Jonathan C.W. Edwards
So It Was Microtubules After All?
Conference Report: Towards a Science of Consciousness, Helsinki 2015
I will start with some whingeing, but then rapidly move on to the good bits and finally to the very good bits.
Despite the admirable efforts and warm hospitality of Paavo Pylkkänen, Stuart Hameroff, and the organizing team, by the end of the second day I was sipping coffee with a delegate I met in Tucson in 2004 bemoaning the quality of what we had heard so far. Elder statesmen barely able to speak had given plenaries; postdocs had overrun time having forgotten to rehearse their papers; Hameroff, like Tigger, had, after each presentation, reduced all coherence to a wave function collapse comment (Bing!); we found ourselves asking the question — was it all getting a bit tired and repetitive? The free-for-all atmosphere of TSC is great, and better than the alternative, but is it getting threadbare?
By the fifth day it was clear that this was a misdiagnosis. TSC is a victim of its own success. Like all successful conference series it has got big, and at any big conference you expect to have to seek out the gems amongst a lot of stuff of more interest to someone else (at best). Some fine-tuning might not go amiss but we did not go home disappointed.
A lot of the good stuff was at the interface between fundamental physics, metaphysics, and philosophy of mind. Basil Hiley opened the proceedings with a keynote talk on weak values. What are weak values? I was not entirely sure either, but the gist of the talk was that physicists are beginning to understand and measure more subtle aspects of the mathematical structure of quantum systems and that it is interesting to see how these fit with the alternative mathematical formulations for which Hiley and Bohm are famous, in particular the concept of the Bohm potential. Two general messages for the non-specialist appeared to be confirmation of the complex yet indivisible nature of quantum dynamic units — of interest to those struggling with the binding problem — and further confirmation that quantum dynamics apply to large as well as small things. Hiley illustrated the latter with an experiment in which Buckminsterfullerine (C60) molecules were made to show quantum interference when passed through a tiny aperture.
A recurring theme linking this sort of fundamental physics to metaphysics and mind was the death of intuitive materialism as typified by the sort of causation that James Ladyman refers to as ‘microbanging’ (atoms colliding like billiard balls). There was some clearing of the air in two philosophy sessions in this regard. John Heil set the tone with a very entertaining talk opened with the comment that most philosophers, including himself, probably should not be philosophers because they are not in fact experts in the material they are discussing. Nevertheless, Heil ably demonstrated the emptiness of the traditional concept of ‘physical’, leaving the ‘hard problem’ with little foundation. He seemed to carry the entire audience with him until he ended with the comment that the redness of a tomato really is intrinsically in the tomato. Acting as audience spokesperson, Susan Blackmore asked him if he really meant that, since neurophysiology showed that it was wrong. Heil replied that Blackmore was wrong. Well, at least people seemed to agree with everything else he said.
Alyssa Ney and David Papineau gave two versions of the same talk (they declared this) on the difficulty that the causal closure of physics is no good as an argument for physicalism if cause has been abolished. A number of intriguing arguments were adduced but both ended up admitting that although they had no way to restore ‘causal closure’ they still thought physicalism was right. We were left with a slight sense that if a straw emperor has no clothes maybe it doesn’t really matter — or something to that effect. I was also a little disappointed by the lack of reference to various ways physicists, including Leibniz, have replaced ‘causation’ with something that seems about as good on the ‘closure’ front, just without the microbanging; or have suggested that the apparent time symmetry of fundamental equations is not necessarily a sound basis for abolishing a more general concept of sequence or becoming.
A further session dealt with structuralism directly, opening with a friendly but impassioned attempt by Philip Goff at demolition. Goff argued that structuralism fails to address experience and so is as incomplete a metaphysics as Galileo’s mechanics. He claimed that since Galileo the subjective aspect of metaphysics had been ignored, and just as Galileo moved on from the murky mysteries of the Scholastics it was now time to move on to a ‘post-Galilean metaphysics’ that brought subjectivity back into the story.
