Ruth de Diego-Balaguer

Statistical learning (SL), the ability to extract regularities from sensory input, is hypothesized to play a major role in many aspects of higher-level cognition. While earlier investigations centered on group-level effects, a growing body of research now focuses on individual differences in this ability. This consensus paper presents the collective insights of 27 researchers worldwide who are active in the field of individual differences in SL, with the goal of reflecting on the field’s progress so far while outlining key challenges going forward. Our discussion covers possible theoretical and operational definitions of SL, the trade-offs between controlled laboratory tasks and ecologically valid paradigms, considerations in interpreting correlations between SL performance and other cognitive abilities, issues of measurement reliability and statistical power, and rationales for adopting individual-differences versus group-level approaches. Based on these discussions, we outline nine actionable recommendations for the field to move forward, which are meant to lead to a new body of evidence that is more theoretically informed and more methodologically robust.

In language processing, word meanings are retrieved from memory and converted into sentential expressions that have referential meaning at the level of utterances. Cognitive processes related to lexical retrieval versus the construction of referential meaning through grammar have not been properly disentangled yet. We addressed this with an experimental design aimed to link well-known N400 effects and late positive components (LPCs) to this distinction. We tested the hypothesis that semantic conflicts derived from a collision with stored (= ‘offline’) knowledge would elicit N400 effects, while semantic conflicts arising ‘online’ during the grammatical construction of referential meaning would exclusively trigger LPCs and also differ in their latency, amplitude or topography from formal-syntactic violations. Three anomalous conditions were compared against a baseline of correct sentences, across five word-positions within stimulus sentences: semantic anomalies that are lexically- but not grammatically-driven (I arrived in a button last week), semantic anomalies that are grammatically- but not lexically-driven (I arrived in a London last week), and formal-syntactic anomalies (I arrived in a cities last week). Results showed, in line with predictions, that (i) only the lexically-driven anomalies elicited N400 components, while (ii) all three anomalous conditions elicited LPCs, extending from the target word to subsequent word positions in the sentence; and (iii) the LPCs were all distinct in latency, amplitude and spatiotemporal topographies. These results demonstrate for the first time that grammar-level meaning, which is referential rather than lexical-conceptual in nature, has its own distinctive ERP, while confining the role of the N400 to effects of stored, or else contextually updated, memory.

Sentence comprehension requires identifying the meanings associated with specific words, but also the referential content of the sentence as determined by grammar. We aimed to address the contribution of grammar to referential meaning by constructing semantic anomalies where word meanings were correct but referential meaning was anomalous due to the grammatical structure involved: e.g., I visited a London this morning. Using a self-paced reading task, we targeted how these anomalies affected reading time increases, and the location of this effect in the sentences. Seventy native speakers of Spanish read a total of 640 sentences and judged them as either correct or anomalous. Anomalous sentences could be: (ii) Semantic anomalies due to lexical but not grammatical factors (e.g., I arrived in a button), (iii) Semantic anomalies due to grammatical but not lexical factors, or (iv) formal-syntactic anomalies (gender and number agreement mismatches). Responses and response times of these judgements and a confidence rating were collected. We hypothesized that if referential meaning is calculated at the grammatical level, the location of reading time increases would differ between the two semantic anomaly types. We predicted that lexical-semantic anomalies (ii) would cause local reading time increases, whereas grammatical-semantic anomalies (iii) would cause more global effects. In line with our prediction, grammatical-semantic anomalies showed global effects, while lexical-semantic and syntactic anomalies showed local effects. Moreover, lexical-semantic and syntactic anomalies led to shorter reaction times in grammaticality judgments. In this way, semantic anomalies that are neither lexically driven nor syntactically illicit, provide a distinctive index of grammatical meaning: meaning that is referential and integrated at, and through, the grammatical level.

Background: Despite its impact on daily life, impulsivity in Huntington's disease (HD) is understudied as a neuropsychiatric symptom. Our aim is to characterize temporal impulsivity in HD and to disentangle the white matter correlate associated with impulsivity.Methods: Forty‐seven HD individuals and 36 healthy controls were scanned and evaluated for temporal impulsivity using a delay‐discounting (DD) task and complementary Sensitivity to Punishment and Sensitivity to Reward Questionnaire. Diffusion tensor imaging was employed to characterize the structural connectivity of three limbic tracts: the uncinate fasciculus (UF), the accumbofrontal tract (NAcc‐OFC), and the dorsolateral prefrontal cortex connectig the caudate nucleus (DLPFC‐cn). Multiple linear regression analyses were applied to analyze the relationship between impulsive behavior and white matter microstructural integrity.Results: Our results revealed altered structural connectivity in the DLPC‐cn, the NAcc‐OFC and the UF in HD individuals. At the same time, the variability in structural connectivity of these tracts was associated with the individual differences in temporal impulsivity. Specifically, increased structural connectivity in the right NAcc‐OFC and reduced connectivity in the left UF were associated with higher temporal impulsivity scores.Conclusions: The present findings highlight the importance of investigating the spectrum of temporal impulsivity in HD. As, while less prevalent than other psychiatric features, this symptom is still reported to significantly impact the quality of life of patients and caregivers. This study provides evidence that individual differences observed in temporal impulsivity may be explained by variability in limbic frontostriatal tracts, while shedding light on the role of sensitivity to reward in modulating impulsive behavior through the selection of immediate rewards.

