¡Descarga Working Memory y más Apuntes en PDF de Psicología solo en Docsity!
C H A P T E R
Working
Memory
1. Using Working Memory 1.1. A Computer Metaphor 1.2. Implications of the Nature of Working Memory 2. From Primary Memory to Working Memory: A Brief History 2.1. William James: Primary Memory, Secondary Memory, and Consciousness 2.2. Early Studies: The Characteristics of Short-Term Memory 2.2.1. Brevity of Duration 2.2.2. Ready Accessibility 2.3. The Atkinson-Shiffrin Model: The Relationship of Short-Term and Long- Term Memory 2.4. The Baddeley-Hitch Model: Working Memory 3. Understanding the Working Memory Model
3.1. The Phonological Loop: When It Works and When It Doesn’t 3.2. The Visuospatial Scratchpad 3.3. The Central Executive 3.4. Are There Really Two Distinct Storage Systems?
4. How Working Memory Works 4.1. Mechanisms of Active Maintenance DEBATE BOX: How Are Working Memory Functions Organized in the Brain? 4.2. The Role of the Prefrontal Cortex in Storage and Control A CLOSER LOOK: Mechanisms of Working Memory Storage in the Monkey Brain 5. Current Directions 5.1. The Episodic Buffer 5.2. Person-to-Person Variation 5.3. The Role of Dopamine Revisit and Reflect
Learning Objectives
Y ou’re in the middle of a lively conversation about movies, one in particular. You and your
friends have all seen it and have come away with different views. One friend says he felt that one of the leads was not convincing in the role; you disagree—you think the failing was in the screenplay, and want to make your case. But before you have a chance to get going, another friend jumps in and says she doesn’t think this actor was miscast, just that he’s not very good, and is prepared to argue chapter and verse. You think your point is a good one, and you want to make it; but you’ll only offend this friend, who’s now arguing her point with enthusi- asm. Moreover, you find yourself agreeing with some of what she’s saying. Your challenge is
240 CHAPTER 6 Working Memory
to manage two tasks at once: pay attention to what your friend is saying, both out of courtesy and to follow her argument so you don’t repeat or overlook her points when you speak; and hold on to your own argument, which is forming in your head as you listen. Your working mem- ory is getting a workout! Working memory is widely thought to be one of the most important mental faculties, critical for cognitive abilities such as planning, problem solving, and reasoning. This chapter describes current conceptions regarding the nature of working memory, its internal compo- nents, and the way it works. We specifically address five questions:
1. How is working memory used in cognition? 2. How did the modern view of working memory arise? 3. What are the elements of working memory? 4. How does working memory “work” in the brain? 5. How might views of working memory change in the future? 1. USING WORKING MEMORY
Every day we have occasion to keep particular pieces of critical information briefly in mind, storing them until the opportunity to use them arrives. Here are some ex- amples: remembering a phone number between the time of hearing it and dialing it (“1 646 766-6358”); figuring a tip (the bill is $28.37, call it $30; 10 percent of that is $3.00, half of that is $1.50, $3.00 plus $1.50 is $4.50, the 15 percent you’re aim- ing for); holding driving directions in mind until you get to the landmarks you’ve been told to watch for (“take the first left, continue for one mile, past the school, bear right, left at the four-way intersection, then it’s the third building on the left— you can pull into the driveway”). Sometimes a problem offers multiple possible so- lutions, such as when you must look ahead along various possible sequences of moves in a chess game, and sometimes, as when you must untangle the structure of a complex sentence like this one, it is straightforward but nonetheless requires hold- ing bits of information in mind until you can put it all together. In situations like these, not only do we need to keep certain bits of information ac- cessible in mind, but also we need to perform cognitive operations on them, mulling them over, manipulating or transforming them. These short-term mental storage and manipulation operations are collectively called working memory. Think of working memory as involving a mental blackboard—that is, as a workspace that provides a temporary holding store so that relevant information is highly accessible and available for inspection and computation. When cognitive tasks are accomplished, the informa- tion can be easily erased, and the process can begin again with other information.
