Hebbian learning can explain rhythmic neural entrainment to statistical regularities.

In many domains, learners extract recurring units from continuous sequences. For example, in unknown languages, fluent speech is perceived as a continuous signal. Learners need to extract the underlying words from this continuous signal and then memorize them. One prominent candidate mechanism is statistical learning, whereby learners track how predictive syllables (or other items) are of one another. Syllables within the same word predict each other better than syllables straddling word boundaries. But does statistical learning lead to memories of the underlying words-or just to pairwise associations among syllables? Electrophysiological results provide the strongest evidence for the memory view. Electrophysiological responses can be time-locked to statistical word boundaries (e.g., N400s) and show rhythmic activity with a periodicity of word durations. Here, I reproduce such results with a simple Hebbian network. When exposed to statistically structured syllable sequences (and when the underlying words are not excessively long), the network activation is rhythmic with the periodicity of a word duration and activation maxima on word-final syllables. This is because word-final syllables receive more excitation from earlier syllables with which they are associated than less predictable syllables that occur earlier in words. The network is also sensitive to information whose electrophysiological correlates were used to support the encoding of ordinal positions within words. Hebbian learning can thus explain rhythmic neural activity in statistical learning tasks without any memory representations of words. Learners might thus need to rely on cues beyond statistical associations to learn the words of their native language. RESEARCH HIGHLIGHTS: Statistical learning may be utilized to identify recurring units in continuous sequences (e.g., words in fluent speech) but may not generate explicit memory for words. Exposure to statistically structured sequences leads to rhythmic activity with a period of the duration of the underlying units (e.g., words). I show that a memory-less Hebbian network model can reproduce this rhythmic neural activity as well as putative encodings of ordinal positions observed in earlier research. Direct tests are needed to establish whether statistical learning leads to declarative memories for words.

Hebbian, correlational learning provides a memory-less mechanism for Statistical Learning irrespective of implementational choices: Reply to Tovar and Westermann (2022).

Statistical learning relies on detecting the frequency of co-occurrences of items and has been proposed to be crucial for a variety of learning problems, notably to learn and memorize words from fluent speech. Endress and Johnson (2021) (hereafter EJ) recently showed that such results can be explained based on simple memory-less correlational learning mechanisms such as Hebbian Learning. Tovar and Westermann (2022) (hereafter TW) reproduced these results with a different Hebbian model. We show that the main differences between the models are whether temporal decay acts on both the connection weights and the activations (in TW) or only on the activations (in EJ), and whether interference affects weights (in TW) or activations (in EJ). Given that weights and activations are linked through the Hebbian learning rule, the networks behave similarly. However, in contrast to TW, we do not believe that neurophysiological data are relevant to adjudicate between abstract psychological models with little biological detail. Taken together, both models show that different memory-less correlational learning mechanisms provide a parsimonious account of Statistical Learning results. They are consistent with evidence that Statistical Learning might not allow learners to learn and retain words, and Statistical Learning might support predictive processing instead.

Generic learning mechanisms can drive social inferences: The role of type frequency.

How do we form opinions about typical and morally acceptable behavior in other social groups despite variability in behavior? Similar learning problems arise during language acquisition, where learners need to infer grammatical rules (e.g., the walk/walk-ed past-tense) despite frequent exceptions (e.g., the go/went alternation). Such rules need to occur with many different words to be learned (i.e., they need a high type frequency). In contrast, frequent individual words do not lead to learning. Here, we ask whether similar principles govern social learning. Participants read a travel journal where a traveler observed behaviors in different imaginary cities. The behaviors were performed once by many distinct actors (high type frequency) or frequently by a single actor (low type frequency), and could be good, neutral or bad. We then asked participants how morally acceptable the behavior was (in general or for the visited city), and how widespread it was in that city. We show that an ideal observer model estimating the prevalence of behaviors is only sensitive to the behaviors' type frequency, but not to how often they are performed. Empirically, participants rated high type frequency behaviors as more morally acceptable more prevalent than low type frequency behaviors. They also rated good behaviors as more acceptable and prevalent than neutral or bad behaviors. These results suggest that generic learning mechanisms and epistemic biases constrain social learning, and that type frequency can drive inferences about groups. To combat stereotypes, high type frequency behaviors might thus be more effective than frequently appearing individual role models.

