Redefining "Learning" in Statistical Learning: What Does an Online Measure Reveal About the Assimilation of Visual Regularities?

Acute Stress Effects on Statistical Learning and Episodic Memory

Stress is widely considered to negatively impact hippocampal function, thus impairing episodic memory. However, the hippocampus is not merely the seat of episodic memory. Rather, it also (via distinct circuitry) supports statistical learning. On the basis of rodent work suggesting that stress may impair the hippocampal pathway involved in episodic memory while sparing or enhancing the pathway involved in statistical learning, we developed a behavioral experiment to investigate the effects of acute stress on both episodic memory and statistical learning in humans. Participants were randomly assigned to one of three conditions: stress (socially evaluated cold pressor) immediately before learning, stress ∼15 min before learning, or no stress. In the learning task, participants viewed a series of trial-unique scenes (allowing for episodic encoding of each image) in which certain scene categories reliably followed one another (allowing for statistical learning of associations between paired categories). Memory was assessed 24 hr later to isolate stress effects on encoding/learning rather than retrieval. We found modest support for our hypothesis that acute stress can amplify statistical learning: Only participants stressed ∼15 min in advance exhibited reliable evidence of learning across multiple measures. Furthermore, stress-induced cortisol levels predicted statistical learning retention 24 hr later. In contrast, episodic memory did not differ by stress condition, although we did find preliminary evidence that acute stress promoted memory for statistically predictable information and attenuated competition between statistical and episodic encoding. Together, these findings provide initial insights into how stress may differentially modulate learning processes within the hippocampus.

Perceptual Expectations Are Reflected by Early Alpha Power Reduction

The predictability of a stimulus can be characterized by its transitional probability. Perceptual expectations derived from the transitional probability of the stimulus were found to modulate the early alpha oscillations in the sensory regions of the brain when neural responses to expected versus unexpected stimuli were compared. The objective of our study was to find out the extent to which this low-frequency oscillation reflects stimulus predictability. We aimed to detect the alpha-power difference with smaller differences in transitional probabilities by comparing expected stimuli with neutral ones. We studied the effect of expectation on perception by applying an unsupervised visual statistical learning paradigm with expected and neutral stimuli embedded in an image sequence while recording EEG. Time-frequency analysis showed that expected stimuli elicit lower alpha power in the window of 8-12 Hz and 0-400 msec after stimulus presentation, appearing in the centroparietal region. Comparing previous findings of expectancy-based alpha-band modulation with our results suggests that early alpha oscillation shows an inverse relationship with stimulus predictability. Although current data are insufficient to determine the origin of the alpha power reduction, this could be a potential sign of expectation suppression in cortical oscillatory activity.

Assessing the impact of attention fluctuations on statistical learning

Attention fluctuates between optimal and suboptimal states. However, whether these fluctuations affect how we learn visual regularities remains untested. Using web-based real-time triggering, we investigated the impact of sustained attentional state on statistical learning using online and offline measures of learning. In three experiments (N = 450), participants performed a continuous performance task (CPT) with shape stimuli. Unbeknownst to participants, we measured response times (RTs) preceding each trial in real time and inserted distinct shape triplets in the trial stream when RTs indicated that a participant was attentive or inattentive. We measured online statistical learning using changes in RTs to regular triplets relative to random triplets encountered in the same attentional states. We measured offline statistical learning with a target detection task in which participants responded to target shapes selected from the regular triplets and with tasks in which participants explicitly re-created the regular triplets or selected regular shapes from foils. Online learning evidence was greater in high vs. low attentional states when combining data from all three experiments, although this was not evident in any experiment alone. On the other hand, we saw no evidence of impacts of attention fluctuations on measures of statistical learning collected offline, after initial exposure in the CPT. These results suggest that attention fluctuations may impact statistical learning while regularities are being extracted online, but that these effects do not persist to subsequent tests of learning about regularities.

Dissociation Between Linguistic and Nonlinguistic Statistical Learning in Children with Autism

Statistical learning (SL), the ability to detect and extract regularities from inputs, is considered a domain-general building block for typical language development. We compared 55 verbal children with autism (ASD, 6-12 years) and 50 typically-developing children in four SL tasks. The ASD group exhibited reduced learning in the linguistic SL tasks (syllable and letter), but showed intact learning for the nonlinguistic SL tasks (tone and image). In the ASD group, better linguistic SL was associated with higher language skills measured by parental report and sentence recall. Therefore, the atypicality of SL in autism is not domain-general but tied to specific processing constraints related to verbal stimuli. Our findings provide a novel perspective for understanding language heterogeneity in autism.

Reliability of individual differences in distractor suppression driven by statistical learning

A series of recent studies has demonstrated that attentional selection is modulated by statistical regularities, even when they concern task-irrelevant stimuli. Irrelevant distractors presented more frequently at one location interfere less with search than distractors presented elsewhere. To account for this finding, it has been proposed that through statistical learning, the frequent distractor location becomes suppressed relative to the other locations. Learned distractor suppression has mainly been studied at the group level, where individual differences are treated as unexplained error variance. Yet these individual differences may provide important mechanistic insights and could be predictive of cognitive and real-life outcomes. In the current study, we ask whether in an additional singleton task, the standard measures of attentional capture and learned suppression are reliable and stable at the level of the individual. In an online study, we assessed both the within- and between-session reliability of individual-level measures of attentional capture and learned suppression. We show that the measures of attentional capture, but not of distractor suppression, are moderately stable within the same session (i.e., split-half reliability). Test-retest reliability over a 2-month period was found to be moderate for attentional capture but weak or absent for suppression. RT-based measures proved to be superior to accuracy measures. While producing very robust findings at the group level, the predictive validity of these RT-based measures is still limited when it comes to individual-level performance. We discuss the implications for future research drawing on inter-individual variation in the attentional biases that result from statistical learning.

Specificity of Motor Contributions to Auditory Statistical Learning

Statistical learning is the ability to extract patterned information from continuous sensory signals. Recent evidence suggests that auditory-motor mechanisms play an important role in auditory statistical learning from speech signals. The question remains whether auditory-motor mechanisms support such learning generally or in a domain-specific manner. In Experiment 1, we tested the specificity of motor processes contributing to learning patterns from speech sequences. Participants either whispered or clapped their hands while listening to structured speech. In Experiment 2, we focused on auditory specificity, testing whether whispering equally affects learning patterns from speech and non-speech sequences. Finally, in Experiment 3, we examined whether learning patterns from speech and non-speech sequences are correlated. Whispering had a stronger effect than clapping on learning patterns from speech sequences in Experiment 1. Moreover, whispering impaired statistical learning more strongly from speech than non-speech sequences in Experiment 2. Interestingly, while participants in the non-speech tasks spontaneously synchronized their motor movements with the auditory stream more than participants in the speech tasks, the effect of the motor movements on learning was stronger in the speech domain. Finally, no correlation between speech and non-speech learning was observed. Overall, our findings support the idea that learning statistical patterns from speech versus non-speech relies on segregated mechanisms, and that the speech motor system contributes to auditory statistical learning in a highly specific manner.