1.1. A Computer Metaphor
The computer, so useful a metaphor in cognitive psychology, offers an intuitively appealing model for thinking about the nature and structure of working memory.
242 CHAPTER 6 Working Memory
success, we might be able to develop methods to train and exercise working memory in a manner that could improve its function, and consequently enhance a person’s cognitive repertoire. Today’s conceptions of working memory have evolved from earlier ideas in cog- nitive psychology, and current research stands, as so often in science, on the shoul- ders of predecessors. What the earliest workers did not have were the tools provided by modern neuroscience. Nonetheless, their work is a good place to begin.
Comprehension Check:
1. Give an example of an everyday situation in which you would need to use work- ing memory. 2. If working memory were a capacity of a computer, what component might it correspond to, and why? 2. FROM PRIMARY MEMORY TO WORKING MEMORY: A BRIEF HISTORY
The notion that there is a distinct form of memory that stores information tem- porarily in the service of ongoing cognition is not new, but ideas regarding the nature and function of short-term storage have evolved considerably during the
FIGURE 6–1 A standard test of working memory capacity To take this test yourself, cut out a window in a blank sheet of paper so that it exposes only one line at a time. For each line, determine whether the arithmetic is correct or not: say, out loud, “yes” or “no.” Then look at the word that follows the problem and memorize it. Move through each line quickly. After you have finished all the lines, try to recall the words in order. The number you get correct is an estimate of your working memory capacity. Very few people have a working memory as high as 6; the average is around 2 or 3.
(5 3) 4 17?
(6 2) 3 8?
(4 4) 4 12?
(3 7) 6 27?
(4 8) 2 31?
( 9 2) 6 24?
BOOK
HOUSE
JACKET
CAT
PEN
WATER
IS
IS
IS
IS
IS
IS
last hundred years. The very terms for this storage system have changed over the years, from primary memory to short-term memory to working memory. How and why did this happen?
2.1. William James: Primary Memory, Secondary Memory,
and Consciousness
The first discussion of a distinction between short-term and long-term storage sys- tems was put forth by the pioneering American psychologist William James in the late nineteenth century. James called these two forms of memory primary memory and secondary memory, using these terms to indicate the degree of the relationship of the stored information to consciousness (James, 1890). In James’s view, primary memory is the initial repository in which information can be stored and made avail- able to conscious inspection, attention, and introspection. In this way, such infor- mation would be continually accessible. In James’s words, “an object of primary memory is thus not brought back; it never was lost.” He contrasted primary mem- ory with a long-term storage system, or secondary memory, from which informa- tion cannot be retrieved without initiating an active cognitive process. The link between working memory and consciousness that James sought to describe remains a central part of most current thinking; the question of whether or not we are con- scious of the entire contents of working memory is still open to debate. Some current models suggest that only a subset of working memory is consciously experienced (Cowan, 1995).