Memory and Proactive Interference for spatially distributed items.

Our ability to briefly retain information is often limited. Proactive Interference (PI) might contribute to these limitations (e.g., when items in recognition tests are difficult to reject after having appeared recently). In visual Working Memory (WM), spatial information might protect WM against PI, especially if encoding items together with their spatial locations makes item-location combinations less confusable than simple items without a spatial component. Here, I ask (1) if PI is observed for spatially distributed items, (2) if it arises among simple items or among item-location combinations, and (3) if spatial information affects PI at all. I show that, contrary to views that spatial information protects against PI, PI is reliably observed for spatially distributed items except when it is weak. PI mostly reflects items that appear recently or frequently as memory items, while occurrences as test items play a smaller role, presumably because their temporal context is easier to encode. Through mathematical modeling, I then show that interference occurs among simple items rather than item-location combinations. Finally, to understand the effects of spatial information, I separate the effects of (a) the presence and (b) the predictiveness of spatial information on memory and its susceptibility to PI. Memory is impaired when items are spatially distributed, but, depending on the analysis, unaffected by the predictiveness of spatial information. In contrast, the susceptibility to PI is unaffected by either manipulation. Visual memory is thus impaired by PI for spatially distributed items due to interference from recent memory items (rather than test items or item-location combinations).

Socio-Cultural Values Are Risk Factors for COVID-19-Related Mortality.

To assess whether socio-cultural values are population-level risk factors for health, I sought to predict COVID-19-related mortality between 2 weeks and 6 months after the first COVID-19-related death in a country based on values extracted from the World Values Survey for different country sets, after controlling for various confounding variables. COVID-19-related mortality was increased in countries endorsing political participation but decreased in countries with greater trust in institutions and materialistic orientations. The values were specific to COVID-19-related mortality, did not predict general health outcomes, and values predicting increased COVID-19-related mortality predicted decreased mortality from other outcomes (e.g., environmental-related mortality).

When forgetting fosters learning: A neural network model for statistical learning.

Learning often requires splitting continuous signals into recurring units, such as the discrete words constituting fluent speech; these units then need to be encoded in memory. A prominent candidate mechanism involves statistical learning of co-occurrence statistics like transitional probabilities (TPs), reflecting the idea that items from the same unit (e.g., syllables within a word) predict each other better than items from different units. TP computations are surprisingly flexible and sophisticated. Humans are sensitive to forward and backward TPs, compute TPs between adjacent items and longer-distance items, and even recognize TPs in novel units. We explain these hallmarks of statistical learning with a simple model with tunable, Hebbian excitatory connections and inhibitory interactions controlling the overall activation. With weak forgetting, activations are long-lasting, yielding associations among all items; with strong forgetting, no associations ensue as activations do not outlast stimuli; with intermediate forgetting, the network reproduces the hallmarks above. Forgetting thus is a key determinant of these sophisticated learning abilities. Further, in line with earlier dissociations between statistical learning and memory encoding, our model reproduces the hallmarks of statistical learning in the absence of a memory store in which items could be placed.

Statistical learning and memory.

Learners often need to identify and remember recurring units in continuous sequences, but the underlying mechanisms are debated. A particularly prominent candidate mechanism relies on distributional statistics such as Transitional Probabilities (TPs). However, it is unclear what the outputs of statistical segmentation mechanisms are, and if learners store these outputs as discrete chunks in memory. We critically review the evidence for the possibility that statistically coherent items are stored in memory and outline difficulties in interpreting past research. We use Slone and Johnson's (2018) experiments as a case study to show that it is difficult to delineate the different mechanisms learners might use to solve a learning problem. Slone and Johnson (2018) reported that 8-month-old infants learned coherent chunks of shapes in visual sequences. Here, we describe an alternate interpretation of their findings based on a multiple-cue integration perspective. First, when multiple cues to statistical structure were available, infants' looking behavior seemed to track with the strength of the strongest one - backward TPs, suggesting that infants process multiple cues simultaneously and select the strongest one. Second, like adults, infants are exquisitely sensitive to chunks, but may require multiple cues to extract them. In Slone and Johnson's (2018) experiments, these cues were provided by immediate chunk repetitions during familiarization. Accordingly, infants showed strongest evidence of chunking following familiarization sequences in which immediate repetitions were more frequent. These interpretations provide a strong argument for infants' processing of multiple cues and the potential importance of multiple cues for chunk recognition in infancy.

