In this article, we develop a hierarchical Bayesian model of learning in a general type of artificial language‐learning experiment in which learners are exposed to a mixture of grammars representing the variation present in real learners’ input, particularly at times of language change. The modeling goal is to formalize and quantify hypothesized learning biases. The test case is an experiment (Culbertson, Smolensky, & Legendre, 2012) targeting the learning of word‐order patterns in (...) the nominal domain. The model identifies internal biases of the experimental participants, providing evidence that learners impose (possibly arbitrary) properties on the grammars they learn, potentially resulting in the cross‐linguistic regularities known as typological universals. Learners exposed to mixtures of artificial grammars tended to shift those mixtures in certain ways rather than others; the model reveals how learners’ inferences are systematically affected by specific prior biases. These biases are in line with a typological generalization—Greenberg's Universal 18—which bans a particular word‐order pattern relating nouns, adjectives, and numerals. (shrink)

How do speakers learn the social meaning of different linguistic variants, and what factors influence how likely a particular social–linguistic association is to be learned? It has been argued that the social meaning of more salient variants should be learned faster, and that learners' pre‐existing experience of a variant will influence its salience. In this paper, we report two artificial‐language‐learning experiments investigating this. Each experiment involved two language‐learning stages followed by a test. The first (...) stage introduced the artificial language and trained participants in it, while the second stage added a simple social context using images of cartoon aliens. The first learning stage was intended to establish participants' experience with the artificial language in general and with the distribution of linguistic variants in particular. The second stage, in which linguistic stimuli were accompanied by images of particular aliens, was intended to simulate the acquisition of linguistic variants in a social context. In our first experiment, we manipulated whether a particular linguistic variant, associated with one species of alien in the second learning phase, had been encountered in the first learning phase. In the second experiment, we manipulated whether the variant had been encountered in the same grammatical context. In both cases we predicted that the unexpectedness of a new variant or a new grammatical context for an old variant would increase the variant's salience and facilitate the learning of its social meaning. This is what we found, although in the second experiment, the effect was driven by better learners. Our results suggest that unexpectedness increases the salience of variants and makes their social distribution easier to learn, deepening our understanding of the role of individual language experience in the acquisition of sociolinguistic meaning. (shrink)

Artificial language learning experiments have become an important tool in exploring principles of language and language learning. A persistent question in all of this work, however, is whether ALL engages the linguistic system and whether ALL studies are ecologically valid assessments of natural language ability. In the present study, we considered these questions by examining the relationship between performance in an ALL task and second language learning ability. Participants enrolled in a (...) Spanish language class were evaluated using a number of different measures of Spanish ability and classroom performance, which was compared to IQ and a number of different measures of ALL performance. The results show that success in ALL experiments, particularly more complex artificial languages, correlates positively with indices of L2 learning even after controlling for IQ. These findings provide a key link between studies involving ALL and our understanding of second language learning in the classroom. (shrink)

General principles of human cognition can help to explain why languages are more likely to have certain characteristics than others: structures that are difficult to process or produce will tend to be lost over time. One aspect of cognition that is implicated in language use is working memory—the component of short‐term memory used for temporary storage and manipulation of information. In this study, we consider the relationship between working memory and regularization of linguistic variation. Regularization is a well‐documented process (...) whereby languages become less variable (on some dimension) over time. This process has been argued to be driven by the behavior of individual language users, but the specific mechanism is not agreed upon. Here, we use an artificial language learning experiment to investigate whether limitations in working memory during either language learning or language production drive regularization behavior. We find that taxing working memory during production results in the loss of all types of variation, but the process by which random variation becomes more predictable is better explained by learning biases. A computational model offers a potential explanation for the production effect using a simple self‐priming mechanism. (shrink)

Natural languages contain many layers of sequential structure, from the distribution of phonemes within words to the distribution of phrases within utterances. However, most research modeling language acquisition using artificial languages has focused on only one type of distributional structure at a time. In two experiments, we investigated adult learning of an artificial language that contains dependencies between both adjacent and non-adjacent words. We found that learners rapidly acquired both types of regularities and that the (...) strength of the adjacent statistics influenced learning of both adjacent and non-adjacent dependencies. Additionally, though accuracy was similar for both types of structure, participants’ knowledge of the deterministic non-adjacent dependencies was more explicit than their knowledge of the probabilistic adjacent dependencies. The results are discussed in the context of current theories of statistical learning and language acquisition. (shrink)

Over half a century ago, George Zipf observed that more frequent words tend to be older. Corpus studies since then have confirmed this pattern, with more frequent words being replaced and regularized less often than less frequent words. Two main hypotheses have been proposed to explain this: that frequent words change less because selection against innovation is stronger at higher frequencies, or that they change less because stochastic drift is stronger at lower frequencies. Here, we report the first experimental test (...) of these hypotheses. Participants were tasked with learning a miniature language consisting of two nouns and two plural markers. Nouns occurred at different frequencies and were subjected to treatments that varied drift and selection. Using a model that accounts for participant heterogeneity, we measured the rate of noun regularization, the strength of selection, and the strength of drift in participant responses. Results suggest that drift alone is sufficient to generate the elevated rate of regularization we observed in low‐frequency nouns, adding to a growing body of evidence that drift may be a major driver of language change. (shrink)

