Memorization and recall of very long lists accounted for within the Long-Term Working Memory framework

In a recent paper, Hu, Ericsson, Yang, and Lu (2009) found that an ability to memorize very long lists of digits is not mediated by the same mechanisms as exceptional memory for rapidly presented lists, which has been the traditional focus of laboratory research. Chao Lu is the holder of the Guinness World Record for reciting the most decimal positions of pi, yet he lacks an exceptional memory span for digits. In the first part of this paper we analyzed the reliability and structure of his reported encodings for lists of 300 digits and his application of the story mnemonic. Next, his study and recall times for lists of digits were analyzed to test hypotheses about his detailed encoding processes, and cued-recall performance was used to assess the structure of his encodings. Three experiments were then designed to interfere with the uniqueness of Chao Lu's story encodings, and evidence was found for his remarkable ability to adapt his encoding processes to reduce the interference. Finally, we show how his skills for encoding and recalling long lists can be accounted for within the theoretical framework of Ericsson and Kintsch's (1995) Long-Term Working Memory.     

Memory skills mediating superior memory in a world-class memorist

Laboratory studies have investigated how individuals with normal memory spans attained digit spans over 80 digits after hundreds of hours of practice. Experimental analyses of their memory skills suggested that their attained memory spans were constrained by the encoding time, for the time needed will increase if the length of digit sequences to be memorised becomes longer. These constraints seemed to be violated by a world-class memorist, Feng Wang (FW), who won the World Memory Championship by recalling 300 digits presented at 1 digit/s. In several studies we examined FW’s memory skills underlying his exceptional performance. First FW reproduced his superior memory span of 200 digits under laboratory condition, and we obtained his retrospective reports describing his encoding/retrieval processes (Experiment 1). Further experiments used self-paced memorisation to identify temporal characteristics of encoding of digits in 4-digit clusters (Experiment 2), and explored memory encoding at presentation speeds much faster than 1 digit/s (Experiment 3). FW’s superiority over previous digit span experts is explained by his acquisition of well-known mnemonic techniques and his training that focused on rapid memorisation. His memory performance supports the feasibility of acquiring memory skills for improved working memory based on storage in long-term memory.     

Dissociating retention and access in working memory: an age-comparative study of mental arithmetic

In two experiments, young and older adults solved arithmetic chain tasks with single-digit operands, with or without a concurrent memory load of three or six digits. Variables in the arithmetic tasks had to be replaced by digits from the screen or from the memory set. A task-irrelevant concurrent load impaired neither speed nor accuracy of arithmetic in younger adults. In Experiment 2, this was also true for older adults. A large decrease in arithmetic performance was observed, however, when variables in the arithmetic task had to be substituted by digits from the memory list. Older adults had specific problems with this condition in Experiment 1, where the substitution involved two successive steps, but not in Experiment 2, where the substitution from memory could be done in a single step. The results are difficult to reconcile with models assuming a common resource for storage and processing. Rather, they are compatible with the hypothesis that a concurrent memory load interferes with a processing task only during the points of access to working memory. Further, even though access to working memory was found to be the critical source of concurrent-load interference, it was found to be insensitive to the effects of adult aging.     

The Role of Implicit Memory in Controlling a Dynamic System

The relationship between implicit memory and implicit learning is explored. Dienes and Fahey (1995) showed that learning to control a dynamic system was mediated by a look-up table consisting of previously successful responses to specific situations. The experiment reported in this paper showed that facilitated performance on old situations was independent of the subjects' ability to recognize those situations as old, suggesting that memory was implicit. Further analyses of the Dienes and Fahey data replicated this independence of control performance on recognition. However, unlike the implicit memory revealed on fragment completion tasks, successful performance on the dynamic control tasks was remarkably resilient to modality shifts. The results are discussed in terms of models of implicit learning and the nature of implicit memory.     

