ABSTRACT
It seems intuitively obvious that when faced with the task of holding information actively in mind during a delay, participants will verbally rehearse this information wherever possible, in order to maximize their chances of remembering it successfully at the point of recall. For example, Spiker (1956) tested young children’s memory during a delay by hiding a coin in one of two differently patterned boxes that were both fi xed to a rotating disk. The disk was then rotated for 25 seconds at a speed at which it was impossible to visually track the baited box. Participants could therefore only fi nd the coin by remembering which box had been baited initially. While some participants were required to do this solely on the basis of their memory of the visual pattern shown on the correct box, half of the participants were pre-trained to associate different verbal labels with each location. Children in the latter group showed superior recall, leading Spiker to suggest that “the possession of verbal names for the stimulus permits [the participant] to produce a representation of the absent stimuli during the delay period” (p. 111) or in other words, to verbally rehearse the identity of the target box. Although one certainly can maintain information in a visual form in memory tasks (Logie, 1995; Pavio, 1971), participants tend to elect to maintain information verbally wherever possible. In part this may follow from the benefi ts that can be gained from a reduction of information by coding stimuli with simpler verbal labels; for example Glanzer and Clark (1962, 1963) proposed the “verbal loop hypothesis” which stated that participants engage in covert naming of visual stimuli in order to improve recall. However, it also appears that verbal rehearsal of information within immediate memory tasks is a relatively effi cient way of maintaining this information for subsequent recall (Peterson & Peterson, 1959). Models of immediate, or short-term memory have therefore often included verbal rehearsal as a means of item maintenance (Atkinson & Shiffrin, 1968; Baddeley & Hitch, 1974), perhaps most explicitly in Baddeley’s (1986) model of working memory. In this model the phonological loop supports the short-term maintenance of verbal information, and consists of a limited capacity phonological store and a subvocal, phonologically mediated rehearsal process. According to Baddeley, rehearsal offsets forgetting of information that would otherwise be caused by
trace decay. Two characteristic phenomena of the phonological loop result from this conception. First, items that take longer to rehearse, such as words of a long spoken duration, are subject to greater forgetting than items that take less time to rehearse, leading to the “word length effect” (Baddeley, Thomson, & Buchanan, 1975). Indeed, if adults are required to engage in concurrent articulation of other material, which is presumed to block rehearsal, the word length effect is eliminated (Baddeley, Lewis, & Vallar, 1984; though see Romani, McAlpine, Olson, Tsouknida, & Martin, 2005). Second, the faster an individual can rehearse material the less forgetting they experience. Subvocal rehearsal rates are assumed to relate to the rate at which individuals can overtly repeat verbal material (Landauer, 1962; Lovelace, Powell, & Brooks, 1973), and a correlation between speech rate and immediate memory span is therefore predicted by this model (Baddeley et al., 1975). A third potential marker of rehearsal is the presence of a “phonological similarity effect” for visually presented material. The standard phonological similarity effect is the fi nding of poorer recall for phonologically confusable verbal items (e.g., hat, cat, mat, map, tap) than phonologically distinct items (Conrad & Hull, 1964). When such an effect is seen for visually presented stimuli it indicates that the participant has recoded the presented information into a phonological form. A number of authors have argued that the process of verbal recoding is analogous to rehearsal, as it involves the participant subvocally naming, or re-presenting, the information to themselves (Gathercole, 1998; Howard & Franklin, 1990; Vallar & Papagno, 1995; though see Crowder, 1976, p. 86). The current chapter examines the development of rehearsal in two ways. First, it evaluates the evidence for each of the three potential markers of rehearsal in the short-term memory performance of children of different ages; and, in particular, focuses on evidence from studies from our research group that have been carried out among children showing atypical development. Second, it considers the impact of changes in the extent of children’s rehearsal on their performance on working memory measures. While short-term memory refers to individuals’ ability to maintain information actively in mind, working memory involves doing this in the face of additional potential distraction from processing requirements associated with the task (Jarrold & Towse, 2006). For example, while short-term memory is typically assessed using so-called “simple span” tasks that require the immediate recall of just presented storage items, working memory is often measured using “complex span” procedures that combine the presentation of to-beremembered storage items with the requirement to complete additional processing operations (e.g., Daneman & Carpenter, 1980). As will be discussed below, working memory measures therefore tap a greater range of abilities than do short-term memory tasks, but both share the need to maintain storage materials (Bayliss, Jarrold, Gunn, & Baddeley, 2003; see also Unsworth & Engle, 2007a), and are therefore both open to the potential infl uence of rehearsal processes.
