Tuesday, 21 August 2012

Gerald Edelman on consciousness

Anyone interested in the origins of art needs to study how human beings evolved their distinctive creative intelligence. One of the leading writers and researchers in the field of human consciousness and neuroscience is the Nobel Prize-winning biologist Gerald M. Edelman. In this article we shall take a very brief walk through his theory of consciousness and try to keep this incredibly complicated topic accessible to the general reader.

Edelman’s goal is simply to answer the question: What is consciousness? “How can the firing of neurons give rise to subjective sensations, thoughts and emotions?... A scientific explanation must provide a causal account of the connection between these two domains”[1]. In the theory he calls ‘Neural Darwinism’, Edelman argues that consciousness – “what you lose when you fall into a deep dreamless sleep and what you regain when you wake up” – is a product of natural selection. His ideas emphasise the brain’s plasticity in response to the environment, and he rejects reductionism, metaphysics and wrong-headed analogies with computers.


Consciousness is rooted in the operations of an individual body, above all the brain, and in its history and experiences. Importantly, “consciousness is a process, not a thing”, the “dynamic accomplishment of the distributed activities of populations of neurons in many different areas of the brain.” It is individual, continuous, intentional and unitary or integrated. At any given moment,

The scene is not just wider than the sky, it can contain many disparate elements – sensations, perceptions, images, memories, thoughts, emotions, aches, pains, vague feelings, and so on. Looked at from the inside, consciousness seems continually to change, yet at each moment it is all of a piece.

Human beings are conscious of being conscious. So we must make a distinction between ‘primary consciousness’ – the “state of being mentally aware of things in the world, of having mental images in the present”, which we have in common with many animals – and higher order consciousness – allowing “the recognition by a thinking subject of his or her own acts and affections”. The latter includes the ability to have intentions for the future and requires the use of symbols, which in its most advanced form means language capability. Besides ourselves, the only animals thought to possess higher order consciousness to a debateable degree are the higher primates.

When conscious, individuals experience qualia. The term ‘quale’ refers to our particular experience of a property: such as redness, or warmth, or pain. Edelman describes qualia as “high order discriminations that constitute consciousness… experienced as parts of the unitary and integrated conscious scene”. All conscious events involve a complex of qualia – a quale cannot be experienced in isolation.

Neural basis of consciousness

To develop a theory of consciousness we must first understand how the brain works. This is no easy matter, as the human brain is the most complicated object in the known universe and is still poorly understood. Its dominant feature is the cerebral cortex, a convoluted structure making up about two-thirds of the brain mass which lies over and around most of the brain. It contains 30 billion or more neurons (nerve cells) and a million billion synapses (connections), and most of the brain’s information processing takes place there. The cerebral cortex is divided into regions with different functions, e.g. areas involved in sight, hearing, touch, movement, and smell.

Neurons are connected to each other to form a dense network to pass signals around the nervous system. They are very diverse, but a typical neuron has a long extension called an axon, which connects the neuron to other neurons at gaps called synapses. The synapse allows the neuron to pass an electrical or chemical signal to another cell.

A region essential to consciousness is the thalamus, located at the centre of the brain and equivalent in size to a pair of walnuts. It serves to relay signals from the nerves (e.g. in your eyes, ears or skin) to the cerebral cortex, acting like a kind of central switchboard. For example, information from the retina of the eye is sent to a nucleus of the thalamus, which forwards it to the part of the cerebral cortex responsible for processing visual information. Another subcortical region is the hippocampus, important for short-term to long-term memory.

The brain’s motor functions regulate not only movement but also assist the forming of images and concepts. The primary motor cortex sends signals down the spinal column to the muscles, and the cerebellum, a structure at the base of the brain, helps coordinate our physical actions. Located in the centre of the brain are the basal ganglia, which connect to the cortex via the thalamus. They are associated with voluntary movements and regulation of motor systems.

Edelman concludes that there are three neuro-anatomical ‘motifs’ in our brains. The first is the thalamus and cortex. The second is the inhibitory circuits of the basal ganglia. The third are the ascending systems: nuclei of the brain stem that release neuromodulators such as serotonin and dopamine.

We should not think simplistically of specific areas of the brain controlling specific functions. Certain activities tend to be region-specific, but the regions are connected up in a complex and integrated system. This integration is essential to consciousness.

