Search

The present study investigated whether lexical codes contribute to retention of verbal information in working memory. We used event-related brain potentials (ERPs) recorded while participants were performing a serial recall task to show differences in brain activity during retention of words or pseudowords. The effects of lexical status and memory load (task difficulty) upon ERP activity during retention also differed. with memory load affecting ERP indices of phonological processing. The timing of the ERPs suggested that the influence of lexical status upon retention began during encoding. The contrast between ERPs in the serial recall task and in a control task with negligible memory demand indicated that the effect of lexical status in the retention interval of the memory task was specific to consciously controlled memory operations. These results support the view that lexical codes actively contribute to retention of words in working memory. (c) 1999 Academic Press [ABSTRACT FROM AUTHOR]

The Implicit Association Test (IAT) is a popular behavioral measure that assesses the associative strength between outgroup members and stereotypical and counterstereotypical traits. Less is known, however, about the degree to which the IAT reflects automatic processing. Two studies examined automatic processing contributions to a gender-IAT using a data driven, social neuroscience approach. Performance on congruent (e.g., categorizing male names with synonyms of strength) and incongruent (e.g., categorizing female names with synonyms of strength) IAT blocks were separately analyzed using EEG (event-related potentials, or ERPs, and coherence; Study 1) and lesion (Study 2) methodologies. Compared to incongruent blocks, performance on congruent IAT blocks was associated with more positive ERPs that manifested in frontal and occipital regions at automatic processing speeds, occipital regions at more controlled processing speeds and was compromised by volume loss in the anterior temporal lobe (ATL), insula and medial PFC. Performance on incongruent blocks was associated with volume loss in supplementary motor areas, cingulate gyrus and a region in medial PFC similar to that found for congruent blocks. Greater coherence was found between frontal and occipital regions to the extent individuals exhibited more bias. This suggests there are separable neural contributions to congruent and incongruent blocks of the IAT but there is also a surprising amount of overlap. Given the temporal and regional neural distinctions, these results provide converging evidence that stereotypic associative strength assessed by the IAT indexes automatic processing to a degree.

The present study supports activation models of verbal short-term memory that include a semantic contribution to the retention process. Event-related brain potentials were used to probe the level of activation of semantic representations of a series of words in a delay interval following their presentation. The levels of activation were compared in two tasks: (1) a short-term memory task that involved a semantic judgment in the recall phase following the delay interval, and (2) a nonmemory control task. The level of semantic activation during the delay interval was higher in the short-term memory task, indicating that enhanced activation of semantic representations is involved in the short-term storage of verbal information. This result implies that activated long-term memory provides a representational basis for semantic verbal short-term memory, and hence supports theories that postulate that short- and long-term stores are not separate.

High temporal resolution event-related brain potential and electroencephalographic coherence studies of the neural substrate of short-term storage in working memory indicate that the sustained coactivation of both prefrontal cortex and the posterior cortical systems that participate in the initial perception and comprehension of the retained information are involved in its storage. These studies further show that short-term storage mechanisms involve an increase in neural synchrony between prefrontal cortex and posterior cortex and the enhanced activation of long-term memory representations of material held in short-term memory. This activation begins during the encoding/comprehension phase and evidently is prolonged into the retention phase by attentional drive from prefrontal cortex control systems. A parsimonious interpretation of these findings is that the long-term memory systems associated with the posterior cortical processors provide the necessary representational basis for working memory, with the property of short-term memory decay being primarily due to the posterior system. In this view, there is no reason to posit specialized neural systems whose functions are limited to those of short-term storage buffers. Prefrontal cortex provides the attentional pointer system for maintaining activation in the appropriate posterior processing systems. Short-term memory capacity and phenomena such as displacement of information in short-term memory are determined by limitations on the number of pointers that can be sustained by the prefrontal control systems.

