Search

ABSTRACT This article focuses on familiarity, the form of memory allowing humans to recognize stimuli that have been encountered before. We aim to emphasize its complex nature which includes representational and phenomenological dimensions. The former implies that its neural correlates depend on the type and complexity of the cue stimulus, as different classes of stimuli are represented in distributed ventral visual and medial temporal regions. The second dimension relates to the subjective feeling of familiarity, which results from a fluency signal that is attributed to past encounters with the stimulus. We review mnemonic and non‐mnemonic sources of fluency that can induce a sense of familiarity, as well as cases where fluency is not attributed to memory, among which the phenomenological experience of déjà‐vu. Across these two dimensions, we highlight key questions to be answered by future studies to improve our understanding of the underpinnings of this form of memory and contribute to building an integrative neurocognitive model of familiarity. Essential to this aim is the clarification of the computational, cognitive, and neural mechanisms involved, namely global matching, fluency attribution, and sharpening. Furthermore, future research is needed to unravel the relationships between these mechanisms. We argue that to achieve these goals, researchers must use appropriate behavioral paradigms and clearly define which dimension of familiarity they investigate. [ABSTRACT FROM AUTHOR]

Familiarity is the strange feeling of knowing that something has already been seen in our past. Over the past decades, several attempts have been made to model familiarity using artificial neural networks. Recently, two learning algorithms successfully reproduced the functioning of the perirhinal cortex, a key structure involved during familiarity: Hebbian and anti‐Hebbian learning. However, performance of these learning rules is very different from one to another thus raising the question of their complementarity. In this work, we designed two distinct computational models that combined Deep Learning and a Hebbian learning rule to reproduce familiarity on natural images, the Hebbian model and the anti‐Hebbian model, respectively. We compared the performance of both models during different simulations to highlight the inner functioning of both learning rules. We showed that the anti‐Hebbian model fits human behavioral data whereas the Hebbian model fails to fit the data under large training set sizes. Besides, we observed that only our Hebbian model is highly sensitive to homogeneity between images. Taken together, we interpreted these results considering the distinction between absolute and relative familiarity. With our framework, we proposed a novel way to distinguish the contribution of these familiarity mechanisms to the overall feeling of familiarity. By viewing them as complementary, our two models allow us to make new testable predictions that could be of interest to shed light on the familiarity phenomenon. [ABSTRACT FROM AUTHOR]

Competitive trace theory holds that semanticization following reactivation is characterised by a fidelity loss in the memory representation due to the competition between different traces formed after each occurrence of a given stimulus. This is manifested in the Mnemonic Similarity Task as an increase in hits and in false recognition of similar lures. We tested this account across two encoding conditions emphasising the perceptual versus semantic features of the items, which were presented either once or three times. Our results supported the hypothesis that semanticization following repetition occurs regardless of the type of encoding induced. [ABSTRACT FROM AUTHOR]

Unitization, that is, the encoding of an association as one integrated entity, has been shown to improve associative memory in populations presenting with associative memory deficit due to hippocampal dysfunction, such as amnesic patients with focal hippocampal lesions and healthy older adults. One reason for this benefit is that encoding of unitized associations would rely on the perirhinal cortex (PrC) and thus minimize the need for hippocampal recruitment. Mild cognitive impairment (MCI) is accompanied by a deficit in associative memory. However, unitization has never been studied to explore the potential benefit in associative memory in MCI, maybe because MCI is characterized by PrC pathology. However, the PrC may potentially still function sufficiently to allow for the successful adoption of unitization. In this study, we aimed at assessing whether unitization could attenuate MCI patients' associative memory deficit, and whether the ability to remember unitized associations would be modulated by the integrity of the PrC in MCI patients. Unitization was manipulated at a conceptual level, by encouraging participants to encode unrelated word pairs as new compound words. Participants also underwent a structural MRI exam, and measures of PrC were extracted (Brodmann Areas [BA] 35 and 36). Results showed that, contrary to healthy controls, MCI patients did not benefit from unitization. Moreover, their memory performance for unitized associations was related to the measure of PrC integrity (BA35), while it was not the case in controls. This finding thus suggests that unitization does not help to attenuate the associative deficit in MCI patients, and brings support to the literature linking unitization to the PrC function. [ABSTRACT FROM AUTHOR]