It has been suggested that external and/or internal limitations paradoxically may lead to superior learning, that is, the concepts of starting small and less is more (Elman, ; Newport, ). In this paper, we explore the type of incremental ordering during training that might help learning, and what mechanism explains this facilitation. We report four artificial grammar learning experiments with human participants. In Experiments 1a and 1b we found a beneficial effect of starting small using two (...) types of simple recursive grammars: right‐branching and center‐embedding, with recursive embedded clauses in fixed positions and fixed length. This effect was replicated in Experiment 2 (N = 100). In Experiment 3 and 4, we used a more complex center‐embedded grammar with recursive loops in variable positions, producing strings of variable length. When participants were presented an incremental ordering of training stimuli, as in natural language, they were better able to generalize their knowledge of simple units to more complex units when the training input “grew” according to structural complexity, compared to when it “grew” according to string length. Overall, the results suggest that starting small confers an advantage for learning complex center‐embedded structures when the input is organized according to structural complexity. (shrink)

We investigated the emergence of sociolinguistic indexicality using an artificial-language-learning paradigm. Sociolinguistic indexicality involves the association of linguistic variants with nonlinguistic social or contextual features. Any linguistic variant can acquire “constellations” of such indexical meanings, though they also exhibit an ordering, with first-order indices associated with particular speaker groups and higher-order indices targeting stereotypical attributes of those speakers. Much natural-language research has been conducted on this phenomenon, but little experimental work has focused on how indexicality emerges. (...) Here, we present three miniature artificial-language experiments designed to break ground on this question. Results show ready formation of first-order indexicality based on co-occurrence alone, with higher-order indexicality emerging as a result of extension to new speaker groups, modulated by the perceived practical importance of the indexed social feature. (shrink)

Parallel language activation in bilinguals leads to competition between languages. Experience managing this interference may aid novel language learning by improving the ability to suppress competition from known languages. To investigate the effect of bilingualism on the ability to control native-language interference, monolinguals and bilinguals were taught an artificial language designed to elicit between-language competition. Partial activation of interlingual competitors was assessed with eye-tracking and mouse-tracking during a word recognition task in the novel (...) language. Eye-tracking results showed that monolinguals looked at competitors more than bilinguals, and for a longer duration of time. Mouse-tracking results showed that monolinguals’ mouse movements were attracted to native-language competitors, whereas bilinguals overcame competitor interference by increasing the activation of target items. Results suggest that bilinguals manage cross-linguistic interference more effectively than monolinguals. We conclude that language interference can affect lexical retrieval, but bilingualism may reduce this interference by facilitating access to a newly learned language. (shrink)

Linguistic category learning has been shown to be highly sensitive to linear order, and depending on the task, differentially sensitive to the information provided by preceding category markers (premarkers, e.g., gendered articles) or succeeding category markers (postmarkers, e.g., gendered suffixes). Given that numerous systems for marking grammatical categories exist in natural languages, it follows that a better understanding of these findings can shed light on the factors underlying this diversity. In two discriminative learning simulations and an artificial (...) language learning experiment, we identify two factors that modulate linear order effects in linguistic category learning: category structure and the level of abstraction in a category hierarchy. Regarding category structure, we find that postmarking brings an advantage for learning category diagnostic stimulus dimensions, an effect not present when categories are non‐confusable. Regarding levels of abstraction, we find that premarking of super‐ordinate categories (e.g., noun class) facilitates learning of subordinate categories (e.g., nouns). We present detailed simulations using a plausible candidate mechanism for the observed effects, along with a comprehensive analysis of linear order effects within an expectation‐based account of learning. Our findings indicate that linguistic category learning is differentially guided by pre‐ and postmarking, and that the influence of each is modulated by the specific characteristics of a given category system. (shrink)

Linguistic category learning has been shown to be highly sensitive to linear order, and depending on the task, differentially sensitive to the information provided by preceding category markers (premarkers, e.g., gendered articles) or succeeding category markers (postmarkers, e.g., gendered suffixes). Given that numerous systems for marking grammatical categories exist in natural languages, it follows that a better understanding of these findings can shed light on the factors underlying this diversity. In two discriminative learning simulations and an artificial (...) language learning experiment, we identify two factors that modulate linear order effects in linguistic category learning: category structure and the level of abstraction in a category hierarchy. Regarding category structure, we find that postmarking brings an advantage for learning category diagnostic stimulus dimensions, an effect not present when categories are non‐confusable. Regarding levels of abstraction, we find that premarking of super‐ordinate categories (e.g., noun class) facilitates learning of subordinate categories (e.g., nouns). We present detailed simulations using a plausible candidate mechanism for the observed effects, along with a comprehensive analysis of linear order effects within an expectation‐based account of learning. Our findings indicate that linguistic category learning is differentially guided by pre‐ and postmarking, and that the influence of each is modulated by the specific characteristics of a given category system. (shrink)

In language learning, learners engage with their environment, incorporating cues from different sources. However, in lab‐based experiments, using artificial languages, many of the cues and features that are part of real‐world language learning are stripped away. In three experiments, we investigated the role of positive, negative, and mixed feedback on the gradual learning of language‐like statistical regularities within an active guessing game paradigm. In Experiment 1, participants received deterministic feedback (100%), whereas probabilistic (...) feedback (i.e., 75% or 50%) was introduced in Experiment 2. Finally, Experiment 3 explored the impact of mixed probabilistic feedback (33% positive, 33% negative, 33% no feedback). The results showed that cross‐situational learning of words was observed without feedback, but participants were able to learn structural regularities of the miniature language only when feedback was provided. Interestingly, the presence of positive feedback was particularly helpful for the learner, promoting more in‐depth learning of the artificial language. (shrink)