The dynamics of multiword sequence extraction

Being able to process multiword sequences is central for both language comprehension and production. Numerous studies support this claim, but less is known about the way multiword sequences are acquired, and more specifically how associations between their constituents are established over time. Here we adapted the Hebb naming task into a Hebb lexical decision task to study the dynamics of multiword sequence extraction. Participants had to read letter strings presented on a computer screen and were required to classify them as words or pseudowords. Unknown to the participants, a triplet of words or pseudowords systematically appeared in the same order and random words or pseudowords were inserted between two repetitions of the triplet. We found that response times (RTs) for the unpredictable first position in the triplet decreased over repetitions (i.e., indicating the presence of a repetition effect) but more slowly and with a different dynamic compared with items appearing at the predictable second and third positions in the repeated triplet (i.e., showing a slightly different predictability effect). Implicit and explicit learning also varied as a function of the nature of the triplet (i.e., unrelated words, pseudowords, semantically related words, or idioms). Overall, these results provide new empirical evidence about the dynamics of multiword sequence extraction, and more generally about the role of statistical learning in language acquisition.

The successor representation subserves hierarchical abstraction for goal-directed behavior

Humans have the ability to craft abstract, temporally extended and hierarchically organized plans. For instance, when considering how to make spaghetti for dinner, we typically concern ourselves with useful "subgoals" in the task, such as cutting onions, boiling pasta, and cooking a sauce, rather than particulars such as how many cuts to make to the onion, or exactly which muscles to contract. A core question is how such decomposition of a more abstract task into logical subtasks happens in the first place. Previous research has shown that humans are sensitive to a form of higher-order statistical learning named "community structure". Community structure is a common feature of abstract tasks characterized by a logical ordering of subtasks. This structure can be captured by a model where humans learn predictions of upcoming events multiple steps into the future, discounting predictions of events further away in time. One such model is the "successor representation", which has been argued to be useful for hierarchical abstraction. As of yet, no study has convincingly shown that this hierarchical abstraction can be put to use for goal-directed behavior. Here, we investigate whether participants utilize learned community structure to craft hierarchically informed action plans for goal-directed behavior. Participants were asked to search for paintings in a virtual museum, where the paintings were grouped together in "wings" representing community structure in the museum. We find that participants' choices accord with the hierarchical structure of the museum and that their response times are best predicted by a successor representation. The degree to which the response times reflect the community structure of the museum correlates with several measures of performance, including the ability to craft temporally abstract action plans. These results suggest that successor representation learning subserves hierarchical abstractions relevant for goal-directed behavior.

Rhythmically Modulating Neural Entrainment during Exposure to Regularities Influences Statistical Learning

The ability to discover regularities in the environment, such as syllable patterns in speech, is known as statistical learning. Previous studies have shown that statistical learning is accompanied by neural entrainment, in which neural activity temporally aligns with repeating patterns over time. However, it is unclear whether these rhythmic neural dynamics play a functional role in statistical learning or whether they largely reflect the downstream consequences of learning, such as the enhanced perception of learned words in speech. To better understand this issue, we manipulated participants' neural entrainment during statistical learning using continuous rhythmic visual stimulation. Participants were exposed to a speech stream of repeating nonsense words while viewing either (1) a visual stimulus with a "congruent" rhythm that aligned with the word structure, (2) a visual stimulus with an incongruent rhythm, or (3) a static visual stimulus. Statistical learning was subsequently measured using both an explicit and implicit test. Participants in the congruent condition showed a significant increase in neural entrainment over auditory regions at the relevant word frequency, over and above effects of passive volume conduction, indicating that visual stimulation successfully altered neural entrainment within relevant neural substrates. Critically, during the subsequent implicit test, participants in the congruent condition showed an enhanced ability to predict upcoming syllables and stronger neural phase synchronization to component words, suggesting that they had gained greater sensitivity to the statistical structure of the speech stream relative to the incongruent and static groups. This learning benefit could not be attributed to strategic processes, as participants were largely unaware of the contingencies between the visual stimulation and embedded words. These results indicate that manipulating neural entrainment during exposure to regularities influences statistical learning outcomes, suggesting that neural entrainment may functionally contribute to statistical learning. Our findings encourage future studies using non-invasive brain stimulation methods to further understand the role of entrainment in statistical learning.

Cross-modal implicit learning of random time patterns

Perception is sensitive to statistical regularities in the environment, including temporal characteristics of sensory inputs. Interestingly, implicit learning of temporal patterns in one modality can also improve their processing in another modality. However, it is unclear how cross-modal learning transfer affects neural responses to sensory stimuli. Here, we recorded neural activity of human volunteers using electroencephalography (EEG), while participants were exposed to brief sequences of randomly timed auditory or visual pulses. Some trials consisted of a repetition of the temporal pattern within the sequence, and subjects were tasked with detecting these trials. Unknown to the participants, some trials reappeared throughout the experiment across both modalities (Transfer) or only within a modality (Control), enabling implicit learning in one modality and its transfer. Using a novel method of analysis of single-trial EEG responses, we showed that learning temporal structures within and across modalities is reflected in neural learning curves. These putative neural correlates of learning transfer were similar both when temporal information learned in audition was transferred to visual stimuli and vice versa. The modality-specific mechanisms for learning of temporal information and general mechanisms which mediate learning transfer across modalities had distinct physiological signatures: temporal learning within modalities relied on modality-specific brain regions while learning transfer affected beta-band activity in frontal regions.

The relation between implicit statistical learning and proactivity as revealed by EEG

Environmental events often occur on a probabilistic basis but can sometimes be predicted based on specific cues and thus approached proactively. Incidental statistical learning enables the acquisition of knowledge about probabilistic cue-target contingencies. However, the neural mechanisms of statistical learning about contingencies (SLC), the required conditions for successful learning, and the role of implicit processes in the resultant proactive behavior are still debated. We examined changes in behavior and cortical activity during an SLC task in which subjects responded to visual targets. Unbeknown to them, there were three types of target cues associated with high-, low-, and zero target probabilities. About half of the subjects spontaneously gained explicit knowledge about the contingencies (contingency-aware group), and only they showed evidence of proactivity: shortened response times to predictable targets and enhanced event-related brain responses (cue-evoked P300 and contingent negative variation, CNV) to high probability cues. The behavioral and brain responses were strictly associated on a single-trial basis. Source reconstruction of the brain responses revealed activation of fronto-parietal brain regions associated with cognitive control, particularly the anterior cingulate cortex and precuneus. We also found neural correlates of SLC in the contingency-unaware group, but these were restricted to post-target latencies and visual association areas. Our results document a qualitative difference between explicit and implicit learning processes and suggest that in certain conditions, proactivity may require explicit knowledge about contingencies.

Conscious awareness and memory systems in the brain

The term "memory" typically refers to conscious retrieval of events and experiences from our past, but experience can also change our behaviour without corresponding awareness of the learning process or the associated outcome. Based primarily on early neuropsychological work, theoretical perspectives have distinguished between conscious memory, said to depend critically on structures in the medial temporal lobe (MTL), and a collection of performance-based memories that do not. The most influential of these memory systems perspectives, the declarative memory theory, continues to be a mainstay of scientific work today despite mounting evidence suggesting that contributions of MTL structures go beyond the kinds or types of memory that can be explicitly reported. Consistent with these reports, more recent perspectives have focused increasingly on the processing operations supported by particular brain regions and the qualities or characteristics of resulting representations whether memory is expressed with or without awareness. These alternatives to the standard model generally converge on two key points. First, the hippocampus is critical for relational memory binding and representation even without awareness and, second, there may be little difference between some types of priming and explicit, familiarity-based recognition. Here, we examine the evolution of memory systems perspectives and critically evaluate scientific evidence that has challenged the status quo. Along the way, we highlight some of the challenges that researchers encounter in the context of this work, which can be contentious, and describe innovative methods that have been used to examine unconscious memory in the lab. This article is categorized under: Psychology > Memory Psychology > Theory and Methods Philosophy > Consciousness.

