Consciousness theories

by Sanjay Manohar, 2001

A "linked list" of common theories of consciousness

Phenomena such as the effect of brain damage or drugs on conscious states, and considerations that humans are conscious where amoebae (maybe) are not, motivate many varied theories of consciousness. They all aim to explain these things and perhaps also determine some unknowns, e.g. whether intermediate animals are conscious, or whether computers can be conscious.
Although the theories come from diverse branches of knowledge, there is much to suggest that we should break down today's notion of consciousness into a few different kinds. We can divide the evidence into a groups according how much they interact in a causal way. The numbers and letters (e.g. 2b) refer to connections between theories that may be linked - they could be different facets of the same theory.
  1. Neuroanatomical

    1. Thalamus
    2. The thalamus is a deep nucleus of the prosencephalon which connects to most areas of the cerebral cortex. It straddles the midline, and all fibres entering the cortex synapse here first. Knowledge about this area has grown immensely in the last 10 year, but is probably still in its very early infancy.
      • 2b
    3. Brain stem
    4. The brain stem is an evolutionarily conserved area which is necessary for vital functions like breathing, projects axons widely into the cortex from the 'reticular activating system', whose activity is required for many vertebrate organisms to remain alert.
    5. Basal ganglia
    6. Multiple interconnected nuclei deep within each hemisphere receive fibres from the cerebral cortex, and send fibres variously to the thalamus (and thus back to the cortex), brainstem, cerebellum and spinal cord.
      • 2c
    7. Cerebral cortex
    8. The cortex is a sheet of grey matter convoluted into multiple folds. It appears to be organised into 6 layers with cells of different shapes and sizes, and into 'columns' with all cells from all the layers in a small region having a similar specialisation in representing information.
      • 2c
    9. Motor cortex
    10. Prefrontal cortex
    11. The prefrontal cortex is the most-recently-evolved cortex. The cells are similar to elsewhere in the cortex, but their function does not appear to be related in a simple direct way to sensory or motor information. Prefrontal cortex has extensive projections to and from most other cortical areas.
      • 3d
  2. Neurophysiological

    1. Neural synchrony
    2. In the 1980s several experiments showed that brain cells receiving the same signal from two sources are more likely to pass the information on if the two inputs are synchronised, in terms of action potential timing. It has become clear that spike-timing is critical in what information is "amplified or attenuated" by a cell.
      • 4b
    3. 40-Hz Thalamocortical oscillation
    4. The thalamus is probably involved in generating many of the oscillatory features on the EEG. Such spatio-temporal wave-like activity appears to correlate with wakefulness and also to what extent sensory information is perceived.
      • 3a
    5. (inter-)Connectivity
    6. Regions that receive more information from many brain areas, and send outputs to many areas, may be integrating higher-level features of the environment, or producing higher-level instructions for action. Thus highly interconnected areas may operate with greater degrees of abstraction.
      • 5a
  3. Psychological

    1. Attention
    2. Attention is the ability to select certain features of the environment, at exclusion of others, for further processing. Further processing can mean, for example, the production or priming of an action, or storage in memory. Although attention is obviously confounded with just those things that we can observe psychologically, it correlates well with common descriptions of how conscious representations should behave.
    3. Linguistic capability
    4. The ability to manipulate the order of symbols to change meanings allows a potentially infinite set of token combinations, and therefore meanings. This faculty, which appears all-or-none, means that a single mental token can be put to any use, and can be involved in any semantic operation. So far it seems that humans are the only organisms that can uncontraversially do this.
      • 6b
    5. Awareness/Global workspace
    6. According to supporters of this theory, representations are conscious when their information becomes accessible by many cognitive systems; in particular when sensory information is in a form where it is verbally articulable, can be stored in long term memory, and can be used to guide flexible actions, it has reached this state. It is said to be in the 'global workspace'. Although accessibility across cognitive 'modules' seems to be confounded with our ability to probe that information in day-to-day situations, it does correlate well with our descriptions of what a conscious representation should be like.
    7. Executive control
    8. The description of certain behaviours as 'automatic' and others as 'controlled' emphasises the distinction between effortful and effortless processes, and volition, as opposed to processes we are or are not aware of. Unlike sensory awareness, it is relatively easy to know whether subjects are 'in control' because we can measure how well they can override automatic actions. There seem to be critical differences between executive control and awareness as markers of consciousness.
    9. Short term memory
    10. At any moment, the entities that are 'at the forefront of the mind' are said to be in working memory. Those items are ephemerally present to us, and can be manipulated actively in the mind. Such items are subject to characteristic patterns of interference and degradation, and maintainence or filtering appears to require certain limited cognitive resources. This kind of active representation may characterise conscious representations.
    11. Binding
    12. When multiple features are present in the environment, registering to a certain degree of depth seems to require associating that feature with its context, for example its location in time and space. According to certain representation models, the product of this association has a unitary nature that is best described as an 'object' or 'object file'. It has been suggested that this process of binding features together into objects is how conscious representations are generated.
      • 2a
  4. Quantum

    1. Wave function collapse
    2. According to a common interpretation of quantum mechanicals, the wave function of a system descripbes superpositions of multiple states of the world - and these all have some objective existence, with varying probabilities. When measurements are made, the wave function appears to collapse stochastically to a reduced set of states. From a subjective viewpoint, the moment that collapse occurs must be at the moment I know the result of the measurement. Consciousness therefore appears to infect every world state that can be measured, suggesting that what I know determines how physical processes proceed.
      • 6a
    3. Electron tunnelling
    4. It has been suggested that quantum computations can be devised that perform highly parallel algorithms almost instantaneously. One theory implicates electron movements in microtubules within neurones as quantum computational devices.
  5. Computational

    1. Algorithmic complexity
    2. If a person's brain were replaced cell-by-cell by electronic components that perform the same function, it seems that their thought processes (and consenquentially consciousness) is preserved. Such arguments indicate that computers that emulate thought have this kind of equivalence with real minds. Strong artificial intelligence arguments go one step further and suggest that this equivalence is sufficient for consciousness: as long as the algorithms of the mind are being actively performed by any device, that device can be said to be conscious. The Turing test could be seen as a practical spin-off from this view.
    3. Topological geometrodynamics
    4. ??
  6. Philosophical

    1. Free will
    2. The feeling of being free (to act, choose, or think) seems tightly bound to being conscious. Compatibilists often take this sense in which we are free as a datum, although psychologists treat it as a datum to be explained in a critical manner.
    3. Communication
    4. To be conscious may require a degree of self-representation, which can only arise by acquiring the linguistic tokens to be able to represent distinctions such as self and other.
    5. Time
    6. Relating the tensed perspective (looking out from the subjective present) and untensed perspective (events being before and after one another) requires a mechanism by which the subjective now is placed in relation to particular temporal events. Such a mechanism would need to explain why there is only ever a 'here and now', but would thereby go a good way to explaining the subjectivity of consciousness.
    7. Intentionality/functionalism
    8. The strong claim that there is no more to consciousness than how we represent ourselves as being conscious has understandably been met with some disappointment. However such an argument does answer some of a dualist's questions, in particular it may be able to provide a critical assessment of why we experience consciousness as such a peculiar thing. It is, of course, not easy to argue against functionalism, even when one's intuitions may oppose it.