The Brain is wider than the sky
For put them side by side
The one the other will contain
With ease ____ and you ____ beside’ Emily Dickinson
This book really falls into two halves. First there is an interesting discussion of brain function, in which Edelman attacks the popular brain/computer analogy and argues for the importance of re-entrant connections between the cortex and the thalamus, notably as a possible solution to the binding problem. The second aspect is an unconvincing attempt to explain consciousness. Edelman seems vaguely aware that he has not quite made his consciousness argument stick, since the later chapters of the book appear to repeat much of the earlier material along with sideswipes at belief in spirits, the after life and quantum gravity. This is a pity as the earlier parts of the book contain thoughtful material.
Edelman starts by suggesting that we should not look for consciousness in specific neurons or even particular parts of the cortex, but should view it as a process involving neurons in different parts of the brain. He suggests that particular neurons may be involved in the conscious process at some points but not at others. Importantly, he argues that some areas of the brain may be necessary for consciousness, but this does not mean that they are sufficient for consciousness. This looks to be a way of dismissing naive arguments that because areas such as the brain stem are necessary for consciousness that is the only place we need to look for it. In contrast, Edelman thinks that consciousness requires the engagement of multiple brain regions.
Attention & consciousness: He makes the best distinction, I have so far read, between attention and consciousness. Attention can direct consciousness, but it is not the same thing, notably because we are consciousness of peripheral objects and events that are not the focus of attention. Further to this, consciousness, at least in humans, involves being conscious of being conscious, recreating past episodes through memory, and forming future intentions, plans and scenarios. In humans, it also involves the assignment of meaning to symbols, a function at its most developed in language. Finally, consciousness involves qualia, the basic subjective experience, for instance the greenness of green. It is stressed that qualia are not usually experienced in isolation but as part of a complex.
Cortex & thalamus: The aspect of the brain that particularly attracts Edelman is the relationship between the cortex and the thalamus. Most incoming sensory data passes through the thalamus and is relayed on to the cortex. A layered structure called the reticular nucleus surrounds the thalamus, and is thought to inhibit some of the sensory input. However, Edelman stresses that there are also many axons running back from the cortex to the thalamus. A set of thalamic nuclei called the intralaminar nuclei receive connections from structures in the brainstem and then project to the cortex. The intralaminar may set thresholds for cortical response, and may be necessary for consciousness. The author emphasises the importance of the ascending systems also referred to as the value systems. These systems send axons through the brain in order to distribute neurotransmitters. Each system handles a different neurotransmitter. The locus coeruleus, located in the brain stem, distributes noradrenalin, the raphé nucleus distributes serotonin, and the cholinergic nuclei distribute acetylcholine. The dopamine nuclei release dopamine, while the histamine system is based in the hypothalamus. The widespread distribution of axons from these nuclei means that they can affect large numbers of neurons. The release of these neurotransmitters increases the chance that neurons will fire after receiving glutamine input. The systems affected included memory, learning and body responses. P. Edelman thinks that the understanding of consciousness requires an understanding of perception, memory, action and intention. He rejects the idea of the brain as a computer, but instead thinks in terms of the selection of particular states from a large repertoire of states. The idea is that neurons that are well matched to the value systems of neurotransmitters are the ones that will survive.
Not a computer: Edelman further criticises the brain/computer model. The computer model assumes a one-to-one relationship with sensory input that is unambiguous, and to which there can be a straightforward appropriate response. He argues that in fact the brains input is not unambiguous. The real world is not like a piece of computer tape, and the visual system in particular makes adjustments to its input in order to try and make sense of it. Thus even in adults the somatosensory cortex is constantly changing in response to how the body is used. Edelman points out that while there are at least 30 functionally specialised areas scattered across the brain, there is no central coordinating area, as there would be if it was a conventional computer system. The problem of how these areas are coordinated into a single conscious experience is known as the binding problem. His point is that brains construct responses to the environment, and do not, as computers do, work on the basis of formal rules and unambiguous instructions.
TNGS: Edelman calls his concept the theory of neuronal group selection (TNGS). He thinks that brains develop first according to genetic and epigenetic pressures and then according to experience. Beyond these patterns of nature and nurture, reciprocal connections coordinate different parts of the brain. Signals are exchanged between parallel brain regions coordinating the brain in space and time. The main consequence of this reciprocal system is the synchronisation of the activity of widely dispersed neuronal groups. Functionally different circuits are bound into a coherent output. Edelman first proposed this version of brain function in 1978. He claims that evidence has accumulated in favour of his theory since that date.
Degeneracy: He also stresses the importance of so-called degeneracy in the brain. This means that there are structurally different elements of the system that are able to produce the same output. In particular, this gets rid of bugbears such as the so-called grandmother cell. This was the idea that there could be a neuron with a particular function, such as recognising one’s grandmother. In Edelman’s scheme there is degeneracy, which allows different cells, or more likely groups of them, to deal with recognising grandmother at different times.