We introduce a set of biologically and computationally motivated design choices for modeling the learning of language, or of other types of sequential, hierarchically structured experience and behavior, and describe an implemented system that conforms to these choices and is capable of unsupervised learning from raw natural-language corpora. Given a stream of linguistic input, our model incrementally learns a grammar that captures its statistical patterns, which can then be used to parse or generate new data. The (...) grammar constructed in this manner takes the form of a directed weighted graph, whose nodes are recursively (hierarchically) defined patterns over the elements of the input stream. We evaluated the model in seventeen experiments, grouped into five studies, which examined, respectively, (a) the generative ability of grammar learned from a corpus of natural language, (b) the characteristics of the learned representation, (c) sequence segmentation and chunking, (d) artificial grammar learning, and (e) certain types of structure dependence. The model's performance largely vindicates our design choices, suggesting that progress in modeling language acquisition can be made on a broad front—ranging from issues of generativity to the replication of human experimental findings—by bringing biological and computational considerations, as well as lessons from prior efforts, to bear on the modeling approach. (shrink)

We introduce a set of biologically and computationally motivated design choices for modeling the learning of language, or of other types of sequential, hierarchically structured experience and behavior, and describe an implemented system that conforms to these choices and is capable of unsupervised learning from raw natural‐language corpora. Given a stream of linguistic input, our model incrementally learns a grammar that captures its statistical patterns, which can then be used to parse or generate new data. The (...) grammar constructed in this manner takes the form of a directed weighted graph, whose nodes are recursively (hierarchically) defined patterns over the elements of the input stream. We evaluated the model in seventeen experiments, grouped into five studies, which examined, respectively, (a) the generative ability of grammar learned from a corpus of natural language, (b) the characteristics of the learned representation, (c) sequence segmentation and chunking, (d) artificial grammar learning, and (e) certain types of structure dependence. The model's performance largely vindicates our design choices, suggesting that progress in modeling language acquisition can be made on a broad front—ranging from issues of generativity to the replication of human experimental findings—by bringing biological and computational considerations, as well as lessons from prior efforts, to bear on the modeling approach. (shrink)

This study investigates the joint influences of three factors on the discovery of new word‐like units in a continuous artificial speech stream: the statistical structure of the ongoing input, the initial word‐likeness of parts of the speech flow, and the contextual information provided by the earlier emergence of other word‐like units. Results of an experiment conducted with adult participants show that these sources of information have strong and interactive influences on word discovery. The authors then examine the ability (...) of different models of word segmentation to account for these results. PARSER (Perruchet & Vinter, 1998) is compared to the view that word segmentation relies on the exploitation of transitional probabilities between successive syllables, and with the models based on the Minimum Description Length principle, such as INCDROP. The authors submit arguments suggesting that PARSER has the advantage of accounting for the whole pattern of data without ad‐hoc modifications, while relying exclusively on general‐purpose learning principles. This study strengthens the growing notion that nonspecific cognitive processes, mainly based on associative learning and memory principles, are able to account for a larger part of early language acquisition than previously assumed. (shrink)

Over the last decade, iterated learning studies have provided compelling evidence for the claim that linguistic structure can emerge from non‐structured input, through the process of transmission. However, it is unclear whether individuals differ in their tendency to add structure, an issue with implications for understanding who are the agents of change. Here, we identify and test two contrasting predictions: The first sees learning as a pre‐requisite for structure addition, and predicts a positive correlation between learning accuracy (...) and structure addition, whereas the second maintains that it is those learners who struggle with learning and reproducing their input who add structure to it. This prediction is hard to test in standard iterated learning paradigms since each learner is exposed to a different input, and since structure and accuracy are computed using the same test items. Here, we test these contrasting predictions in two experiments using a one‐generation artificial language learning paradigm designed to provide independent measures of learning accuracy and structure addition. Adults (N = 48 in each study) were exposed to a semi‐regular language (with probabilistic structure) and had to learn it: Learning was assessed using seen items, whereas structure addition was calculated over unseen items. In both studies, we find a strong positive correlation between individuals' ability to learn the language and their tendency to add structure to it: Better learners also produced more structured languages. These findings suggest a strong link between learning and generalization. We discuss the implications of these findings for iterated language models and theories of language change more generally. (shrink)

It is a reasonable assumption that universal properties of natural languages are not accidental. They occur either because they are underwritten by genetic code, because they assist in language processing or language learning, or due to some combination of the two. In this paper we investigate one such language universal: the suffixing preference across the world’s languages, whereby inflections tend to be added to the end of words. A corpus analysis of child‐directed speech in English found (...) that suffixes were more accurate at cuing the grammatical category of the root word than were prefixes. An artificial language experiment found that there was a learning advantage for suffixes over prefixes in terms of grammatical categorization within an artificial language. The results are consistent with an account of language universals that originate in general purpose learning mechanisms. (shrink)