Sequential Presentation Protects Working Memory From Catastrophic Interference.

Neural network models of memory are notorious for catastrophic interference: Old items are forgotten as new items are memorized (French, 1999; McCloskey & Cohen, 1989). While working memory (WM) in human adults shows severe capacity limitations, these capacity limitations do not reflect neural network style catastrophic interference. However, our ability to quickly apprehend the numerosity of small sets of objects (i.e., subitizing) does show catastrophic capacity limitations, and this subitizing capacity and WM might reflect a common capacity. Accordingly, computational investigations (Knops, Piazza, Sengupta, Eger & Melcher, 2014; Sengupta, Surampudi & Melcher, 2014) suggest that mutual inhibition among neurons can explain both kinds of capacity limitations as well as why our ability to estimate the numerosity of larger sets is limited according to a Weber ratio signature. Based on simulations with a saliency map-like network and mathematical proofs, we provide three results. First, mutual inhibition among neurons leads to catastrophic interference when items are presented simultaneously. The network can remember a limited number of items, but when more items are presented, the network forgets all of them. Second, if memory items are presented sequentially rather than simultaneously, the network remembers the most recent items rather than forgetting all of them. Hence, the tendency in WM tasks to sequentially attend even to simultaneously presented items might not only reflect attentional limitations, but also an adaptive strategy to avoid catastrophic interference. Third, the mean activation level in the network can be used to estimate the number of items in small sets, but it does not accurately reflect the number of items in larger sets. Rather, we suggest that the Weber ratio signature of large number discrimination emerges naturally from the interaction between the limited precision of a numeric estimation system and a multiplicative gain control mechanism.

A Simple, Biologically Plausible Feature Detector for Language Acquisition.

Language has a complex grammatical system we still have to understand computationally and biologically. However, some evolutionarily ancient mechanisms have been repurposed for grammar so that we can use insight from other taxa into possible circuit-level mechanisms of grammar. Drawing upon recent evidence for the importance of disinhibitory circuits across taxa and brain regions, I suggest a simple circuit that explains the acquisition of core grammatical rules used in 85% of the world's languages: grammatical rules based on sameness/difference relations. This circuit acts as a sameness detector. "Different" items are suppressed through inhibition, but presenting two "identical" items leads to inhibition of inhibition. The items are thus propagated for further processing. This sameness detector thus acts as a feature detector for a grammatical rule. I suggest that having a set of feature detectors for elementary grammatical rules might make language acquisition feasible based on relatively simple computational mechanisms.

Duplications and domain-generality.

Although specialized, adaptive behavioral traits are ubiquitous in the animal kingdom, at least in humans, there are considerable debates on whether the mind is primarily characterized by various special-purpose, domain-specific mechanisms or by a few general-purpose, domain-general mechanisms. Drawing from research on artificial language learning, associative learning, serial learning, executive control, and formal linguistics, I argue that neither domain-specificity nor domain-generality provide satisfactory descriptions when considering how cognitive mechanisms are implemented. I suggest that some cognitive mechanisms are domain-bound-they are available in multiple domains (and thus not domain-specific), but not in other domains (and thus not domain-general). Hence, these computations can be performed in many domains but not in others, can be recruited simultaneously by multiple domains, and, across domains, individual abilities with a given computation are relatively uncorrelated. Domain-bound mechanisms have a straightforward evolutionary interpretation: Analogously to the evolution of molecular and morphological structures, cognitive mechanisms can become duplicated over evolution, with independent copies in different domains. This and previous evidence for the importance of duplications for our cognitive abilities call for a revision of the concept domain-generality, suggesting that, in many cases, mechanisms traditionally seen as domain-general might really reflect a collection of local copies of specialized mechanisms. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Interference and memory capacity limitations.