2.2. Early Studies: The Characteristics of Short-Term Memory
Despite James’s early work regarding the system for short-term information storage, there were no experimental studies of the characteristics of this system until the 1950s. Part of the reason for this neglect was the dominance of behaviorist views in the first half of the twentieth century, which shifted the focus of investigation away from cognitive studies. Then George Miller, an early and influential cognitive theo- rist, provided detailed evidence that the capacity for short-term information storage is limited. In what has to be one of the most provocative opening paragraphs of a cognitive psychology paper, Miller declared: “My problem is that I have been perse- cuted by an integer.... [T]his number has followed me around, has intruded in my most private data, and has assaulted me from the pages of our most public journals. This number assumes a variety of disguises, being sometimes a little larger and some- times a little smaller than usual, but never changing so much as to be unrecogniz- able” (Miller, 1956, p. 81). In this paper, titled “The Magical Number Seven, Plus or Minus Two,” Miller suggested that people can keep only about seven items active in short-term storage, and that this limitation influences performance on a wide range of mental tasks. What data supported Miller’s claim? Tests of short-term memorization, such as repeating a series of digits, showed that regardless of how long the series is, correct recall of digits appears to plateau at about seven items (though for some people this
- From Primary Memory to Working Memory: A Brief History 243
- From Primary Memory to Working Memory: A Brief History 245
interval showed the time course of forgetting. After a delay as short as 6 seconds, re- call accuracy declined to about 50 percent, and by about 18 seconds recall was close to zero (Figure 6–2). These findings suggested the shortness of short-term storage. (About this time investigations were also being conducted on an even briefer form of storage—termed sensory memory —that serves to keep a perceptual representation of a stimulus persisting for a few hundred milliseconds after the sensory input is gone; Sperling, 1960.) However, in work that followed, a controversy arose as to whether the forgetting of information was truly due to a passive decay over time, or rather due to interference from other, previously stored information (similar to the controversy regarding long- term memory, discussed in the previous chapter). The argument favoring the role of in- terference was bolstered by the fact that participants’ recall performance tended to be much better in the first few trials of the task (when proactive interference from the ear- lier trials had not yet built up). Moreover, if a trial was inserted that tested memory for a different type of information than that sought in the previous trials (for example, switching from consonants to vowels), participants’ recall performance greatly in- creased on the inserted trial (Wickens et al., 1976). The debate over whether informa- tion is lost from short-term memory because of decay, in addition to interference, has not been resolved, and the question is still studied today (Nairne, 2002).
FIGURE 6–2 Short-term recall related to delay interval in the Brown-Peterson task
Typically, accuracy in recalling short consonant strings decays to about 50 percent by 6 seconds and almost to zero by 18 seconds if rehearsal is blocked. (Peterson, L. R., and Peterson, M. J. (1959). Short-term retention of individual verbal items. Journal of Experimental Psychology, 58, 193–198, Fig. 3, p. 195 [adapted].)
100
80
60
40
20
0 3 6 9 12 15 18
Accuracy of recall (%)
Recall interval (sec)
246 CHAPTER 6 Working Memory
2.2.2. Ready Accessibility
The high level of accessibility of information stored in short-term memory was demonstrated in a classic set of studies conducted by Saul Sternberg (1966, 1969a), which we briefly considered in Chapter 1. We now consider these findings in greater detail. A variable number of items, such as digits (the memory set), were presented briefly to participants at the beginning of a trial and then removed for a minimal delay. Following the delay, a probe item appeared and participants were to indicate whether or not the probe matched an item in the memory set. The time required to respond should reflect the sum of four quantities: (1) the time required to process the probe item perceptually, (2) the time required to access and compare an item in short-term memory against the probe item, (3) the time required to make a binary response decision (match–nonmatch), and (4) the time required to execute the nec- essary motor response. Sternberg hypothesized that as the number of items in the memory set increased, the second quantity—the total time required for access and comparison—should increase linearly with each additional item, but the other three quantities should remain constant. Thus, Sternberg hypothesized that when the re- sponse time was plotted against the number of memory set items, the result would be a straight line on the graph. Moreover, the slope of that line should reveal the av- erage time needed to access and compare an item held in short-term memory. The re- sults were as predicted—the plotted data formed an almost perfect straight line, and the slope indicated an access plus-comparison time of approximately 40 milliseconds (Figure 6–3). The hypothesis that information held in short-term memory could be accessed at high speed was certainly borne out by these findings.
FIGURE 6–3 Recognition time related to memory set size in the Sternberg item recognition task As the number of items to be memorized—the memory set size—increases from one to six, the time to evaluate a probe increases in a linear manner with about a 40-millisecond increase per additional item. The best-fitting line for the data obtained is plotted here; it is very close to the actual data points. (Sternberg, S. (1969). The discovery processing stages: Extension of Donders’ method. In W. G. Koster (ed.), Attention and Performance II. Amsterdam: North-Holland.)