In filler-gap sentences, a phrase ('filler') is separated by intervening words from a subsequent phrase ('gap') with which it is integrated. The filler-gap interval provides a useful model for the study of short-term retention processes during sentence comprehension. Kluender and Kutas [J. Cogn. Neurosci. 5 (1993) 29] used event-related brain potentials (ERPs) to show that a filler phrase places a demand on short-term retention processes in the filler-gap interval, but left the processing level at which this demand arises unspecified. Here we use ERPs to address the issue of whether the filler places a demand on the semantic component of short-term retention processes in the filler-gap interval. Participants read filler-gap sentences, which began with a filler phrase and, in the filler-gap interval, contained a subject and object that were either semantically related or unrelated. There was also a control condition in which the filler phrase was absent (i.e. less memory demand). The main result was that during the filler-gap interval, bilateral posterior electrodes displayed a larger positivity for unrelated than related words. Moreover, during the same interval, posterior electrodes displayed a larger positivity for filler-gap sentences than for control sentences. In the control condition (non-filler gap sentences), manipulation of semantic relatedness did not produce differences in ERP activity. Our results suggest that a filler phrase places a demand on the semantic component of verbal working memory during on-line sentence comprehension., (Copyright 2003 Elsevier Science B.V.)

This study investigated cognitive and neural processes involved in gap filling during on-line sentence comprehension. Electroencephalogram (EEG) coherences were used to demonstrate that increases in the synchronization of neural activity in different cortical regions occur during gap filling when load in semantic working memory is high due to semantically unrelated words in the filler-gap interval. Sentences could either require gap filling at a verb or not, and the nouns preceding the verb could be either semantically related or unrelated. In the unrelated but not related condition, coherences in the beta band were larger during verb processing for sentences requiring gap filling compared to sentences not requiring gap filling. The coherence changes involved linkages between frontal and posterior temporal-parietal sites in both hemispheres. These results further indicate that semantic working memory is involved in the process of gap filling.

Event-related potentials (ERPs) recorded from the human scalp can provide important information about how the human brain normally processes information and about how this processing may go awry in neurological or psychiatric disorders. Scientists using or studying ERPs must strive to overcome the many technical problems that can occur in the recording and analysis of these potentials. The methods and the results of these ERP studies must be published in a way that allows other scientists to understand exactly what was done so that they can, if necessary, replicate the experiments. The data must then be analyzed and presented in a way that allows different studies to be compared readily. This paper presents guidelines for recording ERPs and criteria for publishing the results.

A. R. Haig, E. Gordon, and S. Hook (1997) disputed G. McCarthy and C. C. Wood's (1985) contention that scaling should be used when assessing the statistical significance of between condition (or group) differences in the shapes of event-related potential (ERP) scalp topographies. Haig et al. based their contention upon the lack of empirical realism in McCarthy and Wood's model of within-group ERP noise, claiming that McCarthy and Wood's results could not be generalized to realistic ERP data. We argue, on both empirical and theoretical grounds, that Haig et al. do not make a compelling case against generalization of McCarthy and Wood's results. Moreover, Haig et al.'s conclusion is based upon a misconception of how scaling should be used. We conclude that when a quantitative measure of differences between topographic shapes is needed, scaling is not an option--it is a requirement.

The present study was concerned with whether there are separate, modality-specific processing "streams" in verbal working memory for information that is heard or read. We used event-related brain potentials (ERPs) recorded from scalp of normal humans to show between-modality differences in spatio-temporal patterns of brain activity during retention in working memory of aurally or visually presented verbal information. The ERP patterns suggested that a sustained, automatically maintained auditory store was activated by auditory presentation and a transient, visual-verbal store was activated by visual presentation. In addition to these modality-specific differences, the ERPs indicated that the phonological loop was activated in both modalities and further suggested that the onset of phonological loop activation was earlier for auditory presentation.

Numerous studies have shown that the visual system involves different cortical pathways in the perception of object (ventral visual pathway) and spatial (dorsal visual pathway) information. The present study was concerned with whether human visuospatial working memory divides along similar lines. We used event-related brain potentials (ERPs) recorded from scalp of normal humans to show the existence of different buffering systems for the retention of object and spatial visual information. Subjects were presented with object or spatial stimuli to be retained for a 3.6-sec interval. The ERPs isolated brain activity associated with retention from earlier storage and later retrieval processes. The ERP scalp topographies indicated that the underlying patterns of brain activation were different during retention of object and spatial information.

