Pupil size changes during recognition memory

Pupils dilate to a greater extent when participants view old compared to new items during recognition memory tests. We report three experiments investigating the cognitive processes associated with this pupil old/new effect. Using a remember/know procedure, we found that the effect occurred for old items that were both remembered and known at recognition, although it was attenuated for known compared to remembered items. In Experiment 2, the pupil old/new effect was observed when items were presented acoustically, suggesting the effect does not depend on low-level visual processes. The pupil old/new effect was also greater for items encoded under deep compared to shallow orienting instructions, suggesting it may reflect the strength of the underlying memory trace. Finally, the pupil old/new effect was also found when participants falsely recognized items as being old. We propose that pupils respond to a strength-of-memory signal and suggest that pupillometry provides a useful technique for exploring the underlying mechanisms of recognition memory.     

Immediate and delayed stimulus repetitions evoke different ERPs in a serial-probe recognition task

In this study we examined whether event-related potentials (ERPs) associated with stimulus repetition and recognition in a serial-probe recognition task were comparable to ERPs in other tasks that are more typically used to investigate old/new ERP effects. The experiment consisted of 320 trials in which a recognition probe followed a four-item memory set; 160 trials consisted of images depicting common objects that were easy to label (EL task), and 160 trials consisted of images depicting abstract patterns that were difficult to label (DL task). Nineteen participants indicated whether a probe that followed each memory set was or was not presented in the memory set. Half of the probes matched, and half did not match, an item in the preceding memory set. ERPs appeared to reflect two processes--one that differentiated between recently presented stimuli and other stimuli and another that distinguished between repeated stimuli and new stimuli. ERPs to recent probes were more positive than ERPs to other probes in the EL and DL tasks. ERPs to match (old) probes were more positive than ERPs to nonmatch (new) probes only in the EL task.     

Interfacing mind and brain: a neurocognitive model of recognition memory

A variety of processes contribute to successful recognition memory, some of which can be associated with spatiotemporally distinct event-related potential old/new effects. An early frontal and a subsequent parietal old/new effect are correlated with the familiarity and recollection subcomponents of recognition memory, respectively, whereas a late, postretrieval old/new effect seems to reflect an ensemble of evaluation processes that are set by the task context in which retrieval occurs. Both the early frontal and the parietal old/new effects are differentially modulated by the informational content (e.g., object forms and spatial locations) of recognition and seem to rely on brain systems damaged in amnesia. The late frontal effect appears to reflect prefrontal cortex activation. A neurophysiologically based model of recognition memory retrieval is presented and it is shown that coupling recognition memory subprocesses with distinct old/new effects allow examination of the time course of the processes that contribute to correct and to illusory memories. In conjunction with event-related functional magnetic resonance imaging activation patterns the brain systems recruited by various aspects of episodic memory retrieval can be identified.     

EEG power changes reflect distinct mechanisms during long‐term memory retrieval

The roles of theta and alpha oscillations for long‐term memory (LTM) retrieval are still under debate. Both are modulated by LTM retrieval demands, but it is unclear what specific LTM functions they are related to. Here, different oscillatory correlates of LTM retrieval could be obtained for theta and alpha with a paradigm that is suited to monitor the activation of a varying number of material‐specific LTM representations. Both frequency bands responded parametrically to the number of retrieved items. However, only the alpha effect dissociated topographically for material type, indicating that the activation of material‐specific representations became systematically modulated. For theta, this effect was material‐unspecific with mid‐frontal topography. These results suggest that alpha is functionally related to the activation of stored information, whereas theta is a sign of retrieval‐related control processes.     

An electrophysiological measure of visual short‐term memory capacity within and across hemifields

Recent electrophysiological research has identified a specific contralateral correlate of the contents of visual short‐term memory (VSTM). This posterior contralateral delay activity (CDA) persists during the retention period, and its amplitude is modulated by the number of items held at any one time. Here we show that, despite the contralateral nature of this activity, its amplitude is modulated by the number of objects from both hemifields. When objects from one side of the visual field are held in memory, CDA activity increased for arrays of one, two, and three objects, but stopped getting larger for arrays of four objects. However, when objects from both sides are memorized at the same time, CDA activity reached its asymptotic limit for arrays of two objects per side. These results suggest that, in spite of being contralaterally organized, VSTM is limited by the number of objects from both hemifields.     

Object‐substitution masking modulates spatial attention deployment and the encoding of information in visual short‐term memory: Insights from occipito‐parietal ERP components

If object‐substitution masking (OSM) arises from mask representations replacing target representations, OSM should impede the formation of representations in visual short‐term memory (VSTM). We utilized event‐related potentials to examine the effect of OSM on target processing. An N2pc was observed on trials with delayed‐offset masks, indicating that focused attention was directed to the target. The sustained posterior contralateral negativity (SPCN), an index of VSTM storage, was observed in delayed‐offset trials only on trials with correct responses. This supports the hypothesis that inaccurate performance on delayed‐offset trials arises from a failure to encode the target in VSTM. On co‐termination trials, accuracy was high and neither the N2pc nor SPCN was observed. This indicates that, in the absence of masking, the task was accomplished by maintaining a diffuse attentional state that enabled the joint encoding of the potential target items.     

Two stages of directed forgetting: Electrophysiological evidence from a short‐term memory task

In this study, a short‐term memory test was used to investigate the temporal course and neural mechanism of directed forgetting under different memory loads. Within each trial, two memory items with high or low load were presented sequentially, followed by a cue indicating whether the presented items should be remembered. After an interval, subjects were asked to respond to the probe stimuli. The ERPs locked to the cues showed that (a) the effect of cue type was initially observed during the P2 (160–240 ms) time window, with more positive ERPs for remembering relative to forgetting cues; (b) load effects were observed during the N2‐P3 (250–500 ms) time window, with more positive ERPs for the high‐load than low‐load condition; (c) the cue effect was also observed during the N2‐P3 time window, with more negative ERPs for forgetting versus remembering cues. These results demonstrated that directed forgetting involves two stages: task‐relevance identification and information discarding. The cue effects during the N2 epoch supported the view that directed forgetting is an active process.     

The role of inhibition for working memory processes: ERP evidence from a short‐term storage task

Human working memory is the central unit for short‐term storage of information. In addition to the selection and adequate storage of relevant information, the suppression of irrelevant stimuli from the environment seems to be of importance for working memory processes. To learn more about the interplay of information uptake and inhibition of irrelevant information, the present study used ERP measures and a short‐term storage and retrieval task, in which pairs of either numbers or letters had to be compared. Random sequences of four stimuli (two numbers and two letters) were presented, with either the numbers or the letters being relevant for comparison. The analysis of ERPs to each of the four stimuli indicated more pronounced P2 and P3b amplitudes for relevant than irrelevant stimuli. In contrast, the N2 (reflecting inhibitory control) was only elicited by irrelevant stimuli. Moreover, the N2 amplitude of the second irrelevant stimulus was associated with behavioral performance, indicating the importance of inhibition of task‐irrelevant stimuli for working memory processes. In sum, the findings demonstrate the role of cognitive control mechanisms for protecting relevant contents in working memory against irrelevant information.     

Spatial layout of letters in nonwords affects visual short‐term memory load: Evidence from human electrophysiology

The sustained posterior contralateral negativity (SPCN) was used to investigate the effect of spatial layout on the maintenance of letters in VSTM. SPCN amplitude was measured for words, nonwords, and scrambled nonwords. We reexamined the effects of spatial layout of letters on SPCN amplitude in a design that equated the mean frequency of use of each position. Scrambled letters that did not form words elicited a larger SPCN than either words or nonwords, indicating lower VSTM load for nonwords presented in a typical horizontal array than the load observed for the same letters presented in spatially scrambled locations. In contrast, prior research has shown that the spatial extent of arrays of simple stimuli did not influence the amplitude of the SPCN. Thus, the present results indicate the existence of encoding and VSTM maintenance mechanisms specific to letter and word processing.     

