This paper introduces a special issue of Cognitive Science initiated on the 25th anniversary of the publication of Parallel Distributed Processing (PDP), a two-volume work that introduced the use of neural network models as vehicles for understanding cognition. The collection surveys the core commitments of the PDP framework, the key issues the framework has addressed, and the debates the framework has spawned, and presents viewpoints on the current status of these issues. The articles focus on both historical roots and contemporary (...) developments in learning, optimality theory, perception, memory, language, conceptual knowledge, cognitive control, and consciousness. Here we consider the approach more generally, reviewing the original motivations, the resulting framework, and the central tenets of the underlying theory. We then evaluate the impact of PDP both on the field at large and within specific subdomains of cognitive science and consider the current role of PDP models within the broader landscape of contemporary theoretical frameworks in cognitive science. Looking to the future, we consider the implications for cognitive science of the recent success of machine learning systems called “deep networks”—systems that build on key ideas presented in the PDP volumes. (shrink)

In rebellion against low‐dimensional (e.g., two‐factor) theories in psychology, the authors make the case for high‐dimensional theories. This change in perspective requires a shift towards a focus on computation and quantitative reasoning.

Successful behavior requires actively acquiring and representing information about the environment and people, and manipulating and using those acquired representations flexibly to optimally act in and on the world. The frontal lobes have figured prominently in most accounts of flexible or goal-directed behavior, as evidenced by often-reported behavioral inflexibility in individuals with frontal lobe dysfunction. Here, we propose that the hippocampus also plays a critical role by forming and reconstructing relational memory representations that underlie flexible cognition and social behavior. There (...) is mounting evidence that damage to the hippocampus can produce inflexible and maladaptive behavior when such behavior places high demands on the generation, recombination, and flexible use of information. This is seen in abilities as diverse as memory, navigation, exploration, imagination, creativity, decision-making, character judgments, establishing and maintaining social bonds, empathy, social discourse, and language use. Thus, the hippocampus, together with its extensive interconnections with other neural systems, supports the flexible use of information in general. Further, we suggest that this understanding has important clinical implications. Hippocampal abnormalities can produce profound deficits in real-world situations, which typically place high demands on the flexible use of information, but are not always obvious on diagnostic tools tuned to frontal lobe function. This review documents the role of the hippocampus in supporting flexible representations and aims to expand our understanding of the dynamic networks that operate as we move through and create meaning of our world. (shrink)

This paper reviews the fate of the central ideas behind the complementary learning systems (CLS) framework as originally articulated in McClelland, McNaughton, and O’Reilly (1995). This framework explains why the brain requires two differentially specialized learning and memory systems, and it nicely specifies their central properties (i.e., the hippocampus as a sparse, pattern-separated system for rapidly learning episodic memories, and the neocortex as a distributed, overlapping system for gradually integrating across episodes to extract latent semantic structure). We review the application (...) of the CLS framework to a range of important topics, including the following: the basic neural processes of hippocampal memory encoding and recall, conjunctive encoding, human recognition memory, consolidation of initial hippocampal learning in cortex, dynamic modulation of encoding versus recall, and the synergistic interactions between hippocampus and neocortex. Overall, the CLS framework remains a vital theoretical force in the field, with the empirical data over the past 15 years generally confirming its key principles. (shrink)

The remember/know procedure is the most widely used method to investigate recollection and familiarity. It uses trial-by-trial reports to determine how much recollection and familiarity contribute to different kinds of recognition. Few other methods provide information about individual memory judgements and no alternative allows such direct indications of recollection and familiarity influences. Here we review how the RK procedure has been and should be used to help resolve theoretical disagreements about the processing and neural bases of components of recognition memory. (...) Emphasis is placed on procedural weaknesses and a possible confound of recollection and familiarity with recognition memory strength. Recommendations are made about how to minimise these problems including using modified versions of the procedure. The proposals here are important for improving behavioural and lesion research, and vital for brain imaging work. (shrink)

Barzykowski and Moulin's view on involuntary autobiographical memory focuses on automatic activation of representations and inhibitory control mechanisms. We discuss how and when a known neural mechanism – pattern completion – may result in involuntary autobiographical memories, the types of cues that may elicit this phenomenon and consider interactions with future-oriented cognition.

We fully support dissociating the subjective experience from the memory contents in recognition memory, as Bastin et al. posit in the target article. However, having two generic memory modules with qualitatively different functions is not mandatory and is in fact inconsistent with experimental evidence. We propose that quantitative differences in the properties of the memory modules can account for the apparent dissociation of recollection and familiarity along anatomical lines.

It is timely to assess Bayesian models, but Bayesianism is not a religion. Bayesian modeling is typically used as a tool to explain human data. Bayesian models are sometimes equivalent to other models, but have the advantage of explicitly integrating prior hypotheses with new observations. Any lack of representational or neural assumptions may be an advantage rather than a disadvantage.

Eye movement desensitization and reprocessing therapy is a well-established therapeutic method to treat post-traumatic stress disorder. However, how EMDR exerts its therapeutic action has been studied in many types of research but still needs to be completely understood. This is in part due to limited knowledge of the neurobiological mechanisms underlying EMDR, and in part to our incomplete understanding of PTSD. In order to model PTSD, we used a biologically inspired computational model based on firing rate units, encompassing the cortex, (...) hippocampus, and amygdala. Through the modulation of its parameters, we fitted real data from patients treated with EMDR or classical exposure therapy. This allowed us to gain insights into PTSD mechanisms and to investigate how EMDR achieves trauma remission. (shrink)