The age-related associative memory deficit can be alleviated, under some conditions, when to-be-remembered associations are semantically related. In this study, we explored the experimental conditions in which older adults benefited from semantic relatedness and those that hindered any associative memory improvement. We did so by manipulating the level of semantic support within the associations presented at encoding and within the recombined pairs (i.e., the lures) at retrieval, such that pairs with high semantic support at encoding were recombined into pairs with equally high or with lower level of semantic support, and vice versa. We predicted that semantic relatedness would benefit older adults' associative memory when there was a decrease in semantic support from encoding to retrieval. Conversely, older adults' associative memory would be hindered when a recombination was equally or more familiar than the studied association. In Experiment 1, we manipulated the presence versus absence of semantic relatedness within associations both at encoding and at retrieval. In Experiment 2, we manipulated the frequency of related associations at encoding and at retrieval. Taken together, the results showed that older adults' associative memory was better in conditions in which associations closely matched semantic knowledge at encoding and were recombined into associations with no or less semantic support at retrieval. In contrast, older adults' performance was worse for semantically poorer associations at encoding that were recombined into associations with greater semantic support at retrieval. This suggests that older adults' associative memory can be improved by semantic support under specific experimental conditions only. (PsycINFO Database Record (c) 2019 APA, all rights reserved). [ABSTRACT FROM AUTHOR]

Memory for episodic associations declines in aging, ostensibly due to decreased recollection abilities. Accordingly, associative unitization - the encoding of associated items as one integrated entity - may potentially attenuate age-related associative deficits by enabling familiarity-based retrieval, which is relatively preserved in aging. To test this hypothesis, we induced bottom-up unitization by manipulating semantic relatedness between memoranda. Twenty-four young and 24 older adults studied pairs of object pictures that were either semantically related or unrelated. Participants subsequently discriminated between intact, recombined and new pairs. We found that semantic relatedness increased the contributions of both familiarity and recollection in young adults, but did not improve older adults' performance. Instead, they showed associative deficits, driven by increased recollection-based false recognition. This may reflect a "misrecollection" phenomenon, in which older adults make more false alarms to recombined pairs with particularly high confidence, due to poorer retrieval monitoring regarding semantically-related associative probes. [ABSTRACT FROM AUTHOR]

Aging is accompanied by a decline in associative memory that can, however, be attenuated when associations are unitized at encoding, that is, when they form an integrated entity. Unitization is thought to promote familiarity-based recognition memory, which is preserved in aging. We examined whether preexperimentally unitized associations (compound words (CWs)) do indeed reduce age differences in memory, and whether preexperimental unitization promotes familiarity. In Experiment 1, we assessed the memory of 20 young and 20 older participants for compound versus unrelated word pairs using a yes/no recognition test with Remember/Know/Guess judgments. In Experiment 2, we tested 20 young and 20 older participants using the same procedure, except for the use of a two-alternative forced-choice recognition paradigm, which is thought to enhance the contribution of familiarity. The results of both experiments corroborated the greater contribution of familiarity to recognition of unitized associations. In Experiment 1, however, the use of CWs did not attenuate the age-related associative decline. We suggest that preexisting knowledge associated with recombined compounds induced high absolute familiarity and illusory recollection, leading to high false-recognition rates for the older adults. By contrast, the two groups performed similarly across both conditions in Experiment 2. Thus, the forced-choice procedure facilitates the use of familiarity in such a way that it improves older adults' associative memory to the level of young participants. These results suggest that the modulation of associative memory in aging by preexisting unitization varies according to methodological parameters, such as the nature of the lures and the test format. [ABSTRACT FROM AUTHOR]

Recent advances in multivariate neuroimaging analyses have made possible the examination of the similarity of the neural patterns of activations measured across participants, but it has not been investigated yet whether such measure is age-sensitive. Here, in the scanner, young and older participants viewed scene pictures associated with labels. At test, participants were presented with the labels and were asked to recollect the associated picture. We used Pattern Similarity Analyses by which we compared patterns of neural activation during the encoding or the remembering of each picture of one participant with the averaged pattern of activation across the remaining participants. Results revealed that across-participants neural similarity was higher in young than in older adults in distributed occipital, temporal and parietal areas during encoding and retrieval. These findings demonstrate that an age-related reduction in specificity of neural activation is also evident when the similarity of neural representations is examined across participants.

The integrative memory model formalizes a new conceptualization of memory in which interactions between representations and cognitive operations within large-scale cerebral networks generate subjective memory feelings. Such interactions allow to explain the complexity of memory expressions, such as the existence of multiples sources for familiarity and recollection feelings and the fact that expectations determine how one recognizes previously encountered information.