Ladyman, in typical fashion, refused to admit the claim of ‘touché!’. Rather like Eeyore in the episode when his tail was used by Owl as a bellpull, the response was ‘I really don’t mind’. Ladyman’s trump card was that if an anti-physicalist has trouble with materialism at least structuralism makes things quite a lot easier. He makes no claim to a fundamental level, and can point out that if nobody else has fitted subjectivity into a dynamic story, he cannot be blamed for not doing so. William Seager followed with some further nuances on the advantages and disadvantages of structuralism, but the main message seemed to be that structuralism is the most helpful position on physics for the philosopher of mind (since it is the position of physics) but cannot be expected to answer questions it does not set out to.
A further foray into the physics/philosophy debate was made by David Chalmers in a keynote talk devoted to, rather unexpectedly, collapse of the wave function (Bing!?). Chalmers presented with characteristic panache, but it was not terribly clear to delegates I spoke to how far he had moved things forward in physics or philosophy. Patricia Churchland gave a defence of a biological approach against the ‘bamboozling modal logic’ of the zombie conceivability argument. Again it was an excellent piece of theatre but Chalmers didn’t really seem to mind either, maybe because things have moved on.
Maybe the philosophy was just turning the usual handle, but I think there was a useful sense that various ghosts were finally being laid to rest, leaving us with the opportunity to address the problem of consciousness in a scientific way, but hopefully with a breadth of context that could kick-start Philip Goff’s post-Galilean comprehensive metaphysics (even if Descartes and Leibniz had made a promising start 300 years ago, but others had chosen not to follow).
So, what about the very good bits?
I was never very impressed by Christof Koch’s claim that the hard problem would solve itself once the ‘easy’ biology had been worked through, but I do have a feeling that the science may be homing in on the right place to slot in the metaphysical considerations.
One particular concurrent session presentation I would single out as an hors d’oeuvre to the gourmet section was that by Ausaf Farooqui from Cambridge.
Jean-Pierre Changeux had, earlier in the programme, described work done by Stanislas Dehaene and colleagues suggesting that visual consciousness is characterized by a particular pattern of electrical activity that spreads forward from occipital lobes to produce activation of frontoparietal regions. Subliminal presentation of target images produce a local activation at the back of the brain but when images are reported as consciously seen, firing increases further forward in a particular way. However, Farooqui had done the obvious control experiment in which subjects were presented with both ‘target’ images (numbers) and ‘irrelevant’ images (letters) in a sequence much in the way that target images are presented in sequence with prior or later ‘masks’. What he found was that, although subjects had no trouble reporting having seen the ‘irrelevant’ letters, their presentation was only associated with local activation at the back, whereas the ‘target’ numbers produced activation at the front.
The clear conclusion from this study was that the spreading activation in frontoparietal areas often said to be a hallmark of consciousness could not be. It was much more likely to be a hallmark of some thought process related to being a subject in a psychology experiment concentrating on carrying out a predefined task. The activation in the front was not a sign of conscious seeing. Nor was it even a sign of attention, because the subjects attended to the irrelevant letters at least for long enough to work out they were not numbers. An additional experiment using null images further emphasized this conclusion — that the activation in the front had more to do with thinking about doing a job than it did with seeing anything.
The implications of this study would seem considerable. Claims that distributed activity in frontal and parietal lobes might be pretty close to the neural correlates of consciousness in the desired sense of ‘where the experience is’ look to be premature at best. The options are still open. Maybe it is in the thalamus after all. Maybe it is not distributed in that sense. It looks as if a different methodology is needed to find the NCC.
That methodology could be anaesthesia after all — the most obvious way to intervene in consciousness. Travis Craddock reviewed the evidence for mechanisms for anaesthetics. In recent years emphasis has moved away from thermodynamic or colligative property effects of lipid soluble molecules in biomembranes to a presumption that specific interactions with hydrophobic regions of proteins are involved. The evidence here is, however, still conflicting. Solubility in lipid correlates extremely well with anaesthetic concentration for a remarkably wide range of molecules. Enantiomers (right- as opposed to left-handed versions) of certain anaesthetic molecules are not anaesthetic and this runs against a simple colligative property effect, yet the lack of structure activity relationship otherwise, together with the suggestion that it does not terribly much matter which way round anaesthetics ‘hide’ in hydrophobic pockets, continues to weaken the argument for a fine-grained steric mode of action. The jury may still have to be out.
