baddeley_2000 - Short-Term and Working Memory ALAN...

Info iconThis preview shows pages 1–16. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 16
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Short-Term and Working Memory ALAN BADDEL‘EY The study of short-term memory, the retention of small amounts of information over brief time intervals, formed a major component of the development of cognitive psychology dur- ing the 19605. It had a strong theoretical com- ponent, derived from the increasingly influen— tial computer metaphor, combined in Britain at least with a concern for application to prob- lems such as those of air traffic communica— tion (Broadbent, 1958) and of coding in tele- phony and postal systems (Conrad, 1964). The attempt to develop information—processing models of short—term memory (STM) led to some major controversies (see below). Unfor- tunately, resolving these issues unequivocally proved beyond the capability of the methods available at the time, resulting in a decline of interest in STM during the 1970s, and subse— quently even to a declaration of its demise (Crowder, 1982). However, as the old concept of STM was losing favor, it became incorporated within a more complex framework, working memory (WM), which proposed that the older concept 61? 77a unitary ’sfir’rflié' replaced by ‘a’multico’m'f ponent system that utilized storage as part of its function of facilitating complex cognitive activities such as learning, comprehending, and reasoning (Baddeley 8: Hitch, 1974). Inter- est in WM continued to develop through the 19805, though with somewhat different em— phases on different sides of the Atlantic. Dur- ing the 19905, the Whole area has received a further boost from the development of func- tional imaging techniques, with the compo- nents of working memory offering an appro- priate level of complexity for the developing techniques of brain scanning. This develop- ment was facilitated by the very fruitful rela- tionship between cognitive psychology and the neuropsychology of working memory, which provided hypotheses as to which areas of the brain might be most likely to be in- volved in particular tasks, together with con— cepts that facilitate the linking of the neuro- anatomy to a coherent cognitive framework. Finally, and coincidentally, some of the old applied problems are now beginning to resur- face. In both the United States and Britain, for example, there is currently considerable con- cern about the best way to extend the ever—ex— panding series of telephone codes so as to op- timize capacity without unduly increasing length, while new areas such as pharmaceuti— cal prescribing errors are beginning to high- light the need for an understanding of the pro- vcessesvinvolvedjvvanditor*draw fiupon ethe“ empirical work of the 19605 (Lambert, 1997). As a result of these variousdevelopments, there is a growing’intere’stintthetfieldfiof 51PM and WM fiom scientists Whose principal train- ing has been in other areas, but who wish to incorporate measures of short-term and work- ing memory into their work. Finally, the area The strer memory h thinkers ft decades it gence of a 'first on be] on the ne\ These rece that are iii their impli The C out these t the theori discover-h described, professio’r ing, and c Endei world—cla: handbook exciting f: chapters : who prov and outlil ahead. The l examinin: of memo ' ' ory in th memory memory, ideal for hi? We, W e flineurosdu psycholo 78 MEMORY IN THE LABORATORY continues to attract good young researchers who see the study of working memory as an important interface between research on mem— ory, perception, and attention. While there are many overviews of the area, ranging from the relatively brief (Badde- ley, 1992) to the chapter length (Baddeley, 1996) and the book length (Gathercole, 1996; Miyake 8: Shah, in press), neither these nor journal articles tend to include thewsort of, practical detail thatmis so important if one wishes to carry out or evaluate experiments. The present chapter aims to go some way to- -ward filling this gap, while bearing in mind that the only way to fully understand a tech- nique is to use it. Here STM and WM are treated separately, since the relevant tech- niques, driven by the specific theoretical is- sues of the time, tend to be somewhat differ- ent. However, it is important to bear in mind that they do form part of the same tradition, and that it is increasingly common for 19605 techniques and methods to find new uses in ' the 1990s. It may be useful, however, to begin with some terminology. Terminology The division of memory into two or more sys— tems was proposed by William James (1890), who distinguished between primary memory, which he regarded as closely associated with conscious awareness, and secondary memory, which referred to more durable memories. When the interest in fractionating memory re- vived in the late 19505, the term STM was used to refer to tasks in which small amounts of material were retained over brief intervals, in contrast to LTM, which involved retention over more than a few seconds. It subsequently became clear that performance on STM tasks was not a pure reflection of the hypothetical underlying system, but was also influenced by LTM. To avoid confusion, some investigators used different terms to refer to the hypotheti— cal underlying theoretical memory systems, such as short—term and long-term store [STS and LTS: Atkinson 8: Shiffrin, 1968), or revert~ ingwtb primafy’and'secondaiy memoiytwaag W & Norman, 1965). In recent years, the term workingrmremprz proposed, by iMillerjfilialanWter, and Pribram [1960) has been developed to emphasize the functional role of STM as part of an integrated system for holding and manipulating informa- tion during the performance of complex cogni— tive tasks (Baddeley & Hitch, 1974). Unf( tunately, the same term has been usl independently within the animal learning 1 erature, where it refers to situations in Whit the animal needs to retain information acro several trials during the same day [Olto Walker, & Gage, 1978), almost certainly i volving different mechanisms from these i volved in the typical human WM task. Final] the productionisystem approach to comput tional modeling proposed by Newell and 5 mon (1972) postulates a working memory unlimited capacity, although this is not a sumed to be related to the limited capaci' STM system proposed by experimentalis such as Baddeley and Hitch (1974). Fort nately, the context is usually sufficient ' avoid too much confusion between the var ous users of the term. Short-Term Memory Methods and Techniques Before going on to discuss recent theoretic: developments in the area, it may be useful 1 describe some of the rich armament of met} ads and techniques that have been develope to study verbal and visual STM. Verbal 5 TM Memory Span Subjects are presented with a sequence c items, which they attempt to reproduce in th presented order. Typically, digits, consonant: or words are used. Presentation may be visue or auditory, with auditory presentation tend ing to give a slight advantage, particularl‘ over the last one or two items, the so—calleu modality efiect. Rates of presentation typicall; range from 0.53 to 2.5 per item, with 15 proba bly being the commonest. Presentation rate i not a major variable ‘within this range, bu faster rates run the risk of errors owing to fail ure to perceive, while slower rates give suffi cient time for subjects to engage in comple: and often highly 'vari'ableflrehearsal’strategies. Recall may be spoken or written. It is usua to require, the subject to recall in the order 0 presentation, and to monitor that this is ithl 7 case. Performance is typically measured it terms of the maximum level achieved, witl span formally being the point at which tht subject recalls the ordered sequence of item: SHORT-TERM AND WORKING MEMORY on 50% of occasions. This is not easy to deter- mine directly, and hence a number of approxi- mations are used. One simple method is to take the mean of the length of the three longest sequences correctly recalled, so a subject be— ing correct on three out of four 7-item se- quences, and one at length eight, would have