Dembski, Decoherence and the brain

Over at Dembski’s blog you will find him commenting on neuroscience.

Dembski wrote:

My good friend and colleague Jeffrey Schwartz (along with Mario Beauregard and Henry Stapp) has just published a paper in the Philosophical Transactions of the Royal Society that challenges the materialism endemic to so much of contemporary neuroscience. By contrast, it argues for the irreducibility of mind (and therefore intelligence) to material mechanisms.

Unfortunately for Dembski, this is completely wrong. The paper, “Quantum physics in neuroscience and psychology: a neurophysical model of mind-brain interaction” Jeffrey M. Schwartz, Henry P. Stapp, Mario Beauregard, Philosophical Transactions: Biological Sciences, 2005 argues for a quantum mechanical approach to the problem of mind-brain interaction. Quantum mechanics may seem really weird to the non-physicist, and involve things like “spooky action at a distance” but quantum mechanics is part of the material world in the sense that both scientists use it and Schwartz et al., are using the this paper [1].

Schwartz wrote:

…brain is made up entirely of material particles and fields, and that all causal mechanisms relevant to neuroscience can therefore be formulated solely in terms of properties of these elements. (Emphasis added)

An electron is no less material in quantum mechanics for it being described as a probability distribution. What Schwartz et al. are arguing for is a non-mechanistic description (in the classical physics sense) of mind-brain interaction, not a non-materialist one (in Dembski’s sense). Furthermore, it is not “irreducible” in Dembski’s sense either.

Now, a few caveats. Firstly, I’m a neuropharmacologist, I grow pretend nerve cells in dishes and try to unravel the molecular basis of nerve function and survival. So I’m at the “reductionist” end of the neuroscience spectrum (and the Paleyists have a thing about reductionism) and my comments on psychology are those of an interested lay-person. On the other hand I’ve spent a lot of my professional career working on neuronal calcium channels in one way or another, so when I talk about ion channels, it’s in a professional capacity. Secondly, the mind-body problem is hard; really, really hard. And there have been no end of books by eminent philosophers and neuroscientists on it (see the end of this post for some suggested reading). We are far from understanding the biological basis of consciousness, and it is one of the top 25 questions in the journal Science’s 125th anniversary issue.

However, there is a general consensus that the “mind” [2] is intimately associated with the brain. Brain damage affects the mind. Stroke can affect personality, the ability to associate words with images. Brain tumours can induce extreme personality changes that are reversed when the tumour is removed. A wide variety of drugs, acting solely on brain structures, influence our minds. What is contentious is whether the “mind” is solely generated by the brain (either directly or emergently), or whether “mind” exists in some sense separately from the brain. Also, how can the “mind” influence the brain if it is a construct of the brain?

The latter seems to be the starting point of Schwartz et al. They observe that people can be trained to regulate their emotions or overcome phobias. “Change the mind and you change the brain” is the title of a paper from one of the authors. One of their claims seems to be that as “mental effort” is experiential and cannot be described exclusively in material terms, we cannot use “classical physical” explanations to describe how “mind” can change the brain. Thus they turn to quantum mechanics. This is not new; Roger Penrose articulated a quantum mechanical view of consciousness some time ago. However they have looked at a quantum mechanical description of brain action in more depth than Penrose did.

I have two issues with their approach. Firstly, there is no need for some new principle to describe what happens when people learn to overcome social phobias. The key is that it is learning. We have known for a long time that learning changes brain structure. Nerve firing rates are changed through use dependent changes in nerve chemistry, new connections between nerves are forged and existing ones re-enforced. New firing circuits are produced. This happens in all learning, from unconscious learning of motor skills to conscious learning of more complicated cognitive tasks. Directed attention of the sort used in the phobia paper is also seen in non-human primates as well: it is not only a human domain. Interestingly, there is a type of mental retardation associated with fragile X syndrome that involves “executive attention”, one of the processes Schwartz et al. talk about. A single gene disorder, it causes these people’s brains to be normal
macroscopically, but with fewer nerve connections than normal. Learning has a basis in remodeling the brain.

Learning in stroke patients can produce new nerve pathways to replace the damaged ones and restore some degree of function. This doesn’t require quantum mechanics to explain, so why should learning which circumvents phobias be any different to learning that circumvents stroke damage? And it is learning. In the spider phobia paper, people are repeatedly exposed to spiders in an environment where they learn that spiders can’t harm them. The title of the paper should have been “Change the brain and you change the mind”.

Secondly, the way they describe the quantum mechanical processes in the brain is problematic. For nerve cells to fire, calcium must enter the nerve cell in response to a stimulus. They correctly state that the ion channels that let calcium into nerve cells are very narrow (0.086-0.158 nm). They claim this will result in a calcium ion to be laterally confined, so that its velocity must become large by the quantum uncertainty principle, a cloud of probability spreads out from the ion channel. The spreading of the ion wavepacket means that it may or may not interact with the calcium-binding proteins that will result in neurotransmitter release, which will mean that the nerve may or may not fire and so on until the brain is one mass of probability and requires a quantum process to collapse it.

