ABSTRACT
The ability to apply strategic control is commonly used as a criterion for distinguishing between knowledge which is consciously available and knowledge which is not. For instance, methods based on the Process Dissociation Procedure (Jacoby, 1991) – in which performance is compared under conditions where participants attempt to apply versus withhold knowledge – are frequently used as criteria for determining the extent to which learning is implicit or explicit. In this chapter we first present a brief overview of existing methods that have been developed for measuring strategic control in the serial reaction time (SRT; Nissen and Bullemer, 1987) and artificial grammar learning (AGL; Reber, 1967) tasks. Even though the described methods are traditionally used to determine the conscious status of acquired rule knowledge, it has also been shown that strategic control can occur in the absence of detailed conscious awareness of the learned regularities (Fu, Dienes and Fu 2010; Norman, Price, Duff, and Mentzoni, 2007; Norman, Price, and Jones, 2011; Wan, Dienes, and Fu, 2008). This challenges the assumption that strategic control requires conscious awareness (Baars, 1988; Jacoby, 1991). We present the results from an experiment that specifically addresses whether knowledge of learned grammars can be applied in a strategically controlled manner even when structural knowledge is not conscious. The procedure combines two different measures of consciousness of structural knowledge of grammar rules. These are the trial-by-trial evaluation of decision strategy (Dienes and Scott, 2005) and verbal report of the nature of the grammars in a situation where the grammar rules can logically be related to any of three stimulus dimensions (Norman et al., 2011). The results indicate that strategic control may occur even when participants express global unawareness of the nature of the rule that governs letter strings, when this is measured by self-report after the experiment.