Illusory conjunctions in visual short‐term memory: Individual differences in corpus callosum connectivity and splitting attention between the two hemifields

Overloading the capacity of visual attention can result in mistakenly combining the various features of an object, that is, illusory conjunctions. We hypothesize that if the two hemispheres separately process visual information by splitting attention, connectivity of corpus callosum—a brain structure integrating the two hemispheres—would predict the degree of illusory conjunctions. In the current study, we assessed two types of illusory conjunctions using a memory‐scanning paradigm; the features were either presented across the two opposite hemifields or within the same hemifield. Four objects, each with two visual features, were briefly presented together followed by a probe‐recognition and a confidence rating for the recognition accuracy. MRI scans were also obtained. Results indicated that successful recollection during probe recognition was better for across hemifields conjunctions compared to within hemifield conjunctions, lending support to the bilateral advantage of the two hemispheres in visual short‐term memory. Age‐related differences regarding the underlying mechanisms of the bilateral advantage indicated greater reliance on recollection‐based processing in young and on familiarity‐based processing in old. Moreover, the integrity of the posterior corpus callosum was more predictive of opposite hemifield illusory conjunctions compared to within hemifield illusory conjunctions, even after controlling for age. That is, individuals with lesser posterior corpus callosum connectivity had better recognition for objects when their features were recombined from the opposite hemifields than from the same hemifield. This study is the first to investigate the role of the corpus callosum in splitting attention between versus within hemifields.     

The coupling of emotion and cognition in the eye: introducing the pupil old/new effect

The study presented here investigated the effects of emotional valence on the memory for words by assessing both memory performance and pupillary responses during a recognition memory task. Participants had to make speeded judgments on whether a word presented in the test phase of the experiment had already been presented ("old") or not ("new"). An emotion-induced recognition bias was observed: Words with emotional content not only produced a higher amount of hits, but also elicited more false alarms than neutral words. Further, we found a distinct pupil old/new effect characterized as an elevated pupillary response to hits as opposed to correct rejections. Interestingly, this pupil old/new effect was clearly diminished for emotional words. We therefore argue that the pupil old/new effect is not only able to mirror memory retrieval processes, but also reflects modulation by an emotion-induced recognition bias.     

Explicit and implicit memory in female college students with schizotypal traits: an event-related potential study

The explicit and implicit memory of nonclinical individuals with schizotypal traits was investigated using event-related potentials. Explicit and implicit memory was measured with continuous recognition and categorization tasks, respectively. On the recognition task, the control group demonstrated a greater old/new effect in response to the old than to the new words during the 250-750 ms post-stimulus period, whereas schizotypal trait group did not exhibit an old/new effect during the 550-650 ms period. The control group demonstrated faster response times to the old than to the new words, whereas the schizotypal group demonstrated longer response times to the old than to the new words. On the categorization task, both groups showed old/new effects during the 250-550 ms after stimulus onset and responded more rapidly and with fewer errors to the old than to the new words. These results suggest that individuals with schizotypal traits have impaired explicit but preserved implicit memory.     

The effects of stimulus novelty and familiarity on neuronal activity in the amygdala of monkeys performing recognition memory tasks

The function of the amygdala in behavioural responses to novel stimuli and its possible function in recognition memory were investigated by recording the responses of 659 amygdaloid neurons in monkeys performing recognition memory and visual discrimination tasks. The aim was to determine the contribution of the amygdala in the encoding of familiarity and therefore its role in supporting memory-related neuronal mechanisms in the basal forebrain. The responses of three groups of neurons reflected different forms of memory. One group (n= 10) responded maximally to novel stimuli and significantly less so to the same stimuli when they were familiar. The calculated memory spans of these neurons were in the range of 2–10 intervening trials, and this shortterm retention of information may reflect the operation of a neural mechanism encoding memory for the recency of stimulus presentation. Two other groups responded to the sight of particular categories of familiar stimuli: to foods (n = 6) or to faces (n = 10). The responses of some of these stimulus-selective neurons declined with repeated presentations of foods (3/4 tests) and faces (2/6 tests). The activity of these latter two groups of neurons may be involved in behavioural responses to familiar visual stimuli, particularly when such stimuli have affective or motivational significance. We conclude that the neurophysiological data provide evidence of amygdaloid mechanisms for the recognition of recently seen visual stimuli. However, these amygdaloid mechanisms do not appear to be sufficient to support the performance of long-term recognition memory tasks without additional and complementary functions carried out by other ventromedial temporal, prefrontal and diencephalic structures which also project to the basal forebrain.     