Statistical learning of distractor locations is dependent on task context

Through statistical learning, humans can learn to suppress visual areas that often contain distractors. Recent findings suggest that this form of learned suppression is insensitive to context, putting into question its real-life relevance. The current study presents a different picture: we show context-dependent learning of distractor-based regularities. Unlike previous studies which typically used background cues to differentiate contexts, the current study manipulated task context. Specifically, the task alternated from block to block between a compound search and a detection task. In both tasks, participants searched for a unique shape, while ignoring a uniquely colored distractor item. Crucially, a different high-probability distractor location was assigned to each task context in the training blocks, and all distractor locations were made equiprobable in the testing blocks. In a control experiment, participants only performed a compound search task such that the contexts were made indistinguishable, but the high-probability locations changed in exactly the same way as in the main experiment. We analyzed response times for different distractor locations and show that participants can learn to suppress a location in a context-dependent way, but suppression from previous task contexts lingers unless a new high-probability location is introduced.

Chunking Versus Transitional Probabilities: Differentiating Between Theories of Statistical Learning

There are two main approaches to how statistical patterns are extracted from sequences: The transitional probability approach proposes that statistical learning occurs through the computation of probabilities between items in a sequence. The chunking approach, including models such as PARSER and TRACX, proposes that units are extracted as chunks. Importantly, the chunking approach suggests that the extraction of full units weakens the processing of subunits while the transitional probability approach suggests that both units and subunits should strengthen. Previous findings using sequentially organized, auditory stimuli or spatially organized, visual stimuli support the chunking approach. However, one limitation of prior studies is that most assessed learning with the two-alternative forced-choice task. In contrast, this pre-registered experiment examined the two theoretical approaches in sequentially organized, visual stimuli using an online self-paced task-arguably providing a more sensitive index of learning as it occurs-and a secondary offline familiarity judgment task. During the self-paced task, abstract shapes were covertly organized into eight triplets (ABC) where one in every eight was altered (BCA) from the canonical structure in a way that disrupted the full unit while preserving a subunit (BC). Results from the offline familiarity judgment task revealed that the altered triplets were perceived as highly familiar, suggesting the learned representations were relatively flexible. More importantly, results from the online self-paced task demonstrated that processing for subunits, but not unit-initial stimuli, was impeded in the altered triplet. The pattern of results is in line with the chunking approach to statistical learning and, more specifically, the TRACX model.

Age-related effects on online and offline learning in visuo-spatial working memory

Learning results from online (within-session) and offline (between-sessions) changes. Heterogeneity of age-related effects in learning may be ascribed to aging differentially affecting these two processes. We investigated the contribution of online and offline consolidation in visuo-spatial working memory (vWM). Younger and older participants performed a vWM task on day one and after nine days, allowing us to disentangle online and offline learning effects. To test whether offline consolidation needs continuous practice, two additional groups of younger and older adults performed the same vWM task in between the two assessments. Similarly to other cognitive domains, older adults improved vWM through online (during session one) but not through offline learning. Practice was necessary to improve vWM between sessions in older participants. Younger adults instead exhibited only offline improvement, regardless of practice. The findings suggest that while online learning remains efficient in aging, practice is instead required to support more fragile offline mechanisms.

What sticks after statistical learning: The persistence of implicit versus explicit memory traces

Statistical learning is a powerful mechanism that extracts even subtle regularities from our information-dense worlds. Recent theories argue that statistical learning can occur through multiple mechanisms-both the conventionally assumed automatic process that precipitates unconscious learning, and an attention-dependent process that brings regularities into conscious awareness. While this view has gained popularity, there are few empirical dissociations of the hypothesized implicit and explicit forms of statistical learning. Here we provide strong evidence for this dissociation in two ways. First, we show in healthy adults (N = 60) that implicit and explicit traces have divergent consolidation trajectories, with implicit knowledge of structure strengthened over a 24-h period, while precise explicit representations tend to decay. Second, we demonstrate that repeated testing strengthens the retention of explicit representations but that implicit statistical learning is uninfluenced by testing. Together these dissociations provide much needed support for the reconceptualization of statistical learning as a multi-component construct.

No evidence for spatial suppression due to across-trial distractor learning in visual search

Previous studies have shown that during visual search, participants are able to implicitly learn across-trial regularities regarding target locations and use these to improve search performance. The present study asks whether such across-trial visual statistical learning also extends to the location of salient distractors. In Experiments 1 and 2, distractor regularities were paired so that a specific distractor location was 100% predictive of another specific distractor location on the next trial. Unlike previous findings that employed target regularities, the current results show no difference in search times between predictable and unpredictable trials. In Experiments 3-5 the distractor location was presented in a structured order (a sequence) for one group of participants, while it was presented randomly for the other group. Again, there was no learning effect of the across-trial regularities regarding the salient distractor locations. Across five experiments, we demonstrated that participants were unable to exploit across-trial spatial regularities regarding the salient distractors. These findings point to important boundary conditions for the modulation of visual attention by statistical regularities and they highlight the need to differentiate between different types of statistical regularities.

Can adults with developmental dyslexia apply statistical knowledge to a new context?

We investigated transfer of artificial grammar learning in adults with and without dyslexia in 3 experiments. In Experiment 1, participants implicitly learned an artificial grammar system and were tested on new items that included the same symbols. In Experiment 2, participants were given practice with letter strings and then tested on strings created with a different letter set. In Experiment 3, participants were given practice with shapes and then tested on strings created with different shapes. Results show that in Experiment 1, both groups demonstrated utilization of pre-trained instances in the subsequent grammaticality judgement task, while in Experiments 2 (orthographic) and 3 (nonorthographic), only typically developed participants demonstrated application of knowledge from training to test. A post hoc analysis comparing between the experiments suggests that being trained and tested on an orthographic task leads to better performance than a nonorthographic task among typically developed adults but not among adults with dyslexia. Taken together, it appears that following extensive training, individuals with dyslexia are able to form stable representations from sequential stimuli and use them in a subsequent task that utilizes strings of similar symbols. However, the manipulation of the symbols challenges this ability.

Modality, presentation, domain and training effects in statistical learning

While several studies suggest that the nature and properties of the input have significant effects on statistical learning, they have rarely been investigated systematically. In order to understand how input characteristics and their interactions impact statistical learning, we explored the effects of modality (auditory vs. visual), presentation type (serial vs. simultaneous), domain (linguistic vs. non-linguistic), and training type (random, starting small, starting big) on artificial grammar learning in young adults (N = 360). With serial presentation of stimuli, learning was more effective in the auditory than in the visual modality. However, with simultaneous presentation of visual and serial presentation of auditory stimuli, the modality effect was not present. We found a significant domain effect as well: a linguistic advantage over nonlinguistic material, which was driven by the domain effect in the auditory modality. Overall, the auditory linguistic condition had an advantage over other modality-domain types. Training types did not have any overall effect on learning; starting big enhanced performance only in the case of serial visual presentation. These results show that input characteristics such as modality, presentation type, domain and training type influence statistical learning, and suggest that their effects are also dependent on the specific stimuli and structure to be learned.

