Abstract

The concept of automaticity, central to an understanding of skilled performance, has been invoked when accurate performance is faster than expected by Cognitive Psychologists. Unfortunately, there has been an inherent risk that the construct of automaticity is inferred from the very same performance to be explained. To avoid such circularity, theorists had suggested two distinct modes of processing which possessed properties additional to those to be explained. Controlled processes were slow, intentional and aware, while automatic processes were fast, obligatory and beyond awareness. Nevertheless, such distinctions are only of use if they are internally consistent, and the observation of exceptions to a neat two-process model of processing led to a search for other ways to conceptualise the automaticity of skilled performance. Failures in the automatic execution of behaviour dramatically illustrate the existence of automatic processes. Examples of failures of automaticity from the domains of neurology or engineering psychology, substantiate the existence of automatic processes, illustrate the importance of automaticity, and may provide insights into associated mechanisms. Conditions such as Parkinson's disease serve as neuropsychological demonstrations of the inability to automatically execute movements, while similar problems are observed during interactions within artificial environments (e.g. graphical user interfaces or simulators). A better delineation of automatic processes provides insights as to the nature of information processing and skilled performance. Skilled performance can reflect superior algorithms, or may be a property of the mechanism (akin to analog computation), but it is the flexibility associated with the evolved facility to use tools that should be of greater interest to psychologists. A consideration of the human factors limiting performance in designed environments provides future directions for research.