Memory & Cognition 1984,12 (1),71-83 Developing the theory of working memory JOHN T. E. RICHARDSON Brunei University, Uxbridge, Middlesex, England The theory of working memory was devised to explain the effects of a concurrent memory load in various experimental situations in terms of the operation of a central executive processor and a phonemic response buffer. It also explains the effects of phonemic similarity, articula- tory suppression, word length, and unattended speech. Experiment 1 demonstrated that a con- current memory load markedly reduced the phonemic similarity effect in immediate serial recall, which was taken to support the concept of a limited-capacity phonemic response buffer. A more detailed analysis of the results suggested that a concurrent memory load may affect the storage capacity of the central executive processor and the translation of orthographic stimuli into phonological representations, as well as the storage capacity of the phonemic response buffer. Experiment 2 showed that a concurrent free-recall task reduced the phonemic similarity effect in immediate serial recall, but only in the case of visually presented sequences of items. More- over, unattended speech was found to have no effect upon performance in immediate free recall. These results were taken to imply that the phonemic response buffer contributes only to perfor- mance in cognitive tasks that require the accurate retention of serial-order information. A series of experiments reported by Baddeley and Hitch (1974) investigated the effects of a concurrent serial-recall task upon performance in tests of reasoning, comprehension, and free recall. Such a "concurrent memory load" was expected to "absorb some of the stor- age capacity of a limited capacity working memory sys- tem" (p. 50), and thus to impair performance in any cri- terion task that relied upon such a system. Baddeley and Hitch found that a concurrent memory load of six items reduced performance in all three sorts of task. They also showed that phonemic similarity among the stimulus items impaired both reasoning and comprehension, and that suppressing any relevant articulatory activity by requiring the subjects to produce irrelevant vocalizations impaired performance in free recall and, to a lesser extent, in reasoning. These results were taken to support the notion of working memory as a short-term information- storage system that had access to phonemic coding. Such a notion was entirely compatible with the prevalent information-processing approach to human cognition, and in itself involved no radical departure from previous theoretical accounts of short-term mem- A presentation on which this paper is based was given to an International Conference on "Memory" organized by the Cog- nitive Psychology Section of the British Psychological Society at Plymouth Polytechnic on September 9, 1981. I am indebted to Alan Baddeley, Graham Hitch, Stephen Monsell, and two anon- ymous reviewers for their comments, advice, and encourage- ment, and also to colleagues and students of Brunei University and the Open University. I am grateful to Shareen Courtney, Tom Marden, and Nick Abbot for their assistance in running Experiment 1, and to Irene Short, Janet Pearson, and Linda Burman for their assistance in running Experiment 2. My mail- ing address is: Department of Psychology, Brunei University, Uxbridge, Middlesex UB8 3PH, England. 71 ory. In fact, it is now widely accepted that the span of immediate memory is determined by the available storage capacity of working memory, conceived as a general information-processing system. Nevertheless, Klapp, Marshburn, and Lester (1983) pointed out that it is difficult to reconcile this assumption with the rather poor correlations that are typically found between digit span and full-scale general intelligence, or with the existence of neurological patients who demonstrate a selective impairment of verbal short-term memory in the absence of any wider impairment of overall cognitive abilities. These findings entail that there exists a limited- capacity memory component that is not involved in general cognitive processing. Klapp et al. reported a series of experimental investigations (to be mentioned later in this paper) that cast further doubt upon the existence of a unitary mechanism of working memory that has both processing and storage functions in human cognition. A more viable alternative is to provide an articulated theory of working memory as a system of interrelated mechanisms that are differentially responsible for these functions. In fact, Baddeley and Hitch (1974) elaborated their account by distinguishing between a relatively active, flexible, central executive processor and an optionally employed, although relatively passive, pho- nemic response buffer. The former was assumed to be responsible for information processing and decision making, whereas the latter was regarded as a slave sys- tem that "is able to store a limited amount of speech- like material in the appropriate serial order" (see also Baddeley, 1976, p. 176, and Baddeley & Lewis, 1981). In other words, although the properties of the central executive processor were left somewhat unclear, the phonemic response buffer was a distinctive hypothetical Copyright 1984 Psychonomic Society, Inc.