Mind-body problem: New approaches

• John G. Taylor, King's College London

John: You need 1- or 2-sentence definition here. I have put one at the beginning of section 1 - why up before the introduction?

A New Look at the Mind-Body Problem

The mind is composed of mental fragments - sensations, feelings, thoughts, imaginations, all flowing now in an ordered sequence, now in a chaotic fashion. There are also non-conscious components involved in early brain processing of stimuli (as in lower level processing in vision, such as in V1) or in emotions not yet in consciousness. On the other hand the body is constructed under the underlying laws of physics, and its components obey the well-enumerated laws of physiology. The Mind-Body problem has been in existence for several thousand years - going back to Plato, Aristotle The Buddha and many other ancient Greek and Eastern thinkers. The problem is simple to state: the mind and the body seem to be entities of very different kinds (as their earlier descriptions indicated), so how do they interact so as to produce in a person a mind able to have effects on their body (as when the person wills the body to perform some act), whilst also their body can affect their mind (as in the experience of pain)? Although the problem is simple it has as yet no satisfactory solution.

Much has been written on the variety of solutions to the mind-body problem. There are the dualist solutions (mind and body are distinct, although then the problem of how they interact becomes even more embarassing, and is so far unresolved in any satisfactory manner); the idealist solution (there is only mind, and body/matter is merely a manifestation of mind, although how mind has created the beautiful subtlety of the unified forces of nature of U(1)xSU(2)xSU(3) form, with a proposed extension to superstrings, is far removed from such a scenario); a third position is that there is only body, so this is a reductive physicalist approach (but then how the amazing mental world full of our experiences is thereby created from matter has to be faced). There are also nuanced versions of one or other of these three initial positions for solving the mind/body problem, some having great sophistication and subtlety in themselves.

There is also the question of what exactly is the mind? It is certainly composed of conscious components, but it would also seem to contain non-conscious ones as noted earlier, such as unconscious emotions and low-level processing as yet out of consciousness; automatic motor responses are also below the radar of consciousness. For some, these unconscious or pre-conscious components are not problematic, since they do not seem to acquire the expanded non-spatial aspects of conscious thought and experience that the conscious components appear so to do. These non-conscious processes can thus be more readily accepted as arising purely from suitable brain processing, so more easily understood as components of the body (especially with the great advances in brain science tracking down and modelling in detail this pre-conscious neural activity). More especially these non-conscious components of mind obviously do not have any 'consciousness' component, and so there is no difficulty in expecting them to arise solely from brain activity: they can therefore be seen as on the body side of the mind-body duo. Thus it is possible to consider the inexplicable part of the mind - that part to which the epithet 'mind-body problem' most closely applies - to be that of consciousness. I will take that position here because it addresses what appears to be the most difficult part of the overall problem of mind-body interaction: how conscious experience can interpenetrate and fuse with bodily activity to affect each other in the way mentioned in the first paragraph. Detailed neural models of the other (non-conscious) components of brain activity are increasingly convincing, for predictive value coded by dopamine, for motor responses guided by internal motor models, and for early models of visual and other sensory processing through a hierarchy of increasingly complex feature detectors; they clearly support the lack of any 'nonconscious mind'-body gap. In support of the approach used here is that it is more in line with modern thinking on the mind-body problem, especially that brought to the fore by Descartes and since emphasised in the notions of the 'hard problem' and the 'explanatory gap'; these specifically emphasise the gap between consciousness and brain activity.

Beyond restricting the notion of 'mind' to 'conscious mind', I also propose to take an approach to the mind-body problem that does not list all of the various phsilosphical solutions (dualism, monism, idealism, interactionism, etc, etc - there are many places to find the standard list trotted out and expounded) and explore them in toto, but instead to attempt to relate the problem to the most recent developments in science. Only in this way, I suggest, can real progress on this difficult problem be made. Only then can we attempt to refine models by testing against their predictions in the real world, and thereby narrow down the possibilities. hopefully ultimately approaching a solution to the mind-body problem.

