Naomi Langerock

Time causes forgetting from working memory

Although forgetting in the short term is a ubiquitous phenomenon, its exact causes remain undecided. The aim of the present study was to test the temporal decay hypothesis according to which memory traces fade away with time when attention is diverted by concurrent activities. In two experiments involving complex span tasks, adults were asked to remember series of items (either letters or spatial locations) while verifying multiplications. The duration of processing was manipulated by presenting multiplications either in word (three    ×    four    =    twelve) or digit (3    ×    4    =    12) format, the former taking longer to solve, while the time available to restore memory traces after each operation was kept constant across conditions. In line with the temporal decay hypothesis, the longer solution times elicited by solving word multiplications resulted in poorer recall performance. The fact that longer processing times had a comparable effect on both verbal and visuospatial memory and that the difference between conditions remained stable from the first to the last trials makes it difficult to account for these findings by assuming that forgetting is exclusively due to representation-based interference or buildup of proactive interference.

Evidence for a central pool of general resources in working memory

The present study addresses the existence of a central pool of domain-general resources in working memory. For this purpose, we examined interference between processing and storage activities involving information pertaining to different domain (verbal vs. visuo-spatial) while explicitly minimising representation-based interference at the peripheral level of working memory. Experiment 1 required maintenance of auditorily presented letters for further oral recall while concurrently judging visually presented spatial configurations by pressing keys. Experiment 2 required maintenance of visually presented random locations for further manual recall while concurrently judging auditorily presented words by giving oral responses. In both experiments, the cognitive load of processing was manipulated, a manipulation that clearly affected recall performance. This suggests strongly that working memory comprises a pool of domain-general attentional resources at the central level.

The Maintenance of Cross-Domain Associations in the Episodic Buffer

The episodic buffer has been described as a structure of working memory capable of maintaining multimodal information in an integrated format. Although the role of the episodic buffer in binding features into objects has received considerable attention, several of its characteristics have remained rather underexplored. This is the case for its maintenance capacity limits and its separability from domain-specific maintenance buffers. The present study addressed these questions, making use of a complex span paradigm in which participants were asked to maintain cross-domain (i.e., verbal-spatial) associations. The 1st experiment showed that the capacity limit for these cross-domain associations proved to be lower than the capacity limit for single features, and did not exceed 3. Cross-domain associations and single features depended, however, to the same extent on attentional resources for their maintenance. The 2nd experiment showed that domain-specific (verbal or spatial) resources were not involved in the maintenance of cross-domain information, revealing a clear distinction between the episodic buffer and the domain-specific buffers. Overall, in line with the episodic buffer hypothesis, these findings support the existence of a central system of limited capacity for the maintenance of cross-domain information.

Maintaining Information in Visual Working Memory: Memory for Bindings and Memory for Features Are Equally Disrupted by Increased Attentional Demands

This study examined the role of attention in maintaining information between visual features in visual working memory. In a change detection paradigm, two different memory conditions were created: one that required the maintenance of features and one that required the maintenance of how the features were bound together. During the short retention interval that separated the study display and test display, a tone discrimination task was to be performed. The attentional demand of the tone discrimination task was manipulated to test whether memory for binding was more disrupted than memory for features when the proportion of time during which attention is unavailable for maintenance is increased. We observed that memory for features and memory for bindings were equally disrupted by increasing the attentional demands of the tone discrimination task. This suggests that attention does not play a special role in the maintenance of feature bindings in visual working memory.

Working memory still needs verbal rehearsal.

The causal role of verbal rehearsal in working memory has recently been called into question. For example, the SOB-CS (Serial Order in a Box-Complex Span) model assumes that there is no maintenance process for the strengthening of items in working memory, but instead a process of removal of distractors that are involuntarily encoded and create interference with memory items. In the present study, we tested the idea that verbal working memory performance can be accounted for without assuming a causal role of the verbal rehearsal process. We demonstrate in two experiments using a complex span task and a Brown-Peterson paradigm that increasing the number of repetitions of the same distractor (the syllable ba that was read aloud at each of its occurrences on screen) has a detrimental effect on the concurrent maintenance of consonants whereas the maintenance of spatial locations remains unaffected. A detailed analysis of the tasks demonstrates that accounting for this effect within the SOB-CS model requires a series of unwarranted assumptions leading to undesirable further predictions contradicted by available experimental evidence. We argue that the hypothesis of a maintenance mechanism based on verbal rehearsal that is impeded by concurrent articulation still provides the simplest and most compelling account of our results.