Binding serial order to representations in working memory: a spatial/verbal dissociation

Verbal information is coded naturally as ordered representations in working memory (WM). However, this may not be true for spatial information. Accordingly, we used memory span tasks to test the hypothesis that serial order is more readily bound to verbal than to spatial representations. Removing serial-order requirements improved performance more for spatial locations than for digits. Furthermore, serial order was freely reproduced twice as frequently for digits as for locations. When participants reordered spatial sequences, they minimized the mean distance between items. Participants also failed to detect changes in serial order more frequently for spatial than for verbal sequences. These results provide converging evidence for a dissociation in the binding of serial order to spatial versus verbal representations. There may be separable domain-specific control processes responsible for this binding. Alternatively, there may be fundamental differences in how effectively temporal information can be bound to different types of stimulus features in WM.     

Memory span as a measure of individual differences in memory capacity

Two experiments were carried out to investigate whether the immediate digit span measure traditionally used in the assessment of individual differences in cognition is a good predictor of performance on other memory tasks. In the first experiment, it was found that subjects’ digit spans were not significantly related to their performances on either short-term or longterm memory tasks, or to theoretical measures of their memory store capacities. Memory for the temporal occurrence of events, however, proved to be positively correlated with digit span. A second experiment confirmed that digit span was correlated with memory for the temporal occurrence of events, but not with item memory. Thus it was concluded that an individual’s digit span reflects his ability to retain information about the order of a sequence of events rather than the capacity of his short- or long-term memory.     

Competing with reasoning: A test of the working memory hypothesis

An experiment is reported in which subjects attempted conditional reasoning problems while concurrently articulating a series of digits, with or without memory load. Logical performance was not impaired by the competing tasks and the latency of responding was actually faster under concurrent articulation, without memory load, than in a control group. The results are discussed with reference to the Baddeley and Hitch (1974) model of working memory.     

Cross-task cross talk in memory and perception

The application of the additive factors method depends on finding factors that selectively influence processing stages. When all the processes for a task are in series, a factor directly influencing a process might change its output and thereby have indirect influence on succeeding processes. We investigate whether such indirect influence is possible between processes associated with different tasks being performed together. In two dual-task memory scanning and arithmetic experiments with digits as the stimuli for both tasks, information relevant for only one of the tasks nonetheless affected performance of the other. When the same digit was relevant for the two tasks, cross-task facilitation and interference were observed in some cases. Displaying the same digit for both tasks led to relatively fast response times, paralleling the effect of flankers in the response competition paradigm. But repetition of digits in memory slowed responses. It is suggested that the need for control processes to keep task information segregated is responsible for the pattern of effects.     

Comparison of memory skills in learning disabled, low-reading, and nondisabled adolescents

Concerns are raised regarding the samples used in studies of learning disabled children. In this study, subjects were 56 high school students; 10 exhibited perceptual or processing difficulties ("learning disabled"), 15 performed poorly on reading achievement tests ("low readers"), and 31 showed no scholastic problems ("nondisabled"). Scores of the three groups were compared on nine memory tasks involving either auditory or visual input and encompassing a wide variety of content (digits, pictures, related and unrelated words, paired-associates, or sentences). The nondisabled group performed significantly better than the learning disabled group on all but one of the tasks. The low readers' performance was similar to that of the learning disabled group for some tasks, but significantly different on other tasks, most notably on tasks involving visual factors. Implications for research with learning disabled populations are discussed.     

Adult counting is resource demanding

Several recent studies on both the development of counting and working‐memory span tasks have provided results that could be interpreted as ruling out any cognitive resource model for counting. The aim of this study was to test the hypothesis that, even in adults, counting is a demanding task that requires the allocation of cognitive resources. In a first experiment, we asked adults to count arrays of dots while maintaining 5 items in memory (either digits or letters). As we predicted, the concurrent memory load did not increase the rate of errors but induced longer counting times. In a second experiment, we asked adults to count using either the numeric chain or the alphabet while they maintained 1, 3 or 5 items in memory (digits or letters). First, we replicated the load effect observed in Experiment 1. Second, though both types of counting required similar amounts of time, counting with the less automatized chain (i.e. the alphabet) resulted in a poorer recall performance. Finally, this detrimental effect in recall was all the more pronounced the greater the number of items to be recalled. These results are interpreted within theoretical frameworks that consider cognitive resources as attentional capacities.     

Is mental rotation controlled or automatic?

