Finally, although neurons firing or not firing seems digital in nature, the brain is far more analog then digital. Each synapse is analog, the integration of all the synapses on the dendrite branches is analog, the action of the glial, chemical and electromagnetic environments is analog.
    And as for (2) if you accept that the brain is not algorithmic and not digital then there is no reason to think that a Turing machine can simulate its action. There are simulations that use approximations that might be possible, along the lines of Blue Brain once it is more then a single cortical module and more like a whole brain. But this approximation is not what is meant by being in the family of computers that can be replaced with the universal computer. We have a metaphor here rather than a description or a definition.
    I do not basically disagree with you on freewill (3). It really depends on how freewill and determinism are defined. I am more comfortable with their hard philosophical definitions (ie predicable materialism or magical non-materialism) and I reject both. Others may use a different meaning for one or both and then accept both. What bothers me is the continuation of the opposition of the two and the attempt to force people to choose between them.

  2. <p>Janet,</p>
    <p>Thank you again for taking the time to write such a detailed reply.</p>
    <p>In your reply to point (1) you give a very lucid and detailed description of the brain as being a massive dynamical system. This, surely, is the most powerful descriptive paradigm we have for making sense of its structure and operation. However, I think that the point I was trying to make about Turing machines still stands. I shall try to explain it slightly differently, and see if I can make myself clearer.</p>
    <p>The point behind universal computability is that even the most massive, parallel, dynamical system can be simulated by a Turing machine of sufficient size. A simple dynamical system, for example the Lorenz oscillator, can be simulated on a home PC. The Lorenz oscillator is dynamical and chaotic, but can be simulated perfectly. Next, consider the 3-body problem: it too can be perfectly simulated. With a supercomputer one can scale up the 3-body problem, adding body after body to the total system. There is no theoretical impediment to simulating the n-body problem perfectly.</p>
    <p>The brain contains 10^11 neurons. I can’t even begin to estimate how many atoms it contains, but, assuming that brain behaviour is independent of sub-atomic effects, then simulating every atom in the brain (and therefore neural feedback loops, gas diffusion, synaptic junctions, magnetic fields and all!) is a theoretical possibility.</p>
    <p>What would result from such a simulation? Well, in my (materialist) opinion, nothing less than a brain…</p>
    <p>I think that this thought experiment tells us two things. </p>
    <p>The first is that the brain is in essence a pattern, independent of physical implementation. It might well turn out that in practice, it is impossible to implement a brain using a serial Turing machine, because the algorithmic complexity of the n-body problem grows exponentially with each additional body. There might just not be enough time or enough molecules available to do it. If this is the case then the brain will be, for all intents and purposes, algorithmically incompressible (this is one of the criteria used to measure complexity)</p>
    <p>The second is the point I was originally trying to make: whether a brain can be simulated in practice is irrelevant; because in theory it has the property of being simulable by a Turing machine, and this means it is deterministic.</p>
    <p>I ought to reiterate the additional premise I smuggled in above: that the function of the brain is independent of sub-atomic effects. If there is genuine indeterminacy at the sub-atomic level, then no Turing-computability results hold.</p>
    <p>Just a quick note regarding point (3). Many compatibilist philosophers have proposed other notions of free will so I wouldn’t say that there was a single, philosophical definition of it. (Take, for example, the work of Harry Frankfurt, who identifies it with ‘second-order desires’.)</p>
    <p>Kind regards,</p>
    Will, Good reply.
    I think our differences are basically just semantic.
    If we (1) worked out an equation for the whole brain and (2) if that equation is solved with a Turing machine that carried sufficient significant figures in its operation and (3) if an instantaneous starting point could be measured for a brain ‘state’ and its environment – than it would be theoretically possible to solve for the next instant with a Turing machine. It would probably take more time and material than we have available to do any one of the three. I think this bends the idea of the Turing machine and simulation to its breaking point. We agree that there is no non-material aspect to the brain but that it is not predicable in practice.
    However it is really the definitions of determinism and freewill that we differ about. I do not want to choose – I don’t want to make them compatible – I want to abandon the concepts and the question. My reason is because they carry too much awful baggage and are dangerous to communication.
    You raise the question of what are brains. I have a very biological viewpoint. I find it difficult to call something that is not part of a living organism a brain. My way to try to understand consciousness (the subject of the blog) is very biological and any other description (for example a purely mathematical one) would not be as satisfying. It would have to be translated into the biochemistry, biophysics, physiology, genetics, developmental biology, evolutionary biology etc. and then I would feel I understood. So your statement ‘the brain is in essence a pattern, independent of physical implementation’ is true enough if that is the sort of brain you are thinking of – if that is how you use the word brain. I am interested in Blue Brain for what it can illuminate about the biology of the brain but I don’t think of it was a little piece of actual brain to the extent that it shares a pattern with a similar bit of biological brain. Anything that doesn’t ring true biologically just does not ring true to me, period. I find it hard to imagine a brain without being connected to the rest of the nervous system, the muscles, the glands, the sense organs, in fact, the whole body. I like the ’embodied mind’ concept for its obvious simplicity. Again this is a question of semantics. We do actually make decisions and it is a scientific explanation of the mechanics of decision making that interests me and that I look forward to finding.

    Thank you again for your reply Janet.

    I think you are correct: there is no fundamental incoherence between our positions.  It seems to me that any useful account of free will will be biologically driven and nuanced but I still use the concepts of freedom and determinism as ‘stakes in the ground’, points of reference from which a proper theory can be distinguished.

    With regards to the third point: I also find it extremely difficult to think of a brain as anything other than an embodied, situated bit of ‘wetware’, its just that, because its simulable, logic tells me that it’s fundamentally a pattern.  It does raise an interesting point though, which is this: if a scientific description of a phenomenon is so divorced from your mental conceptions of it, how useful is that description, really?  I haven’t got as far as answering this one for the brain-as-pattern picture.

    I am very much enjoying your blog.  Please keep up the excellent work.  It is much appreciated.

    Kind regards,


Leave a Reply

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