On the sources of forgetting in working memory: The test of competing hypotheses:

Whether forgetting from working memory (WM) is only due to interference or is also caused by temporal decay is still a matter of debate. In the present study, this question was examined using complex span tasks in which each memory item was followed by a series of processing episodes, the duration and number of which were varied. It is known that recall performance in these tasks depends on the cognitive load (CL) of concurrent processing conceived as the ratio between processing time and free time, higher CL resulting in lower spans. The decay-and-refresh hypothesis accounts for this effect by assuming that memory traces decay during processing but are refreshed during free time. This hypothesis predicts lower recall performance with longer processing episodes, but no effect of their number as long as CL remains constant. The interference-only hypothesis supposes that free time is used to alleviate the interference created by processing distractors. This hypothesis is potentially compatible with an effect of the duration of processing episodes through increased interference, but predicts a detrimental effect of their number. In three experiments, the recall pattern fitted the predictions of the decay-and-refresh hypothesis for verbal WM, but that of the interference-only hypothesis for visuospatial WM. Although the entire pattern of data is more easily accommodated by the decay-and-refresh hypothesis than by its interference-only contender, our results suggest that it is unwise to aim at identifying a unique source to a complex phenomenon like WM forgetting.

Evidence for spontaneous serial refreshing in verbal working memory

Working memory (WM) keeps information temporarily accessible for ongoing cognition. One proposed mechanism to keep information active in WM is refreshing. This mechanism is assumed to operate by bringing memory items into the focus of attention, thereby serially refreshing the content of WM. We report two experiments in which we examine evidence for the spontaneous occurrence of serial refreshing in verbal WM. Participants had to remember series of red letters, while black probe letters were presented between these memory items, with each probe to be judged present in or absent from the list presented so far, as quickly as possible (i.e., the probe–span task). Response times to the probes were used to infer the status of the representations in WM and, in particular, to examine whether the content of the focus of attention changed over time, as would be expected if serial refreshing occurs spontaneously during inter-item pauses. In sharp contrast with this hypothesis, our results indicate that the last-presented memory item remained in the focus of attention during the inter-item pauses of the probe–span task. We discuss how these findings help to define the boundary conditions of spontaneous refreshing of verbal material in WM, and discuss implications for verbal WM maintenance and forgetting.

Is memory better for objects than for separate single features? The temporal hypothesis.

Working memory, the system allowing for a simultaneous maintenance and processing of information, is typically conceived as a capacity limited system. A proposed method to transcend its standard maintenance capacity is to maintain multifeature objects, instead of isolated features. Several studies have shown that multifeature memory items are stored as objects instead of separate single features in working memory, this object-based maintenance being thought to result in an increase in the number of features that can be maintained. We present a series of 4 experiments that challenge the belief that object-based maintenance per se is at the origin of the better memory for features in case of multifeature item presentation. Instead, we propose an account based on the temporal parameters of the memory item’s presentation, which explains memory performance in terms of the time available for encoding/consolidation per feature.

What do people typically do between list items? The nature of attention-based mnemonic activities depends on task context.

It is commonly assumed that attention-demanding postencoding processes take place during the free time immediately following encoding of each memory item in a list. These processes are thought to prevent loss of information from working memory (WM). We tested whether interitem pauses during presentation of a list are used to focus attention (a) on the last-presented memory item or (b) on all items currently in WM, and (c) whether this changes over time. Here, we presented black probe letters between to-be-remembered red letters. Participants judged whether each probe letter corresponded to the last-presented memory item (last-item match group) or to any of the memory items presented up to that point in the list (any-item match group). To examine mnemonic processing as a function of time, the delay between the to-be-remembered letter and the following probe was manipulated in three experiments. When preprobe delays and interitem intervals were relatively short (Experiment 1), recall performance was observed to be better in the last-item match group and this did not change as a function of the duration of the delay before the probe. When preprobe delays and interitem intervals were longer however (Experiment 2), this disruptive effect of Any-item match instructions was no longer observed. This pattern was found again in Experiment 3 and suggests that the nature of the attention-demanding postencoding processes taking place in between memory items depends on task context in a systematic manner. The results are discussed in terms of previously proposed attention-demanding processes; specifically, consolidation and refreshing.

An examination of refreshing in between‐category sequences

The present study focused on refreshing within a working memory (WM) context. Refreshing refers to the mechanism that brings back information into the focus of attention in order to counteract forgetting of memory traces. Despite some research on this topic, the exact nature of refreshing remains unclear. The present study investigated refreshing by means of the cognitive load (CL) effect. This effect is typically observed in complex span tasks, which combine processing and storage demands. It refers to the observation that WM performance depends on the CL of concurrent processing, defined as the proportion of time between list items that is occupied by concurrent processing and therefore not available to refresh memory items. Traditionally, the CL effect has been demonstrated using within‐category memory sequences in which all memory items are drawn from one category (e.g., all words). Here, we show that the CL effect also applies to between‐category memory sequences in which memory items are drawn from different categories (e.g., words, orientations, faces, etc.). The ensemble of the results adds to the domain‐generality of the CL effect. Implications concerning the specific nature of refreshing and future research directions are discussed.