Some event-related brain potential (ERP) studies of working memory have used delayed match-to-sample designs in which a stimulus (S1) is held in memory for comparison with a subsequent stimulus (S2). During the S1-S2 interval, ERP slow negatives varied with both the type and amount of material held in working memory. One interpretation is that these slow waves index working memory operations. An alternative explanation is that they only reflect general preparatory processing for the response to S2. To decide between these explanations, we used two visual processing tasks that required similar preparation for S2. In one task, visual memory rehearsal operations were required. During the S1-S2 interval, there were clear differences between the amplitudes, topographies, and the effect of information load on the slow waves in two tasks, thus ruling out preparation only as an explanation.

The present report summarizes the results of a simulation study on the accuracy of Scherg's implementation of spatio-temporal analysis (BESA) for estimating the parameters (wave shape, location, orientation) of the intracranial sources of event-related brain potentials (ERPs) recorded from the scalp. In view of the subjective factors that might influence a solution, 10 subjects, ranging from those with much experience with ERPs and extensive background in the use of BESA to those with little experience with BESA and/or no knowledge of ERPs, independently analyzed a set of simulated somatosensory ERP data. The simulation contained wave forms from 32 electrode sites generated by a combination of 10 dipole sources. The primary question was how faithfully the different subjects would depict the source wave shapes, locations and orientations. Based on the 9 subjects who were familiar with ERPs, the grand-average cross-correlation coefficient between subjects' estimated and actual source wave shapes was 0.89 (standard deviation (S.D.) = 0.17). The grand-average location error, based upon a head diameter of 17 cm, was 1.4 cm (S.D. = 1.0 cm). The grand-average orientation error was 24.4 degrees (S.D. = 20 degrees).

Memory deficits are frequently identified in patients with multiple sclerosis. Both working (short-term) and long-term memory appear affected, particularly on free-recall tasks. The focus of our study was to examine neurophysiological correlates of working memory processes and to identify the specific components responsible for the working memory deficits reported in multiple sclerosis. Event-related brain potentials (ERPs) were recorded from the scalp of mildly afflicted multiple sclerosis patients and their matched normal controls during the performance of phonological and visuo-spatial working memory tasks. Neuropsychological test, behavioural performance and ERP data all indicated that verbal working memory is especially susceptible to impairment by multiple sclerosis, while visuospatial working memory is less susceptible. The pattern of results further indicated that the verbal working memory dysfunction in multiple sclerosis is at least partially due to impairment in the phonological loop, a rehearsal mechanism for retaining verbal information in working memory.

Working memory has been conceptualized as consisting of a number of components, such as an articulatory loop for rehearsing verbal material, a visuo-spatial sketch pad for maintaining visual images and a central executive that controls which information is made available for conscious processing. We recorded event-related brain potentials (ERPs) from normal human subjects while they maintained either visuo-spatial or phonological material in short-term memory for a 5-s interval. The results indicated that specialized brain systems for short-term storage of phonological and visuo-spatial information could be identified on the basis of marked differences between the topographies and morphologies of the ERP components elicited during these two types of short-term memory. The differences emerged during early encoding stages and continued through later retention stages.

In separate studies of arithmetic and mental rotation, similar posterior negative slow waves have been found. This similarity was surprising given the difference in cognitive processing required by these tasks. Furthermore, delayed responses were employed in these studies, so that it was not possible to determine the extent to which the slow wave activity was too late to be associated with processing that was specific to performance of the tasks. This experiment was intended to clarify the task-specific and non-specific nature of the slow wave activity. Subjects performed either an arithmetic or mental rotation task at two levels of difficulty on a random, trial-to-trial basis and gave an immediate response. There were a number of late posterior negativities, each with a different timing and topography, which were sensitive to type of task and/or task difficulty. Some components were associated with early task processing that was synchronized to the stimulus while others, revealed by response-synchronized averaging, were associated with later stages of task processing. There also was post-task activity that was sensitive to the difficulty level of the prior operations. In both tasks, there was a pre-frontal positive wave that persisted over most of the pre-response epoch, evidently related to a process that was active throughout the task. There also was centro-frontal phasic negativity, with a large peak at 380 msec in mental rotation, and a smaller, longer latency peak in arithmetic, apparently related to the complexity of the stimulus. Thus we conclude that arithmetic and mental rotation each elicit task-specific slow wave activity.