650
600
550
500
450
400
350 1 2 3 4 5 6 Memory set size
Mean reaction time (msec)
RT 397.2 37.9 msec
Yes No Mean
Response
248 CHAPTER 6 Working Memory
model was highly influential because it laid out a comprehensive view of infor- mation processing in memory. In a nod to the statistical notion of the mode, it is still referred to as the modal model of memory, the model most frequently cited. Yet today the modal model does not have the influence it once had, and most psychologists favor a different conceptualization of short-term storage, one that is not exclusively focused on its relationship to long-term storage and includes a more dynamic role than storage alone. This shift was reflected in the increasing use of the term “ working memory ” which better captures the notion that a temporary storage system might provide a useful workplace in which to engage in complex cognitive activities. What caused this shift in perspective? For one thing, the Atkinson-Shiffrin model is essentially sequential: information passes through short-term memory be- fore entering long-term memory. But neuropsychological data were showing that this assumption is not correct. Some patients with brain damage (typically to the parietal lobe) who showed drastic impairments in short-term memory nevertheless were able to store new information in long-term memory in a fashion comparable to that of neurologically healthy people (Shallice & Warrington, 1970). This find- ing demonstrated that information can gain access to the long-term memory sys- tem even when the short-term memory system was dramatically impaired. The Atkinson-Shiffrin model could not account for this result: with a poorly function- ing short-term memory, according to Atkinson-Shiffrin, long-term storage should also be impaired. Another strand of evidence, from behavioral experiments with neurologically healthy people, suggested that there is not a single system for short-term storage but multiple ones. Alan Baddeley and Graham Hitch (1974) asked participants to make simple true–false decisions about spatially arrayed letters: for example, shown “B A” they were to decide whether the statement “B does not follow A” was true or false. Before each trial, the participants were also given a string of six to eight dig- its (which according to Miller should fill the capacity of short-term memory) to re- peat immediately after each true–false task. If the short-term memory store is critical for performing complex cognitive tasks and there is only one short-term store available, then performance on the reasoning task should drastically decline with the addition of the digit-memorization task. However, this was not the case. The participants took slightly longer to answer questions but made no more errors when also holding digit strings in short-term memory. From these results Baddeley and Hitch argued that there are multiple systems available for short-term storage and that these storage systems are coordinated by the actions of a central control system that flexibly handles memory allocation and the balance between processing and storage.
2.4. The Baddeley-Hitch Model: Working Memory
The dynamic concept of “working memory”—as opposed to the passive nature of a simple information store—is at the heart of the Baddeley-Hitch model , a system that
- From Primary Memory to Working Memory: A Brief History 249
consists of two short-term stores and a control system. Three important characteris- tics differentiate this model from the Atkinson-Shiffrin model. First, the function of short-term storage in the Baddeley-Hitch model is not pri- marily as a way station for information to reside en route to long-term memory. In- stead, the primary function of short-term storage is to enable complex cognitive activities that require the integration, coordination, and manipulation of multiple bits of mentally represented information. Thus, in the “A–B” reasoning problem de- scribed earlier, working memory is required to (1) hold a mental representation of the two letters and their spatial relationship to each other, (2) provide a workspace for analyzing the statement “B does not follow A” and deciding that it implies that “A follows B,” and (3) enable comparison of the mental representations of the letters and statement. Second, in the Baddeley-Hitch model there is an integral relationship between a control system—a central executive —that governs the deposition and removal of in- formation from short-term storage and the storage buffers themselves. This tight level of interaction is what enables the short-term stores to serve as effective work- places for mental processes. Third, the model proposes (as implied earlier) at least two distinct short-term memory buffers, one for verbal information (the phonological loop ) and the other for visuospatial information (the visuospatial scratchpad ). Because these short-term stores are independent, there is greater flexibility in memory storage. Thus, even if one buffer is engaged in storing information, the other can still be utilized to full ef- fectiveness. The supervision of these storage systems by a central executive suggests that information can be rapidly shuttled between the two stores and coordinated across them. The three components of the Baddeley-Hitch model interact to provide a com- prehensive workspace for cognitive activity (Figure 6–5). Applying the terms of the Baddeley-Hitch model to the “A–B” task, the phonological loop was occupied stor- ing the digits, and the visuospatial scratchpad did much of the cognitive work in evaluating the spatial relationships in the true–false task. Coordination was supplied by the central executive, which transformed information from reading the statement (essentially in the verbal store) into a mental image on the visuospatial scratchpad. These interactions meant that performance on the reasoning task did not decline greatly when digit memorization was added.