Temporal Dynamics of Competition between Statistical Learning and Episodic Memory in Intracranial Recordings of Human Visual Cortex

The function of long-term memory is not just to reminisce about the past, but also to make predictions that help us behave appropriately and efficiently in the future. This predictive function of memory provides a new perspective on the classic question from memory research of why we remember some things but not others. If prediction is a key outcome of memory, then the extent to which an item generates a prediction signifies that this information already exists in memory and need not be encoded. We tested this principle using human intracranial EEG as a time-resolved method to quantify prediction in visual cortex during a statistical learning task and link the strength of these predictions to subsequent episodic memory behavior. Epilepsy patients of both sexes viewed rapid streams of scenes, some of which contained regularities that allowed the category of the next scene to be predicted. We verified that statistical learning occurred using neural frequency tagging and measured category prediction with multivariate pattern analysis. Although neural prediction was robust overall, this was driven entirely by predictive items that were subsequently forgotten. Such interference provides a mechanism by which prediction can regulate memory formation to prioritize encoding of information that could help learn new predictive relationships.SIGNIFICANCE STATEMENT When faced with a new experience, we are rarely at a loss for what to do. Rather, because many aspects of the world are stable over time, we rely on past experiences to generate expectations that guide behavior. Here we show that these expectations during a new experience come at the expense of memory for that experience. From intracranial recordings of visual cortex, we decoded what humans expected to see next in a series of photographs based on patterns of neural activity. Photographs that generated strong neural expectations were more likely to be forgotten in a later behavioral memory test. Prioritizing the storage of experiences that currently lead to weak expectations could help improve these expectations in future encounters.

The role of the hippocampus in statistical learning and language recovery in persons with post stroke aphasia

Although several studies have aimed for accurate predictions of language recovery in post stroke aphasia, individual language outcomes remain hard to predict. Large-scale prediction models are built using data from patients mainly in the chronic phase after stroke, although it is clinically more relevant to consider data from the acute phase. Previous research has mainly focused on deficits, i.e., behavioral deficits or specific brain damage, rather than compensatory mechanisms, i.e., intact cognitive skills or undamaged brain regions. One such unexplored brain region that might support language (re)learning in aphasia is the hippocampus, a region that has commonly been associated with an individual's learning potential, including statistical learning. This refers to a set of mechanisms upon which we rely heavily in daily life to learn a range of regularities across cognitive domains. Against this background, thirty-three patients with aphasia (22 males and 11 females, M = 69.76 years, SD = 10.57 years) were followed for 1 year in the acute (1-2 weeks), subacute (3-6 months) and chronic phase (9-12 months) post stroke. We evaluated the unique predictive value of early structural hippocampal measures for short-term and long-term language outcomes (measured by the ANELT). In addition, we investigated whether statistical learning abilities were intact in patients with aphasia using three different tasks: an auditory-linguistic and visual task based on the computation of transitional probabilities and a visuomotor serial reaction time task. Finally, we examined the association of individuals' statistical learning potential with acute measures of hippocampal gray and white matter. Using Bayesian statistics, we found moderate evidence for the contribution of left hippocampal gray matter in the acute phase to the prediction of long-term language outcomes, over and above information on the lesion and the initial language deficit (measured by the ScreeLing). Non-linguistic statistical learning in patients with aphasia, measured in the subacute phase, was intact at the group level compared to 23 healthy older controls (8 males and 15 females, M = 74.09 years, SD = 6.76 years). Visuomotor statistical learning correlated with acute hippocampal gray and white matter. These findings reveal that particularly left hippocampal gray matter in the acute phase is a potential marker of language recovery after stroke, possibly through its statistical learning ability.

Online measurement of learning temporal statistical structure in categorization tasks

The ability to grasp relevant patterns from a continuous stream of environmental information is called statistical learning. Although the representations that emerge during visual statistical learning (VSL) are well characterized, little is known about how they are formed. We developed a sensitive behavioral design to characterize the VSL trajectory during ongoing task performance. In sequential categorization tasks, we assessed two previously identified VSL markers: priming of the second predictable image in a pair manifested by a reduced reaction time (RT) and greater accuracy, and the anticipatory effect on the first image revealed by a longer RT. First, in Experiment 1A, we used an adapted paradigm and replicated these VSL markers; however, they appeared to be confounded by motor learning. Next, in Experiment 1B, we confirmed the confounding influence of motor learning. To assess VSL without motor learning, in Experiment 2 we (1) simplified the categorization task, (2) raised the number of subjects and image repetitions, and (3) increased the number of single unpaired images. Using linear mixed-effect modeling and estimated marginal means of linear trends, we found that the RT curves differed significantly between predictable paired and control single images. Further, the VSL curve fitted a logarithmic model, suggesting a rapid learning process. These results suggest that our paradigm in Experiment 2 seems to be a viable online tool to monitor the behavioral correlates of unsupervised implicit VSL.

Statistical Learning of Language: A Meta-Analysis Into 25 Years of Research

Statistical learning is a key concept in our understanding of language acquisition. Ample work has highlighted its role in numerous linguistic functions-yet statistical learning is not a unitary construct, and its consistency across different language properties remains unclear. In a meta-analysis of auditory-linguistic statistical learning research spanning the last 25 years, we evaluated how learning varies across different language properties in infants, children, and adults and surveyed the methodological trends in the literature. We found robust learning across stimuli (syllables, words, etc.) in infants, and across stimuli and structures (adjacent dependencies, non-adjacent dependencies, etc.) in adults, with larger effect sizes when multiple cues were present. However, the analysis also showed significant publication bias and revealed a tendency toward using a narrow range of simplified language properties, including in the strength of the transitional probabilities used during training. Bayes factor analyses revealed prevalent data insensitivity of moderators commonly hypothesized to impact learning, such as the amount of exposure and transitional probability strength, which contradict core theoretical assumptions in the field. Methodological factors, such as the tasks used at test, also significantly impacted effect sizes in adults and children, suggesting that choice of task may critically constrain current theories of how statistical learning operates. Collectively, our results suggest that auditory-linguistic statistical learning has the kind of robustness needed to play a foundational role in language acquisition, but that more research is warranted to reveal its full potential.

Emotional Faces Facilitate Statistical Learning

Detecting regularities and extracting patterns is a vital skill to organize complex information in our environments. Statistical learning, a process where we detect regularities by attending to relationships between cues in our environment, contributes to knowledge acquisition across myriad domains. However, less is known about how emotional cues-specifically facial configurations of emotion-influence statistical learning. Here, we tested two pre-registered aims to advance knowledge about emotional signals and statistical learning: (1) we examined statistical learning in the context of emotional compared to non-emotional information, and (2) we assessed how emotional congruency (i.e., whether facial stimuli conveyed the same, or different emotions) influenced regularity extraction. We demonstrated statistical learning in the context of emotional signals. Further, we showed that statistical learning occurs more efficiently in the context of emotional faces. We also established that congruent cues benefited an online measure of statistical learning, but had varied effects when statistical learning was assessed via post-exposure recognition test. The results shed light on how affective signals influence well-studied cognitive skills and address a knowledge gap about how cue congruency impacts statistical learning, including how emotional cues might guide predictions in our social world.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42761-022-00130-9.