a score of 7.33. Memory span has the disadvantage that many of the data collected come from se- quence lengths at which performance is’per~ fect, hence providing little information. More information may be gained from using a pro— cedure in which all sequences are presented at the same length, which should be at or slightly beyond span. Performance can then be scored either in terms of number of sequences com— pletely correct or of number of items recalled in the correct serial position. In his classic‘paper The magic number seven, George Miller (1956) speculated that in the area of absolute perceptual judgments, subjects could typically distinguish about 7 separate categories, while a typical digit span was about 7 items. .He went on to emphasise that this latter conclusion was a gross over- simplification since it was possible to, increase this substantially by chunking, a process whereby several items are aggregated into a larger super item. Perhaps the clearest demon— stration of this is in immediate memory for prose material; memory span for unrelated words is about 5 or 6, whereas With meaning— ful sentences, spans of 16 words or more are not unusual (Baddeley, Vallar, 8: Wilson, 198 7). Syntax and meaning make prose highly redundant, and an early paper by Miller and Selfiidge (1950) showed that the more closely a string of words approximates to English prose, the longer the memory span. However, although absolute number of words increases with approximation to English, Tulving and Patkau (1982) showed that the number of chunks remains constant. Free Recall. This simple task involves presenting the sub- Wijrect with ailist, typically of words, that sub- jects attempt to recalliin Order they Wish. The classic serial position curve shows excel- lent recal1,,9f,,th§11§31fewiifigéithe recency efiect), somewhat better recall of thefirst oil—e7 or two items (the primacy efiect), and a rela- tively flat function between. A brief filled delay will wipe out the re- cency effect While having little effect on ear- 79 lier items. Virtually any variable that will in— fluence long-term learning (e.g., rate of presentation, familiarity of material, the pres— ence of a secondary task, or the age of the sub— ject) will influence earlier items but have little or no impact on the recency effect [Glanzer, 1972). The recency effect reflects a strategy of first recalling the earlier items, and is abol— ished if subjects are dissuaded from this. It ap- ' pears to be a very basic and robust strategy that is found'in young childreniamnesi’cma- tients, and even patients suffering from Alz- heimer’s disease (Glanzer, 1972; Baddeley & Hitch, 1993). The primacy effect is less marked and less robust. It may reflect a number of variables, but in particular the tendency to give more at- tention and possibly more rehearsal to the ini- tial item (Hockey, 1973). While the typical serial position function operates across a Wide range of lengths and presentation rates, most experimenters avoid sequences of less than 10 items, since there is a tendency for subjects to attempt to recall short sequences in serial order. Presentation rate is typically slower than in memory span, since this increases the amount of recall from the earlier long-term part of the curve, with 23 per item being the most common presentation rate. It is also not uncommon to use semanti- cally categorized material, since this again in— creases performance and also gives some indi— cation as to Whether the subject is able to take advantage of meaning (Tulving 8: Pearlstone, 1965). Short—Term Forgetting The classic paradigm here was developed by Peterson and Peterson (1959); their subjects were presented with three consonants and re— quired to retain them over a delay ranging from 0—185, during which they counted back— wards in threes. Performance reflects an STM component that declines over about 5 seconds [Baddeley 8: Scott, 1971), and an LTM compo- nent reflecting the extent to Which items can be discriminated from prior items, the result of proactive interference (PI; Keppel 8: Under— wood; "1962.) . ’PI’can’be'prevented by’changing the type of material to be remembered—for ex- ample, switching from animals to flowers ’[Wifikéfi’s'j’ ’19'70)76r7by i118 ertin‘g"’a’delay”be:""* ""‘TTTWW W Tii'fi tween successive trials (Loess 8: Waugh, 1957), resulting in a recovery of performance (release from PI), followed by a further buildup of PI. The mem think decal genc first on t] WTheé that their out 1 the disc desc ‘ prof I mg, wor han- exc: chaj wh< and ahe exa‘ i 80 MEMORY IN THE LABORATORY Memory Probe Techniques The act of recalling an item can itself produce forgetting, either because the time taken to re- call allows further trace decay or because the recall process disrupts the memory trace. One way of avoiding both of these is through probe techniques, whereby only part of the remem- bered material is sampled. For example, Sper- ling [1963] presented subjects with 3 rows of posal of alternative models (Anderson, 1973; Theios, 1973). With the growth in number of models and a lack of crucial experimental evi- dence, the technique became unfashionable, although it is still quite extensively used as a measure of cognitive deficit following drugs or stressors, for which it provides a neat and rea- sonably sensitive measure. In the absence of any broadly accepted theoretical interpreta- tion, it continues to offer a theoretical chal- 4’l’etter'37At’re’c’all, one of’th’e 3 rows is erred by a tone. Since the subject does not know in advance which row, one can legitimately mul- tiply that score by the number of rows to esti- mate the capacity of the memory system, which is typically greater than that obtained using more standard total recall methods. In a variant of this, Waugh and Norman (1965) presented their subjects with a series of digit strings varying in length. The experimenter then provided one item from the string and re— quired the subject to produce the next in se- quence. Recall performance showed a very clear recency effect, which was minimally af- fected by rate of presentation, suggesting that forgetting was principally due to the limited capacity of the short—term store rather than to temporal trace decay. A variant of the memory probe technique was developed by Sternberg (1966), who used speed of response as a means of investigating the storage of items within the memory span. A digit list ranging in length from 1 to 6 was presented, followed by a probe digit. The task was to decide whether the probe digit had been part of the previously presented se- quence. Reaction time increased linearly with the length of the presented sequence. This oc- curred not only for positive probes but also for negative probes, where‘the item had not been in the list. Sternberg proposed a model based on the analogy of a computer serially scanning its memory store, with the slope of the func- tion relating RT to number items in store pro- viding a measure of hypothetical scanning rate, typically about 40 ms per item. The fact that slopes for “yes” and “no” responses were the same prompted Sternberg to suggest that the search was exhaustive. If subjects could léIigE 7 PM ' '7 '7 Nonverbal 5 TM Research on STM was dominated by verbal tasks, probably because the material is so easy to manipulate and record. However, analogous effects have been shown for visual memory. Dale [1973) required subjects to remember the location of a single point on an open field over a delay filled by verbal counting, finding that accuracy declined steadily over time. Phillips [1974) presented subjects with a matrix of which half the cells were filled, presenting a second matrix for recognition after a filled de- lay varying in length. Performance remained high over the delay for simple 2 X 2 matrices, _ with forgetting becoming steeper as the com— plexity of the matrix increased. In a subse— quent study, Phillips and Christie (1977) pre- sented subjects with a sequence of matrix patterns, observing that only the last pattern showed evidence of excellent initial perfor— mance followed by rapid decay, while earlier matrices showed a low level of performance. This pattern