There are a number of problems with this. Firstly, lots of excitable tissues have narrow calcium channels and multiple connections. Exactly the same process occurs in the heart, where clouds of ions spread out from a calcium channel until a large mass of cells are firing, the same goes with blood vessels. We need no appeal to quantum mechanics to understand the heart beat, so why is the brain in principle any different (the brain will have more quantum superimposed states, but the heart will have several billion as well). Schwartz et al. claim there is a minimum complexity where quantum effects will begin to dominate, but don’t provide an indication of what this minimum size is. We can have a stab at it by looking at the minimum brain size in a conscious organism. New Caledonian Crows are tool makers and users. When presented with a unique problem, they can create a new tool to help them solve it. By all definitions of the word consciousness, New Caledonian Crows are conscious entities like chimps and us.

Yet they have a brain the size of a walnut. So Schwartz et al.’s quantum processes must take place at these levels of cell number and connectivity. This means that the quantum mechanical processes do occur in the heart if Schwartz et al’s interpretation is right. They will also occur in the enteric nervous system, a thick layer of nerves that lie between two muscle layers in the gut. Highly branching and interconnected, the enteric nervous system has been termed a “second brain”. The quantum processes that underlie Schawrtz et al.’s model of mind-brain interaction must underlie enteric nervous system-gut interaction.

The other implication is that these processes are not confined to humans (pace the New Caledonian Crows above), and must apply to monkeys, marmosets and mollusks. So the quantum mechanical processes cannot be an “irreducible” barrier between humans and animals, as Dembski hopes. Furthermore, it is not “irreducible” in Dembski’s sense. In Schwartz et al.’s model, the conscious mind-brain interaction is an emergent property that occurs when the number of connections are high enough for quantum properties to dominate. Biologists are perfectly happy with emergent phenomena, and connectivity-related emergence has been suggested to explain brain phenomena before.

Finally there is the problem of quantum decoherence. Schwatz et al. largely dismiss it.

Schwartz wrote:

The brain matter is warm and wet and is continually interacting intensely with its environment. It might be thought that the strong quantum decoherence effects associated with these conditions would wash out all quantum effects, beyond localized chemical processes that can be conceived to be imbedded in an essentially classic world. Strong decoherence effects are certainly present, but they are automatically taken into account in the von Neumann formulation employed here. ….

I think the decoherence effects are a lot stronger than they suspect. A calcium ion has to run the gauntlet of many, many molecules before it reaches a binding site, it repeatedly bounces off water molecules and protein molecules. If there is any meaningful quantum effects left by the time calcium binds to synaptogamin, I’d be very surprised. I’ve measured calcium transients in nerve cells (and so have many other people), the spread of the calcium in the nerve terminal is at the standard diffusion rate, so it looks like the quantum effects have been largely removed (allowing of course for the fact that we are observing these systems, which collapses their quantum properties). Also, Schwartz et al. talk of a single ion channel and a single calcium ion and a single calcium binding target. But in realty in a single nerve terminal, there are many ion channels that will be activated, letting in many calcium ions (typical nerve terminal concentrations of calcium during a nerve impulse is around 100 nM), which will bind to many binding sites. The statistical effects of these many interacting calcium ions should wipe out any quantum indeterminacy.

There are many aspects of this paper that don’t seem to hang together for me outside of the issues outlined above. However, I must emphasize again that I am a neuropharmacologist, not a physicist (I don’t even play one on TV). Even though I have forced my way to the end of both “The Emperors New Mind” and “Shadows of the Mind” my grasp of quantum mechanics remains very basic.

But the main issue is that, even if Schwartz et al. are completely correct, this is still a physical theory, and is still “materialist” in the sense that scientists use the word.

To summarise: 1) Schwartz et al.’s model is a materialistic model; it uses a quantum mechanical rather than a classical approach, but it is no less materialistic for that. 2) Schwartz et al.’s model applies to all large concentrations of interacting, excitable cells, not just conscious brains. Consciousness is not unique in this model. 3) Schwartz et al.’s model applies to conscious non-humans. It provides no distinguishing barrier between humans and non-humans. 4) Schwartz et al.’s model is not “irreducible” in Dembski’s sense, it is a version of emergence. 5) It is not clear if Schwartz et al’s model is really needed to explain the phenomena they need to explain.

[1] It needs to be noted that there are several different interpretations of Quantum Mechanics. The most familiar will be the “many worlds” interpretation. Another common one is a Bayesian statistical approach. The interpretation used in this paper is Stapp’s own, and is not very widespread.

Further reading: The Skeptic’s Dictionary entry on Mind Blackmore, Susan. Consciousness: An Introduction (Oxford University Press 2003). Dennett, Daniel Clement. Consciousness explained illustrated by Paul Weiner (Boston : Little, Brown and Co., 1991). Penrose Roger, The Emperor’s New Mind: Concerning Computers, Minds and the Laws of Physics, (Oxford Paperbacks. 1997)

Acknowledgements: Many thanks to the Panda’s Thumb crew for help discussion, particularly Erik for helping me with some Quantum Mechanical concepts.