Detecting non-adjacent dependencies is the exception rather than the rule

Statistical learning refers to our sensitivity to the distributional properties of our environment. Humans have been shown to readily detect the dependency relationship of events that occur adjacently in a stream of stimuli but processing non-adjacent dependencies (NADs) appears more challenging. In the present study, we tested the ability of human participants to detect NADs in a new Hebb-naming task that has been proposed recently to study regularity detection in a noisy environment. In three experiments, we found that most participants did not manage to extract NADs. These results suggest that the ability to learn NADs in noise is the exception rather than the rule. They provide new information about the limits of statistical learning mechanisms.

What to expect where and when: how statistical learning drives visual selection

While the visual environment contains massive amounts of information, we should not and cannot pay attention to all events. Instead, we need to direct attention to those events that have proven to be important in the past and suppress those that were distracting and irrelevant. Experiences molded through a learning process enable us to extract and adapt to the statistical regularities in the world. While previous studies have shown that visual statistical learning (VSL) is critical for representing higher order units of perception, here we review the role of VSL in attentional selection. Evidence suggests that through VSL, attentional priority settings are optimally adjusted to regularities in the environment, without intention and without conscious awareness.

Visual statistical learning in deaf and hearing infants and toddlers

Congenital hearing loss offers a unique opportunity to examine the role of sound in cognitive, social, and linguistic development. Children with hearing loss demonstrate atypical performance across a range of general cognitive skills. For instance, research has shown that deaf school-age children underperform on visual statistical learning (VSL) tasks. However, the evidence for these deficits has been challenged, with mixed findings emerging in recent years. Here, we used a novel approach to examine VSL in the action domain early in development. We compared learning between deaf and hearing infants, prior to cochlear implantation (pre-CI), and a group of toddlers post implantation (post-CI). Findings revealed a significant difference between deaf and hearing infants pre-CI, with evidence for learning only in the hearing infants. However, there were no significant group differences between deaf and hearing toddlers post-CI, with both groups demonstrating learning. Further, VSL performance was positively correlated with language scores for the deaf toddlers, adding to the body of evidence suggesting that statistical learning is associated with language abilities. We discuss these findings in the context of previous evidence for group differences in VSL skills, and the role that auditory experiences play in infant cognitive development.

Statistical learning in children with a family risk of dyslexia

The assumption that statistical learning is affected in dyslexia has generally been evaluated in children and adults with diagnosed dyslexia, not in pre-literate children with a family risk (FR) of dyslexia. In this study, four-to-five-year-old FR children (n = 25) and No-FR children (n = 33) completed tasks of emerging literacy (phoneme awareness and RAN). They also performed an online non-adjacent dependency learning (NADL) task, based on the Serial Reaction Time (SRT) task paradigm. Children's accuracy (hits), signal sensitivity (d') and reaction times were measured. The FR group performed marginally more poorly on phoneme awareness and significantly more poorly on RAN than the No-FR group. Regarding NADL outcomes, the results were less straightforward: the data suggested successful statistical learning for both groups, as indicated by the hit and reaction time curves found. However, the FR group was less accurate and slower on the task than the No-FR group. Furthermore, unlike the No-FR group, performance in the FR group varied as a function of the specific stimulus presented. Taken together, these findings fail to show a robust difference in statistical learning between children with and without an FR of dyslexia at preschool age, in line with earlier work on older children and adults with dyslexia.

Learning Words While Listening to Syllables: Electrophysiological Correlates of Statistical Learning in Children and Adults

From an early age, exposure to a spoken language has allowed us to implicitly capture the structure underlying the succession of speech sounds in that language and to segment it into meaningful units (words). Statistical learning (SL), the ability to pick up patterns in the sensory environment without intention or reinforcement, is thus assumed to play a central role in the acquisition of the rule-governed aspects of language, including the discovery of word boundaries in the continuous acoustic stream. Although extensive evidence has been gathered from artificial languages experiments showing that children and adults are able to track the regularities embedded in the auditory input, as the probability of one syllable to follow another syllable in the speech stream, the developmental trajectory of this ability remains controversial. In this work, we have collected Event-Related Potentials (ERPs) while 5-year-old children and young adults (university students) were exposed to a speech stream made of the repetition of eight three-syllable nonsense words presenting different levels of predictability (high vs. low) to mimic closely what occurs in natural languages and to get new insights into the changes that the mechanisms underlying auditory statistical learning (aSL) might undergo through the development. The participants performed the aSL task first under implicit and, subsequently, under explicit conditions to further analyze if children take advantage of previous knowledge of the to-be-learned regularities to enhance SL, as observed with the adult participants. These findings would also contribute to extend our knowledge of the mechanisms available to assist SL at each developmental stage. Although behavioral signs of learning, even under explicit conditions, were only observed for the adult participants, ERP data showed evidence of online segmentation in the brain in both groups, as indexed by modulations in the N100 and N400 components. A detailed analysis of the neural data suggests, however, that adults and children rely on different mechanisms to assist the extraction of word-like units from the continuous speech stream, hence supporting the view that SL with auditory linguistic materials changes through development.

Surprisingly inflexible: Statistically learned suppression of distractors generalizes across contexts

The present study investigates the flexibility of statistically learned distractor suppression between different contexts. Participants performed the additional singleton task searching for a unique shape, while ignoring a uniquely colored distractor. Crucially, we created two contexts within the experiments, and each context was assigned its own high-probability distractor location, so that the location where the distractor was most likely to appear depended on the context. Experiment 1 signified context through the color of the background. In Experiment 2, we aimed to more strongly differentiate between the contexts using an auditory or visual cue to indicate the upcoming context. In Experiment 3, context determined the appropriate response ensuring that participants engaged the context in order to be able to perform the task. Across all experiments, participants learned to suppress both high-probability locations, even if they were not aware of these spatial regularities. However, these suppression effects occurred independent of context, as the pattern of suppression reflected a de-prioritization of both high-probability locations which did not change with the context. We employed Bayesian analyses to statistically quantify the absence of context-dependent suppression effects. We conclude that statistically learned distractor suppression is robust and generalizes across contexts.

Unraveling the Interconnections Between Statistical Learning and Dyslexia: A Review of Recent Empirical Studies

One important aspect of human cognition involves the learning of structured information encountered in our environment, a phenomenon known as statistical learning. A growing body of research suggests that learning to read print is partially guided by learning the statistical contingencies existing between the letters within a word, and also between the letters and sounds to which the letters refer. Research also suggests that impairments to statistical learning ability may at least partially explain the difficulties experienced by individuals diagnosed with dyslexia. However, the findings regarding impaired learning are not consistent, perhaps partly due to the varied use of methodologies across studies - such as differences in the learning paradigms, stimuli used, and the way that learning is assessed - as well as differences in participant samples such as age and extent of the learning disorder. In this review, we attempt to examine the purported link between statistical learning and dyslexia by assessing a set of the most recent and relevant studies in both adults and children. Based on this review, we conclude that although there is some evidence for a statistical learning impairment in adults with dyslexia, the evidence for an impairment in children is much weaker. We discuss several suggestive trends that emerge from our examination of the research, such as issues related to task heterogeneity, possible age effects, the role of publication bias, and other suggestions for future research such as the use of neural measures and a need to better understand how statistical learning changes across typical development. We conclude that no current theoretical framework of dyslexia fully captures the extant research findings on statistical learning.