The New Knowledge Base

The knowledge base we can bring to attempt to solve the mind-body problem has changed enormously over the hundreds of years of the problem's existence. Matter has been probed down to very short distances through the development of ever higher-energy particle accelerators, so that we now have an understanding of matter in terms of the gauge theories unifying the electromagnetic and weak forces and further fusion with the strong nuclear forces by means of the quark-gluon gauge theory. Together with gravity, understood through Einstein's general relativity, the comprehension of matter has improved down to distances of about seventeen orders of magnitude. This year (2007) this distance will decrease even more by the coming on-stream of the large hadron collider, the LHC, at CERN. Hopefully it will find support for the basic features of the present level of unification, in terms of the discovery of the Higgs boson (as the basis for mass in the Universe) as well as providing us with new clues as to the next steps of probing ever deeper into the material world, such as by finding supersymmetric partners (or even supergravitational partners) to the present spectrum of 'elementary' particles (for which there are even now claimed 'bumps' in particle processes best explained, it has been claimed by their discoverers, by these exotic possibilities).

This progressive and impressive unlocking of the secrets of matter has only recently begun to be caught up by new ideas and new tools in brain science. These have allowed an ever better understanding of the manner in which the brain creates and controls experience. This knowledge is allowing ever deeper appreciation of how defects in brain processing brought about by various sorts of damage to the brain can enormously modify the overall experience of the person. There is still a large distance to travel to bring the knowledge of the brain up to the level of our enormously more detailed understanding of matter. But that gap appears to be closing fast, with a possible hiatus in the physics of the elementary particles due to the presently intractible character of the mathematics of superstrings (the only way known to sensibly unify gravity with the other forces of matter); no such hiatus appears in brain science, except, of course, for the mind-body problem. Will that be the rock on which neuroscience founders in a similar way, it is claimed by some physicists, as that of unifying the forces of nature?

The tools of brain science have been put to good use in probing the brain both locally and globally. The machines of PET and fMRI, measuring changes in blood flow in local regions of the brain due to neural activity, and EEG and MEG, measuring the electric and magnetic activities related more directly to neural activity in the brain, are allowing an increasingly precise view of the networks active across a range of functions. Combined with trans-cranial magnetic stimulation (TMS), the timing of the flow of crucial activity across the functional networks in the brain are beginning to give up their secrets. At the same time correlated activities of small groups of single neurons in different regions of the brain are indicating how these various regions process activity at a local level, but as part of a larger network of such regions in interaction. Ideas from dynamical systems theory have begun to help us understand better the brain at this intermediate mesoscopic level. In all, brains are becoming ever more understandable across all scales.

Global Principles of the Brain's Action

An important part of this understanding is that of the extraction of basic principles at a global level so as to help comprehend the awesome power possessed by these coupled circuits of local neurons. Several global principles can be extracted from the welter of possibilities. Three of these, I would suggest, are the most important:

• Attention, as a control system to filter lower-level brain activity so as to allow very few stimulus representations to enter the higher-level arena of thought and manipulation of neural activities; these filters are mainly in parietal and pre-frontal cortices;
• Emotion, in terms of value maps learnt in orbito-frontal cortex (and also coded in associated amygdala sites) so as to bias what is to be processed and to guide choice of task goals (by their associated predicted rewards);
• Long-term memory created on-line, so as to allow for incremental wisdom about the environment for use as a guide for further interactions.

Other global principles can be added to these three possiblities, such as the use of hierarchical processing (as noted especially in vision), so as to create flexible codes for complex objects which can be used at a variety of scales. Another possibility involves the use of synchronization of neurons over long distances, such as at 40Hz and other frequencies, so as to solve the binding problem of combining the different codes for objects (as occur in multi-modal hierarchical coding schemes described earlier. However we will explore only the three main principles I adumbrated above in order to see how we can face the mind/body problem starting from the material side but with a modicum of structure (as given by the principles) to work from.