Recent research has begun to explore individual differences in statistical learning, and how those differences may be related to other cognitive abilities, particularly their effects on language learning. In this research, we explored a different type of relationship between language learning and statistical learning: the possibility that learning a new language may also influence statistical learning by changing the regularities to which learners are sensitive. We tested two groups of participants, Mandarin (...) Learners and Naïve Controls, at two time points, 6 months apart. At each time point, participants performed two different statistical learning tasks: an artificial tonal language statistical learning task and a visual statistical learning task. Only the Mandarin-learning group showed significant improvement on the linguistic task, whereas both groups improved equally on the visual task. These results support the view that there are multiple influences on statistical learning. Domain-relevant experiences may affect the regularities that learners can discover when presented with novel stimuli. (shrink)

Languages spoken in larger populations are relatively simple. A possible explanation for this is that languages with a greater number of speakers tend to also be those with higher proportions of non‐native speakers, who may simplify language during learning. We assess this explanation for the negative correlation between population size and linguistic complexity in three experiments, using artificial language learning techniques to investigate both the simplifications made by individual adult learners and the potential for such (...) simplifications to influence group‐level language characteristics. In Experiment 1, we show that individual adult learners trained on a morphologically complex miniature language simplify its morphology. In Experiment 2, we explore how these simplifications may then propagate through subsequent learning. We use the languages produced by the participants of Experiment 1 as the input for a second set of learners, manipulating (a) the proportion of their input which is simplified and (b) the number of speakers they receive their input from. We find, contrary to expectations, that mixing the input from multiple speakers nullifies the simplifications introduced by individuals in Experiment 1; simplifications at the individual level do not result in simplification of the population's language. In Experiment 3, we focus on language use as a mechanism for simplification, exploring the consequences of the interaction between individuals differing in their linguistic competence (as native and non‐native speakers might). We find that speakers who acquire a more complex language than their partner simplify their language during interaction. We ultimately conclude that adult learning can result in languages spoken by more people having simpler morphology, but that idiosyncratic simplifications by non‐natives do not offer a complete explanation in themselves; accommodation—by comparatively competent non‐natives to less competent speakers, or by native speakers to non‐natives—may be a key linking mechanism. (shrink)

Human language is a salient example of a neurocognitive system that is specialized to process complex dependencies between sensory events distributed in time, yet how this system evolved and specialized remains unclear. Artificial Grammar Learning (AGL) studies have generated a wealth of insights into how human adults and infants process different types of sequencing dependencies of varying complexity. The AGL paradigm has also been adopted to examine the sequence processing abilities of nonhuman animals. We critically evaluate this (...) growing literature in species ranging from mammals (primates and rats) to birds (pigeons, songbirds, and parrots) considering also cross‐species comparisons. The findings are contrasted with seminal studies in human infants that motivated the work in nonhuman animals. This synopsis identifies advances in knowledge and where uncertainty remains regarding the various strategies that nonhuman animals can adopt for processing sequencing dependencies. The paucity of evidence in the few species studied to date and the need for follow‐up experiments indicate that we do not yet understand the limits of animal sequence processing capacities and thereby the evolutionary pattern. This vibrant, yet still budding, field of research carries substantial promise for advancing knowledge on animal abilities, cognitive substrates, and language evolution. (shrink)

Learners exposed to an artificial language recognize its abstract structural regularities when instantiated in a novel vocabulary (e.g., Gómez, Gerken, & Schvaneveldt, 2000; Tunney & Altmann, 2001). We asked whether such sensitivity accelerates subsequent learning, and enables acquisition of more complex structure. In Experiment 1, pre-exposure to a category-induction language of the form aX bY sped subsequent learning when the language is instantiated in a different vocabulary. In Experiment 2, while naíve learners (...) did not acquire an acX bcY language, in which aX and bY co-occurrence regularities were separated by a c-element, prior experience with an aX bY language provided some benefit. In Experiment 3 we replicated this finding with a 24-hour delay between learning phases, and controlled for prior experience with the aX bY language's prosodic and phonological characteristics. These findings suggest that learners, and the structure they can acquire, change as a function of experience. (shrink)

A word often expresses many different morphological functions. Which part of a word contributes to which part of the overall meaning is not always clear, which raises the question as to how such functions are learned. While linguistic studies tacitly assume the co-occurrence of cues and outcomes to suffice in learning these functions (Baer-Henney, Kügler, & van de Vijver, 2015; Baer-Henney & van de Vijver, 2012), error-driven learning suggests that contingency rather than contiguity is crucial (Nixon, 2020; Ramscar, (...) Yarlett, Dye, Denny, & Thorpe, 2010). In error-driven learning, cues gain association strength if they predict a certain outcome, and they lose strength if the outcome is absent. This reduction of association strength is called unlearning. So far, it is unclear if such unlearning has consequences for cue–outcome associations beyond the ones that get reduced. To test for such consequences of unlearning, we taught participants morphophonological patterns in an artificial language learning experiment. In one block, the cues to two morphological outcomes—plural and diminutive—co-occurred within the same word forms. In another block, a single cue to only one of these two outcomes was presented in a different set of word forms. We wanted to find out, if participants unlearn this cue's association with the outcome that is not predicted by the cue alone, and if this allows the absent cue to be associated with the absent outcome. Our results show that if unlearning was possible, participants learned that the absent cue predicts the absent outcome better than if no unlearning was possible. This effect was stronger if the unlearned cue was more salient. This shows that unlearning takes place even if no alternative cues to an absent outcome are provided, which highlights that learners take both positive and negative evidence into account—as predicted by domain general error-driven learning. (shrink)