Dynamics of nonlinguistic statistical learning: From neural entrainment to the emergence of explicit knowledge

Humans are highly attuned to patterns in the environment. This ability to detect environmental patterns, referred to as statistical learning, plays a key role in many diverse aspects of cognition. However, the spatiotemporal neural mechanisms underlying implicit statistical learning, and how these mechanisms may relate or give rise to explicit learning, remain poorly understood. In the present study, we investigated these different aspects of statistical learning by using an auditory nonlinguistic statistical learning paradigm combined with magnetoencephalography. Twenty-four healthy volunteers were exposed to structured and random tone sequences, and statistical learning was quantified by neural entrainment. Already early during exposure, participants showed strong entrainment to the embedded tone patterns. A significant increase in entrainment over exposure was detected only in the structured condition, reflecting the trajectory of learning. While source reconstruction revealed a wide range of brain areas involved in this process, entrainment in areas around the left pre-central gyrus as well as right temporo-frontal areas significantly predicted behavioral performance. Sensor level results confirmed this relationship between neural entrainment and subsequent explicit knowledge. These results give insights into the dynamic relation between neural entrainment and explicit learning of triplet structures, suggesting that these two aspects are systematically related yet dissociable. Neural entrainment reflects robust, implicit learning of underlying patterns, whereas the emergence of explicit knowledge, likely built on the implicit encoding of structure, varies across individuals and may depend on factors such as sufficient exposure time and attention.

Online Statistical Learning in Developmental Language Disorder

Purpose: The vulnerability of statistical learning (SL) in developmental language disorder (DLD) has mainly been demonstrated with metacognitive offline measures which give little insight into the more specific nature and timing of learning. Our aims in this study were to test SL in children with and without DLD with both online and offline measures and to compare the efficiency of SL in the visual and acoustic modalities in DLD. Method: We explored SL in school-age children with and without DLD matched on age and sex (n = 36). SL was investigated with the use of acoustic verbal and visual nonverbal segmentation tasks relying on online (reaction times and accuracy) and offline (two-alternative forced choice, 2AFC and production) measures. Results: In online measures, learning was evident in both groups in both the visual and acoustic modalities, while offline measures showed difficulties in DLD. The visual production task showed a significant learning effect in both groups, while the visual two-alternative forced choice (2AFC) and the two acoustic offline tasks only showed evidence of learning in the control group. The comparison of learning indices revealed an SL impairment in DLD, which is present in both modalities. Conclusions: Our findings suggest that children with DLD are comparable to typically developing (TD) children in their ability to extract acoustic verbal and visual nonverbal patterns that are cued only by transitional probabilities in online tasks, but they show impairments on metacognitive measures of learning. The pattern of online and offline measures implies that online tests can be more sensitive and valid indices of SL than offline tasks, and the combined use of different measures provides a better picture of learning efficiency, especially in groups where metacognitive tasks are challenging.

Language statistical learning responds to reinforcement learning principles rooted in the striatum

Statistical learning (SL) is the ability to extract regularities from the environment. In the domain of language, this ability is fundamental in the learning of words and structural rules. In lack of reliable online measures, statistical word and rule learning have been primarily investigated using offline (post-familiarization) tests, which gives limited insights into the dynamics of SL and its neural basis. Here, we capitalize on a novel task that tracks the online SL of simple syntactic structures combined with computational modeling to show that online SL responds to reinforcement learning principles rooted in striatal function. Specifically, we demonstrate—on 2 different cohorts—that a temporal difference model, which relies on prediction errors, accounts for participants’ online learning behavior. We then show that the trial-by-trial development of predictions through learning strongly correlates with activity in both ventral and dorsal striatum. Our results thus provide a detailed mechanistic account of language-related SL and an explanation for the oft-cited implication of the striatum in SL tasks. This work, therefore, bridges the long-standing gap between language learning and reinforcement learning phenomena.

Statistical learning is the ability to extract regularities from the environment; in the domain of language, this ability is fundamental in the learning of words and structural rules. This study uses a combination of computational modelling and functional MRI to reveal a fundamental link between online language statistical learning and reinforcement learning at the algorithmic and implementational levels.

Statistically Induced Chunking Recall: A Memory-Based Approach to Statistical Learning

The computations involved in statistical learning have long been debated. Here, we build on work suggesting that a basic memory process, chunking, may account for the processing of statistical regularities into larger units. Drawing on methods from the memory literature, we developed a novel paradigm to test statistical learning by leveraging a robust phenomenon observed in serial recall tasks: that short-term memory is fundamentally shaped by long-term distributional learning. In the statistically induced chunking recall (SICR) task, participants are exposed to an artificial language, using a standard statistical learning exposure phase. Afterward, they recall strings of syllables that either follow the statistics of the artificial language or comprise the same syllables presented in a random order. We hypothesized that if individuals had chunked the artificial language into word-like units, then the statistically structured items would be more accurately recalled relative to the random controls. Our results demonstrate that SICR effectively captures learning in both the auditory and visual modalities, with participants displaying significantly improved recall of the statistically structured items, and even recall specific trigram chunks from the input. SICR also exhibits greater test-retest reliability in the auditory modality and sensitivity to individual differences in both modalities than the standard two-alternative forced-choice task. These results thereby provide key empirical support to the chunking account of statistical learning and contribute a valuable new tool to the literature.

Optimizing steady-state responses to index statistical learning: Response to Benjamin and colleagues

Neural entrainment refers to the tendency of neural activity to align with an ongoing rhythmic stimulus. Measures of neural entrainment have been increasingly leveraged as a tool to understand how the brain tracks different types of regularities in sensory input. However, the methods used to quantify neural entrainment are varied, with numerous analytic decision points whose consequences have not been well-characterized. In a valuable contribution to this field, Benjamin, Dehaene-Lambertz and Flo (submitted) systematically compare various methodological approaches for studying neural entrainment. They demonstrate that the use of overlapping epochs, in which sliding time windows are extracted and analyzed, results in an artifactual inflation of entrainment estimates at the frequency of overlap. Here, in response to this updated best practice recommendation, we reanalyzed three previously published datasets that had been previously analyzed with overlapping epochs. Although our main results and conclusions are unaltered from those originally reported, we agree with Benjamin and colleagues that overlapping epochs should generally be avoided in classic analyses of steady-state experiments, which aim to quantify overall peaks in phase or power across an entire experimental duration. However, we present a case that overlapping epochs may be beneficial in fine-grained analyses of neural entrainment over time. The use of overlapping epochs in such analyses could improve temporal resolution without complicating interpretability of the results in cases where the question of interest relates to relative changes in neural entrainment over time within a given frequency.