The brain is not a computer

Although he sometimes uses metaphors such as brain ‘circuitry’, Edelman makes a strong case against describing the brain as a computer. There is rich variation within the formation and movements of cells during the brain’s development, meaning no two brains are alike. The brain is not hard-wired, but develops patterns of neural activity, captured in the phrase ‘neurons that fire together wire together’. Although there are programmed stages of development, the behaviour of cells is always variable or plastic. “The result is a pattern of constancy and variation leading to highly individual networks in each animal.” This is no way to build a computer, which demands precise wiring and predictable programming. Inputs to the brain are not a sequence of ones and zeros – they are ambiguous. The computer analogy is too rigid to describe the organic, dynamic processes of the mind, which has to deal with a world that is unpredictable and is based on pattern recognition rather than logic.

An example of this pattern recognition is the so-called ‘binding problem’, i.e. the question of how brains combine elements of complex patterns of information. When we see a red car drive past, there are separate processes to register colour, movement, orientation, and so on. A perception emerges in various contexts, and theory must find a mechanism to explain how it works.

Another complication is degeneracy, which in this context means the ability of structurally different parts to perform similar functions under certain conditions, while performing different functions in other conditions. Again, this really doesn’t resemble a computer.

Neural Darwinism

The brain evolved – it was not designed. Darwin argued that new organisms emerge from selection among the variant individuals in a population, based upon their fitness for survival within a particular environment. One of the tasks of neuroscience is to work out how precisely this process created the human brain.

Just like any population of animals, brains show a huge amount of variation between individuals. Edelman sees this variation as fundamental:

selection from such a population of variants could lead to patterns even under unpredictable circumstances, provided that some constraint of value or fitness was satisfied. In evolution, fitter individuals survive and have more progeny. In the individual brain, those synaptic populations that match value systems or rewards are more likely to survive or contribute more to the production of future behaviour.

Edelman calls his selectionist theory the theory of neuronal group selection, or TNGS. This has three basic tenets.

Developmental selection: selection creates a wide variety of brain ‘circuitry’ within individuals during their growth and development. No two people will have exactly the same synaptic structures in comparable areas of brain tissue – a bit like unique fingerprints.
Experiential selection: overlapping that first phase and after the major neuroanatomy is built, variations in environmental input continue to create variations in synaptic strengths, favouring some pathways and weakening others.
Reentry: ‘reentry’ is an interchange of signals that continuously relates parts of the brain to each other, relying on networks of connections between groups of neurons that have arisen out of the other two processes above. Reentry is not sequential but involves many paths acting simultaneously; it is the means by which bits of the brain communicate directly with each other. If a computer is organised by logic, a brain is organised by the process of reentry.

The consequence of this process is the binding of neuronal groups with different functions into a coherent system. “How can it be,” asks Edelman, “that… up to thirty-three functionally segregated and widely distributed visual maps in the brain can nevertheless yield perception that coherently binds edges, orientations, colours, and movement into one perceptual image?” His answer is: through reentry. Degeneracy is also important, as it allows different neurons and neuronal groups to yield similar outputs despite their different structures. “Different cells can carry out the same function and the same cell can, at two different times, carry out different functions in different neuronal groups.” The TNGS means that we do not need any fixed, computer-like plan to explain what happens in consciousness.

During natural selection, neuronal groups (rather than individual neurons) are selected for fitness from among the available variations.

Mechanisms of consciousness

How do these workings of the brain give rise to consciousness?

One of the most basic processes is the ability to categorise information from outside to make sense of the world. For example, we continually process various signals to categorise them as stable objects – chairs, cars, cats and so on. For Edelman, this categorisation is carried out by ‘global mappings’, i.e. sensory maps linked by reentry, and linked in turn to other systems such as the cerebellum and basal ganglia. Global mappings sample the world of signals and categorise them through the connections between neuronal groups.

However, these signals could not help an animal learn without memory, which Edelman defines as “the capacity to repeat or suppress a specific mental or physical act”. Memory is essential to a theory of consciousness.

Global mappings, concept formation and memory, along with the three neuro-anatomical ‘motifs’ of thalamus-cortex, subcortical organs and ascending value systems – these are the necessary evolutionary precursors of conscious activity. Then, Edelman argues, at some point in evolution, a new connectivity developed in the system. The critical development that allowed primary consciousness was the linking of memory to perceptual categorisation, granting an animal the ability to construct complex scenes and discriminate between elements of those scenes by referring to its memory of previous experience. This construction of a ‘remembered present’ improves the animal’s survival chances: it can make better choices about how to respond to its environment, for example by remembering that the last time it heard a particular growl, a predator appeared shortly after.

