Academic journal article Psychonomic Bulletin & Review

Task Rules, Working Memory, and Fluid Intelligence

Academic journal article Psychonomic Bulletin & Review

Task Rules, Working Memory, and Fluid Intelligence

Article excerpt

Published online: 18 July 2012

© The Author(s) 2012. This article is published with open access at

Abstract Many varieties of working memory have been linked to fluid intelligence. In Duncan et al. (Journal of Experimental Psychology:General 137:131-148, 2008), we described limited working memory for new task rules: When rules are complex, some may fail in their control of behavior, though they are often still available for explicit recall. Unlike other kinds of working memory, load is determined in this case not by real-time performance demands, but by the total complexity of the task instructions. Here, we show that the correlation with fluid intelligence is stronger for this aspect of working memory than for several other, more traditional varieties-including simple and complex spans and a test of visual short-term memory. Any task, we propose, requires construction of a mental control program that aids in segregating and assembling multiple task parts and their controlling rules. Fluid intelligence is linked closely to the efficiency of constructing such programs, especially when behavior is complex and novel.

Keywords Attention * Executive control * Working memory * Individual differences * Memory capacity

Tests of "fluid intelligence," such as Raven's Progressive Matrices (Raven, Court, & Raven, 1988) and Cattell's Culture Fair (Institute for Personality and Ability Testing, 1973), are important for their broad ability to predict success in many different kinds of cognitive activity, from laboratory tests to educational and work achievements. Typically, fluid-intelligence tests involve novel reasoning, using geometrical, verbal, or other materials (Marshalek, Lohman, & Snow, 1983). A large research literature has investigated what basic cognitive mechanisms are measured in tests of this sort. In particular, strong links have been suggested between fluid intelligence and working memory (Ackerman, Beier, & Boyle, 2005; Kane & Engle, 2002; Kyllonen & Christal, 1990).

Working memory, however, is a complex concept. It is often proposed that working memory can be fractionated into distinct components, including somewhat separate short-term stores for materials of different kinds (Baddeley, 1986), as well as distinct processing control functions, such as resistance to interference (Kane & Engle, 2002) and cognitive updating (Miyake et al., 2000). Not surprisingly, correlations with fluid intelligence vary widely from one test of working memory to another (Ackerman et al., 2005).

Previously, we have investigated a form of working memory concerning the learning and use of new task rules. The ability to follow new task rules shows substantial individual differences (Duncan et al., 2008). In some cases, failures take a striking form that we have called "goal neglect" (Duncan, Emslie, Williams, Johnson, & Freer, 1996): Though rules can be explicitly recalled, they exert no apparent control over behavior. Critically, neglect increases with the task complexity-that is, the number of rules that the task requires (Duncan et al., 1996; Duncan et al., 2008). This limit, however, seems rather different from capacity limits that have been found in many, more traditional aspects of working memory. In particular, neglect concerns not the processing requirements of an individual trial, but rather the total complexity of the whole set of rules described in the initial instructions. Neglect of a given rule increases with the number of other rules described, even if participants know that, for a given block of trials, these other rules will not be required and can be ignored (Duncan et al., 2008).

Any multistep behavior must be controlled by some internal mental program defining and focusing on separate task parts. At each stage of such a program, relevant information and operations must be assembled, and irrelevant material disregarded, creating an organized series of distinct attentional episodes (Duncan, 2010a). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.