Statistical Learning and the Effect of Starting Small in Developmental Dyslexia

Purpose Impairments in statistical learning abilities of individuals with developmental dyslexia (DD) have been demonstrated in word segmentation and in visual artificial grammar learning (AGL) tasks, but so far, little attention has been devoted to the AGL abilities of this population in the acoustic verbal domain. This study aimed to test whether adolescents with dyslexia have difficulties in extracting abstract patterns from auditory sequences of nonsense syllables based on a finite state grammar relative to typically developing (TD) peers. We also tested whether incremental presentation of stimuli of different lengths (starting small) has a facilitating effect on learning complex structures in dyslexia (and in TD) as opposed to presenting strings in random order. Method Thirty-one adolescents with DD and 31 age-matched control participants completed an AGL task. Participants passively listened to acoustic sequences of nonsense syllables generated by an artificial grammar in the training phase. In the test phase, they were presented with pairs of novel grammatical and nongrammatical sequences and were required to decide which member of a sequence pair was more similar to the material heard during training. Results Performance levels and the proportion of learners were smaller in participants with DD than in the control group. While the starting small effect was nominally present both in performance levels and in the number of learners in participants with DD, but not in the group with TD, the presentation of strings in incremental order did not statistically improve learning performance in either group. Conclusion Our results suggest that (a) statistical learning of abstract sequences in the acoustic domain is less efficient in people with dyslexia than in TD controls and (b) while incremental presentation of stimuli of different length did not improve learning in our study, the observed pattern of results suggests that the effects of different training designs should be explored further in developmental disorders.

The contribution of individual differences in statistical learning to reading and spelling performance in children with and without dyslexia

Using an individual differences approach in children with and without dyslexia, this study investigated the hypothesized relationship between statistical learning ability and literacy (reading and spelling) skills. We examined the clinical relevance of statistical learning (serial reaction time and visual statistical learning tasks) by controlling for potential confounds at the participant level (e.g., non-verbal reasoning, attention and phonological skills including rapid automatized naming and phonological short-term memory). A 100 Dutch-speaking 8- to 11-year-old children with and without dyslexia participated (50 per group), see also van Witteloostuijn et al. (2019) for a study with the same participants. No evidence of a relationship between statistical learning and literacy skills is found above and beyond participant-level variables. Suggestions from the literature that the link between statistical learning and literacy attainment, and therefore its clinical relevance, might be small and strongly influenced by methodological differences between studies are not contradicted by our findings.

Visual object recognition is facilitated by temporal community structure

Humans and others primates are highly attuned to temporal consistencies and regularities in their sensory environment and learn to predict such statistical structure. Moreover, in several instances, the presence of temporal structure has been found to facilitate procedural learning and to improve task performance. Here we extend these findings to visual object recognition and to presentation sequences in which mutually predictive objects form distinct clusters or "communities." Our results show that temporal community structure accelerates recognition learning and affects the order in which objects are learned ("onset of familiarity").

Tracking statistical learning online: Word segmentation in a target detection task

Despite the essential role of statistical learning in shaping human behavior, there are still controversies concerning its measurement. In this paper, we present a novel online target-detection task in an acoustic word segmentation paradigm, which is able to track the process of learning and does not build on deliberation and decision making. Beside testing the novel online task, we also examined its relationship with two offline measures: the traditional two-alternative forced choice (2AFC) task, and the statistically-induced chunking recall (SICR) task (Isbilen et al., 2017). Participants showed a significant learning effect on the online task, reflected in the decrease of reaction times during training and in the differences between reaction times to predictable versus unpredictable targets. Online learning scores correlated with the 2AFC scores, but this association was only present when participants did not have explicit knowledge about stimuli. SICR scores were not associated with any of the other measures. The internal consistency was higher for online learning measures than for the other two tasks. These findings show that the online target detection task is a good tool for assessing statistical learning, and invite further research on its psychometric properties.

Not All Words Are Equally Acquired: Transitional Probabilities and Instructions Affect the Electrophysiological Correlates of Statistical Learning

Statistical learning (SL), the process of extracting regularities from the environment, is a fundamental skill of our cognitive system to structure the world regularly and predictably. SL has been studied using mainly behavioral tasks under implicit conditions and with triplets presenting the same level of difficulty, i.e., a mean transitional probability (TP) of 1.00. Yet, the neural mechanisms underlying SL under other learning conditions remain largely unknown. Here, we investigated the neurofunctional correlates of SL using triplets (i.e., three-syllable nonsense words) with a mean TP of 1.00 (easy "words") and 0.50 (hard "words") in an SL task performed under incidental (implicit) and intentional (explicit) conditions, to determine whether the same core mechanisms were recruited to assist learning. Event-related potentials (ERPs) were recorded while participants listened firstly to a continuous auditory stream made of the concatenation of four easy and four hard "words" under implicit instructions, and subsequently to another auditory stream made of the concatenation of four easy and four hard "words" drawn from another artificial language under explicit instructions. The stream in each of the SL tasks was presented in two consecutive blocks of ~3.5-min each (~7-min in total) to further examine how ERP components might change over time. Behavioral measures of SL were collected after the familiarization phase of each SL task by asking participants to perform a two-alternative forced-choice (2-AFC) task. Results from the 2-AFC tasks revealed a moderate but reliable level of SL, with no differences between conditions. ERPs were, nevertheless, sensitive to the effect of TPs, showing larger amplitudes of N400 for easy "words," as well as to the effect of instructions, with a reduced N250 for "words" presented under explicit conditions. Also, significant differences in the N100 were found as a result of the interaction between TPs, instructions, and the amount of exposure to the auditory stream. Taken together, our findings suggest that triplets' predictability impacts the emergence of "words" representations in the brain both for statistical regularities extracted under incidental and intentional instructions, although the prior knowledge of the "words" seems to favor the recruitment of different SL mechanisms.

Integrating when and what information in the left parietal lobe allows language rule generalization

A crucial aspect when learning a language is discovering the rules that govern how words are combined in order to convey meanings. Because rules are characterized by sequential co-occurrences between elements (e.g., “These cupcakes are unbelievable”), tracking the statistical relationships between these elements is fundamental. However, purely bottom-up statistical learning alone cannot fully account for the ability to create abstract rule representations that can be generalized, a paramount requirement of linguistic rules. Here, we provide evidence that, after the statistical relations between words have been extracted, the engagement of goal-directed attention is key to enable rule generalization. Incidental learning performance during a rule-learning task on an artificial language revealed a progressive shift from statistical learning to goal-directed attention. In addition, and consistent with the recruitment of attention, functional MRI (fMRI) analyses of late learning stages showed left parietal activity within a broad bilateral dorsal frontoparietal network. Critically, repetitive transcranial magnetic stimulation (rTMS) on participants’ peak of activation within the left parietal cortex impaired their ability to generalize learned rules to a structurally analogous new language. No stimulation or rTMS on a nonrelevant brain region did not have the same interfering effect on generalization. Performance on an additional attentional task showed that this rTMS on the parietal site hindered participants’ ability to integrate “what” (stimulus identity) and “when” (stimulus timing) information about an expected target. The present findings suggest that learning rules from speech is a two-stage process: following statistical learning, goal-directed attention—involving left parietal regions—integrates “what” and “when” stimulus information to facilitate rapid rule generalization.