Primary consciousness is experienced by many animals besides ourselves. Animals with only primary consciousness have no real sense of past or future or of a socially defined, named self, and they are not conscious of being conscious. This doesn’t mean they don’t have a self, or don’t have memory. The difference between them and us, according to the TNGS, is that they have no semantic abilities, i.e. “they are not able to use symbols as tokens to lend meaning to acts and events and to reason about events not unfolding in the present moment.” This doesn’t quite mean that language is necessary for higher order consciousness – some apes have semantic abilities, including the ability to use symbols, without their being able to talk. But our real reference for higher order consciousness is ourselves. At some point, we realised that an arbitrary token, such as a gesture or word, can stand for a thing or event.

When a sufficiently large lexicon of such tokens is subsequently accumulated, higher-order consciousness can greatly expand in range. Associations can be made by metaphor, and with ongoing activity, early metaphor can be transformed into more precise categories of intrapersonal and interpersonal experience. The gift of narrative and an expanded sense of temporal succession then follow. While the remembered present is, in fact, a reflection of true physical time, a higher-order consciousness makes it possible to relate a socially constructed self to past recollections and future imaginations. The Heraclitean illusion of a point in the present moving from the past into the future is constructed by these means. This illusion, mixed with the sense of a narrative and metaphorical ability, elevates higher-order consciousness to new heights.

We later evolved additional ‘circuitry’ – hand in hand with the evolution of the vocal tract, increase in brain size, bipedal posture and other developments – that made large-scale connections between conceptual systems, allowing symbolic communication and language, and for the higher order consciousness characteristic of the human mind. The heart of this was the dynamic core, a huge network of neurons that maintain a continual and integrated picture from a range of possibilities despite being constantly re-arranged; it is not a specific brain area but a constant process. Semantic and linguistic ability required new reentry pathways and circuits and greatly expanded the range of conscious thought. We could now invent narratives and fantasies.

Because the reentrant circuitry of our minds is degenerate, Edelman doubts that there is a one-to-one correlation between a representation of an image or thought with any particular circuit or neurons. A neuron may help a representation one moment and not help at all the next. Representation is created by a complex network of neurons, synapses, environment, history and other contexts in which there are many ways to make the same meaning.

There are no functional states that can be uniquely equated with defined or coded computational states in individual brains and no processes that can be equated with the execution of algorithms. Instead, there is an enormously rich set of selectional repertoires of neuronal groups whose degenerate responses can, by selection, accommodate the open-ended richness of environmental input, individual history, and individual variation.

Just as every organism has a unique biological identity, each consciousness has a unique history.


To summarise: Consciousness is rooted in the brain, but the brain is embedded both in a body and in an environment. Consciousness is unitary, while at the same time it shifts and changes. Our earliest interactions with the world involve information from motor areas and emotional responses, and therefore create a self which acts as a reference for memory. In primary consciousness, this self exists in a ‘remembered present’ constructed around an integrated scene over a short time period. Even an animal with only primary consciousness and very little understanding of past and future can make many conscious discriminations between states, experienced as qualia. Primary consciousness depends on parallel, recursive activity within and between areas of the thalamus and the cortex.

With the evolution of higher order consciousness based on semantic ability, concepts of self, past and future emerge. Human beings have a self acting in a remembered present, but also a defined self; we are conscious of being conscious, have awareness of the past and can imagine the future. We have language, i.e. not only semantic ability but full syntactic ability as well. We can use symbols to divorce ourselves from the remembered present by acts of attention.

Neuroscience is still in its infancy. Scientists dispute whether there is any need to introduce Darwinism into the connecting of neurons, and Edelman does not give enough emphasis to consciousness as belonging to active people rather than brain processes. But Edelman’s account, with its correct emphasis on the mind as dynamic, plastic and organic rather than rigid or machine-like, may yet prove seminal for our understanding of consciousness.

[1] Quotes are from Edelman’s succinct and relatively accessible book Wider Than the Sky: The Phenomenal Gift of Consciousness (2004).

No comments: