Is "memory-for-when" universal? Group and individual variability in temporal position memory for words, faces, and classrooms

Many contemporary theories of memory assume that everyone automatically stores temporal contextual information about all types of encountered information, yet most studies on this topic have used words and ignored individual differences. Five experiments accumulated evidence that explicit storage of temporal context information does not appear to occur automatically for all people and types of memoranda. We collected judgments of temporal position (memory-for-when) for words (Experiments 1 & 3), faces (Experiments 2A, 3, 4, and 5), and classrooms (Experiments 2B & 3). At the group level, for each of these memoranda memory-for-when was sensitive to the original input position and showed a temporal primacy effect reflecting better memory for position for items near the beginning of the list, indicating some automatic storage of temporal context information. However, memory-for-when was significantly better for words than classrooms, with faces in the middle. Moreover, individuals varied dramatically in their ability to indicate memory-for-when, especially for classrooms where many people performed at or near chance. Taken together, the data suggest that explicit memory-for-when may be dissociable from the more implicit use of temporal contextual information that is theorised to occur during free recall.

On the sensitivity of event-related fields to recollection and familiarity

The sensitivity of event-related potentials (ERPs) to the processes of recollection and familiarity has been explored extensively, and ERPs have been used subsequently to infer the contributions these processes make to memory judgments under a range of different circumstances. It has also been shown that event-related fields (ERFs, the magnetic counterparts of ERPs) are sensitive to memory retrieval processes. The links between ERFs, recollection and familiarity are, however, established only weakly. In this experiment, the sensitivity of ERFs to these processes was investigated in a paradigm used previously with ERPs. An early frontally distributed modulation varied with memory confidence in a way that aligns it with the process of familiarity, while a later parietally distributed modulation tracked subjective claims of recollection in a way that aligns it with this process. These data points strengthen the argument for employing ERFs to assess the contributions these processes can make to memory judgments, as well as for investigating the nature of the processes themselves.

ERP evidence for hemispheric asymmetries in exemplar-specific explicit memory access

The right cerebral hemisphere (RH) appears to be more effective in representing visual objects as distinct exemplars than the left hemisphere (LH) which is presumably biased towards coding objects at the level of abstract prototypes. As of yet, relatively little is known about the role that asymmetries in exemplar-specificity play at the level of explicit memory retrieval. In the present study, we addressed this issue by examining hemispheric asymmetries in the putative event-related potential (ERP) correlates of familiarity (FN400) and recollection (LPC). In an incidental study phase, pictures of familiar objects were presented centrally. At test, participants performed a memory inclusion task on identical repetitions and different exemplars of study items as well as new items which were presented in only one visual hemifield using the divided visual field technique. With respect to familiarity, we observed exemplar-specific FN400 old/new effects that were more pronounced for identical repetitions than different exemplars, irrespective of the hemisphere governing initial stimulus processing. In contrast, LPC old/new effects were subject to some hemispheric asymmetries indicating that exemplar-specific recollection was more extensive in the RH than in the LH. This further corroborates the idea that hemispheric asymmetries should not be generalized but need to be distinguished not only in different domains but also at different levels of processing.

Memory in time: electrophysiological comparison between reality filtering and temporal order judgment

Orbitofrontal reality filtering (ORF) denotes a little known but vital memory control mechanism, expressed at 200-300ms after stimulus presentation, that allows one to sense whether evoked memories (thoughts) refer to present reality and can be acted upon, or not. Its failure induces reality confusion evident in confabulations that patients act upon and disorientation. In what way ORF differs from temporal order judgment (TOJ), that is, the conscious knowledge about when something happened in the past, has never been explored. Here we used evoked potential analysis to compare ORF and TOJ within a combined experimental task and within a comparable time frame, close to the experienced present. Seventeen healthy human subjects performed an experiment using continuous recognition tasks that combined the challenges of ORF and TOJ. We found that the two mechanisms dissociated behaviorally: subjects were markedly slower and less accurate in TOJ than ORF. Both mechanisms evoked similar potentials at 240-280ms, when ORF normally occurs. However, they rapidly dissociated in terms of amplitude differences and electrical source from 310 to 360ms and again from 530 to 560ms. We conclude that the task of consciously ordering memories in the immediate past (TOJ) is effortful and slow in contrast to sensing memories' relation with the present (ORF). Both functions invoke similar early electrocortical processes which then rapidly dissociate and engage different brain areas. The results are consistent with the different consequences of the two mechanisms' dysfunction: while failure of ORF has a known clinical manifestation (reality confusion as evident in confabulation and disorientation), the failure of TOJ, as tested here, has no such known clinical correlate.