of results, therefore, suggests a short-term visual memory store that is limited to one pattern, with performance on that pat- tern being a function of its complexity. This has been used to develop a measure of pattern span in which the subject is shown a pattern and attempts to reproduce it on a matrix. The test begins with a Z x 2 matrix with half the cells filled, increasing to a point at which the subject 'is no longer able to accurately repro— duc‘e the pattern, which for a normal adult is typically around 16 cells (Della Sala, Gray, Baddeley, & Wilson, 1997). An alternative measure of visuo~spatial *respondra'srs oon'asthey’detecte d’arrnatch’witlr”sp'an ’is ’th’e’Corsi’bl'ock'tapping ’ta’sk’[Mi’lner""’ 7’“ 7 the probe, then the “yes” response slope should be shallower than the “no.” This led to "" 'inte’nsive'experime'ntal Wofk’that'uno’ovefedv phenomena inconsistent with the scanning model, such as effects of recency (Corballis, Kirby, 8: Miller, 1972] and repetition effects (Baddeley 8: Ecob, 1973], leading to the pro- ) 1968), in which the subject is faced with an array of 9 quasi—randomly arranged blocks. The" experimenter "téipg El pelitlfiilfif'Séqfiéfiégwrrfim ""7 A of blocks and asks the subject to imitate, start— ing with just 2 and building up to a point at which performance breaks down, typically around 5 taps. This task has a sequential and 1 SHORT—TERM AND WORKING MEMORY 81 motor component missing from the pattern span, and appears to measure a different as— pect of visuo-spatial memory, since patients can be impaired on one but not the other; fur- thermore, spatial activity interferes with Corsi span, while intervening abstract pictures dif- ferentially interferes with pattern span [Della Sala, Gray, Baddeley, Allamano, & Wilson, in press). Memory for location using a technique somewhat similar’tovthattdeveloped by Dale" suggests that visual and verbal STM involve different brain regions (Smith 8: Ionides, 1995) and also that the maintenance of even a single item involves an active process involving the frontal lobes [Goldman-Rakic, 1996; Haxby, Ungerleider, Horwitz, Rapoport, 8c Grady, 1995). Research on other nonverbal retention is less well developed, but studies of memory for kinaesthetic stimuli (Adams 8: Dijkstra, 1966] and tactile stimuli [Gilson 8: Baddeley, 1969) show rapid forgetting over a short delay, whereas memory for odors (Eugen, Kuisma, 8: Eimas, 1973) does not. , Theoretical Issues Despite earlier suggestions that there might be more than one kind of memory (Hebb, 1949; James, 1890), the issue was largely ignored un- til the discovery of the short—term forgetting of small amounts of information over filled inter~ vals by Brown (1958) and Peterson and Pe~ terson [1959), which led the investigators to propose separate LTM and STM memory sys— tems, with short-term forgetting reflecting the spontaneous decay of the memory trace. This view was resisted, notably by Melton (1963), who argued strongly for a unitary memory sys- tem in which forgetting reflected associative interference between the items retained, rather than trace decay. The importance of PI in the STM paradigm (Keppel 8: Underwood, 1962] suggested that interference effects certainly occur in STM, although these in turn could be interpreted as reflecting limited capacity, rath— er than classic associative interference (Waugh 8: Norman, 1965). The issue of Whether short— remains unresolved. During the mid—19605 proponents of a di— 7 r 7 WirifirAiiu W: : chotomyebetweenr~STMiand: LTM: generated" memory: evidence from a range of sources, including: Two Component Tasks: Tasks such as free recall appear to have separate components, with the recency effect reflecting STM, while earlier items appear to depend upon LTM (Glanzer, 1972). Acoustic and Semantic Coding: Conrad (1962) showed that errors in recalling Visually presented consonants tended to be similar in sound to the correct items (e.g., B is remem— bered as V), suggesting that recall is based on an acoustic code. Baddeley (1966a, 1966b) showed that immediate recall sequences of 5 unrelated words was highly susceptible to acoustic similarity but insensitive to semantic similarity, while delayed recall of 10-word lists showed exactly the opposite pattern. Us- ing a probe technique, Kintsch and Buschke [1969) showed that the recency part of the function reflected acoustic similarity effects, while performance on earlier items reflected semantic coding. These studies, therefore, ap— peared to suggest a predilection for acoustic coding in STM and semantic coding in LTM. Neuropsychological Evidence: Amnesic pa- tients such as the classic case HM (Milner, 1966) showed grossly impaired LTM, together with preserved span. Such patients also showed preserved recency, and if intellectu- ally otherwise intact, normal performance on the Peterson Short—Term Forgetting Task (Bad~ deley 8: Warrington, 1970). In contrast, a sec- ond class of patient appeared to show the op- posite pattern with digit spans of 1 or 2 items, very poor Peterson performance, and little or no recency, coupled with apparently normal LTM [Shallice 8: Warrington, 1970). This dou— ble dissociation strongly supported a separa~ tion of LTM and STM. By the late 19605, a range of models began to appear in which STM and LTM were con— ceptualized as separate systems. The most in- fluential of these was the Atkinson and Shif~ frin (1968) model, which became known as the moda] mode]. As shown in figure 5.1, it assumes that information comes in from the environment through a parallel series of sen- sory memory systems into a limited-capacity short-term store, which forms a crucial bottle- was also assumed to be necessary for recall, and to act as a limited—capacity working In the early 19705, the mddal model en- countered two major problems. The first con—- earned its assumptions regarding long-term learning, while the second involved its capac- \ v swim—W a" W. an Wfimfitermi forgettingsreflectsidecaysoriinterferenceifineck between perceptioniandiLTM-iTh97313~774~~ was s wives 82 MEMORY IN THE LABORATORY Environmental input Sensory registers Auditory Short-term store (STS) Temporary working memory , "aan;;ot,;e2;s;e:""r'~ Response output Rehearsal Coding Decision u _ _ _ . . . _ . _ _ -— Long-term store (LTS) Permanent memory store Figure 5.1 Atkinson & Shiffrin’s (1968) influential model of STM. ity to explain the neuropsychological evi— to the best retention. While the detailed appli- dence. cation of this model can be criticized [Badde— The modal model assumed that the longer ley, 1978), there is no doubt that it represents an item was held in STS, the greater the a good account of a considerable amount of 7V, 3118999 Qfl’gs being transferred to the LTS. This data, and that the underdevelopment of its assumptioniw’as challenged[el'gZJCraiknga 7-7etreatrnentvofecodinglepllesemailigité’éql} 0f ‘ I _ kins, 1973), leading Craik and Lockhart (1972) the modal model. WW" ' ‘J ***** fl 7 r, a, to propose their levels of processing hypothe— The second problem with the modal model ‘ ‘ sis. This proposes that the duraliil'ity of‘mem— ’"stems from its apparentnpredictinLtliajjaéwi cry increases With depth of processing, hence tients with a grossly impaired STS should eh- VA’WW'A'A’ processing a word in terms of its Visual ap— counter associated problems in long—term ' pearance leads to little learning. Phonological learning. Furthermore, since the STS was as- processing in terms of sound is somewhat bet— sumed to act as working memory, allowing ter, whereas deeper semantic processing leads complex information processing to proceed, SHORT-TERM AND WORKING MEMORY 83 then such patients should also have major general information-processing deficits. How- ever, the few relatively pure cases studied ap— peared to have normal long-term memory and to lead largely normal lives [Shallice 8: War- rington,’1970; Vallar & Shallice, 1990]. Working Memory - In order to tackle this problem, Baddeley and Hitch [1974) proposed that the concept of a single unitary STM be replaced by a multi- component system, focusing on three subsys- tems. These comprised two slave systems; one, the phonological