A task requiring the mental rotation of letters was performed either alone or during the retention interval of two short-term memory tasks. The retention of eight digits or random configurations of eight dots slowed overall reaction time on the mental-rotation task, but did not significantly alter the estimated rate of mental rotation. Conversely, performance on the memory tasks did not vary with the angle of mental rotation. These results support the view that attentional control is required to set up the mental structures required in mental rotation, but that the actual execution of the mental-rotation component can be relegated to subordinate systems that do not compete for attentional resources.     

Differential constraints on the working memory and reading abilities of individuals with learning difficulties and typically developing children

This study examined the factors that constrain the working memory span performance and reading ability of individuals with generalized learning difficulties. In the study, 50 individuals with learning difficulties (LD) and 50 typically developing children (TD) matched for reading age completed two working memory span tasks. Participants also completed independent measures of the processing and storage operations involved in each working memory span task and Raven's Coloured Progressive Matrices. The results showed that despite an equivalent level of working memory span, the relative importance of the constraints on working memory differed between the groups. In addition, working memory span was not closely related to word recognition or sentence comprehension performance in the LD group. These results suggest that the working memory span performance of LD and TD individuals may reflect different working memory limitations and that individuals with generalized learning difficulties may approach cognitive tasks in a qualitatively different way from that of typically developing individuals.     

Decomposing adult age differences in symbol arithmetic

A componential analysis was conducted to determine the locus of adult age differences in symbol arithmetic. Measures of the duration of two proposed components, substitution of digits for symbols and the addition or subtraction of the digits resulting from these substitutions, were obtained from 52 young adults and 52 older adults. Tests of working memory, perceptual speed, motor speed, and associative learning were also administered to all subjects. The results were most consistent with an interpretation postulating that the speed of many different cognitive processes decreases with increased age. Considerable age-related variance remained in the measures of symbol arithmetic performance after statistical control of working memory and associative learning performance, casting doubt on alternative hypotheses of the source of age-related differences in this task.     

Working memory, psychiatric symptoms, and academic performance at school

Previous studies of the relationship among working memory function, academic performance, and behavior in children have focused mainly on clinical populations. In the present study, the associations of the performance in audio- and visuospatial working memory tasks to teacher reported academic achievement and psychiatric symptoms were evaluated in a sample of fifty-five 6-13-year-old school children. Working memory function was measured by visual and auditory n-back tasks. Information on incorrect responses, reaction times, and multiple and missed responses were collected during the tasks. The children's academic performance and behavioral and emotional status were evaluated by the Teacher Report Form. The results showed that good spatial working memory performance was associated with academic success at school. Children with low working memory performance, especially audiospatial memory, were reported to have more academic and attentional/behavioral difficulties at school than children with good working memory performance. An increased number of multiple and missed responses in the auditory and visual tasks was associated with teacher reported attentional/behavioral problems and in visual tasks with teacher reported anxiety/depressive symptoms. The results suggest that working memory deficits may underlie some learning difficulties and behavioral problems related to impulsivity, difficulties in concentration, and hyperactivity. On the other hand, it is possible that anxiety/depressive symptoms affect working memory function, as well as the ability to concentrate, leading to a lower level of academic performance at school.     

When visual and verbal memories compete: Evidence of cross-domain limits in working memory

Recently, investigators have suggested that visual working memory operates in a manner unaffected by the retention of verbal material. We question that conclusion on the basis of a simple dual-task experiment designed to rule out phonological memory and to identify a more central faculty as the source of a shared limitation. With a visual working memory task in which two arrays of color squares were to be compared, performance was unaffected by concurrent recitation of a two-digit list or a known seven-digit sequence. However, visual working memory performance decreased markedly when paired with a load of seven random digits. This was not a simple tradeoff, inasmuch as errors on the visual array and high digit load tasks tended to co-occur. Working memory for digits and visual information thus are both subject to at least one type of shared limit, not just domain-specific limitations. The nature of the shared limit is discussed.     