Phonological loop
Visuospatial scratchpad
Central executive
FIGURE 6–5 The Baddeley-Hitch model of working memory
Two distinct storage buffers, one for verbal and the other for visuospatial information, interact with a central executive controller. (Baddeley, A. D., and Hitch, G. J. (1974). Working memory. In G. Bower (ed.), The psychology of learning and motivation (Vol. VIII, pp. 47–89). New York: Academic Press. Reprinted with permission from Elsevier.)
- Understanding the Working Memory Model 251
proposed that the phonological loop system involves two subcomponents: a phonological store and an articulatory rehearsal process (Baddeley, 1986). When visually presented verbal information is encoded, the information is transformed into a sound-based, or “auditory-phonological,” code. This code is something like an internal echo-box, a repository for sounds that reverberate briefly before fading away. To prevent complete decay, an active process must refresh the infor- mation, and this is where the idea of a “loop” comes in. The active refreshment comes via articulatory rehearsal , as you voice internally the sounds you heard in- ternally. (The process seems very like our ability to “shadow”; that is, to repeat quickly something that we hear, whether or not we understand it—an indication that the phonological loop may be involved in language learning.) Once the ver- bal information is spoken internally by the mind’s voice in rehearsal, it can then be again heard by the mind’s ear and maintained in the phonological store. In this way a continuous loop plays for as long as the verbal material needs to be main- tained in working memory. The first step of the process—translation into a phonological code—is of course necessary only for visually presented material. For auditory information, such as speech, initial access to the phonological store is automatic. This idea sounds intuitive, because the experience of this kind of internal re- hearsal seems universal, and that has been part of its appeal. For example, in your conversation about the movie, it is likely that you would be using the phonological loop to rehearse the key points you want to make and also time-sharing this same system to help process your friend’s speech. It is significant that this description of the phonological loop component of verbal working memory includes a number of characteristics that should be testable. First, verbal working memory capabilities should depend on the level of difficulty of both “phonological processing” (translating verbal information into a sound-based code) and “articulatory processing” (translating verbal informa- tion into a speech-based code). Second, because working memory is flexible, per- formance on verbal working memory tasks will not be catastrophically disrupted if for some reason the phonological loop component is unusable: in that case, other components, the central executive and the visuospatial scratchpad, kick in. Thus, in your movie conversation, if verbally processing your friends’ ideas tem- porarily uses up too much capacity of the phonological loop, you might be able to use the visuospatial scratchpad to rehearse your ideas, possibly by using visual mental imagery—forming a mental image of your ideas rather than thinking of them in verbal terms. Third, the phonological loop model suggests that the two primary components of verbal working memory—phonological storage and artic- ulatory rehearsal—are subserved by functionally independent systems, and hence should be dissociable. All these hypotheses have been tested in experiments, and all have held up. Behavioral studies have suggested that phonological and articulatory factors significantly affect verbal working memory performance. One example is the phonological similarity effect : when items simultaneously stored in working memory have to be serially recalled, performance is significantly worse when the items to be
252 CHAPTER 6 Working Memory
maintained are phonologically similar—that is, when they sound the same (Conrad & Hull, 1964). The effect is thought to be caused by confusions that arise when sim- ilar sound-based codes are activated for the different items in the phonological loop. This finding can easily be informally appreciated. Try holding these two strings of letters in working memory, one after the other: D B C T P G K F Y L R Q In the first string, the letters all have the “ee” sound; in the second list, all the letter sounds are distinct. Which did you find easier to remember and repeat? In these tasks, the typical error is substituting a phonologically similar item, such as “V” for “G.” The other part of the phonological loop, articulatory processing, or the “speaking” of presented items by the inner voice, is reflected in the word-length ef- fect. Performance on a recall task