loop was concerned with storing acoustic and verbal information, while the second, the Visuojspatial Sketchpad, was its visual equivalent [see figure 5.2). The overall system was assumed to be controlled by a limited—capacity attentional system, the central executive. While the details of this model and its terminology are by no means universally accepted, the last 20 years have seen an increasing tendency for the term work— ing memory to be used, together with a broad general acceptance of the usefulness of postu— lating a system that combines executive con- trol with more specialized storage systems that show important differences between vi- sual and verbal material (Miyake & Shah, 1999). For that reason, the tripartite structure will be used as a basis for the review, while accepting that there may be a subsequent need to postulate other components. Verbal Working Memory This system, labeled by Baddeley and Hitch the articulatory or phonological loop, is clos— est in character to the original concept of a short—term store. It is assumed to be defective in the type of patient studied by Shallice and Warrington (1970]. The general cognitive dis— ruption implied by the modal model does not ,occurhecausethe central executive is intact in such patients. The phonological? loop is’asf sumed to comprise two components, a store in which an acoustic or phonological memory trace is held. The trace is assumed to decay within about two seconds unless performance is maintained by the second component, the process of subvocal articulatory rehearsal. This process is not only able to refresh the memory trace but can also register visually presented but nameable material in the pho- nological store by means of articulation. The principal evidence for phonological coding is the previously described acoustic similarity effect, While the role of the articulatory pro- cess is supported by the word length efiect, whereby the immediate memory span for. words is a direct function of the length of the constituent items. A simple rule of thumb is that subjects can remember as many items as they can say in two seconds (Baddeley, Thompson, 8: Buchanan, 1975). Baddeley and Hitch explained this phenomenon in terms of trace decay, proposing that subvocal mainte- nance rehearsal occurs in real time, hence long words take longer to rehearse, allowing Central Visuo-spatial sketch pad executive Phonological loop Figure 5.2 The Working Memory model proposed by Baddeley 8: HitCh (1974:]. ' ,i, ,,,, , hi,__¥_.,inAinAW 84 MEMORY [N THE LABORATORY more forgetting through trace decay. Cowan et al. [1992) suggest that the word length effect principally is a function of forgetting during the process of recall, with longer words taking longer to produce, hence allowing more de— say. As the effect can also be found, though to a lesser extent, with probed recall, it seems likely that both rehearsal rate and output time contribute to the word—length effect (Avons, Wright,,&,l?ammer,,1994) Articulatory suppression is a procedure whereby the subject is required to utter some repeated redundant sound such as the word “the” While performing another task such as memory span. Murray (1968) showed that suppression reduces performance and also eliminates the phonological similarity effect, with visual, though not with auditory, presen— tation. This is assumed to occur because sup— pression prevents the subject from converting the visual stimulus into a verbal code that is suitable for registering in the phonological store. With auditory presentation, access to the store is assumed to be automatic (Badde- ley, 1986). The effect of suppression on the word—length effect is assumed to be somewhat different. Since the word—length effect is a di- rect function of rehearsal, suppression will re- move the effect, regardless of whether presen— tation is auditory or visual, as indeed is the case (Baddeley, Lewis, 8: Vallar, 1984]. Another area of considerable activity and controversy in connection with the word- length effect relates to individual differences. If trace decay is responsible for the word- length effect, then subjects who rehearse more slowly should show poorer performance. This was indeed found by Baddeley et al. (1975). Nicolson [1981] observed that developmental changes in children’s memory span were asso- ciated with changes in speed of articulation, suggesting that faster rehearsal might be re— sponsible for the increase in span with age. The effect was replicated by subsequent stud- ies [e.g., Hitch, Halliday, Dodd, 8: Littler, 1989], while research on serial recall of pic— tured objects suggested that verbal coding was a strategy that children begin to adopt be- tween theragesnf Zandvlu ,asneflspted,,by the influence on performance of the acoustic simi- larity of the names in the set and of their spo— When material is presented auditorily, pho- nological similarity and word—length effects appear at a much earlier age, a result which was initially taken to suggest that rehearsal be— gins at this early stage. However, opinion is now shifting toward the assumption that this very early rehearsal reflects a different and rel~ atively automatic process—more like a spon— taneous internal echoing of the stimulus than as is found in older children and adults [Gath— ercole & Hitch, 1998). Finally, there has been considerable inter— est in recent years in the possible evolutionary function of the phonological loop; if patients can show groSs impairment in memory span with little impact on everyday functioning, can the loop be of much biological signifi— cance? Initial work focused on the possible role of the loop in language comprehension (Vallar & Baddeley, 1984]. Although there are some differences among patients, the general consensus. is that most have difficulty only when syntactic structures require the literal maintenance of the first part of the sentence until it is disambiguated at the end, as for ex— ample in the case of self—embedded sentences (see Vallar 8: Shallice, 1990, for a review). A much stronger case for the importance of the phonological loop can be made in the case of new phonological learning. For example, PV, a patient with a very pure STM deficit, showed no difficulty in learning to associate pairs of words in her own language, but was grossly impaired in capacity to learn the vo— cabulary of an unfamiliar language, Russian [Baddeley, Papagno, 8: Vallar, 1988]. In a sub— sequent study, Gathercole and Baddeley (1990) found that children with a specific lan— guage disability were particularly impaired in their capacity to hear and repeat back unfamil- iar sound sequences. This deficit was more pronounced than their language impairment and did not appear to be attributable to either perceptual or speech production problems. This work led to the development of a non- word repetition test in which the subject at- tempts to repeat spoken nonwords ranging in voltularity). Nonword repetition performance proved to correlate with level of vocabulary length from 2 syllables (e-gglzgllqr) relies-i. kenlengtthitchJ—Ialliday,.Schaafstad,,& Heb, ,developmentacross ,avziderangegf figBfiLQXtinAdifi#n, ferman, 1991]. Younger children appear to use some form of visual code, and hence perform more poorly when the items depicted are sim— ilar in shape—for example, a spoon, a pen, and a twig. the 4- to 5-year range, cross-lagged correlation suggested that nonword repetition was caus— ally related to the subsequent development of vocabulary, rather than the reverse (see Bad— deley, Gathercole, 8c Papagno, 1998, for a re— SHORT-TERM AND WORKING MEMORY 85 View). Finally, the phonological short-term ate learning, Logie (1986) showed that perfor— store appears to be related to the capacity for mance could be disrupted by the simple second—language acquisition in both children requirement to observe patterns or patches of (Service, 1992) and adults (Papagno, Valen- color, a visual rather than a spatial task. tine, & Baddeley, 1991), with variables such as Most disrupting tasks tend to involve both articulatory suppression, phonological simi— visual and spatial processing, and may also larity, and word length all influencing the