Computational models have shown that purely statistical knowledge about words’ linguistic contexts is sufficient to learn many properties of words, including syntactic and semantic category. For example, models can infer that “postman” and “mailman” are semantically similar because they have quantitatively similar patterns of association with other words. In contrast to these computational results, artificial language learning experiments suggest that distributional statistics alone do not facilitate learning of linguistic categories. However, experiments in this paradigm expose participants (...) to entirely novel words, whereas real language learners encounter input that contains some known words that are semantically organized. In three experiments, we show that the presence of familiar semantic reference points facilitates distributional learning and this effect crucially depends both on the presence of known words and the adherence of these known words to some semantic organization. (shrink)

This paper studies the relationship of Artificial Intelligence to the study of language and the representation of the underlying knowledge which supports the comprehension process. It develops the view that intelligence is based on the ability to use large amounts of diverse kinds of knowledge in procedural ways, rather than on the possession of a few general and uniform principles. The paper also provides a unifying thread to a variety of recent approaches to natural language comprehension. We (...) conclude with a brief discussion of how Artificial Intelligence may have a radical impact on education if the principles which it utilizes to explore the representation and use of knowledge are made available to the student to use in his own learning experiences. (shrink)

To what degree can language be acquired from linguistic input alone? This question has vexed scholars for millennia and is still a major focus of debate in the cognitive science of language. The complexity of human language has hampered progress because studies of language–especially those involving computational modeling–have only been able to deal with small fragments of our linguistic skills. We suggest that the most recent generation of Large Language Models (LLMs) might finally provide the (...) computational tools to determine empirically how much of the human language ability can be acquired from linguistic experience. LLMs are sophisticated deep learning architectures trained on vast amounts of natural language data, enabling them to perform an impressive range of linguistic tasks. We argue that, despite their clear semantic and pragmatic limitations, LLMs have already demonstrated that human-like grammatical language can be acquired without the need for a built-in grammar. Thus, while there is still much to learn about how humans acquire and use language, LLMs provide full-fledged computational models for cognitive scientists to empirically evaluate just how far statistical learning might take us in explaining the full complexity of human language. (shrink)

According to classical arguments, language learning is both facilitated and constrained by cognitive biases. These biases are reflected in linguistic typology—the distribution of linguistic patterns across the world's languages—and can be probed with artificial grammar experiments on child and adult learners. Beginning with a widely successful approach to typology (Optimality Theory), and adapting techniques from computational approaches to statistical learning, we develop a Bayesian model of cognitive biases and show that it accounts for the detailed pattern (...) of results of artificial grammar experiments on noun-phrase word order (Culbertson, Smolensky, & Legendre, 2012). Our proposal has several novel properties that distinguish it from prior work in the domains of linguistic theory, computational cognitive science, and machine learning. This study illustrates how ideas from these domains can be synthesized into a model of language learning in which biases range in strength from hard (absolute) to soft (statistical), and in which language-specific and domain-general biases combine to account for data from the macro-level scale of typological distribution to the micro-level scale of learning by individuals. (shrink)

Studies of language evolution in the lab have used the iterated learning paradigm to show how linguistic structure emerges through cultural transmission—repeated cycles of learning and use across generations of speakers. However, agent-based simulations suggest that prior biases crucially impact the outcome of cultural transmission. Here, we explored this notion through an iterated learning study of English-French bilingual adults (mostly sequential bilinguals dominant in English). Each participant learned two unstructured artificial languages in a counterbalanced fashion, (...) one resembling English, another resembling French at the phono-orthographic level. The output of each participant was passed down to the next participant, forming diffusion chains of 10 generations per language. We hypothesized that artificial languages would become easier to learn and exhibit greater structure when they were aligned with participants’ bilingual experience (i.e., English languages being easier to learn overall), or as a function of practice (i.e., languages learned second being easier to learn overall). Instead, we found that English-like languages became more structured over generations, but only when they were learned first. In contrast, French-like languages became more structured regardless of the order of learning, suggesting the presence of an asymmetric switch cost during artificial language learning. Moreover, individual differences in language usage modulated the amount of structure produced by the participants. Overall, these data suggest that bilingual experience impacts how novel languages are learned at an individual level, which can then scale up to cultural transmission of novel language at a group level. (shrink)

Purpose: Yerkish is an artificial language created in 1971 for the specific purpose of exploring the linguistic potential of nonhuman primates. The aim of this paper is to remind the research community of some important issues and concepts related to Yerkish that seem to have been forgotten or appear to be distorted. These are, particularly, its success, its promising aspects for future research and last but not least that it was Ernst von Glasersfeld who invented Yerkish: he coined (...) the term “lexigrams,” created the first 120 of them and designed the grammar that regulated their combination. Design: The first part of this paper begins with a short outline of the context in which the Yerkish language originated: the original LANA project. It continues by presenting the language itself in more detail: first, its design, focusing on its “lexigrams” and its “correlational” grammar, and then its use by the chimpanzee Lana in formulating sentences. The second part gives a brief introduction to the foundation of Yerkish in Silvio Ceccato’s Operational Methodology, particularly his idea of the correlational structure of thought and concludes with the main insights that can be derived from the Yerkish experiment seen in the light of Operational Methodology. Findings: Lana’s success in language learning and the success of Yerkish during the past decades are probably due to the characteristics of Yerkish, particularly its foundation in operational methodology. The operation of correlation could be what constitutes thinking in a chimpanzee and an attentional system could be what delivers the mental content that correlation assembles into triads and networks. Research implications: Since no other assessment or explanation of Lana’s performances has considered these foundational issues, a new research project or program should validate the above-mentioned hypotheses, particularly the correlational structure of chimpanzee thinking. (shrink)

