(1) Neurological Function
(a) Supervenience
Behavioural decisions in brains are the result of lots of smaller local ‘decisions’ (selections of alternatives), not unlike like the behaviour of an ant colony is the result of lots of smaller local ‘decisions’ by individual ants. Upper organisational levels do not control lower organisational levels — they are different views of the same system.
(b) Consciousness
Behavioural choice does not require that the monitoring that is consciousness affect the selection process at the time of choosing. Instead, the monitoring of past decisions alters the probabilities of future selections.
(2) UnPredictability
(a) Causation
The question of whether one event causes another can be understood as an assessment of usuality or probability. A necessary, though not sufficient, condition of causation is correlation. For example, in terms of temporal relations, the proposition ‘event x is always followed by event y’ is a candidate for a causal relation, as is the conditional proposition ‘if event x occurs, then event y always occurs’, and the accompaniment proposition ‘event x and event y always accompany each other’.[1]
However, correlation is insufficient in itself, since two correlated events may be the joint effects of another event, or series of events. A second necessary condition of causation, then, is a chain of interactions between two events. Combining the two necessary conditions, it can be said that causation involves a chain of interactions from event x such that event y eventuates with a probability of 1.0.[2] Causation can thus be seen as located within a larger scheme whereby events alter the probabilities of other events occurring.
(b) NonLinearity
As argued in previous chapters, behaviour — semiotic or otherwise — is nonlinear. The dynamics of behavioural systems involve feedback, and so enable outputs (motor expressions of potential) to be disproportionate to inputs (ongoing value-categorising of sensory experience). This means that behavioural sequences can be ‘sensitively dependent on initial conditions’ such that slight variations in at given phase can lead to vastly different endpoints.[3] That is, behavioural processes are more like open-ended meteorological processes than developmental processes, since the latter move towards endpoints already established in the evolution of the species. With sensitive dependence on initial conditions, according to Ruelle (1993: 45), even though a trajectory is precisely determined by the initial condition:
there is a fundamental limitation in predicting the trajectory. We have determinism, yet long-term unpredictability. That is because we know the initial condition with a certain imprecision…we don’t know, therefore, which of the possible predictions is correct.
Nonlinear systems are unpredictable in the long-term, but are explainable in hindsight. The system has to play itself out to reveal ‘what its future was’.[4] Long-term unpredictability subsumes both chance and determinism, as Ruelle (1993: 48) points out:
An essential point made by Poincaré is that chance and determinism are reconciled by long-term unpredictability. Here it is in one crisp sentence: A very small cause, which escapes us, determines a considerable effect which we cannot ignore, and we then say that this effect is due to chance.
(3) Modality
The ability to choose specific behaviours can be described in terms of the two systems of modality: modalisation and modulation.[5] Modalisation includes the notions of probability and usuality, while modulation includes those of obligation and inclination. In a given situation, the selection of a specific behaviour might be certain (high probability), probable (median probability) or possible (low probability). A choice might occur always (high usuality), usually (median usuality) or sometimes (low usuality). A choice might be required (high obligation), expected (median obligation) or allowed (low obligation).[6] A chooser might be determined (high inclination), keen (median inclination) or willing (low inclination). Obligatory behaviours may, of course, be either consistent or inconsistent with the inclination of the chooser.
(4) Fitting To Context
(a) Somatic Contexts
Value systems, which link the behavioural potential of neurological systems with homeostatic systems responsible for maintaining equilibrium states in the body, bias behavioural choices. In doing so, they reduce the number of options available from moment to moment. That is, the range of choice varies with the degree of somatic disequilibrium during behaviour. The less their needs are met, the fewer choices behavers have.
For example, someone in a state of extreme hunger has fewer choices than the well-fed; someone in a state of fear[7] has fewer choices than the unafraid; someone consumed by a grievance has fewer choices than those without grievances.
Because of this, the behaviour of individuals is less variable, and so more predictable, in situations where bodies are at far-from-equilibrium states. The smaller the range of options, the more robotic behaviour appears.[8]
(b) Social Contexts
In a specific social context, the range of behavioural options can vary with the degree to which metafunctional consistency is required. Ideational consistency includes behaving only in ways that are appropriate to the situational field (what is happening); interpersonal consistency includes behaving only in ways that are appropriate to the situational tenor (who are taking part); and textual consistency includes behaving only in ways that are appropriate to the situational mode (what rôle the behaviour is playing).
Metafunctional consistency also applies to acting in a self-consistent manner. Ideational consistency includes acting in ways that are experientially and logically self-consistent; interpersonal consistency includes acting in ways that are self-consistent in relation to social values and group loyalties; textual consistency includes acting in ways that are self-consistent in terms of what is foregrounded and backgrounded with regards to attention.
In any given situation, the range of appropriate choices also depends on where the behavioural turn is temporally located in relation to other behaviours the sequence of social interactions. That is, the behavioural choices of others can expand or contract the range of choices of each individual.
Footnotes:
[1] ‘If…then’ is the recognition relation: if perceptual categorisation x, then perceptual categorisation y follows; this enables anticipation of future events on the basis of past experience, across species.
[2] Note that this probability does not hold at the quantum level.
[3] And an individual does not necessarily have the same probability of behaving the same way in the same context on different occasions.
[4] The future state of persons, like all evolving systems, is indeterminate.
[5] An animal does what it can do and what it can’t help doing.
[6] Social obligations include the informal and formal, the implicit and explicit, the personal and institutional, the moral, the legal, and so on. Access to making choices depends on power relations (tenor) in social contexts.
[7] As politicians seeking election know only too well.
[8] Other animals’ behaviours will appear robotically predictable when examined by putting them in extreme situations — value disequilibria requiring immediate satisfaction — such as experimental conditions. Their complexity only becomes obvious to humans when examined ethologically across a range of behavioural contexts and value states.