This continuing uncertainty notwithstanding, there are now clear indications that certain anaesthetics can interact with tubulin molecules in microtubules. One might say ‘not microtubules again’, but maybe there always were good reasons for taking microtubules seriously. Microtubules form the skeleton of the dendritic tree and they are anchored directly to receptors for neurotransmitters such as GABA. Altering microtubular function would be very likely to alter the fine-tuning of any global dynamics of a dendritic tree, just as much as the membrane.
UnCheol Lee then described how in terms of effects on brain-wide neural activity there are important differences between anaesthetic agents and between effects on different areas. It seems that anaesthetics do not abolish neural electrical activity in a blanket fashion. Cortex shows reduced activity more than other areas. But most specifically the action of anaesthetics seems to correlate with an inhibition of feedback activity from frontoparietal to primary sensory areas. The basic message seems to be that anaesthesia is achieved by disrupting coordination of neural activity rather than simply abolishing it. That might fit with some sort of ‘detuning’ action.
From my perspective much the most exciting material at the conference was that relating to microtubule electrodynamics. Cameron Keys presented data published in a series of recent papers by Bandyopadhyay’s group. The key finding was that whole microtubules appear to offer no more resistance to electrical current than individual tubulin molecules under certain conditions, in particular that of oscillating current with a frequency in the megahertz range. The implication is that an unprecedented form of electrodynamic behaviour is manifest. In addition, Hameroff’s group have found that ~megahertz ultrasonic vibrations applied to the frontal cortex can alter mood.
So far it remains very unclear how high-frequency electrical and mechanical oscillations might be involved in consciousness, but it raises the possibility that some form of electromechanical coupling (piezoelectricity is a familiar but not the only form) could be involved. Energy from post-synaptic potentials could be converted into elastic/ acoustic energy and this could then in return influence the integration of potentials into somatic spikes by a mechanical effect on stretch sensitive ion channels. Complicated integrational or ‘computational’ events might then be considered to occur within the microtubule structure itself. We do not know what sort of computational maths might be involved and whether this needs to invoke entanglement (I am personally sceptical). But if modes in microtubules are quantized this might help solve the binding problem in that all potentials entering the dendritic tree would be co-available to an extended resonant microtubular mode.
The paradigm shift for me is the recognition that microtubules are likely to be just as integral to the electrical computational role of dendrites as the membrane. Fröhlich (1968) focused on membrane, suggesing that high-frequency oscillatory modes might occur in the lipid bilayer of the dendritic membrane through electromechanical coupling of dipoles. Anaesthetics were for a long time thought most likely to act on the membrane itself. However, the intimate anchorage of microtubules to membrane may suggest that the mechanical modes involved in any coupling may depend chiefly on the properties of the microtubules for parameters such as frequency, wavelength, and velocity of propagation. If anaesthetics ‘de-tuned’ those parameters then it would be reasonable to expect stable conscious thoughts to dissolve if such thoughts depend on synchronized feedback loops for their integrity (or at least reportability), as seems very likely.
So perhaps it was microtubules after all, but maybe the precise physics is yet to be established. Even so, we may be a bit closer to resolving some of the philosophical issues if it transpires that consciousness of the human sort arises in dendritic electro-acoustic modes that may be very unusual in their form. Although these would be within physics, maybe it is not quite as absurd as everyone assumes to consider there to be ‘special life forces’ involved. Higher-order electro-acoustic forces may not belong to the ‘fundamental four’, but their quantized modes can be just as real. Moreover, it may be that higher-level electromechanical effects in microtubules represent subspecies of force that are more or less unique to living organisms and which are ‘not yet within physics’ in the sense that nobody has written the equations yet. That might make life easier for the physicalists who find indiscriminate panpsychism hard to swallow. And since such modes would relate indivisibly and have no ‘extension’ in Descartes’ strict antitypic sense, maybe all he really got wrong was suggesting the pineal instead of the thalamus or cingulate.
Next year we may be able to get going with what Philip Goff wanted.
Fröhlich, H. (1968) Long range coherence and energy storage in biological systems, International Journal of Quantum Chemistry, 2, pp. 641–649.