is worse when the items to be maintained are long words, such as university, individual, and operation, than short words, such as yield, item, and brake. The key factor seems not to be the number of syllables per se, but rather the time it takes to pronounce them: performance is worse for two-syllable words that have long vowel sounds, such as harpoon and voodoo, than for two-syllable words with short vowel sounds, such as bishop and wiggle (Baddeley et al., 1975). The phonological loop model accounts for the word-length effect by the assumption that pronunciation time affects the speed of silent re- hearsal, which requires speech-based processing. The longer it takes to rehearse a set of items in working memory, the more likely those items will have been dropped from the phonological store. The relationship between pronunciation time and working memory perform- ance was further tested in a study involving children bilingual in Welsh and English (Ellis & Hennelly, 1980). The names of the digits in Welsh have the same number of vowels as the English names but generally have longer vowel sounds and conse- quently take longer to say. As predicted, when performing digit-span tests in Welsh, the children scored significantly below average norms. However, when they per- formed the tests again in English their scores were normal. Follow-up studies have confirmed that the faster an individual’s speech rate, the more items can be recalled correctly from working memory (Cowan et al., 1992). What happens when the normal operation of the phonological loop is dis- rupted? The Baddeley-Hitch model suggests that the central executive and the vi- suospatial scratchpad take over and with the phonological loop out of operation phonological similarity and word length should no longer have an effect on work- ing memory. Can this hypothesis be tested? Yes, by experiments based on dual- task interference. Participants are asked to maintain visually presented words in working memory while simultaneously producing overt and irrelevant speech, a task that interferes with phonological processing and rehearsal of the informa- tion. (Imagine that, in your movie conversation, while you are trying to keep in mind the point you want to make you also have to say the word the over and over again out loud; you can see how such conditions might make it almost impossible
254 CHAPTER 6 Working Memory
similarity effect on performance was not influenced by word length, and vice versa (Figure 6–6) (Longoni et al., 1993). Of course, behavioral data can provide only one kind of evidence for functional independence. Results from brain-based studies provide a different kind of evidence, showing that separate systems support phonological storage and rehearsal. On the one hand, studies of patients with brain damage have documented a re- lationship between left inferior parietal damage and phonological storage impair- ments, and a relationship between left inferior frontal cortex damage and articulatory rehearsal impairments (Vallar & Papagaro, 1995). (The left inferior frontal cortex, also referred to as Broca’s area, is known to be involved with language.) On the other hand, neuroimaging studies have provided a means to examine these rela- tionships in neurologically healthy participants. Such studies can show whether these brain regions are in fact the ones engaged during normal processing condi- tions. For example, participants in one study were asked to memorize a series of six visually presented items, either six English letters or six Korean language char- acters (none of the participants were speakers of Korean) (Paulesu et al., 1993). The researchers assumed that the phonological loop system would be engaged to main- tain the English letters but not utilized for the Korean characters (because the sounds represented by the characters were unknown to the participants). This assumption was validated by testing the effects of articulatory suppression—as
100
80
60
40
20
0 Short Long
Phonologically dissimilar words
Accuracy of recall (%) Phonologically similar words
Word length
FIGURE 6–6 Independence of the effects of word length and phonological similarity An immediate recall task presented participants with five words that were either phonologically simi- lar (such as FASTER, PLASTER, MASTER, TASTER, and LASTED) or dissimilar (such as FAMOUS, PLASTIC, MAGIC, TEACHER, and STAYED), and were either short (two syllables) or long (four sylla- bles). Both similarity and greater word length decreased recall performance, but the parallel slopes of the lines indicate that the two effects are independent. (Adapted from Longoni, A. M., Richardson, J. T. E., and Aiello, A. (1993). Articulatory rehersal and phonological storage in working memory. Memory and Cognition, 21 (1), 11–22. Reprinted with permission.)