This study uses repetitive transcranial stimulation to show that learning language rules from speech is a two-stage process; following statistical learning, goal-directed attention (involving left parietal regions) integrates "what" and "when" stimulus information to facilitate rapid rule generalization.

Beta-Band Activity Is a Signature of Statistical Learning

Through statistical learning (SL), cognitive systems may discover the underlying regularities in the environment. Testing human adults (n = 35, 21 females), we document, in the context of a classical visual SL task, divergent rhythmic EEG activity in the interstimulus delay periods within patterns versus between patterns (i.e., pattern transitions). Our findings reveal increased oscillatory activity in the beta band (∼20 Hz) at triplet transitions that indexes learning: it emerges with increased pattern repetitions; and importantly, it is highly correlated with behavioral learning outcomes. These findings hold the promise of converging on an online measure of learning regularities and provide important theoretical insights regarding the mechanisms of SL and prediction.SIGNIFICANCE STATEMENT Statistical learning has become a major theoretical construct in cognitive science, providing the primary means by which organisms learn about regularities in the environment. As such, it is a critical building block for basic and higher-order cognitive functions. Here we identify, for the first time, a spectral neural index in the time window before stimulus presentation, which evolves with increased pattern exposure, and is predictive of learning performance. The manifestation of learning that is revealed, not in stimulus processing but in the blank interval between stimuli, makes a direct link between the fields of statistical learning on the one hand and either prediction or consolidation on the other hand, suggesting a possible mechanistic account of visual statistical learning.

Implicit-statistical learning in aphasia and its relation to lesion location

BACKGROUND

Implicit-statistical learning (ISL) research investigates whether domain-general mechanisms are recruited in the linguistic processes that require manipulation of patterned regularities (e.g. syntax). Aphasia is a language disorder caused by focal brain damage in the left fronto-temporal-parietal network. Research shows that people with aphasia (PWA) with frontal lobe lesions manifest convergent deficits in syntax and ISL mechanisms. So far, ISL mechanisms in PWA with temporal or parietal lobe lesions have not been systematically investigated.

AIMS

We investigated two complementary hypotheses: 1) the anatomical hypothesis, that PWA with frontal lesions display more severely impaired ISL abilities than PWA with posterior lesions and 2) the behavioural hypothesis, that the magnitude of impairment in ISL mechanisms correlates to syntactic deficits in aphasia.

METHODS

We tested 13 PWA, 5 with frontal lesions and 8 with posterior lesions, and 11 non-brain-damaged controls on a visual statistical learning (VSL) task. In addition, all PWA completed several linguistic tasks. Reaction times, obtained in the VSL task, were analyzed using linear mixed-effects model. Correlational statistics were used to assess the relationship between VSL task performance and linguistic measures.

RESULTS AND DISCUSSION

We did not find support for the anatomical hypothesis as patients with spared frontal regions also manifested impaired ISL mechanisms. This is attributed to a) ISL mechanisms being vulnerable to other cognitive dysfunctions and/or b) ISL mechanisms anatomically extending to the posterior brain regions. Notably, ISL mechanisms were impaired, but not absent in aphasia. With regards to the behavioural hypothesis, we provide empirical evidence of correlation between ISL mechanisms and syntactic, but not lexical impairment in aphasia. We discuss both the theoretical contributions to the debate of domain-independence of ISL mechanisms and clinical implications for implicit language therapy.

Syllables in Sync Form a Link: Neural Phase-locking Reflects Word Knowledge during Language Learning

Language is composed of small building blocks, which combine to form larger meaningful structures. To understand language, we must process, track, and concatenate these building blocks into larger linguistic units as speech unfolds over time. An influential idea is that phase-locking of neural oscillations across different levels of linguistic structure provides a mechanism for this process. Building on this framework, the goal of the current study was to determine whether neural phase-locking occurs more robustly to novel linguistic items that are successfully learned and encoded into memory, compared to items that are not learned. Participants listened to a continuous speech stream composed of repeating nonsense words while their EEG was recorded and then performed a recognition test on the component words. Neural phase-locking to individual words during the learning period strongly predicted the strength of subsequent word knowledge, suggesting that neural phase-locking indexes the subjective perception of specific linguistic items during real-time language learning. These findings support neural oscillatory models of language, demonstrating that words that are successfully perceived as functional units are tracked by oscillatory activity at the matching word rate. In contrast, words that are not learned are processed merely as a sequence of unrelated syllables and thus not tracked by corresponding word-rate oscillations.

Preverbal Infants Discover Statistical Word Patterns at Similar Rates as Adults: Evidence From Neural Entrainment

The discovery of words in continuous speech is one of the first challenges faced by infants during language acquisition. This process is partially facilitated by statistical learning, the ability to discover and encode relevant patterns in the environment. Here, we used an electroencephalogram (EEG) index of neural entrainment to track 6-month-olds’ ( N = 25) segmentation of words from continuous speech. Infants’ neural entrainment to embedded words increased logarithmically over the learning period, consistent with a perceptual shift from isolated syllables to wordlike units. Moreover, infants’ neural entrainment during learning predicted postlearning behavioral measures of word discrimination ( n = 18). Finally, the logarithmic increase in entrainment to words was comparable in infants and adults, suggesting that infants and adults follow similar learning trajectories when tracking probability information among speech sounds. Statistical-learning effects in infants and adults may reflect overlapping neural mechanisms, which emerge early in life and are maintained throughout the life span.

The multi-faceted nature of visual statistical learning: Individual differences in learning conditional and distributional regularities across time and space

Emerging research has demonstrated that statistical learning is a modality-specific ability governed by domain-general principles. Yet limited research has investigated different forms of statistical learning within modality. This paper explores whether there is one unified statistical learning mechanism within the visual modality, or separate task-specific abilities. To do so, we examined individual differences in spatial and nonspatial conditional and distributional statistical learning. Participants completed four visual statistical learning tasks: conditional spatial, conditional nonspatial, distributional spatial, and distributional nonspatial. Performance on all four tasks significantly correlated with each other, and performance on all tasks accounted for a large portion of the variance across tasks (57%). Interestingly, a portion of the variance of task performance (between 11% and 18%) was also accounted for by performance on each of the individual tasks. Our results suggest that visual statistical learning is the result of the interplay between a unified mechanism for extracting conditional and distributional statistical regularities across time and space, and an individual's ability to extract specific types of regularities.

Detection of regularities in a random environment

Regularity detection, or statistical learning, is regarded as a fundamental component of our cognitive system. To test the ability of human participants to detect regularity in a more ecological situation (i.e., mixed with random information), we used a simple letter-naming paradigm in which participants were instructed to name single letters presented one at a time on a computer screen. The regularity consisted of a triplet of letters that were systematically presented in that order. Participants were not told about the presence of this regularity. A variable number of random letters were presented between two repetitions of the regular triplet, making this paradigm similar to a Hebb repetition task. Hence, in this Hebb-naming task, we predicted that if any learning of the triplet occurred, naming times for the predictable letters in the triplet would decrease as the number of triplet repetitions increased. Surprisingly, across four experiments, detection of the regularity only occurred under very specific experimental conditions and was far from a trivial task. Our study provides new evidence regarding the limits of statistical learning and the critical role of contextual information in the detection (or not) of repeated patterns.