Statistical learning involves extracting statistical regularities from the environment. Despite the ubiquitous crossmodal relations in our environment, previous research has failed in showing crossmodal learning suggesting that statistical learning is modality-specific, occurring within but not between sensory modalities. The present study investigates under which circumstances statistical learning can occur between modalities. In the first experiment, participants viewed a stream of meaningless visual fractals and synthetic auditory stimuli. Importantly, the sequence of stimuli could be grouped into unimodal or crossmodal pairs based on their internal transitional probabilities. Using implicit and explicit measures of learning, we found that participants only learned the unimodal pairs. In the second experiment, pairs were presented in separate unimodal and crossmodal blocks. The crossmodal blocks alternated visual and auditory modalities allowing participants to anticipate the upcoming modality. This manipulation allowed significant statistical learning for the crossmodal pairs, reflected only by implicit measures. This suggests that modality transitions predictability aids correct attention deployment across sensory modalities, crucial for learning crossmodal statistical contingencies. In the third experiment, where audiovisual stimuli with semantic content were used leaving access to an amodal shared representation, participants could explicitly recognize statistical regularities between crossmodal pairs even when the upcoming modality was unpredictable. This finding suggests statistical learning between crossmodal pairs can occur when sensory-level limitations are bypassed, and when learning unfolds at an amodal level of representation. These findings challenge the view that statistical learning is strictly modality-specific, instead indicating that crossmodal statistical learning depends on attentional mechanisms and representational level. Public significance statement: Previous research has established that low-level statistical learning occurs within individual sensory modalities. However, evidence for statistical learning across modalities remains limited. This has led to the hypothesis that statistical learning operates as a modality-specific system, constrained by the perceptual properties and content of each sensory system. In this study, we demonstrate that modality transitions’ predictability enables crossmodal statistical learning, even in the absence of semantic information. Notably, this constraint does not apply to meaningful stimuli, which can be processed at an amodal level. Our findings broaden the classical neurobiological model of statistical learning by highlighting the roles of attentional mechanisms and representational levels in supporting crossmodal statistical learning.

In language, grammatical dependencies often hold between items that are not immediately adjacent to each other. Acquiring these nonadjacent dependencies is crucial for learning grammar. However, there are potentially infinitely many dependencies in the language input. How does the infant brain solve this computational learning problem? Here, we demonstrate that while rudimentary sensitivity to nonadjacent regularities may be present relatively early, robust and reliable learning can only be achieved when convergent statistical and perceptual, specifically prosodic cues, are both present, helping the infant brain detect the building blocks that form a nonadjacent dependency. This study contributes to our understanding of the neural foundations of rule learning that pave the way for language acquisition.

Background Despite its impact on daily life, impulsivity in Huntington’s disease (HD) is understudied as a neuropsychiatric symptom. Our aim is to characterize temporal impulsivity in HD, evaluated through a Delay Discounting (DD) task, and to disentangle the underlying white matter correlates in HD. Methods Forty-seven HD individuals and thirty-six healthy controls conducted a DD task and complementary Sensitivity to Punishment and Sensitivity to Reward (SR) Questionnaire. Diffusion-tensor imaging was employed to characterize the structural connectivity of two limbic tracts: the uncinate fasciculus (UF) and the accumbofrontal tract (NAcc-OFC). Multiple linear regression analyses were applied to analyze the relationship between impulsive behavior and white-matter microstructural integrity. Results Altered structural connectivity in both the NAcc-OFC and UF in HD individuals was observed. Moreover, the variability in structural connectivity of these tracts was associated with the individual differences in temporal impulsivity. Specifically, increased structural connectivity in the right NAcc-OFC predicted increased temporal impulsivity, while reduced connectivity in the left UF was associated with higher temporal impulsivity scores. Limitations Other cognitive mechanisms and white matter tracts may play a role in temporal impulsivity. Conclusions This study provides evidence that individual differences observed in impulsivity may be explained by variability in limbic fronto-striatal tracts. We emphasize the importance of investigating the spectrum of impulsivity in HD, less prevalent than other psychiatric features, but impacting the quality of life of patients and their caregivers.