The effects of stimulus quality on the amplitude, peak latency, onset latency, and duration of the P300 component of the event-related brain potential were studied in patients with either a left or a right anterior temporal lobectomy and in normal controls. Stimulus quality was reduced by adding "noise" letters to words which signalled either a left or a right hand button press. Consistent with an interpretation that stimulus quality affects the subject's degree of equivocation, P300 peak latency, reaction time, and errors were all inversely related to stimulus quality, whereas P300 amplitude was directly related to stimulus quality. There were no significant differences between normal controls and either patient group for any of the ERP parameters or reaction time. Right temporal lobectomy patients made, however, significantly more errors, particularly on the catch trials, which suggests that they did not process the stimuli as thoroughly and accurately as the subjects in the other two groups. The absence of significant group differences in either the lateral symmetry or overall P300 amplitude extends the evidence against the idea that anterior temporal lobe structures make any substantial contribution to the scalp P300 in a visual discrimination paradigm. Because of observed delays in the onset of P300 in the low-quality stimulus condition, procedures were developed to quantify both P300 onset latency and P300 duration. Reduced stimulus quality significantly increased P300 onset latency whereas P300 duration remained unaffected, indicating that stimulus categorization must occur prior to, and not during, P300.

This experiment was concerned with event-related brain potential (ERP) activity related to short-term storage and retention of information in working memory. Our approach was to record the ERPs elicited by a stimulus that had to be memorized while varying the number of items (1, 3 or 6) in the task stimulus. In order to distinguish between ERP effects associated with perceptual complexity and retention of information, there was a second condition in which subjects were required to search the task stimulus for a match with a previously presented item. Thus, in the search condition subjects only had to remember one item (match or mismatch). ERP activity was recorded for 2450 msec after task stimulus offset. Two long-duration components varied as a function of task and memory load: a posterior positive wave and a frontal negative wave. Posterior positive wave amplitude was directly related to information load in the memory task but was negligible in the search task. Following the posterior positive wave was a frontal negative wave which occurred at the highest load level in the memory task but was totally absent in the search task. A P3b was elicited in both tasks. P3b was sensitive to information acquisition processes, but it did not distinguish between memory retention and visual search processes. While P3b amplitude did not vary with task or load, its latency increased with load in both tasks.

This experiment investigated how the P3a, P3b, and Slow Wave components of the event-related brain potential (ERP) respond to manipulations of the nature, timing, and extent of information delivery. There were two experiments in which the total amount of task information was distributed between pairs of successive stimuli (S1 and S2) within each trail. The task was to predict the relation between S1 and S2. In Experiment 1, the S1 could resolve no, partial, or all uncertainty with respect to the prediction outcome (correct or incorrect). Each S1 delivered three types of information: 1) outcome information--which resolved the subjects' uncertainty about the correctness of their prediction; 2) procedural information--which resolved uncertainty about how much outcome information would be delivered by S1; and 3) memory information--the identity of S1, which had to be stored for subsequent comparison with S2. In Experiment 2, the activity of these components was contrasted in two conditions in which the S1 delivered either memory information alone or both memory and procedural information. P3a and Slow Wave were sensitive only to outcome information. P3b was sensitive to all three types of information, and its scalp topography varied as a function of the type of information. The topographic variations indicate that P3b is not a unitary phenomenon but rather is a composite of activity arising from multiple intracranial sources of bioelectric activity.

In summary, what we have tried to do in this paper is to present our view of the ways in which the ERP field has become more complex in recent years. It is not an overstatement to say that, in a certain sense, we know less now than we thought we knew five to ten years ago. New advances have brought with them new problems. But they also point, though not yet with complete clarity, to directions for new solutions. Possibly, as we make further progress toward the definition of various generators for different components, our problems may be simplified. But even this prediction cannot be made with certainty. It should be noted that we have tended in this chapter to emphasize the problems. There has also been a positive side to the recent developments. Because we are more aware of the overlap problem, we now work with methods that attempt to deal with it. We now know, for example, that P3b and SW can relate quite differently to behavioral variables. Decision time, which we had thought occurred at P3 latency, can now be assumed on the basis of recent findings to occur earlier, at the N2 component. The separation of the formerly unitary CNV into several components makes it more possible to develop unique functional roles for each of the components. A similar development has occurred with the separation of the poststimulus negativities into several components. Our constructs for various components are still on the fuzzy side, but the field has better tools at its disposal for making them more precise. Finally, the multiplicity of components, which appears overwhelming initially, provides us with more degrees of freedom in attempting to relate electrophysiological activity to behavior. A number of investigators have commented that the complexity of factors that enter into behavior could not be reflected in the relatively few ERP components we were dealing with. Now that more components are coming to be distinguished, the likelihood increases that we may be able to obtain ERP correlates of more of the dimensions involved in behavior.