Global Principles of the Mind in Action

In order to make progress on how the conscious mind is created by brain activity, we must also outline what it is we consider is the main structure of the mind. This is a highly controversial area, with many possible suggestions. However a certain degree of clarity seems to have arisen from recent ideas about the bipartite nature of consciousness by the influential approach of Block (Block, 1995) as supporting a more general attack given by Western phenomenologists (Zahavi, 1999). These cohere together in terms of consciousness being composed of 1) Content. This is what Block calls Access consciousness, available for report and internal manipulation. This is composed of the colours, hardnesses, the scent of the rose, etc, of external stimuli as experienced in our consciousness. 2) Ownership. This is what Block calls Phenomenological Consciousness and Western phenomenologists term the pre-reflective or inner self. The manner in which these two components of consciousness can interact is highly problematic. It was not resolved by the phenomenlogists, who seemed to take the majority view that the pre-reflective self arose from the body. That is also a highly controversial position, with the danger that subjects who have lost proprioception would lose their inner self; this is known not to happen, so making that position suspect. Nor does this body-based self have the automatic property of 'immunity to error through misidentification of the first person pronoun' (Shoemaker, 1968) and regarded as a crucial component of consciousness. The phrase denotes the fact that I cannot sensibly ask you 'Are you sure it is you in pain?' when you tell me you are feeling pain: you just know it is you. Thus the problem of understanding how content and owner can interact is still open, and in some ways can be seen as hard as that of explaining consciousness more generally. However it has reduced the task to something more specific in terms of the supposed two components of consciousness.

Yet the problem of understanding how the owner of conscious content becomes 'aware' of that content is decidedly non-trivial. This is partly because the owner or 'I' is itself not properly understood. It is possible to consider the nature of the 'I' as expounded in the writings of Eastern meditators. They have spent decades meditating on the nature of their inner self, and have built up a vast literature on their experiences across a range of disciplnes, such as Zen and so on. However this material, whilst potentially very valuable, is also solely descriptive and highly personal. This makes it difficult to relate it to modern brain science. The increasing number of brain imaging experiments performed on meditators is gradually helping to bridge this gap, so that this vast literature from meditors could ultimately help exploration into the inner self.

Attacking the Mind-Body Problem

Thus the mind-body problem still faces brain science and philosophy like a nemesis. Neither the above global principles for the brain nor any additional ones seem to involve its solution at all. That is because they do not explicitly state how and where consciousness arises in the higher-order processing achieved by attention and guided by emotion and long-term memory. More particularly they do not give any handle at all on the neural components that could support the 'I' of the pre-reflective self.

If the mind-body problem cannot be given a satisfactory solution through such an approach as above, possibly expanding on the principles governing the brain and able to be checked by scientific methods, then science itself will have failed in its attempt to explain all of the world. It would not have been able to explain in particular how mental experience is itself created from the activities of the apparently mindless nerve cells in interaction in the brain.

But such a dramatic situation has not been yet met with. Brain science is in its infancy, and even the principles I adumbrated above are controversial. That of the creation of consciousness by brain activity is even more so. Thus we are still not at a point in time where the scientific process towards solving the mind-body problem has been universally seen to have failed, and only a dualistic solution (with mind in some non-physical 'space') would have to be considered.

In order to be able to begin to reach a scientifically testable solution to the mind-body problem, any such solution must itself be couched in scientifically testable form. It must involve entities created by neural activity which are directly related to inner experiences of a subject. But that already puts a strong constraint on any such theories that can be considered. A dualistic theory is automatically discounted, as are those of a weak physicalist form in which no scientific implications arise from the theory. Verifiability is all: models must be proposed which can be falsified.