Computational modeling of the brain holds great promise as a bridge from brain to behavior. To fulfill this promise, however, it is not enough for models to be 'biologically plausible': models must be structurally accurate. Here, we analyze what this entails for so-called psychobiological models, models that address behavior as well as brain function in some detail. Structural accuracy may be supported by (1) a model's a priori plausibility, which comes from a reliance on evidence-based assumptions, (2) fitting existing data, (...) and (3) the derivation of new predictions. All three sources of support require modelers to be explicit about the ontology of the model, and require the existence of data constraining the modeling. For situations in which such data are only sparsely available, we suggest a new approach. If several models are constructed that together form a hierarchy of models, higher-level models can be constrained by lower-level models, and low-level models can be constrained by behavioral features of the higher-level models. Modeling the same substrate at different levels of representation, as proposed here, thus has benefits that exceed the merits of each model in the hierarchy on its own. (shrink)

We distinguish between literal and metaphorical applications of Bayesian models. When intended literally, an isomorphism exists between the elements of representation assumed by the rational analysis and the mechanism that implements the computation. Thus, observation of the implementation can externally validate assumptions underlying the rational analysis. In other applications, no such isomorphism exists, so it is not clear how the assumptions that allow a Bayesian model to fit data can be independently validated.

Existing behavioral, neuropsychological and functional neuroimaging data suggest that at least two major cognitive strategies are used for new word learning: fast mapping via context-dependent inference and explicit encoding via direct instruction. However, these distinctions remain debated at both behavioral and neurophysiological levels, not least due to confounds related to diverging experimental settings. Furthermore, the neural dynamics underpinning these two putative processes remain poorly understood. To tackle this, we designed a paradigm presenting 20 new spoken words in association with pictures (...) in either FM or EE settings, closely matched for auditory and visual features and overall task demands. We tested word acquisition using a range of behavioral measures as well as passive event-related potential responses, an established measure of word memory trace activation, and compared brain activity elicited by novel FM and EE words before and after the learning session. Behavioral data obtained in free recall, recognition and semantic word-picture matching tasks indicated successful acquisition of new words after just 10 exposures. Crucially, we found no behavioral evidence of different acquisition outcomes between FM and EE learning. ERP data, which exhibited the main response peaks at ~170, 250, and 520 ms, also indicated successful learning, with statistically different responses between novel and familiar words present only before, but not after the training, suggesting rapid formation of new neural memory circuits matching in activation those for previously known words. Furthermore, already at the earliest peak, we found different topographic distributions for the two learning types, with left-lateralized FM dynamics, suggestive of core language system involvement, and more diffuse activity for EE items, possibly suggesting the role of attention/executive control network. A similar effect also manifested later, at ~520 ms. Our data suggest that while both EE and FM learning can be successful for rapid word acquisition at the behavioral level, the diverging electrophysiological patterns suggest a dissociation between the neural systems underpinning these learning strategies. (shrink)

In the integrative memory model proposed by Bastin et al., familiarity is thought to arise from attribution of fluency signals. We suggest that, from a computational and anatomical perspective, this conceptualization converges with a global-matching account of familiarity assessment. We also argue that consideration of global matching and evidence accumulation in decision making could help further our understanding of the proposed attribution system.

We extend Mangan's account of fringe consciousness by discussing our work on processing experiences. Our research shows that variations in speed at different stages of perceptual processing can jointly contribute to subjective processing ease, supporting Mangan's notion that different mental processes condense into one subjective experience. We also discuss our studies showing that facilitation of perceptual processing leads to positive affect, supporting Mangan's suggestion that an evaluative component is built into cognitive phenomenology. Finally, we review research demonstrating that people draw (...) on processing experiences to make a variety of judgments, supporting Mangan's emphasis on the functional importance of cognitive feelings. (shrink)

This thesis investigates the relationship between eye movements, mental imagery and memory retrieval in four studies based on eye-tracking experiments. The first study is an investigation of eye movements during mental imagery elicited both visually and verbally. The use of complex stimuli and the development of a novel method where eye movements are recorded concurrently with verbal data enabled the above-mentioned relationship to be studied to an extent going beyond what previous research had been able to do. Eye movements were (...) found to closely reflect content and spatial layout while participants were listening to a spoken scene description, while they were describing the same scene from memory, and while they were describing a picture they had previously seen. This effect was equally strong during recall from memory irrespective of whether the scene visualised had originally been inspected visually by the participants or whether it was constructed whole-cloth from long-term memory. It was also found that eye movements "to nothing" appeared both when the participants were visualising scenes while looking at a blank screen and when they were doing so in complete darkness. The second study explored an effect frequently observed in the first study, involving a "scaling-down" during recall of participants' gaze patterns to an area smaller than that occupied by the stimulus encoded. It was found that this scaling effect correlated with spatial-imagery ability: the gaze patterns of participants with weaker spatial-imagery ability were closer in size to the encoded scene than the gaze patterns of those stronger in spatial-imagery ability. In the third study, the role of eye movements during mental imagery was investigated in four experiments where eye movements were prohibited during either the encoding phase or the recall phase. Experiments 1 and 2 showed that maintaining central fixation during visual or auditory encoding, respectively, had no effect on how eye movements were executed during recall. Thus, oculomotor events during recall are not reproductions of those produced during encoding. In Experiments 3 and 4, central fixation was instead maintained during recall. This turned out to alter and impair scene recollection, irrespective of the modality of encoding. Finally, in the fourth study, the functional role of eye movements in relation to memory retrieval was further investigated by means of direct eye-movement manipulation in the retrieval phase of an episodic-memory task. Four conditions were used: free viewing on a blank screen, maintaining central fixation, viewing within a square congruent with the location of the objects to be recalled, and viewing within a square incongruent with the location of the objects to be recalled. The results obtained show that gaze position plays an active and facilitatory role during memory retrieval. The findings from these studies are discussed in the light of current theories regarding eye movements during mental imagery and memory retrieval. (shrink)