Reference-independent ERP old/new effects of auditory and visual word recognition memory: Joint extraction of stimulus- and response-locked neuronal generator patterns

To clarify polarity, topography, and time course of recognition memory ERP old/new effects during matched visual and auditory continuous word recognition tasks, unrestricted temporal PCA jointly analyzed stimulus- and response-locked, reference-free current source densities (31-channel, N=40). Randomization tests provided unbiased statistics for complete factor topographies. Old/new left parietal source effects were complemented by lateral frontocentral sink effects in both modalities, overlapping modality-specific P3 sources 160 ms preresponse. A mid-frontal sink 45 ms postresponse terminated the frontoparietal generator pattern, showed old/new effects consistent with bilateral activation of anterior cingulate and SMA, and preceded similar activity extending posteriorly along the longitudinal fissure. These methods separated old/new stimulus source (preresponse) and response sink (postresponse) effects from motor and modality-specific ERPs.     

Subjectively‐defined optimal/non‐optimal time of day modulates controlled but not automatic retrieval processes in verbal memory

Performance for controlled, resource‐demanding retrieval in episodic memory has been consistently found to be better at an optimal compared with non‐optimal time of the day, evidencing a synchrony effect. However, performance in memory tasks in which retrieval is mostly based on automatic processes was inconclusively found either to be better at a non‐optimal time of day or independent of synchrony effects. A caveat in most prior studies is that optimal/non‐optimal time of day is based on morningness–eveningness composite scores derived from chronotype questionnaires, which might not efficiently predict subjectively‐defined cognitive efficiency periods. An additional caveat is that separate tasks are used to assess explicit and implicit retrieval in verbal memory. Indeed, no task is process‐pure, and both controlled and automatic retrieval processes may potentially contribute to retrieval scores in different types of memory tasks. In the present study, we investigated the impact of individually defined subjective optimal/non‐optimal time of day on verbal memory retrieval, using an adaptation of the Process‐Dissociation Procedure that allows estimating the respective contributions of automatic and controlled memory retrieval processes within the same memory task. Our results disclose a higher involvement of controlled processes at subjectively optimal compared with non‐optimal time of day, but no differences for automatic processes. Synchrony effects and subjectively‐defined peaks and troughs of performance for controlled components of memory retrieval should be considered in the evaluation of episodic memory.     

Stimulus- and response-locked neuronal generator patterns of auditory and visual word recognition memory in schizophrenia

Examining visual word recognition memory (WRM) with nose-referenced EEGs, we reported a preserved ERP ‘old-new effect’ (enhanced parietal positivity 300–800 ms to correctly-recognized repeated items) in schizophrenia ([Kayser, J., Bruder, G.E., Friedman, D., Tenke, C.E., Amador, X.F., Clark, S.C., Malaspina, D., Gorman, J.M., 1999. Brain event-related potentials (ERPs) in schizophrenia during a word recognition memory task. Int. J. Psychophysiol. 34(3), 249–265.]). However, patients showed reduced early negative potentials (N1, N2) and poorer WRM. Because group differences in neuronal generator patterns (i.e., sink-source orientation) may be masked by choice of EEG recording reference, the current study combined surface Laplacians and principal components analysis (PCA) to clarify ERP component topography and polarity and to disentangle stimulus- and response-related contributions. To investigate the impact of stimulus modality, 31-channel ERPs were recorded from 20 schizophrenic patients (15 male) and 20 age-, gender-, and handedness-matched healthy adults during parallel visual and auditory continuous WRM tasks. Stimulus- and response-locked reference-free current source densities (spherical splines) were submitted to unrestricted Varimax-PCA to identify and measure neuronal generator patterns underlying ERPs. Poorer (78.2 ± 18.7% vs. 87.8 ± 11.3% correct) and slower (958 ± 226 vs. 773 ± 206 ms) performance in patients was accompanied by reduced stimulus-related left-parietal P3 sources (150 ms pre-response) and vertex N2 sinks (both overall and old/new effects) but modality-specific N1 sinks were not significantly reduced. A distinct mid-frontal sink 50-ms post-response was markedly attenuated in patients. Reductions were more robust for auditory stimuli. However, patients showed increased lateral-frontotemporal sinks (T7 maximum) concurrent with auditory P3 sources. Electrophysiologic correlates of WRM deficits in schizophrenia suggest functional impairments of posterior cortex (stimulus representation) and anterior cingulate (stimulus categorization, response monitoring), primarily affecting memory for spoken words.