How do listeners learn about the statistical regularities underlying musical harmony? In traditional Western music, certain chords predict the occurrence of other chords: Given a particular chord, not all chords are equally likely to follow. In Experiments 1 and 2, we investigated whether adults make use of statistical information when learning new musical structures. Listeners were exposed to a novel musical system containing phrases generated using an artificial grammar. This new system contained statistical structure quite different from Western (...) tonal music. Our results suggest that learners take advantage of the statistical patterning of chords to acquire new musical structures, similar to learning processes previously observed for language learning. (shrink)

Studies of AGL have frequently used training and test stimuli that might provide multiple cues for learning, raising the question what subjects have actually learned. Using a selected subset of studies on humans and non‐human animals, Trotter et al. demonstrate how a meta‐analysis can be used to identify relevant experimental variables, providing a first step in asssessing the relative contribution of design features of grammars as well as of species‐specific effects on AGL.

Ramscar, Yarlett, Dye, Denny, and Thorpe (2010) showed how, consistent with the predictions of error‐driven learning models, the order in which stimuli are presented in training can affect category learning. Specifically, learners exposed to artificial language input where objects preceded their labels learned the discriminating features of categories better than learners exposed to input where labels preceded objects. We sought to replicate this finding in two online experiments employing the same tests used originally: A four pictures (...) test (match a label to one of four pictures) and a four labels test (match a picture to one of four labels). In our study, only findings from the four pictures test were consistent with the original result. Additionally, the effect sizes observed were smaller, and participants over‐generalized high‐frequency category labels more than in the original study. We suggest that although Ramscar, Yarlett, Dye, Denny, and Thorpe (2010) feature‐label order predictions were derived from error‐driven learning, they failed to consider that this mechanism also predicts that performance in any training paradigm must inevitably be influenced by participant prior experience. We consider our findings in light of these factors, and discuss implications for the generalizability and replication of training studies. (shrink)

In a society where mobility, globalization and contact with people from other cultures have become its basic descriptors, the enhancement of plurilingualism and intercultural understanding seem to be of the utmost concern. From a Positive Psychology Perspective, agency is the human capacity to affect other people positively or negatively through their actions. This agentic vision can be related to mediation, a concept rooted in the socio-cultural learning theory where social interaction is considered a fundamental cornerstone in the development of (...) cognition. Social interactions in the language learning setting may be facilitated through musical activities due to its social bonding effect. This paper tries to offer insights on how a musical-mediated experience in language learning may develop students’ interpersonal and collaborative competences to become active members of a more inclusive society. Mediation, considered as a paradigm shift in the foreign language classroom and the different out-of-class language learning possibilities, could also provide an environment where learners maximize their emotional intelligence. Our paper focuses on this paradigm shift spearheaded by the CEFR Companion Volume and the considerable repercussions it is bound to have for foreign language didactics, as cooperative tasks become central to foreign language learning. We hypothesize that mediated language learning experiences imply a socioemotional change in learners, focusing on the others, on their needs and interests, by trying to help them understand texts, concepts or facilitating communication with their peers. An intervention with a musical-mediated language learning experience was designed and implemented in the classroom. A self assessment scale with mediation descriptors and the socioemotional expertise scale (SEE) were administered. Results show that students become more mindful of their strengths, of their capacity for collaborating and team working. This leads to more awareness of their mediation skills. Students’ mediation skills correlate significantly with their socio-emotional skills, more concretely with their expressivity. The implementation of a musical-mediated experience also promoted tolerance and enhanced learners’ intrinsic motivation for language learning at the same time as they acknowledged their diversity. (shrink)

Analyzing and evaluating legal case reports are labor-intensive tasks for judges and lawyers, who usually base their decisions on report abstracts, legal principles, and commonsense reasoning. Thus, summarizing legal documents is time-consuming and requires excellent human expertise. Moreover, public legal corpora of specific languages are almost unavailable. This paper proposes a transfer learning approach with extractive and abstractive techniques to cope with the lack of labeled legal summarization datasets, namely a low-resource scenario. In particular, we conducted extensive multi- and (...) cross-language experiments. The proposed work outperforms the state-of-the-art results of extractive summarization on the Australian Legal Case Reports dataset and sets a new baseline for abstractive summarization. Finally, syntactic and semantic metrics assessments have been carried out to evaluate the accuracy and the factual consistency of the machine-generated legal summaries. (shrink)

The cognition of music, like that of language, is partly rooted in enculturative processes of implicit and incidental learning. Musicians and nonmusicians alike are commonly found to possess detailed implicit knowledge of musical structure which is acquired incidentally through interaction with large samples of music. This paper reports an experiment combining the methodology of artificial grammar learning with musical acquisition of melodic structure. Participants acquired knowledge of grammatical melodic structures under incidental learning conditions in (...) both experimental and untrained control conditions. Subsequent analysis indicates a large effect of unsupervised online learning in the experimental and control group throughout the course of the testing phase suggesting an effective ongoing learning process. Musicians did not outperform nonmusicians, indicating that musical expertise is not advantageous for the learning of a new, unfamiliar melodic system. Confidence ratings suggest that participants became aware of the knowledge guiding their classification performance despite the incidental learning conditions. (shrink)