Working memory (WM) is thought to have a fixed and limited capacity. However, the origins of these capacity limitations are debated, and generally attributed to active, attentional processes. Here, we show that the existence of interference among items in memory mathematically guarantees fixed and limited capacity limits under very general conditions, irrespective of any processing assumptions. Assuming that interference (a) increases with the number of interfering items and (b) brings memory performance to chance levels for large numbers of interfering items, capacity limits are a simple function of the relative influence of memorization and interference. In contrast, we show that time-based memory limitations do not lead to fixed memory capacity limitations that are independent of the timing properties of an experiment. We show that interference can mimic both slot-like and continuous resource-like memory limitations, suggesting that these types of memory performance might not be as different as commonly believed. We speculate that slot-like WM limitations might arise from crowding-like phenomena in memory when participants have to retrieve items. Further, based on earlier research on parallel attention and enumeration, we suggest that crowding-like phenomena might be a common reason for the 3 major cognitive capacity limitations. As suggested by Miller (1956) and Cowan (2001), these capacity limitations might arise because of a common reason, even though they likely rely on distinct processes. (PsycINFO Database Record

Transitional probabilities count more than frequency, but might not be used for memorization.

Learners often need to extract recurring items from continuous sequences, in both vision and audition. The best-known example is probably found in word-learning, where listeners have to determine where words start and end in fluent speech. This could be achieved through universal and experience-independent statistical mechanisms, for example by relying on Transitional Probabilities (TPs). Further, these mechanisms might allow learners to store items in memory. However, previous investigations have yielded conflicting evidence as to whether a sensitivity to TPs is diagnostic of the memorization of recurring items. Here, we address this issue in the visual modality. Participants were familiarized with a continuous sequence of visual items (i.e., arbitrary or everyday symbols), and then had to choose between (i) high-TP items that appeared in the sequence, (ii) high-TP items that did not appear in the sequence, and (iii) low-TP items that appeared in the sequence. Items matched in TPs but differing in (chunk) frequency were much harder to discriminate than items differing in TPs (with no significant sensitivity to chunk frequency), and learners preferred unattested high-TP items over attested low-TP items. Contrary to previous claims, these results cannot be explained on the basis of the similarity of the test items. Learners thus weigh within-item TPs higher than the frequency of the chunks, even when the TP differences are relatively subtle. We argue that these results are problematic for distributional clustering mechanisms that analyze continuous sequences, and provide supporting computational results. We suggest that the role of TPs might not be to memorize items per se, but rather to prepare learners to memorize recurring items once they are presented in subsequent learning situations with richer cues.

Learning multiple rules simultaneously: Affixes are more salient than reduplications.

Language learners encounter numerous opportunities to learn regularities, but need to decide which of these regularities to learn, because some are not productive in their native language. Here, we present an account of rule learning based on perceptual and memory primitives (Endress, Dehaene-Lambertz, & Mehler, Cognition, 105(3), 577-614, 2007; Endress, Nespor, & Mehler, Trends in Cognitive Sciences, 13(8), 348-353, 2009), suggesting that learners preferentially learn regularities that are more salient to them, and that the pattern of salience reflects the frequency of language features across languages. We contrast this view with previous artificial grammar learning research, which suggests that infants "choose" the regularities they learn based on rational, Bayesian criteria (Frank & Tenenbaum, Cognition, 120(3), 360-371, 2013; Gerken, Cognition, 98(3)B67-B74, 2006, Cognition, 115(2), 362-366, 2010). In our experiments, adult participants listened to syllable strings starting with a syllable reduplication and always ending with the same "affix" syllable, or to syllable strings starting with this "affix" syllable and ending with the "reduplication". Both affixation and reduplication are frequently used for morphological marking across languages. We find three crucial results. First, participants learned both regularities simultaneously. Second, affixation regularities seemed easier to learn than reduplication regularities. Third, regularities in sequence offsets were easier to learn than regularities at sequence onsets. We show that these results are inconsistent with previous Bayesian rule learning models, but mesh well with the perceptual or memory primitives view. Further, we show that the pattern of salience revealed in our experiments reflects the distribution of regularities across languages. Ease of acquisition might thus be one determinant of the frequency of regularities across languages.