When we turn to possible falsifiability as a criterion for putative solutions to the mind-body problem then we need to be more precise. What is is we are claiming about the relationship between the mental experiences of a subject and the neural activity going on in their brain? Just to look for active brain regions when conscious processing is occurring (which would be silent when it is not going on) does not properly tackle the problem as to the functionality of the network of neural modules observed. It is this last question, of functionality of the relevant 'consciousness creating' networks of the brain, which deserves considerably more analysis than given heretofore. It naturally leads us to consider various attempts of modelling consciousness already presented, to see if we can build on them to create a viable 'mind-from-matter' machine.

Possible Solutions to the Mind-Body Problem

I will consider a range of solutions that have been put forward to solve the mind-body problem with a brain basis. It is only through some such brain-based approach that scientific progress can be achieved, and therefore all possible approaches are of importance to be considered most carefully. There is only space here to briefly survey a few of these approaches:

40 Hz and Competing Coalitions

It was proposed originally (Crick & Koch, 1991) that observed gamma range neural activity during conscious processing creates active bound object neural representations, thus being the basis of consciousness. This proposal has had large popularity more recently, although it apparently cannot answer the question as to the source of the inner perspective in the conscious experience which we each possess. Moreover such synchronisation was also observed in anaesthetised animals. Yet the 40Hz model is an important component in the overall picture, involving as it does a summary of a considerable amount of data indicating that synchronization of activity between various brain sites is an important component of brain processing and more specifically of consciousness.

The 40 Hz approach to consciousness has now been abandoned by Crick and Koch, who replaced it by the concept of 'neural coalitions' (Crick & Koch, 2003) in which 'Consciousness depends on certain conditions that rest on properties of very elaborate networks' (Crick & Koch, p 124). What makes an 'elaborate' network automatically create any conscious experience at all is not further spelt out, except in terms of indicating how attention may play a role in 'biasing the competition among nascent coalitions' (Crick & Koch, p 124). These are certainly only general ideas in this latest 'framework for consciousness', and especially there is no hint of any interest by the authors in the inner or pre-reflective self.

Recurrence

Numbers of authors (for example Lamme, 2003) have suggested that recurrent feedback (observed, for example in visual cortex) can produce the experience of consciousness. However some of these claims are very debatable and need to be considered more carefully.

Strong support for recurrence as crucial in brain processing more generally has arisen, for example, from (Spivey and Dale, 2004) who analysed how recurrent dynamics can lead to an explanation of results from various psychological paradigms involved in classification and language processing, for example (although some of the phenomena they discuss are also non-conscious). Recurrence has also been used as a basis for the continued activity in working memory buffers, as analysed in detail in numerous research papers (for example, Wang, 2001). Such activity has been considered crucial to the generation of conscious experience as providing reportability to other sites of the relevant material, as assumed by many psychologists (see references in Taylor, 1999). However this use of recurrent neural activity to support continued buffer working memory activity does not necessarily provide at the same time any hint as to how any inner perspective might arise in such processing. There is also the difficulty that there are numbers of sites where recurrence is crucial for function (such as hippocampus), but which can be removed without a consequent catastophic failure of consciousness. But it is necessarily the case that recurrence is a crucial component of brain processing as observed by neuroanatomy, so surely will play an important role in consciousness creation, even if not its sole cause.

Quantum Models

These models have been proposed (Penrose, 1989) to avoid strictures about understanding consciousness in computer terms due to Godel's incompleteness theorem. Godel's thereom raises the difficulty that any computational system, if unextended, is unable to be employed to prove all statements in the system, even though we can apparently see these statements as true (outside the system). Thus our ablities, so the argument goes, are always beyond those of a computational system. It is suggested that instead some miracle of quantum mechanics or even of quantum gravity is present to enable the brain to go beyond convential computation. How this occurs has so far not been specified in enough detail to produce a testable model. Moreover yet again the inner perspective is seemingly completely absent from this approach, so would seem unlikely in any case to shed much light on the nature of teh pre-reflective self.