ac~ tend to have an executive component (see Lo— quisition of novel word forms but not affecting gie, 1995, for a review). The technique re— the capacity to associate pairs of already fa— cently developed by Quinn and McConnell iifiii "Wiviim imfliariwordsra’processithat:isiassumeditoidehvwfleee)rappearsitominimizeidisruptionitoianyrw "Wriifi i7irfirfi pend principally on semantic coding (Pa— thing other than the visual component of the pagno & Vallar, 1992). working memory system. Their disrupting task simply requires the subject to fixate on a Neurobiologica/ Evidence screen on which a large matrix of cells is com tinuously flickering on and off. They find that Neuropsychological studies of STM patients this influences performance when subjects are suggested an impaired phonological store learning paired associates using an imagery [Vallar & Baddeley, 1984; Vallar 8: Shallice, mnemonic, while having no effect on rote 1990). The capacity to articulate overtly is not learning performance, in contrast to the effects necessary for rehearsal since dysarthric pa- of irrelevant speech, which produces the op— tients with a peripheral disruption to speech posite pattern. production appear to have normal rehearsal Further evidence for separating visual and capacities (Baddeley & Wilson, 1985). How— spatial aspects of STM come from the observa— ever, dyspraxia, a disruption of the basic ca— tion that pattern span, in which subjects have pacity to program speech output, does inter- to reproduce a pattern of filled and unfilled fere with memory performance (Waters, cells ina matrix, is disrupted by the subse- Rochon, & Caplan, 1992). quent requirement to look at a series of ab— More recently, functional imagery studies stract pictures, but not by a spatial tapping using PET and fMRI have produced clear evi- task, in contrast to the more spatial and serial dence for a phonological short—term store 10- Corsi Block Tapping Task, which shows ex- cated in the perisylvian region of the left actly the opposite pattern (Della Sala et al., in hemisphere, together with a separate rehearsal press). component associated with Broca’s area (Paulesu, Frith, & Frackowiak, 1993; AWh et Neurobiological Evidence al., 1996). . Evidence for separate visual and spatial com— Visuaspaflal Working ponents of the STM system come from neuro- Memory psychological studies, with separate patients capable of per-forming the Corsi Block but not As described earlier, evidence for the storage the Pattern Span tfiSk and Vice VETS?! (D9113 of visual information has been available for 3318 at 31-, in Press), Finally, neuroradiological many years. The use of visuo-spatial coding evidence indicates the separability of visual for Verbal material was demonstrated particu- and Verbal memory (Smith, lonidesy 3: ; larly neatly by Brooks [1967), uSing a tech— Koeppe, 1996), and within that, a distinction i nique in which subjects were induced to store between Spatial and object—based components a sequence of sentences by recoding them in [Smith et 81., 1995). This area continues to C16- terms of a path through a visually presented velop, and further fractionation seems proba- matrix. Using this paradigm, Baddeley,‘Grant, ble (Baddeley, 1993b]- vewvve"Wivwwwfifi Wightr andeThomson~(71973)rrshowedthat vi—W m ~ 7 r e me ~ 7"- ~ , suo-spatial tracking, but not verbal coding, in— _ terfered with visual-imagery-based perfor— EXECUtwe Processes rflivfl'~~irfi *mancerinrrcontrastwtoiarbroadlyrequivalentw ' Wm’wr' ' r ' *A ****************** ~ri verbally coded task. Further work suggested Individual Difference in ' that the coding was specifically spatial (Bad- - deley & Lieberman, 1980). However, using a working Memory Span somewhat different memory paradigm involv— While work utilizing the Baddeley and Hitch ing the use of visual imagery in paired—associ- model has tended to concentrate on the slave a j. 31 2! ii- i: 3 ti; 1, j: l t: l‘. j ii i: jrfiwr were 777 7777*" 777,iimoreestr‘anglyiinfluencBdebyetherpSyChOmen‘iC ,vnfl W. "flair/vellas .the.originalsentencaspan. ' _ n, 86 MEMORY IN THE LABORATORY systems, postponing a more detailed analysis of the central executive, North American re- search on working memory has tended to fol— low the opposite pattern, though with notable exceptions. Furthermore, while neuropsycho- logical evidence has played a particularly im— portant role in European research on working memory, North American research has been tradition with its concern for individual dif— ferences within the normal population. The two approaches are complementary and will be considered in turn. In a classic paper, Daneman and Carpenter (1980) operationally defined working memory as the system responsible for the simultaneous storage and manipulation of information. They developed a measure, working memory span, in which subjects were required to read out a series of sentences and then recall the final word of each. The maximum number of sentences for which all the final words can be correctly recalled is the working memory span, which for normal subjects ranges be- tween 2 and 5. Daneman and Carpenter then demonstrated that span correlated highly with reading comprehension in a sample of student subjects. This finding has been replicated many times [see Daneman 8: Merikle, 1996, for a review). A series of follow-up studies con- trasted subjects who were high and low in span, demonstrating, for example, that there were qualitative differences in the way in which prose is processed by the two groups; for example, high-span subjects are more able to resolve textual ambiguities and to carry in— formation across from one sentence to another in order to do so (Daneman 8: Carpenter, 1983]. Views differ as to whether the measure was concerned with a language-specific system as proposed, for example, by Daneman and Tar— dif [1987] or reflects a more general executive processing capacity, as suggested by Turner and Engle (1989], who showed that a measure they call operation span, based on arithmetic, predicts reading comprehension virtually as Further support for the working memory span measure comes from Kyllonen ,and ChristaletlrQQO), who demonstrate ihatgp erfor; mance on a cluster of working 'memory span tasks correlates highly with more traditional ‘ measures of fluid intelligence while being less subject to the influence of prior knowledge and providing better prediction of success in acquiring practical skills such as programming than more traditional measures. However, despite the apparent success of the working memory span measures, they have recently come under criticism, notably from Waters and Caplan (1996a, 1996b), who question the interpretation of earlier results and also report data from neuropsychological ,7patientsWOf various types that they claim are inconsistent with the theoretical interpreta- tion offered by Carpenter and Inst (1992). The criticism is relatively recent and the issue still unresolved (see just, Carpenter, 8: Keller, 1996]. It seems likely that working memory span probably involves the interaction of sev— eral cognitive subsystems. This highlights the importance of understanding the task if this approach is to continue to be fruitful. Analysis of the WM span task has been one of the major problems tackled by Engle and his group. Engle (1996) showed that high~span subjects are better at generating items from se— mantic categories but, paradoxically, are more impaired than low—span subjects by the re- quirement to perform a concurrent task. This is interpreted as reflecting the successful use of attentional resources by the high—span sub— jects to minimize disruption from already gen- erated items, a strategy that is disrupted by the concurrent load. Low—span subjects are unaf- fected by load because they are incapable or perhaps unwilling to use the inhibitory strat- egy, and are unaffected by a concurrent atten- tional demand. A similar intriguing pattern of results is obtained in studying performance on the Sternberg scanning task, for which there is again evidence for a qualitative difference in performance between high- and low—span sub- jects that is attributed to the capacity to main— tain