Linguistics, cognitive psychology, and the Now-or-Never bottleneck.

Christiansen & Chater (C&C)'s key premise is that "if linguistic information is not processed rapidly, that information is lost for good" (sect. 1, para. 1). From this "Now-or-Never bottleneck" (NNB), C&C derive "wide-reaching and fundamental implications for language processing, acquisition and change as well as for the structure of language itself" (sect. 2, para. 10). We question both the premise and the consequentiality of its purported implications.

The cost of proactive interference is constant across presentation conditions.

Proactive interference (PI) severely constrains how many items people can remember. For example, Endress and Potter (2014a) presented participants with sequences of everyday objects at 250ms/picture, followed by a yes/no recognition test. They manipulated PI by either using new images on every trial in the unique condition (thus minimizing PI among items), or by re-using images from a limited pool for all trials in the repeated condition (thus maximizing PI among items). In the low-PI unique condition, the probability of remembering an item was essentially independent of the number of memory items, showing no clear memory limitations; more traditional working memory-like memory limitations appeared only in the high-PI repeated condition. Here, we ask whether the effects of PI are modulated by the availability of long-term memory (LTM) and verbal resources. Participants viewed sequences of 21 images, followed by a yes/no recognition test. Items were presented either quickly (250ms/image) or sufficiently slowly (1500ms/image) to produce LTM representations, either with or without verbal suppression. Across conditions, participants performed better in the unique than in the repeated condition, and better for slow than for fast presentations. In contrast, verbal suppression impaired performance only with slow presentations. The relative cost of PI was remarkably constant across conditions: relative to the unique condition, performance in the repeated condition was about 15% lower in all conditions. The cost of PI thus seems to be a function of the relative strength or recency of target items and interfering items, but relatively insensitive to other experimental manipulations.

Words, rules, and mechanisms of language acquisition.

We review recent artificial language learning studies, especially those following Endress and Bonatti (Endress AD, Bonatti LL. Rapid learning of syllable classes from a perceptually continuous speech stream. Cognition 2007, 105:247-299), suggesting that humans can deploy a variety of learning mechanisms to acquire artificial languages. Several experiments provide evidence for multiple learning mechanisms that can be deployed in fluent speech: one mechanism encodes the positions of syllables within words and can be used to extract generalization, while the other registers co-occurrence statistics of syllables and can be used to break a continuum into its components. We review dissociations between these mechanisms and their potential role in language acquisition. We then turn to recent criticisms of the multiple mechanisms hypothesis and show that they are inconsistent with the available data. Our results suggest that artificial and natural language learning is best understood by dissecting the underlying specialized learning abilities, and that these data provide a rare opportunity to link important language phenomena to basic psychological mechanisms. For further resources related to this article, please visit the WIREs website.

Something from (almost) nothing: buildup of object memory from forgettable single fixations.

We can recognize thousands of individual objects in scores of familiar settings, and yet we see most of them only through occasional glances that are quickly forgotten. How do we come to recognize any of these objects? Here, we show that when objects are presented intermittently for durations of single fixations, the originally fleeting memories become gradually stabilized, such that, after just eight separated fixations, recognition memory after half an hour is as good as during an immediate memory test. However, with still shorter presentation durations, memories take more exposures to stabilize. Our results thus suggest that repeated glances suffice to remember the objects of our environment.

How are Bayesian models really used? Reply to Frank (2013).

In response to the proposal that cognitive phenomena might be best understood in terms of cognitive theories (Endress, 2013), Frank (2013) outlined an important research program, suggesting that Bayesian models should be used as rigorous, mathematically attractive implementations of psychological theories. This research program is important and promising. However, I show that it is not followed in practice. I then turn to Frank's defense of the assumption that learners prefer more specific rules (the "size principle"), and show that the results allegedly supporting this assumption do not provide any support for it. Further, I demonstrate that, in contrast to Frank's criticisms, there is no circularity in an account of rule-learning based on "common-sense psychology", and that Frank's other criticisms of this account are unsupported. I conclude that the research program outlined by Frank is important and promising, but needs to be followed in practice. Be that as it might, the rule-learning experiments discussed by Frank are still better explained by simple psychological mechanisms.