Global Workspace

This proposes (Baars, 1997) that consciousness arises once neural activity accesses a global workspace of well-connected neural modules (involving at least prefrontal and parietal lobes). It is as if this activity has been 'written' on a blackboard for all to see. Such a feature is by now standard in computer science, and the frontal cortices have been observed to possess the ability to support fast learning, corresponding to fast 'writing' on it. However Baars considers activity on the global workspace to be more active than passive, and is involved in the competitive processes involved in gaining access to the global workspace. This is close to the suggestions in the 'Adaptive Coding Model' of (Duncan, 2001) in which there are competitive processes on the prefrontal cortex to help guide attention selection (in posterior sites). The global workspace therefore has aspects of importance for any future theory of the global brain. It has no specific mechanism to produce any inner perspective, but that could well be added by a more careful inclusion of attention in its full complexity to help lift the global workspace model to a fully testable version, and also to provide inclusion of some form of inner self.

Complexity of Activity

The need for a certain level of complexity of the interaction of activity across the brain in order that consciousness can arise has been emphasised (Edelman & Tononi, 2000) by developments of analyses of brain activity of subjects in various states. This is clearly going to be an important aspect, since consciousness possess that inner complexity of information content as observed experimentally as well as from the 'inside'. If it were to be allied with a more detailed inclusion of the three principles mentioned earlier (attention, emotion value and memory) there may be the beginnings of a viable model of consciousness.

Relational Mind

This proposed (Taylor,1999) that mind arose from the comparison between ongoing activity and activated relevant past memories. Undoubtedly there is such memorial 'filling out' of present activity, for example to give meaning to our current experience. In spite of inclusion of memory and emotion, there was little hint as to how the inner self or perspective could arise; as in Baars' GW model, that could be by further more careful discussion of the way that attention (the missing element of the triumvirate of faculties suggested as a crucial basis for efficient brain processing) is to be incorporated into the whole (simplified) architcture of the brain suggested earlier.

Attention Copy Experience

Attention acts as a control system, and is now known to involve parietal and prefrontal regions involved importantly in creating attention control signals to feedback to lower level inputs. Arguing by analogy with motor control in the brain, it was suggested (Taylor, 2003, 2006) that there is a copy of the attention movement signal that is crucial in providing an early 'wake-up' call for mid-level cortices to enable attended activity to enter awareness by attention amplification, at the same time inhibiting distracters. The detailed mechanisms for using this copy to boost access to awareness of the attended stimulus, at the same time inhibiting distracters, can help explain, given the interpretations of the model, the 'immunity to error of the first person pronoun' (Sheomaker, 1968), an important feature of experience emphasised by Wittgenstein. This is further supported by experimental results indicating that a stimulus representation just accessing awareness can have an inhibitory influence on its relevant activity exerted by the copy signal of a putative distracter created around 200 msecs after the distracter stimulus input (Sergent et al, 2005). This model appears to be a little closer to inner experience than the others listed, although considerable work still needs to be done to fill it out satisfactorily.

There are other models that employ attention formally to help process information more efficiently, as in the GW model of Baars, or in the Competing Coalitions model of Crick & Koch, both mentioned above. However there do not seem to be any models presently available of the attention copy type, in which the copy signal is used directly to provide the experience of ownership, identified with that of 'I'.

Prospects for a Brain-based Solution to the Mind-Body Problem

We repeat again here that we are taking for 'mind' the conscious component of it, and not the many activities of ongoing processing below consciousness. It is clear that we cannot ignore these unconscious processes, since out of them conscious experience will arise. Thus any final model (or class of models) seriously attacking the mind-body problem will have to take careful account of both the conscious and non-conscious activity that is ongoing in the brain.

There are numerous important features from the above models of consciousness that we must take forward to build a more complete model with some chance of success: attention as a filter and possible source of feature binding, activity coming into awareness when accessing sites of working memory, associated attention copy 'ownership', various forms of memory as crucial components to give content to awareness, synchronization as a possible additional source of feature binding for object representations, suitably complex interacting brain states, emotion values as giving the whole system value and source of drive, and so on.