Decreased fronto-temporal interaction during fixation after memory retrieval

Previous studies have revealed top-down control during memory retrieval from the prefrontal cortex to the temporal cortex. In the present functional MRI study, we investigated whether the fronto-temporal functional interaction occurs even during fixation periods after memory retrieval trials. During recency judgments, subjects judged the temporal order of two items in a study list. The task used in the present study consisted of memory trials of recency judgments and non-memory trials of counting dots, and post-trial fixation periods. By comparing the brain activity during the fixation periods after the memory trials with that during the fixation periods after the non-memory trials, we detected heightened brain activity in the lateral prefrontal cortex, the lateral temporal cortex and the hippocampus. Functional interactions during the fixation periods after the memory vs. non-memory trials as examined using a psychophysiological interaction revealed a decreased interaction from the lateral prefrontal cortex to the lateral temporal cortex, but not to the hippocampus. The functional interaction between the same frontal and temporal regions was also present during the memory trials. A trial-based functional connectivity analysis further revealed that the fronto-temporal interaction was positive and decreased during the fixation periods after the memory trials, relative to the fixation periods after the non-memory trials. These results suggest that the fronto-temporal interaction existed during the post-trial fixation periods, which had been present during the memory trials and temporally extended into the fixation periods.

Temporal texture of associative encoding modulates recall processes

Binding aspects of an experience that are distributed over time is an important element of episodic memory. In the current study, we examined how the temporal complexity of an experience may govern the processes required for its retrieval. We recorded event-related potentials during episodic cued recall following pair associate learning of concurrently and sequentially presented object-picture pairs. Cued recall success effects over anterior and posterior areas were apparent in several time windows. In anterior locations, these recall success effects were similar for concurrently and sequentially encoded pairs. However, in posterior sites clustered over parietal scalp the effect was larger for the retrieval of sequentially encoded pairs. We suggest that anterior aspects of the mid-latency recall success effects may reflect working-with-memory operations or direct access recall processes, while more posterior aspects reflect recollective processes which are required for retrieval of episodes of greater temporal complexity.

Dissociable temporo-parietal memory networks revealed by functional connectivity during episodic retrieval

Episodic memory retrieval most often recruits multiple separate processes that are thought to involve different temporal regions. Previous studies suggest dissociable regions in the left lateral parietal cortex that are associated with the retrieval processes. Moreover, studies using resting-state functional connectivity (RSFC) have provided evidence for the temporo-parietal memory networks that may support the retrieval processes. In this functional MRI study, we tested functional significance of the memory networks by examining functional connectivity of brain activity during episodic retrieval in the temporal and parietal regions of the memory networks. Recency judgments, judgments of the temporal order of past events, can be achieved by at least two retrieval processes, relational and item-based. Neuroimaging results revealed several temporal and parietal activations associated with relational/item-based recency judgments. Significant RSFC was observed between one parahippocampal region and one left lateral parietal region associated with relational recency judgments, and between four lateral temporal regions and another left lateral parietal region associated with item-based recency judgments. Functional connectivity during task was found to be significant between the parahippocampal region and the parietal region in the RSFC network associated with relational recency judgments. However, out of the four tempo-parietal RSFC networks associated with item-based recency judgments, only one of them (between the left posterior lateral temporal region and the left lateral parietal region) showed significant functional connectivity during task. These results highlight the contrasting roles of the parahippocampal and the lateral temporal regions in recency judgments, and suggest that only a part of the tempo-parietal RSFC networks are recruited to support particular retrieval processes.

Functional neuroanatomy supporting judgments of when events occurred

The neural substrates of memory for when events occurred are not well described. One reason for this is that the paradigms used to date have permitted isolation of only some of the relevant memory processes. In this experiment, functional magnetic resonance imaging was used to identify for the first time brain regions that support two distinct bases upon which "when" judgments can be made. Seventeen human participants (6 male) completed a continuous recognition memory task where the interval between presentation and re-presentation of words varied between 5 and 25 intervening words (the lag). The task on each trial was to distinguish repeated words from those presented for the first time, and to indicate the lag for repeated words. The inferior parietal lobe showed greater activation for shorter lag judgments, regardless of judgment accuracy. The lingual gyrus, by contrast, was more active for correct than for incorrect lag judgments, regardless of the interval between first and second item presentations. Both of these regions have been linked in previous work to the process of recollection, and the findings described here represent a novel neural dissociation across regions that deploy mnemonic information in fundamentally different ways to support judgments about when events occurred.

