Simulations of Key Experiments

Research activities

View/download

Case Studies

The revised model of working memory – the episodic buffer

Baddeley and Hitch (1974) proposed the model of working memory with three components: the visuo-spatial sketchpad, the phonological loop and the central executive. This model has been successful in accounting for a broad range of data from laboratory studies of immediate recall, as well as data obtained from neuropsychological, developmental and neuroimaging studies. However, more recently, Baddeley (2000) has added a fourth component to the model, the episodic buffer, which is said to play a role in the processes of integrating information between the subsystems and the central executive.

As successful as the earlier formulation of the model has been, Baddeley concedes “there have always been phenomena that did not fit comfortably within the Baddeley and Hitch model” (p. 417). In particular, there are data that are problematic for the phonological loop.

Articulatory suppression

For example, in studies of articulatory suppression, in which the participant repeats the word “the” while learning a visually presented list of numbers, the model predicts that recall will be poor. This is predicted based on the assumptions that (1) visual information has indirect access to the phonological loop (via subvocal rehearsal), and (2) articulatory suppression should prevent visual information from gaining access to the phonological loop. However, the data show that articulatory suppression only results in a small reduction in recall, typically from seven to five digits (Baddeley et al., 1984). In addition, in studies of brain-damaged patients, individuals with impaired short-term memory show better recall for visually presented digits than for auditorily presented digits. The older version of the model is unable to account for these findings, not even by including an explanation based on the visuo-spatial sketchpad. This structure is assumed to capture a single representation of a complex pattern and to be poor at storing serial information.

Recall of prose

When participants are presented with a list of unrelated words, recall is limited to about five or six items. However, when words are related (as in prose) correct recall can be increased to 16 or so items (Baddeley et al., 1987). A recognised explanation for this is that prose utilises information in long-term memory, which has the effect of reducing the amount of information needed to be recalled into smaller chunks, whereas unrelated words do not. Chunks of information are not represented easily in the older version of the model, since it is not a function of the phonological loop, nor do chunks reside in long-term memory. Similarly, in studies of the recall of prose, typical recall is of about 15 to 20 idea units. This amount exceeds the capacity of the phonological loop; it cannot be accounted for by the storage of a pattern in the visuo-spatial sketchpad; and the central executive is assumed to have no storage capacity at all. Retention of this amount of information seems to involve structures within long-term memory.

Subvocal rehearsal

A key component of working memory is that there are separate processes for storage and retrieval. Recall is said to be achieved through subvocal rehearsal of the to-be-remembered items. Support for this comes from the finding that articulatory suppression does not affect the word-length effect (a list of long words is more difficult to recall than a list of short words) when items are presented auditorily, yet it does when items are presented visually. In the case of visual presentation, items cannot be rehearsed subvocally and hence recall is independent of the length of the word. However, what is problematic for the theory is that, although children only begin to show signs that they use subvocal rehearsal by the age of 7, some form of rehearsal occurs in children as young as 3 years old. Such rehearsal cannot be explained by the function of the visuo-spatial sketchpad.

The binding problem

Every object has a number of features, such as its physical shape and size, its location, its colour, whether it is moving and so on. When we perceive more than one object, in most cases we correctly associate the right features with the right object. The binding of an object to its associated features, which is assumed to take place in the brain, serves to avoid incorrect combinations of features of objects. It is not clearly understood how binding is achieved. Baddeley has suggested that working memory plays an important role in binding through the role of the central executive, since the model is essentially concerned with the integration of information from more than one modality. However, since the central executive has no “short-term multi-modal store capable of holding such complex representations” (Baddeley, 2000, p. 421), it is inadequate in explaining the binding problem.

The episodic buffer

The solution to the problems of the previous formulation of working memory is the episodic buffer, which has the following features:

• It is of limited capacity.
• It is a temporary storage system.
• It can integrate information from a variety of sources (perception and long-term memory).
• It is controlled by the central executive through conscious awareness.
• It holds episodes of information that are across space and over time.
• It plays an important role in the storage and retrieval of episodic long-term memory.
• It may reside in the right frontal areas.

In sum, the episodic buffer takes the form of a short-term episodic memory. Whereas the previous formulation of the model focused on separating out the individual components of working memory, the emphasis of the new model is now on the integration of information. The episodic buffer is also suggested to form “the crucial interface between memory and conscious awareness” (Baddeley, 2000, p. 422).

References

Baddeley, A.D. (2000). The episodic buffer: A new component of working memory? Trends in Cognitive Science, 4(11): 417–23.

Baddeley, A.D. & Hitch, G.J. (1974). Working memory. In G.H. Bower (ed.), The psychology of learning and motivation: Advances in research and theory (pp. 47–89). London: Academic Press.

Baddeley, A. D., Lewis, V. & Vallar, G. (1984). Exploring the articulatory loop. Quarterly Journal of Experimental Psychology, 36: 233–52.

Baddeley, A.D., Vallar, G. & Wilson, B.A. (1987). Sentence comprehension and phonological memory: Some neurophysiological evidence. In M. Coltheart (ed.), Attention and performance XII: The psychology of reading (pp. 509–29). Hove: Lawrence Erlbaum Associates.

Flashcards

Multiple Choice Quiz