A comparison was made of evoked and emitted P300 and Slow Wave activity elicited in guessing and signal detection tasks. The stimulus conditions were the same in both tasks. Only the instructions to the subjects were different. Latency corrected averaging was used to compensate for trial-to-trial fluctuation of P300 latency. Principal components-varimax analysis (PCVA) was utilized to deal with overlap among P300, Slow Wave and CNV return-to-baseline. Clear measures (from both conventional and latency corrected averages) of emitted P300 components for correct rejection trials were obtained under the following conditions: (a) stimulus intensity near threshold (detection accuracy approximately equal to 78%, d' approximately equal to 1.9); (b) no simultaneous cueing signal; (c) stimulus presence and absence were equiprobable; (d) estimated contributions of overlap from other event-related potential components were removed from the P300 measurements. A significant difference was found between the Slow Wave topographic profiles elicited in the guessing and detection tasks. For the guessing task Slow Wave was relatively large and positive at vertex whereas in the detection task it was near zero amplitude at vertex. No significant differences due to task were found for the P300s when a PCVA derived estimate of the contribution of Slow Wave overlap was removed from the P300 measures.

We addressed the question of whether the terminal CNV, or E-wave, can be obtained in the absence of a motor response. In our design, the stimuli served only to reduce uncertainty with respect to a prior prediction, so that no response to S2 was required. In one experiment, the location of uncertainty reducing information was manipulated: in S1 alone or in S2 alone. When S2 reduced uncertainty, the pre-S2 E-wave was larger than when S2 did not reduce uncertainty. Similarly, when S1 reduced uncertainty, a pre-S1 negativity (whose topography did not differ from that of the pre-S2 E-wave) was larger than when S1 did not reduce uncertainty. The pre-S1 results also indicated that a prior experimenter-generated warning signal is not necessary for a non-motoric negativity to be obtained. In another experiment, the S1-S2 interval was manipulated (1 sec versus 3 sec). For the pre-S2 E-wave, onset was later and duration was longer for the longer S1-S2 interval. Peak amplitude and time of termination after S2 did not differ for short and long duration E-waves. Apparently, the timing of the E-wave is related to when in time the process it reflects is 'needed.' The dependence of the amplitude and timing of the pre-stimulus negativity on the temporal location of information, in conjunction with its independence of motor response requirement, suggests that the pre-stimulus negativity reflects some operation in the domain of expectancy, anticipation or 'mental preparation' for the informational stimulus.

A prediction paradigm was used to explore the relationship of the amplitude of the scalp-recorded event-related potential to the sequence of preceding signals and to preceding and subsequent behavior. P3b was found to be the only component which related systematically to prior sequence of signals. The CNV, P300E and Slow Wave were not affected by signal sequence. The P3b findings were the same for the emitted and evoked P3b, thus ruling out a sensory interpretation of the effect of signal sequence on P3b amplitude. Furthermore, it was found that signal sequence interacts with the subject's predictions in determining P3b amplitude. For signal discontinuations, P3b was large in amplitude regardless of what had been predicted. However, for signal continuations, P3b was small when continuations had been predicted, but large when discontinuations had been predicted. Finally, we found that for both correctly and incorrectly predicted signal continuations, larger P3bs were more likely than smaller P3bs to be followed by a prediction that the signal for the next trial would be different.

Correlation coefficients were computed between average response waveforms recorded from different brain regions of trained cats, before and after a specific stimulus acquired cue value. Application of signal-analysis techniques to the correlation matrix shows marked increase in the similarity between waveshapes evoked by that stimulus in sensory-specific and nonsensory-specific regions.

Late components of evoked potentials recorded during the occurrence and nonoccurrence of generalization are different. During generalization the evoked response waveshapes resembled those elicited by the conditioned stimulus during correct performance. The differences are statistically significant.