a memory representation against the disruption of potentially interfering items [Conway 8: Engle, 1994]. While Engle’s work is highly creative in linking individual difference measures and more traditional memory measures such as category fluency and the Sternberg task, it ap- pearstto demonstrate qualitative differences in , performance between high—grand low-span sub— jects. These would seem to be at least as likely" to result from differences in strategy as from a qualitative difference inihewayr inwl ich the memory system works. In either case, the dis- continuity casts doubt on using working mem— ory span as a continuous measure. Even more seriously, these results suggest that many of the findings in this area, which are typically ,tentigaal, System (8355),, and 111se,,theSA3 Was, SHORT—TERM AND WORKING MEMORY based on young students who are presumably of above-average intelligence for the popula- tion, may not generalize to samples of older subjects from a wider intellectual range. The success of Kyllonen in using the measures suggests that there is an important core to the work, but the measures are not yet well under- stood. At the very least, it would be useful to have work that separatesiontitherrple,ofithe slave systems from that of executive pro— cesses. The necessity for such a separation is supported both by further psychometric re— search in the working memory span tradition (Shah 8: Miyake, 1996) and by the growing amount of evidence from functional imagery studies (Smith 8: Ionides, 1995). Analysing the Central Executive Work from a multicomponent approach to working memory has tended to use secondary task techniques, contrasting low processing load tasks such as articulatory suppression and spatial tapping that are targeted at the slave systems, with more demanding tasks, such as random generation of digits or key presses. Baddeley [1986) suggested that the ca- pacity to produce a random stream of items such as digits or letters was constrained by the capacity of the central executive to break away from well-learned stereotypes such as the al- phabet by continuously switching to new re— trieval plans. Random generation does indeed dramatically'impair complex tasks such as choosing the appropriate move in chess, in contrast to the simple suppression effect of re- citing the alphabet [Robbins et al., 1996]. Sim— ilarly, a concurrent digit span task can be shown to interfere with manual generation, with randomness decreasing linearly with digit load (Baddeley, Emslie, Kolodny, 8: Dun“ can, 1998). . The initial model of the central executive (Baddeley, 1986) was strongly influenced by Norman and Shallice’s (1986) Supervisory At— assumed to depend on the operation of the hontal lobes. However, a clear distinction was , madelhetweenihe,questionhffinaipfladlm calization and that of the functional analysis 7 of the system assumed to reflect the operation of the central executive. It was suggested that the use of the term frontal syndrome should be avoided; the term 'dysexecutive syndrome 87 was proposed as an alternative (Baddeley 8: Wilson, 1988). One danger with the concept of a central executive is that of postulating a homunculus that is simply assumed to have whatever ca- pacities are necessary to account for the data (Perkin, 1998). One response to this charge (Baddeley, 1998a] is to argue for the value of homunculi as a means of allowingrtheiinvesti— " gator to set aside some of the more intractable problems. The danger occurs when the theo- rist treats the homunculus as a solution rather than as a problem to be solved. The question of how to analyze the central executive remains a difficult one. One ap- proach is, of course, that based on individual differences described above. A second is to at— tempt to understand the breakdown of execu— tive processes following brain damage in iron~ tal lobe patients [e.g., Shallice 8: Burgess, 1996] or in patients suffering from Alzhei— mer’s disease [Baddeley, Bressi, Della Sala, Logic, 8: Spinnler, 1991). Both these ap~ preaches have proposed a number of separa— ble executive subprocesses, such as the capac— ity to focus and switch attention and to divide it among a number of sources. Division of at- tention appears to be particularly impaired in Alzheimer’s disease, for example, while being relatively preserved in normal elderly subjects [Baddeley et al., 1991]. Given the richness and complexity of executive processes, fraction— ation is likely to be a long and complex task. There is evidence to suggest, however, that it will benefit substantially from the develop- ment of functional imagery studies, which are already giving a very clear indication that dif- ferent areas of the frontal lobes may be spe- cialized for different executive functions (for an overview, see the papers included in Rob- erts, Robbins, 8: Weiskrantz, 1998]. Suppose that we are successful in identify- ing a finite array of executive processes, will we then have solved the central executive problem? Clearly not, since a crucial issue is the way in which the constituent processes in— teract. At present we have little evidence to constrain the possibilities, which range from a hierarchical structure with one "Jainism function, to an array of executive processes of approximately equal antes, with 859,11, DEEPER, of interaction from which consensus emerges. If the former, then what is the process that dominates and, if the latter, what are the mechanisms that allow consensus to, be reached? The same question arises within the a. l :5 l. :i. W v mm, ,1 ,, .,,|.,,,,,,i,, ,, 88 more specialized subsidiary systems accessed. We know that verbal memory span is strongly influenced by phonological factors, but is in addition somewhat sensitive to visual similar- ity and can, of course, be strongly influenced by semantic and linguistic factors when sen- tences are retained. As a recent survey of cur- rent models of working memory illustrates (Miyake 8: Shah, 1999], the question of how grated lies at the heart of many approaches to working memory and is likely to offer one of the most important and challenging problems facing the study of working memory in the years to come. Acknowledgment The support provided by grant G9423916 from the Medical Research Council is gratefully acknowledged. References Adams, I. A., 8: Dijkstra, S. (1966). Short-term memory for motor responses. Journal of EX~ perimental Psychology, 71, 314—318. Anderson, I. A. (1973]. A theory for the recog— nition of items from short memorized lists. Psychological Review, 80, 417—438. Atkinson, R. C., 8: Shiffirin, R. M. (1968). Hu~ man memory: A proposed system and its control processes. In K. W. Spence (Ed), The psychology of learning and motiva- tion: Advances in research and theory (pp. 89—195]. New York: Academic Press. Avons, S. E., Wright, K. L., 8: Pammer, K. (1994]. The word-length effect in probed and serial recall. Quarterly Journal of EX- perimental Psychology, 47a, 207—231. Awh, E., Ionides, I., Smith, E. E., Schu— macher, E. 1-1., Koeppe, R. A., 8: Katz, S. (1996). Dissociation of storage and re— trieval in verbal working memory: Evi- dence from positron emission tomography. Psychological Science, 7, 25—31. Baddeley, A. D. (1966a). Short—term memory for word sequences as a function of acous- tic, semantic and formal similarity. Quar- terly Journal of Experimental Psychology, 13: 362—365. "Bad'a’eiey; Aim: (19665; The influéncé'of" “(1984)T'Explming'the’arficulatdrr100 '-‘""""’*"*"“"' t " ' acoustic and semantic similarity on long- ,, term memory fOTWQEd EPQUPESeS- Q1101“ terly Journal of Experimental Psychology, 1 8, 302—309. ~ Baddeley, A. D: (1978).'The trouble with lev— els: A re—examination of Craik and Lock— hart framework for metnory research. Psy— chological Review, 85, 139~152. MEMORY IN THE LABORATORY Baddeley, A. D. (1986]. Working memory. Ox- ford: Oxford University Press. Baddeley, A. D. (1992]. Working memory. Sci— ence, 255, 556—559. Baddeley, A. D. (1996]. The concept of work- ing memory. In S. E. Gathercole (Ed), Mod- els of short-term memory (pp. 1—27]. Hove, England: Psychology, Press. Baddeley, A. D. (1998a). The central execu~ iv'iin’formation”’from’flifferent”source’s""is' ’i’nte—i"itlvei’A’GonGePt’and”some’mlseonceptmnsv’ WW“ ii’yi "W * Journal of the International Neuropsychol— ogy Society, 4, 523—528. Baddeley, A. D. (1998b). Recent develop- ments in working memory. Current Opin- ion in Neurobiology, 8, 234—238. Baddeley, A. D., Bressi, S., Della Sale, 8., Lo- gie, R., 8: Spinnler, H. (1991). The decline of working memory in Alzheimer’s dis- ease: A longitudinal study. Brain, 114, 2521—2542. Baddeley, A. D., 8: Ecob, J. R. (1973). Reaction time and short—term memory. Implications of repetition effects for the high—speed ex- haustive scan hypothesis. Quarterly Jour- nal ofEXperimental Psycholog , 25, 2.29— 240. Baddeley, A. D., Emslie, H., Kolodny, 1., 8: Duncan, I. (1998]. Random generation and the executive control of working memory. Quarterly Journal of Experimental Psychol—- 0g , 51A, 819—852. Baddeley, A., Gathercole, 8., 8: Papagno, C. (1998). The phonological loop as a lan- guage learning device. Psychological He- VIeW, 105, 158—173. Baddeley, A. D., Grant, 8., Wight, E., 8: Thom- son, N. (1973). Imagery and visual working memory. In P. M. A. Rabbitt 8: S. Dornic [Eds], Attention and performance V (pp. 205—217). London: Academic Press. Baddeley, A. D., 8: Hitch, G. I. (1974:). Work— ing memory. In G. A. Bower (Ed), The psy- chology of learning and motivation (pp. 47—89). New York: Academic Press. ' Baddeley, A. D., 8: Hitch, G. I. (1993). The re- cency effect: Implicit learning with ex— , plicit retrieval? Memory and Cognition, 21, 146—155. Baddeley, A. D., Lewis, V. J., 8: Vallar, G. Quarterly Journal of Experimental Psychol- ogy, 36, 233—252. Baddeley, 'AT D'., '8: LiebermaanT (1'980’)?’Sp‘a—AAW"""A “mm” tial working memory. In R. S. Nickerson (Ed), Attention and performance VIII (pp. 521—539]. Hillsdale: NJ: Erlbaum. Baddeley, A. D., Papagno, C., 8: Vallar, G. (1988). When long-term learning depends ........ .. SHORT-TERM AND WORKING MEMORY 39 on short—term storage. Journal of Memory and Language, 27, 586—695. Baddeley, A. D., 8: Scott, D. (1971). Short- term forgetting in the absence of proactive interference. Quarterly Journal of Experi— mental Psychology, 23, 275—283. Baddeley, A. D., Thomson, N2, 8: Buchanan, M. (1975). Word length and the structure of short-term memory. Journal of Verbal Learning and Verbal'Beh’aViohrfiez, 5’75— 589. Baddeley, A. D., Vallar, G., 8: Wilson, B. A. (1987). Sentence comprehension and pho— nological memory: Some neuropsychologi- cal evidence. In M. Coltheart (Ed), Atten- tion and performance XII: The psychology ofreading (pp. 509—529). London: Erl— baum. ~ Baddeley, A. D., 8: Warrington, E. K. (1970). Amnesia and the distinction between long- and short-term memory. Journal of Verbal Learning and Verbal Behavior, 9, 175—189. Baddeley, A. D., 8: Wilson, B. (1985). Phono— logical coding and short—term memory in patients without speech. Journal of Mem- cry and Language, 24, 490—502. Baddeley, A. D., 8: Wilson, B. (1988). Frontal amnesia and the dysexecutive syndrome. Brain 8 Cognition, 7(2), 212—230. Broadbent, D. E. (1958). Perception and com- munication. London: Pergamon Press. Brooks, L. R. (1967). The suppression of visu- alization by reading. Quarterly Journal of Experimental Psychology, 19, 289—299. Brown, I. (1958). Some tests of the decay the— ory of immediate memory. Quarterly Jour- nal of Experimental Psychology, 10, 12—21. Conrad, R. (1962). An association between memory errors and errors due to acoustic masking of speech. Nature, 193, 1314— 1315. Cemad, R. (1964). Acoustic confusion in im- mediate memory. British Journal of Psy— chology, 55, 75~84. Conway, A. R. A., 8: Engle, R. W. (1994). Working memory and retrieval: A re— sourcejiepandentinhihition mo Ciel. Iour: ,1, , nal of Experimental Psychology: General, 123, 354—373. Access to elements of a memorised list. Journal of Experimental Psychology, 94, 185—190. Cowan, N., Day, L., Saults, I. 8., Keller, T. A., Johnson, T., 8: Flores, L. (1992). The role of verbal output time and the effects of word ailnrJeTr EeRichardson, 7R. 7W" Engle, L. Hash— length on immediate memory. Journal of Memory and Language, 31, 1—17. Craik, F. I. M., 8: Lockhart, R. S. (1972). Lev— els of processing: A framework for mem- ory research. Journal of Verbal Learning and Verbal Behavior, 11, 671—684. Craik, F. I. M., & Watkins, M. I. (1973). The role of rehearsal in short-term memory. Journal of Verbal Learning and Verbal Be- havior, 7171’, 671*58’4’. WW 7 Crowder, R. G. (1982). The demise of short- term memory. Acta Psychologica, 50, 291— 323. Dale, H. C. A. (1973). Short—term memory for visual information. British Journal of Psy- chology, 64, 1—8.- Daneman, M., 8: Carpenter, P. A. (1980). Indi- vidual differences in working memory and reading. Journal of Verbal Learning and Verbal Behaviour, 19, 450—466. Daneman, M., & Carpenter, P. A. (1983). Indi- vidual difference in integrating informa- tion between and within sentences. Jour- nal of Experimental Psychology: Learning, Memory and Cognition, 9, 561—584. Daneman, M., 8: Merikle, P. M. (1996). Work- ing memory and language comprehension: A meta—analysis. Psychonomic Bulletin 6' Review, 3, 422—433. Daneman, M., 8: Tardif, T. [1987). Working memory and reading skill re—examined. In M. Coltheart (Ed), Attention and perfor- mance (pp. 491—508). Hove, England: Erl- baum. Della Sala, 8., Gray, C., Baddeley, A.'D., Alla- mano, N., 8: Wilson, L. (in press). Pattern span: A means of unwelding visuo-spatial memory. Neuropsychologia. Della Sala, 3., Gray, C., Baddeley, A. D., 8: Wilson, 1'... (1997). 'The Visual patterns test: A test of short—term Visual recall. Flemp— ton Bury 8: Edomonds, England: Thames Valley Test Company. Eugen, T., Kuisma, I. E., 8: Eirnas, P. D. (1973). Short-term memory-0f odors. Jour~ nal of Experimental Psychology, 99, 222— 225. Engle, R. W. (1996). Working memory and re— trieval: An inhibition—resource approach. er, R. H. Logie, E. R. Stoltfus, 8: R. T. Zacks (Eds), Working memory and human cogni- W7Corba11isaMaC.FEW. J... &,.1\/Iiller,,A.,(19.7.2)aim. tjon,(pp...89—119). .New.¥ork:,OXf0rdiUni—A.4".AAW» versity Press. Gathercole, S. E. (1996). Models of short-term memory Hove, England: Psychology Press. Gathercole, S. E., 8: Baddeley, A. D. (1990). Phonological memory deficits in language— disordered children: Is there a causal con— \J 90 MEMORY IN THE LABORATORY nection? journal of Memory and Language, 29, 336—360. Gathercole, S. E, 8: Hitch, G. J. (1993). Devel— opmental changes in short—term memory: A revised working memory perspective. In A. Collins, S. E. Gathercole, M. A. Con- way, & P. E. Morris (Eds), Theories of memory. Hove, England: Erlbaum. Gilson, A. D., 8: Baddeley, A. D. (1969). Tac- ’ file short-term’memornyuanerly’joarnnl of Experimental Psychology, 21, 180—184. Glanzer, M. (1972). Storage mechanisms in re- call. In G. H. Bower (Ed.), The psychology of learning and motivation: Advances in re- search and theory. New York: Academic Press. Goldman-Rakic, P. S. (1996). The prefrontal landscape: Implications of functional archi- tecture for understanding human menta— tion and the central executive. Philosophi- cal Transactions of the Royal Society [Biological Sciences), 351, 1445-4453. Haxby, J. V., Ungerleider, L, G., Horwitz, B., Rapoport, S. 1., 8: Grady, C. L. (1995). Hemi— spheric differences in neural systems for face working memory: a PET-rCBF study. Human Brain Mapping, 3, 68—82. Hebb, D. O. (1949). Organization of behavior. New York: John Wiley. Hitch, G. J., Halliday, M. 8., Dodd, A., 8: Lit— tler, J. E. (1989). Development of rehearsal in short~term memory: Differences be- tween pictorial and spoken stimuli. British journal of Developmental Psychology, 7, 347—362. / Hitch, G. J., Halliday, M. S., Schaafstal, A. M., 8: Heffernan, T. M. (1991). Speech, “inner speech” and the development of short- term memory: Effects of picture labelling on recall. journal of Experimental Child Psycholog , 51, 220—234. Hockey, G. R. J. (1973). Rate of presentation in running memory and direct manipula- tion of input processing strategies. Quar- terly journal of Experimental Psychology, 25, 104—111. . James, W. (1890). The principles ofpsychol— ogy. New York: Holt, Rinehart 