- Understanding the Working Memory Model 255
expected, articulatory suppression impaired memory performance for the English letters, but had no effect on memory for the Korean letters. PET images revealed in- creased blood flow in both left inferior parietal cortex (storage) and left inferior frontal cortex (rehearsal) only for the English letters. It is interesting that activation was also observed in brain structures associated with motor-related components of speech, even though the task did not require participants to speak overtly. The speech- related brain activity was thus thought to represent “internal speech” or subvocal re- hearsal. In a second experiment, Paulesu and colleagues (1993) attempted to dissociate regions associated with phonological storage from those involved in rehearsal. They asked the same participants to perform rhyme judgments on the English let- ters, deciding whether each letter in turn rhymed with “B.” Here the researchers assumed that the rhyme task would engage rehearsal but not storage, and so it proved. In contrast to the results for the English letter group in the first experi- ment, in which there was increased blood flow in both brain regions, this time only the left frontal cortex was activated; the left parietal cortex was not active above baseline (Figure 6–7b on Color Insert I). Thus, behavioral and neuroimaging re- sults converge to establish the dissociability of the storage and rehearsal compo- nents of verbal working memory. However, additional neuroimaging studies suggest a more complex picture. For example, different subregions of Broca’s area (which is crucially involved in produc- ing speech) appear to be engaged at distinct points in time during the delay period of working memory tasks (Chein & Fiez, 2001). The investigators argue that the more dorsal region of Broca’s area is active only during the first part of the delay period, and is involved in the formation of an articulatory rehearsal program; in contrast, the more ventral region of Broca’s area is active during the remainder of the delay period, and is involved with the act of rehearsal itself. Neuroimaging studies continue to play an important role in refining and reshaping the verbal working memory model. The larger question is what is the true function of the phonological loop in cog- nition? Surely it did not arise just to help us retain letter strings or telephone num- bers! It seems intuitive that the phonological loop would have to play some role in language processing, because it is so clearly integrated with language comprehension and production systems. One hypothesis is that working memory—specifically, the phonological loop—is not critical for comprehension of familiar language, but it is essential for learning new language (Baddeley et al., 1998), a challenge experienced both by children learning their first language and by adults learning a second one or acquiring new vocabulary. It may be that evolution has imbued us with a specific ex- pertise in repeating what we hear, even if we don’t initially understand it. This form of imitation is something that even young infants can do, and it may provide a means for helping us learn new words via a linkage of sound and meaning. Developmental data strongly support this claim: the level of children’s ability to repeat nonwords strongly predicts the size of their vocabulary one year later (Gathercole & Baddeley, 1989). The patient P.V. was found to be completely un- able to learn the Russian equivalent of any words in her native Italian despite ex- tensive practice (Baddeley et al., 1988). Yet she could learn a novel association
- Understanding the Working Memory Model 257
of the letter? To test whether the participants were using visuospatial representations to do the task, some participants were instructed to point to the word YES or NO printed irregularly on a page, and others had to speak the words “yes” or “no.” The hypothesis was that if the classification decision depended on visuospatial represen- tations, then requiring the pointing—a visuospatially based response—would inter- fere with performance. This is exactly what was found; participants took almost three times as long to perform the task when they had to point in response than when they had to speak (Figure 6–8b). These results, and those of many other studies that followed, suggest that mental navigation is an inherently spatial process (Logie, 1995). The subjective experience of moving the mind’s eye from one spatial location to another also suggests the possibility that visuospatial working memory depends on brain systems that plan movements of the eyes (or possibly other parts of the body), just as verbal working memory depends on brain systems involved with planning speech (Baddeley & Lieberman, 1980). Inter- estingly, this movement planning system might also be the basis for spatial rehearsal , the process of mentally refreshing stored locations to keep them highly accessible. The idea is that when you rehearse spatial locations in working memory (think of mentally visualizing driving directions to turn left at the next block, and then right at the stop- light), you are actually utilizing the same systems that would help you move your eyes or body toward that location. And just as rehearsal of verbal information does not re- quire actual speech, it is thought that rehearsal of spatial information does not require actual eye (or body) movements. Instead, spatial rehearsal may involve covert shifts of attention to memorized spatial locations (Awh & Jonides, 2001). In other words, just as we can keep our attention focused on a place in space without actually physically looking at it, we might also be able to keep remembered locations in memory by covertly focusing our attention on those remembered loca- tions. An example: think of being at a party and talking with one friend, keeping your eyes focused on him, while also paying attention, out of the corner of your eye, to the gestures of another friend to your left. This analogy leads to concrete predictions. It is thought that paying attention to a spatial location will enhance perceptual processing at that location. If the systems for spatial working memory are the same as those for spatial attention, then keeping a particular location in spatial working memory should also enhance perceptual pro- cessing of visual information that is physically presented at the remembered location. This prediction was tested behaviorally (Awh et al., 1998). In a spatial working memory task single letters (the cues) were briefly presented in varying loca- tions on a display; after a short delay, another letter (the probe) was presented. In one condition, participants had to remember the location of the cue, and to decide whether the probe was in the same location. In another condition, it was the identity of the letter cue that had to be maintained, and participants had to decide whether the probe had the same identity. Additionally, during the delay participants had a second task—to classify the shape of an object appearing at different locations. On some trials the object appeared in the same location as the letter cue that was being maintained. It was found that the shape-classification decision was made more quickly when the shape’s location matched that of the cue, but only when the
258 CHAPTER 6 Working Memory
information being maintained was the location of the cue. This result suggested that maintaining a location in working memory facilitates the orienting of attention to that location (which is what improved the speed of the shape-classification task). Neuroimaging studies have provided even stronger evidence that rehearsal in spa- tial working memory and spatial selective attention draw on at least some of the same processes, by demonstrating that they both rely on the same right-hemisphere frontal and parietal cortex brain regions. Maintaining a spatial location in working memory produced enhanced brain activity in visual cortex regions of the opposite hemisphere, as expected because of the contralateral organization of these brain regions (Figure 6– on Color Insert J) (Awh & Jonides, 2001; Postle et al., 2004). These results suggest that spatial working memory is accomplished by enhancing processing in brain regions that support visual perceptual processing of those locations. As the compound nature of its name implies, information processed by the visu- ospatial scratchpad is of two sorts: spatial, like the arrangement of your room, and vi- sual, like the face of a friend or the image of a favorite painting. It seems that different types of codes may be required to maintain these two types of nonverbal information on the visuospatial scratchpad. For example, we seem to have the ability to “zoom in” on images like faces and paintings, magnifying particular features (Kosslyn, 1980). And we are able to decompose objects into constituent parts and transform them. We can, for example, imagine how a clean-shaven friend would look with a beard. These men- tal operations seem to be inherently nonspatial, yet nevertheless they require an accu- rate visual representation to be maintained and manipulated within working memory. Thus, visuospatial working memory may be composed of two distinct systems, one for maintaining visual object representations and the other for spatial ones. The distinction between object and spatial processing is clearly in line with obser- vations about the visual system: there is a great deal of evidence for distinct neural pathways involved in processing spatial and object visual features (respectively, the dorsal “where” and ventral “what” pathways) (Ungerleider & Mishkin, 1982; see dis- cussion in Chapter 2). In monkeys it has been found that this distinction is also pres- ent for working memory: neurons in the dorsal region of the prefrontal cortex respond especially strongly to stimuli during a spatial working memory task, whereas neurons in the ventral prefrontal cortex respond especially strongly during an object working memory task (Wilson et al., 1993). In humans, some patients with brain damage have shown selective impairments on nonspatial mental imagery tasks (for example, mak- ing judgments about the shape of a dog’s ears), but not on those involving spatial im- agery (for example, rotating imagined objects) (Farah et al., 1988). The reverse pattern has been observed with other patients, demonstrating a double dissociation (Hanley et al., 1991). Neuroimaging studies have also tended to show dissociations between brain systems involved in spatial and in object working memory (Courtney et al., 1996; Smith et al., 1995), although these dissociations have been most reliable in pos- terior rather than prefrontal cortex (the region identified in monkey studies) (Smith & Jonides, 1999). The specific characteristics of object working memory, such as whether or not it involves a distinct storage buffer or rehearsal system, are not yet well worked out, and the question of a dissociation of object and spatial working memory remains a topic of continued study.