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. Statistical learning, 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 while 5-year-old children and young adults were exposed to a speech stream made of the repetition of eight three-syllable nonsense words presenting different levels of predictability to mimic closely what occurs in natural languages and to get new insights into the changes that the mechanisms underlying auditory statistical learning 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. (shrink)

The article compares the research programs of teaching symbolic language to chimpanzees, pointing on the dichotomy between artificial language vs. ASL, and the dichotomy between researchers who decided to establish emotional relationships between themselves and the apes, and those who have seen apes as instrumental devices. It is concluded that the experiments with the most interesting results have been both with artificial language and ASL, but with strong affiliation between researchers and animal involved in the (...) experiments. The experiments on talking apes are not so much experiments in psycholinguistics (how far can animal learn human language) but wonderful experiments in the communities of communication between human beings and great apes. (shrink)

A prerequisite for spoken language learning is segmenting continuous speech into words. Amongst many possible cues to identify word boundaries, listeners can use both transitional probabilities between syllables and various prosodic cues. However, the relative importance of these cues remains unclear, and previous experiments have not directly compared the effects of contrasting multiple prosodic cues. We used artificial language learning experiments, where native German speaking participants extracted meaningless trisyllabic “words” from a continuous speech stream, to (...) evaluate these factors. We compared a baseline condition to five test conditions, in which word-final syllables were either followed by a pause, lengthened, shortened, changed to a lower pitch, or changed to a higher pitch. To evaluate robustness and generality we used three tasks varying in difficulty. Overall, pauses and final lengthening were perceived as converging with the statistical cues and facilitated speech segmentation, with pauses helping most. Final-syllable shortening hindered baseline speech segmentation, indicating that when cues conflict, prosodic cues can override statistical cues. Surprisingly, pitch cues had little effect, suggesting that duration may be more relevant for speech segmentation than pitch in our study context. We discuss our findings with regard to the contribution to speech segmentation of language-universal boundary cues vs. language-specific stress patterns. (shrink)

In language, recombination of a discrete set of meaningless building blocks forms an unlimited set of possible utterances. How such combinatorial structure emerged in the evolution of human language is increasingly being studied. It has been shown that it can emerge when languages culturally evolve and adapt to human cognitive biases. How the emergence of combinatorial structure interacts with the existence of holistic iconic form-meaning mappings in a language is still unknown. The experiment presented in this (...) paper studies the role of iconicity and human cognitive learning biases in the emergence of combinatorial structure in artificial whistled languages. Participants learned and reproduced whistled words for novel objects with the use of a slide whistle. Their reproductions were used as input for the next participant, to create transmission chains and simulate cultural transmission. Two conditions were studied: one in which the persistence of iconic form-meaning mappings was possible and one in which this was experimentally made impossible. In both conditions, cultural transmission caused the whistled languages to become more learnable and more structured, but this process was slightly delayed in the first condition. Our findings help to gain insight into when and how words may lose their iconic origins when they become part of an organized linguistic system. (shrink)

Recent developments in artificial intelligence based on neural nets—deep learning and large language models which together I refer to as NEWAI—have resulted in startling improvements in language handling and the potential to keep up with changing human knowledge by learning from the internet. Nevertheless, examples such as ChatGPT, which is a ‘large language model’, have proved to have no moral compass: they answer queries with fabrications with the same fluency as they provide facts. I (...) try to explain why this is, basing the argument on the sociology of knowledge, particularly social studies of science, notably ‘studies of expertise and experience’ and the ‘fractal model’ of society. Learning from the internet is not the same as socialisation: NEWAI has no primary socialisation such as provides the foundations of human moral understanding. Instead, large language models are retrospectively socialised by human intervention in an attempt to align them with societally accepted ethics. Perhaps, as technology advances, large language models could come to understand speech and recognise objects sufficiently well to acquire the equivalent of primary socialisation. In the meantime, we must be vigilant about who is socialising them and be aware of the danger of their socialising us to align with them rather than vice-versa, an eventuality that would lead to the further erosion of the distinction between the true and the false giving further support to populism and fascism. (shrink)

Traditional flat-structured bigram and trigram models of phonotactics are useful because they capture a large number of facts about phonological processes. Additionally, these models predict that local interactions should be easier to learn than long-distance ones because long-distance dependencies are difficult to capture with these models. Long-distance phonotactic patterns have been observed by linguists in many languages, who have proposed different kinds of models, including feature-based bigram and trigram models, as well as precedence models. Contrary to flat-structured bigram and trigram (...) models, these alternatives capture unbounded dependencies because at an abstract level of representation, the relevant elements are locally dependent, even if they are not adjacent at the observable level. Using an artificial grammar learning paradigm, we provide additional support for these alternative models of phonotactics. Participants in two experiments were exposed to a long-distance consonant-harmony pattern in which the first consonant of a five-syllable word was [s] or [∫] (“sh”) and triggered a suffix that was either [-su] or [-∫u] depending on the sibilant quality of this first consonant. Participants learned this pattern, despite the large distance between the trigger and the target, suggesting that when participants learn long-distance phonological patterns, that pattern is learned without specific reference to distance. (shrink)