8: Winston. Just, M. A., Carpenter, P. A., 8: Keller, T. A. sion: New frontiers of evidence and argu- ments. Psychological Review, 103, 773—— , 780. W 7 . .7. WW Keppel, G., 8: Underwood, B. J. (1962). Proac— tive inhibition in short-term retention of single items. journal of Verbal Learning and Verbal Behavior, 1, 153—161. Kintsch, W., 8: Buschke, H. (1969). Homo- phones and synonyms in short-term mem— ! I. l .4 2 i cry. journal of Experimental Psychology, 80, 403—407. Kyllonen, P. C., 8: Christal, R. E. (1990). Rea- soning ability is [little more than) working memory capacity. Intelligence, 14, 389— 438. Lambert, B. L. (1997). Predicting look-alike and sound-alike medication errors. Ameri— . can journal of Health-system Pharmacy, 54, 1181—1171. ’ 7 Loess, H., 8: Waugh, N. C. (1967). Short-term memory and inter-trial interval. journal of Verbal Learning and Verbal Behaviour, 6, 445—460. Logie, R. H. (1986). Visuo—spatial processing in working memory. Quarterly journal of Experimental Psychology, 38A, 229—247. Logie, R. H. (1995). Visuo—spatial working memory. Hove, England: Erlbaum. Melton, A. W. (1963). Implications of short- term memory for a general theory of mem- ory. journal of Verbal Learning and Verbal Behavior, 2, 1—21. Miller, G. A. (1956). The magical number . seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81—97. Miller, G. A., Galanter, E., 8: Pribram, K. H. (1960). Plans and the structure of behav- ior. New York: Holt, Reinhart 8: Winston. Miller, G. A., 8: Selfridge, J. A. (1950). Verbal context and the recall of meaningful mate- rial. American journal of Psychology, 63, 176—185. 'Milner, B. (1966). Amnesia following opera- tion on the temporal lobes. In C. W. M. Whitty 8: O. L. Zangwill (Eds), Amnesia (pp. 109—133). London: Butterworths. Milner, B. (1968). Visual recognition and re— call after right temporal-lobe excision in man. Neuropsychologia, 6, 191—209. Miyake, A., 8: Shah, P. (Eds). (1999). Models of working memory: Mechanisms of active maintenance and executive control. New York: Cambridge University Press. Murray, D. J. (1968). Articulation and acous— tic confusability in short—term memory. journal of Experimental Psychology, 78, 879—684. -. 1;“ -1 -1 (1996.). The.capacity.theoryrof.comprehenrrnNewellnA.,,,8:,Simon,.H.;2L.llflZl-liumannnnm, ,, problem solving. Englewood Cliffs, NJ: Prentice—Hall. ,, inmrnNicolson,,R.,,(1,9.8,1)., Tharelationship between memory span and processing speed.'ln M. Friedman, J. P. Das, 8: N. O’Connor (Eds), Intelligence and learning (pp. 179—184). New York: Plenum Press. Norman, D. A., 8: Shallice, T. (1986). Atten— tion to action: Willed and automatic con- SHORT—TERM AND WORKING MEMORY 91. no] of Experimental Psychology: General, 125, 4—27. Shallice, T., & Burgess, P. (1996). The domain of supervisory processes and temporal or— ganization of behaviour. Philosophical Transactions of the Royal Society of Lon- don Series B—Biological Sciences, 351(1346), 1405—1411. Shallice, T., 8: Warrington, E. K. (1970). Inde— Wudent’functioning’ofverbahnemory stores: A neuropsychological study. Quar- terly Journal of Experimental Psychology, 22, 261—273. Smith, E. E., 8: Ionides, I. (1995). Working memory in humans: Neuropsychological Evidence. In M. Gazzaniga (Ed), The cogni— tive neurosciences (pp. 109—1020). Cam— bridge, MA: MIT Press. Smith, E., Ionides, I., 8: Koeppe, R. A. (1996). Dissociating verbal and spatial working memory using PET. Cerebral Cortex, 6, 11—20. Smith, E. E., Ionides, I., Koeppe, R. A., Awh, E., Schumacher,‘ E., 8: Minoshima, S. (1995). Spatial versus object working mem- ory: PET investigations. Journal of Cogni- tive Neuroscience, 7, 337—358. Sperling, G. (1963). A model for visual mem- ory tasks. Human Factors, 5, 19—31. Sternberg, S. (1966). High-speed scanning in human memory. Science, 153, 652—654. Theios, I. (1973). Reaction time measure- ments in the study of memory process: Theory and data. In G. H. Bower (Ed), The Psychology of Learning 8* Motivation. Vol. 7. (pp. 43—85). New York: Academic Press. Tulving, E., 8: Patkau, I. E. (1962). Concurrent effects of contextual constraint and word frequency on immediate recall and learn- ing of verbal material. Canadian Journal of Psychology, 16, 83—95. Tulving, E., 8: Pearlstone, Z. (1966). Availabil— ity versus accessibility of information in ' memory for words. Journal of Verbal Learn— ing and Verbal Behavior, 5, 381—391. Turner, M. L., 8: Engel, R. W. (1989). Is work- ing memory capacity task—dependent? Jour- trol of behaviour. In R. I. Davidson, G. E. 1 Schwarts, 8: D. Shapiro (Eds), Conscious— : ness and self-regulation. Advances in re- search and theory (pp. 1—18). New York: Plenum Press. Olton, D. 8., Walker, ,I. A., 8: Gage, F. H. (1978). Hippocampal connections and spa- tial discrimination. Brain Research, 139, _ 295—308. G’Papagno, C., valentine, T., ’8:’Ba’d’d’e’l'eyTATDi ”’ ’ x (1991). Phonological short-term memory i and foreign language vocabulary learning. Journal of Memory and Language, 30, ' ' 331—347. 5 Papagno, C., 8: Vallar, G. [1992). Phonological l short-term memory and the learning of novel words: The effect of phonological similarity and item length. Quarterly Jour- nal of Experimental Psychology, 44A, 47—67. Perkin, A. I. (1998). The central executive does not exist. Journal of the International Neuropsychology Society, 4, 518—522. Paulesu, E., Frith, C. D., 8: Frackowiak, R. S. I. (1993). The neural correlates of the ver- bal component of working memory. Na- ture, 362, 342—345. Peterson, L. R., 8: Peterson, M. I. (1959). Short—term. retention of individual verbal items. Journal of Experimental Psychology, 58, 193—198. Phillips, W. A. (1974). On the distinction be— tween sensory storage and short—term vi- sual memory. Perception and Psychophys- iCS, 16, 283—290. Phillips, W. A., 8: Christie, D. F. M. (1977). Components of visual memory. Quarterly Journal of Experimental Psychology, 29, 117—133. Quinn, G., 8: McConnell, I. (1996). Irrelevant pictures in visual working memory. Quar— terly Journal of Experimental Psychology, 49A, ZOO—215. Robbins, T., Anderson, E., Barker, D., Bradley, A., Fearneyhough, (3., Henson, R., Hudson, 8., 8: Baddeley, A. D. (1996). Working mem— ory in chess. Memory and Cognition, 1 24(1), 83—93. - Roberts, A. (3., Robbins, T. W., 8: Weiskrantz, “‘11 OfMemOI'Y 011d Language: 28: 127— * ~ —~ 7 — 7 Le (717998).WThepre-frontal cortex: Executivehv ,fi1fi574- and cognitive functions. Oxford: Oxford University Press. AService33.41992).Phonologypworkingrneme ory and foreign-language learning. Quar- terly Journal of Experimental Psychology, 45A, 21—50. Shah, P., 8: Miyake, A. (1996). The separabil- ity of working memory resources for spa- tial thinking and language processing. Jour- ation of working memory. Neuropsycholog- ical Evidence for awphonological short-term store. Journal of Verbal Eearning andIVer- bal Behaviour, 23, 151—161. Vallar, G., 8: Shallice, T. (Eds) (1990). Neuro- psychological impairments of short-term memory. Cambridge: Cambridge University Press. VallarfG:fiddlelexiAfDTl1'7934=lTFTaEfiljfir 7 92 MEMORY IN THE LABORATORY Waters, G. S., 8: Caplan, D. (1996a). The mea— surement of verbal working memory capac— ity and its relation to reading comprehen- sion. Quarterly journal of Experimental Psychology, 49a, 51—79. Waters, G. S., & Caplan, D. (1996b). The ca- pacity theory of sentence comprehension: Critique of Just 8: Carpenter [1992]. Psycho- logical Review, 103, 761—772. Waters, G. 8., Rochon, E., & Caplan, D. [1992). The role of high—level speech plan— ning in rehearsal: Evidence from- patients With apraxia of speech. Journal of Memory and Language, 31, 54—73. Waugh, N. C., & Norman, D. A. (1965). Pri- mary memory. Psychological Review, 72, 89—104. Wickens, D. D. (1970). Encoding categories of words: an empirical approach to meaning. ltsxchologieal Review 77, 1:15- ...
View Full Document

This note was uploaded on 12/01/2011 for the course CLP 7934 taught by Professor Staff during the Fall '08 term at University of Florida.

Page1 / 16

baddeley_2000 - Short-Term and Working Memory ALAN...

This preview shows document pages 1 - 16. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online