Remembering perceptual features unequally bound in object and episodic tokens: Neural mechanisms and their electrophysiological correlates

We present a neurocognitive model of long-term object memory. We propose that perceptual priming and episodic recognition are phenomena based on three distinct kinds of representations. We label these representations types and tokens. Types are prototypical representations needed for object identification. The network of non-arbitrary features necessary for object categorization is sharpened in the course of repeated identification, an effect that we call type trace and which causes perceptual priming. Tokens, on the other hand, support episodic recognition. Perirhinal structures are proposed to bind intrinsic within-object features into an object token that can be thought of as a consolidated perceptual object file. Hippocampal structures integrate object- with contextual information in an episodic token. The reinstatement of an object token is assumed to generate a feeling of familiarity, whereas recollection occurs when the reinstatement of an episodic token occurs. Retrieval mode and retrieval orientation dynamically modulate access to these representations. In this review, we apply the model to recent empirical research (behavioral, fMRI, and ERP data) including a series of studies from our own lab. We put specific emphasis on the effects that sensory features and their study-test match have on familiarity. The type-token approach fits the data and additionally provides a framework for the analysis of concepts like unitization and associative reinstatement.

ERP correlates of target-distracter differentiation in repeated runs of a continuous recognition task with emotional and neutral faces

The emotional salience of faces has previously been shown to induce memory distortions in recognition memory tasks. This event-related potential (ERP) study used repeated runs of a continuous recognition task with emotional and neutral faces to investigate emotion-induced memory distortions. In the second and third runs, participants made more false alarms to distracters (repeated from previous runs). Emotion did not modulate the amount of errors, but the extent to which recollection was employed to maximise performance as reflected in the putative ERP correlate of recollection; the parietal old-new effect. Targets from all stimulus classes (positive, negative, neutral) were associated with parietal ERP memory effects, but this was also the case for correctly rejected negative distracters. This suggests that recollection was strategically used to correctly reject negative distracters (recall-to-reject). This finding is consistent with the view that facilitated recollection of negative stimuli may be used to decrease the susceptibility to memory errors induced by emotional salience.

Differential temporo-parietal cortical networks that support relational and item-based recency judgments

There is a growing interest in the parietal cortical role for episodic memory retrieval. Previous functional magnetic resonance imaging (fMRI) studies of recency judgments, judgments of the relative temporal order of two studied items, have highlighted the involvement of the lateral prefrontal and medial temporal regions. However, the parietal cortical contribution to recency judgments has rarely been highlighted. To examine the parietal involvement, in this study, we conducted a re-analysis to increase the statistical power using three data sets (N=73) from our previous fMRI studies of recency judgments. Recency judgments can be achieved by at least two mechanisms, relational and item-based ones. It has been revealed that the left hippocampus/parahippocampal region is related to relational recency judgments, and that the right anterior temporal region is related to item-based recency judgments. We examined whether the parietal cortex is involved in these two types of recency judgments. Significant brain activity related to relational recency judgments was observed in the left ventral parietal region and, as reported previously, the left parahippocampal region. On the other hand, significant brain activity related to item-based recency judgments was observed in the left dorsal parietal region and, as reported previously, the right anterior temporal region. Furthermore, correlation analyses of resting-state BOLD signals detected significant correlations between the ventral parietal region and the parahippocampal region, as well as between the dorsal parietal region and anterior temporal region. These results suggest that the two temporo-parietal networks differentially contribute to relational and item-based recency judgments.

ERP Evidence for Flexible Adjustment of Retrieval Orientation and Its Influence on Familiarity

The assumption was tested that familiarity memory as indexed by a mid-frontal ERP old–new effect is modulated by retrieval orientation. A randomly cued category-based versus exemplar-specific recognition memory test, requiring flexible adjustment of retrieval orientation, was conducted. Results show that the mid-frontal ERP old–new effect is sensitive to the manipulation of study-test congruency—that is, whether the same object is repeated identically or a different category exemplar is presented at test. Importantly, the effect pattern depends on subjects' retrieval orientation. With a specific orientation, only same items elicited an early old–new effect (same > different = new), whereas in the general condition, the old–new effect was graded (same > different > new). This supports the view that both perceptual and conceptual processes can contribute to familiarity memory and demonstrates that the rather automatic process of familiarity is not only data driven but influenced by top–down retrieval orientation, which subjects are able to adjust on a flexible basis.