Do these various components begin to look as if they could provide a basis for a suitably scientific theory of consciousness? This can only be achieved by a modelling-cum-experimental attack on the many paradigms and associated observations in which consciousness crucially enters, and based on models including the three main principles mentioned earlier and some or all of the features of the various models outlined above. Such phenomena include, for example, the attentional blink, the Stroop test, attention search and associated phenomena such as the Posner benefit paradigm, and numerous others. Only when all of these have been explaind satisfactorily at a quantitative level by such a developed model can we begin to approach the final solution to the mind-body problem. Even then there will be enormous problems of interpretation. But there is still a long way to get to that point yet.

References

• Baars, B (1997) Some essential differences between consciousness and attention, perception, and working memory Consciousness and Cognition 6(2-3):363-71
• Block NJ (1995) On a Confusion About the Function of Consciousness. Behavioural and Brain Sciences 18:227-247
• Crick FHC & Koch C (1991) Towards a neurobiological theory of consciousness. seminars in neuroscience 2:237-249
• Crick FHC & Koch C (2003) A framework for consciousness. Nature Neuroscience6(2):119-126
• Duncan J (2001) An Adaptive Coding Model of Neural Function in Prefrontal Cortex. Nature Reviews Neuroscience 2:820-829
• Edelman G & Tononi G (2000) A Universe of Consciousness: How Matter Becomes IMagination. NY: Basic Books
• Elliott R & Dolan RJ (1998) Neural Response during Preference and Memory Judgements for Subliminally Presented Stimuli: A Functional Neuroimaging Study. Journal of Neuroscience 18(12):4697-4704
• Haynes J-D, Driver J & Rees G (2005) Visbility Reflects Dynamic Changes of Effective Connectivity between V1 and Fusiform Cortex. Neuron 46:811-821
• Lamme V (2003) Why visual attention and awareness are different. Trends in Cognitive Sciences 7:12-18
• Lamme V (2006) Towards a true neural stance on consciousness. Trends in Cognitive Sciences 10(11):494-501
• Moutoussis K & Zeki S (2002) The relationship between cortical activation and perception investigated with invisible stimuli. Proc Nat Acad Sci USA 99:9527-32
• Penrose R (1989) The Emperior's New Mind. Oxford: Oxford University Press
• Sergent C, Baillet S & Dehaene S (2005) Timing of the brain events underlying access to consciosuness during the attentional blink. Nature Neuroscience Sep 11, 2005
• Shoemaker S (1968) Self-reference and self-awareness. The Journal of Philosophy 65:555-567
• Spivey MJ & Dale R (2004) On the Continuity of Mind: Toward a Dynamical Account of Cognition. pp 87-142 in The Psychology of Learning and Motivation, vol 45
• Taylor JG (2006) The Mind: A User's Manual. London: Wiley & Sons
• Taylor JG (2003) Paying Attention to Consciousness. Prog in Neurobiology 71:305-335
• Taylor JG (1999) The Race for Consciousness. Cambridge MA: MIT Press
• Thompson KG & Schall JD (1999) The detection of visual signals by macaque frontl eye fields during masking. nature neuroscience 2(3):283-8
• Wang X-J (2001) Synaptic reverbrations underlying mnemonic persistent memory. Trends in Neuroscience 24:455-463
• Zahavi D (1999) Self-Awareness and Alterity. Evanston Ill: Northwestern University Press

See Also

Attention, Consciousness, Global Workspace, Models of Consciousness

Sponsored by: Anil Seth, University of Sussex, UK

Sponsored by: Eugene M. Izhikevich, Editor-in-Chief of Scholarpedia, the peer-reviewed open-access encyclopedia

Reviewed by: Dr. Igor Aleksander, Imperial College, London, UK

Reviewed by: Anonymous

Accepted on: 2009-08-13 13:13:52 GMT