Gordon B. Moskowitz
Preconscious Control and Compensatory Cognition
It is a general principle in Psychology that consciousness deserts all processes when it can no longer be of use. The tendency of consciousness to a minimum of complication is in fact a dominating law … Now if we analyze the nervous mechanism of voluntary action, we shall see that by virtue of this principle of parsimony in consciousness the motor discharge ought to be devoid of sentience. If we call the immediate psychic antecedent of a movement the latter's mental cue, all that is needed for invariability of sequence on the movement's part is a fixed connection between each several mental cue, and one particular movement. For a movement to be produced with perfect precision, it suffices that it obey instantly its own mental cue and nothing else, and that this mental cue be incapable of awakening any other movement. Now the simplest possible arrangement for producing voluntary movements would be that the memory-images of the movement's distinctive peripheral effects, whether resident or remote, themselves should constitute the mental cues, and that no other psychic facts should intervene or be mixed up with them.
– James (1890/1950, pp. 496–497)
Control is typically equated with deliberate, conscious, and effortful exertion of the will. By its common use it would seem that a controlled process is the opposite of an automatic process. But can control, volition, and goal pursuit occur automatically? The possibility that automaticity and control are not polar opposites is neither a radical nor novel thought (see, Bargh, 1990; James, 1890/1950, above). The cocktail party effect is a good example of individuals preconsciously scanning the environment for information relevant to the self, presumably because this is information they are motivated to detect. Postman, Bruner, and McGinnies (1948) found that words that people valued were perceived more readily (lower thresholds of recognition) than words less valued, despite these words being presented outside of conscious awareness. Similarly, Kelly (1955) discussed individuals as scanning the environment for “blips of meaning.” In the early years of the cognitive revolution in social psychology, Markus (1977) showed words relevant to the self-concept were responded to more quickly on a reaction time task, whereas Higgins, King, and Mavin (1982) and Bargh (1982) found that self-relevant words that were chronically accessible to individuals guided attention and judgment. What these experiments share is a notion of chronicity: Although some deal with the self-concept, others with values, and yet others with volition, they all posit that, over time and with practice, these concepts come to operate prior to conscious awareness and without conscious intent.
The current chapter examines the issue of whether goals are activated and operate preconsciously, extending this discussion to temporary goals (as opposed to those that are chronically accessible). Temporary goals are posited to have preconscious effects when (a) triggered directly by cues in the environment (due to associative links between the goal and its relevant cues; e.g., Ach, 1935; James, 1890/1950) and (b) a discrepancy between one's desired and actual level of goal attainment triggers cognitive activity aimed at reducing this discrepancy and focuses the cognitive system toward goal relevant stimuli (a process labeled compensatory cognition).
The reason the notion of preconscious control does not seem intuitively obvious is partly a problem of language. Definitions of control, volition, intention, and goals (see below) typically mention conscious choice as a central feature. But the belief that control is conscious and effortful is not merely based on the use of these words in the vernacular. This assumption is also made explicit by models in social psychology that place control on the opposing endpoint of a continuum with automatic processing. And this conception of control is not unique to social psychology, but has its roots in cognitive psychology, where social-psychological approaches to automaticity evolved (e.g., Posner & Snyder, 1975).
Before addressing the adequacy of a dichotomous conceptualization of control and automaticity, and the more intriguing question of whether control can be automatic, we must first address what is meant by automaticity and control.