Categorisation: Perceptual categorisation is viewed as being one of the most important functions of brains. This allows the signals from the environment to be split into sequences that are useful for behaviour. So the signals from a room can be split into categories such as ‘chair’ and ‘table’. This involves a so-called global mapping, where maps of different sensory modalities are linked by reciprocal connections. A further step is to generalise between perceptions in order to create concepts. Finally, Edelman’s theory of brain function requires memory. Changes in synaptic function favour the use of neuronal circuits that re-enact particular events. Again there is degeneracy, so more than one set of neurons can re-enact particular memories. Moreover, it is a re-enactment in the sense that it may be somewhat different from the original event. Recall is seen as being influenced by the value system neurotransmitters, thalamocortical mapping and sub cortical structures such as the hippocampus and basal ganglia.
The binding problem: Edelman sees the development of these systems as a necessary precursor of consciousness. The emergence of re-entrant or reciprocal systems is identified with the evolution from reptiles on the one hand, to birds and mammals on the other. In particular, this was the stage at which the reticular nucleus started to select the activity of the nuclei in the thalamus. Edelman stresses that the thalamocortical system interacts mainly within itself, and has relatively few connections with the rest of the brain. He relates this to the distinction between conscious and non-conscious systems in the brain. It is also proposed that the reciprocal action between the neuronal circuits within this part of the brain is what allows perception to be bound together into a single coherent whole, thus solving the binding problem. Edelman further considers that the sense of self originates from the bodily sensations of the infant, to which the experience of other qualia are added as the individual matures. The thalamocortical activity is suggested to be stable over periods of between a few hundred milliseconds and a few seconds and this is identified with the ‘specious present’ originally suggested by William James. This is sometimes referred to as the ‘remembered present’. Edelman also reworks the Dennett approach to zombies, which is also at heart a sleight of hand, in a new form. The philosophical concept of a zombie is that it would be possible to have an entity that performed all the functions of a human without consciousness. Dennett’s (and Edelman’s argument) is that if the zombie is exactly the same as the human it must have consciousness, because that is part of being human. This is obviously true, but in true consciousness studies fashion, it evades the real matter of the zombie argument, which is that the known activities of humans could be performed non-consciously. Edelman has decided that consciousness must arise in the reciprocal setups of the thalamocortical system. These processes are seen as being centred round a self that is a reference point for memory. Edelman refers to the thalamocortical area as the re-entrant dynamic core, and this is where he looks for consciousness. At this point, his account makes a jump or massive assumption that leaves an explanatory gap behind it. He suggests that the dynamic core converts signals from the external world and the brain itself into what he refers to as a ‘phenomenal transform’, which are in fact qualia or consciousness. There is nothing implausible about relating consciousness to this area of the brain, but what is lacking is any explanation of how the signals are suddenly converted into consciousness. In fact, the very use of the phenomenal transform term looks like a bit of a fudge. We are still wondering what this higher order sounding term means, when we discover it means consciousness. If it was simply stated that the signals were turned into consciousness, we might be more inclined to ask how it was done. The whole arrangement is a bit like the cartoon of a consciousness lecture, in which a string of equations is followed by ‘and now a miracle happens’ and then more equations.
More puzzles follow. It is fair enough to say that qualia refer to distinctions, say for instance the distinction between green and red, and that in the brain this is based on neural activity. What is not explained is why we experience this as conscious qualia. A non-conscious machine can just as easily make the distinction between a red and green light, without any need for the additional effort of generating qualia. P. Things get curioser and curioser. It is stated that the ‘phenomenal transform’ is not caused by neural activity but is nevertheless a ‘simultaneous property’. This gives no explanation of how this property arises. If it doesn’t arise from the activity of neurons, what does it arise from? It looks as if consciousness as been simply declared to arise at this point, because this would indeed fit in nicely with the earlier and more convincing discussions of brain function. Edelman doesn’t feel the need to investigate the origin of the ‘phenomenal transform’ any further, and instead moves on to discuss whether it could be causal. For this exercise the phenomenal transform is referred to as C, and the underlying neural processes from which it has mysteriously arisen as C’. It is stressed that C cannot arise in the absence of C’. C is now without previous warning stated to be a relationship, having previously been a property, albeit only a ‘simultaneous property’. It is not stated what it is a relationship between. Perhaps, we are supposed to intuit that it is a relationship with C’ neural activities, except that earlier on it was granted the dignity of being a property alongside the neural activities. At any rate, being a relationship disqualifies it from being causal, according to the laws of physics, which might have been thought to have been rather brutally ignored in the preceding discussion.
Something has happened here that has been seen before in mainstream consciousness studies. While the concept of non-physical is denied, it seems possible, when it’s convenient, to classify consciousness as non-physical, in order to prevent if from being causal, and thus opening up the whole question of freewill. Here consciousness or the ‘phenomenal transform’ was initially described as a property, albeit a simultaneous property, and in the scientific understanding of the universe properties come with a physical description. Later the phenomenal transform, under the guise of C, is demoted from a property to a relationship. However, even this doesn’t justify the non-physical therefore non-causal tag. Relationships are instantiated in matter or energy. The gravitational force between the Earth and a falling rock might be described as a relationship, but for an organism positioned directly beneath the rock, it would be adaptive to regard the relationship as causal.
The conclusion has to be that this book would have been better as a discussion of brain functions, and is marred by the attempt to extend these ideas into an attack on the hard problem of consciousness.