Supporters of eliminative connectionism have argued for a pattern association-based explanation of language learning and language processing. They deny that explicit rules and symbolic representations play any role in language processing and cognition in general. Their argument is based to a large extent on two artificial neural network (ANN) models that are claimed to be able to learn the past tenses of English verbs (Rumelhart & McClelland, 1986, Parallel distributed processing, Vol. 2, Cambridge, MA: MIT (...) Press; MacWhinney & Leinbach, 1991, Cognition, 40, 121-157). In this article we critically review Rumelhart and McClelland's as well as MacWhinney and Leinbach's ANN models and conclude that they do not succeed in the assigned task of learning the past tenses of English verbs. In order to answer their challenge to the symbolic processing approach, we present our symbolic pattern associator (SPA)-a general-purpose pattern associator that can learn to associate arbitrary discrete patterns. We carried out several experiments with the SPA using the same set of verbs that was used in MacWhinney and Leinbach's simulation with more realistic training and testing procedures. The SPA outperformed the connectionist models by a wide margin in the accuracy of learning, and successful inductive generalizations to unseen verbs. Our SPA has very natural and psychologically realistic explanations to many psychological effects such as U-shaped learning curve, and is much closer to human subjects in predicting past tense of the pseudo-verbs. In contrast to ANNs, whose internal representations are entirely opaque, the SPA can represent the acquired knowledge in the form of production rules that allow for further higher-level processing and integration, resulting in linguistically realistic associative templates for irregular verbs and production rules for regular verbs. In the light of these findings, we conclude that eliminative connectionists' vision of cognition as simple pattern association and pattern recognition without symbolic representation is inadequate. Pattern association as such does not imply rule-less or cue-based models of language acquisition or of human learning in general. (shrink)

In this research, patent prosecution is conceptualized as a system of reinforcement learning from human feedback. The objective of the system is to increase the likelihood for a language model to generate patent claims that have a higher chance of being granted. To showcase the controllability of the language model, the system learns from granted patents and pre-grant applications with different rewards. The status of “granted” and “pre-grant” are perceived as labeled human feedback implicitly. In addition, specific (...) to patent drafting, the experiments in this research demonstrate the model’s capability to learn from adjusting claim length and inclusion of limiting terms for narrowing claim scope. As proof of concept, the experiments focus on claim ones only and the training data originates from a patent dataset tailored specifically for artificial intelligence. Although the available human feedback in patent prosecution are limited and the quality of generated patent text requires improvement, the experiments following the 3-stage reinforcement learning from human feedback have demonstrated that generative language models are capable of reflecting the human feedback or intent in patent prosecution. To enhance the usability of language models, the implementation in this research utilizes modern techniques that enable execution on a single consumer-grade GPU. The demonstrated proof of concept, which reduces hardware requirements, will prove valuable in the future as more human feedback in patent prosecution become available for broader use, either within patent offices or in the public domain. (shrink)

Eye-tracking technology has become a powerful tool for biomedical-related applications due to its simplicity of operation and low requirements on patient language skills. This study aims to use the machine-learning models and deep-learning networks to identify key features of eye movements in Alzheimer's Disease (AD) under specific visual tasks, thereby facilitating computer-aided diagnosis of AD. Firstly, a three-dimensional (3D) visuospatial memory task is designed to provide participants with visual stimuli while their eye-movement data are recorded and used (...) to build an eye-tracking dataset. Then, we propose a novel deep-learning-based model for identifying patients with Alzheimer's Disease (PwAD) and healthy controls (HCs) based on the collected eye-movement data. The proposed model utilizes a nested autoencoder network to extract the eye-movement features from the generated fixation heatmaps and a weight adaptive network layer for the feature fusion, which can preserve as much useful information as possible for the final binary classification. To fully verify the performance of the proposed model, we also design two types of models based on traditional machine-learning and typical deep-learning for comparison. Furthermore, we have also done ablation experiments to verify the effectiveness of each module of the proposed network. Finally, these models are evaluated by four-fold cross-validation on the built eye-tracking dataset. The proposed model shows 85% average accuracy in AD recognition, outperforming machine-learning methods and other typical deep-learning networks. (shrink)

Mueller et al. discuss the role of acoustic cues in detecting language structure more generally. Across languages, there are clear links between acoustic cues and syntactic structure. They show that AGL experiments implementing analogous links demonstrate that prosodic cues, as well as various auditory biases, facilitate the learning of structural rules. Some of these biases, e.g. for auditory grouping, are also present in other species.

Information technology is so ubiquitous and AI’s progress so inspiring that also legal professionals experience its benefits and have high expectations. At the same time, the powers of AI have been rising so strongly that it is no longer obvious that AI applications (whether in the law or elsewhere) help promoting a good society; in fact they are sometimes harmful. Hence many argue that safeguards are needed for AI to be trustworthy, social, responsible, humane, ethical. In short: AI should be (...) good for us. But how to establish proper safeguards for AI? One strong answer readily available is: consider the problems and solutions studied in AI & Law. AI & Law has worked on the design of social, explainable, responsible AI aligned with human values for decades already, AI & Law addresses the hardest problems across the breadth of AI (in reasoning, knowledge, learning and language), and AI & Law inspires new solutions (argumentation, schemes and norms, rules and cases, interpretation). It is argued that the study of AI as Law supports the development of an AI that is good for us, making AI & Law more relevant than ever. (shrink)