Basic numerical skills in children with mathematics learning disabilities: a comparison of symbolic vs non-symbolic number magnitude processing

Forty-five children with mathematics learning disabilities, with and without comorbid reading disabilities, were compared to 45 normally achieving peers in tasks assessing basic numerical skills. Children with mathematics disabilities were only impaired when comparing Arabic digits (i.e., symbolic number magnitude) but not when comparing collections (i.e., non-symbolic number magnitude). Moreover, they automatically processed number magnitude when comparing the physical size of Arabic digits in an Stroop paradigm adapted for processing speed differences. Finally, no evidence was found for differential patterns of performance between MD and MD/RD children in these tasks. These findings suggest that children with mathematics learning disabilities have difficulty in accessing number magnitude from symbols rather than in processing numerosity per se.     

Individual differences in phonological learning and verbal STM span

A relationship between phonological short-term memory tasks (e.g., nonword repetition, digit span) and vocabulary learning in both experimental and real-life conditions has been reported in numerous studies. A mechanism that would explain this correlation is, however, not known. The present study explores the possibility that it is the quality of phonological representations that affects both short-term recall and long-term learning of novel wordlike items. In Experiment 1, groups with relatively good and poor span for pseudowords were established. The good group was found to perform better at explicit memory tasks tapping the incidental learning of a limited stimulus pool used in an auditory immediate serial pseudoword recall task. In Experiment 2, the results of Experiment 1 were replicated when experience of correct recall was controlled. In Experiment 3, the immediate recall performance of the good group was found to benefit more than that of the poor group from syllable repetition within stimulus pools. It is concluded that the efficiency of a process that creates phonological representations is related both to short-term capacity for verbal items, and to long-term phonological learning of the structure of novel phonological items.     

Selective sparing of memory functioning in a patient with amnesia following herpes encephalitis

A patient with amnesia subsequent to herpes encephalitis was assessed on a series of tasks which have been shown to result in normal or near normal performance by amnesic subjects. On a pursuit rotor learning task, our patient performed normally, and continued to show evidence of memory for the skill after a period of 30 days. On a perceptual learning task, which involved the identification of line drawings, he also displayed evidence of learning and long-term retention, but his performance remained below that of control subjects. In two priming experiments, which employed a "case-control" paradigm, performance on homonym-spelling and category item-generation tasks was preceded by exposure to a selected list of words. There was no evidence of differential facilitation in respect of those words used in the initial priming task. The findings point to some degree of selectivity of preserved learning capabilities in amnesia. It is hypothesized that this selectivity may be determined by several variables, including the type of task, the etiology and site of pathology, and the severity of amnesia.     

Incidental visual memory for objects in scenes

Three experiments were conducted to investigate the existence of incidentally acquired, long-term, detailed visual memory for objects embedded in previously viewed scenes. Participants performed intentional memorization and incidental visual search learning tasks while viewing photographs of real-world scenes. A visual memory test for previously viewed objects from these scenes then followed. Participants were not aware that they would be tested on the scenes following incidental learning in the visual search task. In two types of memory tests for visually specific object information (token discrimination and mirror-image discrimination), performance following both the memorization and visual search conditions was reliably above chance. These results indicate that recent demonstrations of good visual memory during scene viewing are not due to intentional scene memorization. Instead, long-term visual representations are incidentally generated as a natural product of scene perception.     

The memory of two great mental calculators: Charcot and Binet's neglected 1893 experiments

French neurologist Jean Martin Charcot (1825-1893) and French psychologist Alfred Binet (1857-1911) are almost unknown as investigators who conducted original and fascinating studies in the area of memory. In a series of 1893 experiments, they compared the performance of two expert mental calculators, Jacques Inaudi and Périclès Diamandi, in tasks that consisted of recalling digits. Inspired by Ribot's psychological work (1881), they believed in the existence of not one type of memory but several partial, special, and local memories, each devoted to a particular domain. In all arithmetical prodigies, memory for digits is abnormally developed compared with other memories. Inaudi was considered to be an auditory memory-based mental calculator; when memorizing digits, he did not rely onthe appearance of the items or create visual imagery of any kind. Rather, he remembered digits principally by their sounds. Inaudi's methods of calculation and memorization were original and different from those used by Diamandi, who was a typical visual memory-based mental calculator. The experiments presented in the 1893 article were among the first scientific demonstrations of the importance to psychology of studying different types of memory. The present work gives a translation of this pioneering experimental article on expert calculators by Charcot and Binet, instructive for the comprehension of normal memory.