Dynamic Motion and Human Agents Facilitate Visual Nonadjacent Dependency Learning.

Many events that humans and other species experience contain regularities in which certain elements within an event predict certain others. While some of these regularities involve tracking the co-occurrences between temporally adjacent stimuli, others involve tracking the co-occurrences between temporally distant stimuli (i.e., nonadjacent dependencies, NADs). Prior research shows robust learning of adjacent dependencies in humans and other species, whereas learning NADs is more difficult, and often requires support from properties of the stimulus to help learners notice the NADs. Here, we report on seven experiments that examined NAD learning from various types of visual stimuli, exploring the effects of dynamic motion on adults' NAD learning from visual sequences involving human and nonhuman agents. We tested adults' NAD learning from visual sequences of human actions, object transformations, static images of human postures, and static images of an object in different postures. We found that dynamic motion aids the acquisition of NADs. We also found that learning NADs in sequences involving human agents is more robust compared to sequences involving nonhuman objects. We propose that dynamic motion and human agents both independently result in richer representations that provide a stronger signal for NAD learning.

The importance of modeling pragmatic syntactic bootstrapping.

Syntactic bootstrapping is based on the premise that there are probabilistic correspondences between the syntactic structure in which a word occurs and the word's meaning, and that such links hold, with some degree of generality, cross-linguistically. The procedure has been extensively discussed with respect to verbs, where it has been proposed as a mechanism for constraining the massive ambiguity that arises when inferring the meaning of a verb that is used to describe an event (Fisher, Hall, Rakowitz & Gleitman, 1994; Gleitman, 1990; Gleitman, Cassidy, Nappa, Papafragou & Trueswell, 2005). In her keynote paper (Hacquard, 2022), Hacquard focuses on classes of verbs for which inferences about meaning are arguably even harder, because they involve concepts that have no observable counterparts: these are attitude verbs such as think and want, and modals such as must and can. She walks us through, in meticulous detail, the limits of a purely syntactic bootstrapping mechanism, and she describes how augmenting syntactic information with pragmatic information, via pragmatic syntactic bootstrapping (Hacquard, 2022; Hacquard & Lidz, 2019), might address these limitations. The proposal is exciting, and the detail with which Hacquard works through these examples is impressive; she supports her arguments with behavioral experiments, corpus analyses, and two very targeted computational analyses. In this commentary I suggest that Hacquard's proposal is laid out in sufficient detail such that a comprehensive computational modeling effort would be fruitful for evaluating and further developing her account.

Learning non-adjacent rules and non-adjacent dependencies from human actions in 9-month-old infants.

Seven month old infants can learn simple repetition patterns, such as we-fo-we, and generalize the rules to sequences of new syllables, such as ga-ti-ga. However, repetition rule learning in visual sequences seems more challenging, leading some researchers to claim that this type of rule learning applies preferentially to communicative stimuli. Here we demonstrate that 9-month-old infants can learn repetition rules in sequences of non-communicative dynamic human actions. We also show that when primed with these non-adjacent repetition patterns, infants can learn non-adjacent dependencies that involve memorizing the dependencies between specific human actions-patterns that prior research has shown to be difficult for infants in the visual domain and in speech. We discuss several possible mechanisms that account for the apparent advantage stimuli involving human action sequences has over other kinds of stimuli in supporting non-adjacent dependency learning. We also discuss possible implications for theories of language acquisition.

Non-adjacent Dependency Learning in Humans and Other Animals.

Learning and processing natural language requires the ability to track syntactic relationships between words and phrases in a sentence, which are often separated by intervening material. These nonadjacent dependencies can be studied using artificial grammar learning paradigms and structured sequence processing tasks. These approaches have been used to demonstrate that human adults, infants and some nonhuman animals are able to detect and learn dependencies between nonadjacent elements within a sequence. However, learning nonadjacent dependencies appears to be more cognitively demanding than detecting dependencies between adjacent elements, and only occurs in certain circumstances. In this review, we discuss different types of nonadjacent dependencies in language and in artificial grammar learning experiments, and how these differences might impact learning. We summarize different types of perceptual cues that facilitate learning, by highlighting the relationship between dependent elements bringing them closer together either physically, attentionally, or perceptually. Finally, we review artificial grammar learning experiments in human adults, infants, and nonhuman animals, and discuss how similarities and differences observed across these groups can provide insights into how language is learned across development and how these language-related abilities might have evolved.

Top-down grouping affects adjacent dependency learning.

A large body of research has demonstrated that humans attend to adjacent co-occurrence statistics when processing sequential information, and bottom-up prosodic information can influence learning. In this study, we investigated how top-down grouping cues can influence statistical learning. Specifically, we presented English sentences that were structurally equivalent to each other, which induced top-down expectations of grouping in the artificial language sequences that immediately followed. We show that adjacent dependencies in the artificial language are learnable when these entrained boundaries bracket the adjacent dependencies into the same sub-sequence, but are not learnable when the elements cross an induced boundary, even though that boundary is not present in the bottom-up sensory input. We argue that when there is top-down bracketing information in the learning sequence, statistical learning takes place for elements bracketed within sub-sequences rather than all the elements in the continuous sequence. This limits the amount of linguistic computations that need to be performed, providing a domain over which statistical learning can operate.

Successfully learning non-adjacent dependencies in a continuous artificial language stream.

Much of the statistical learning literature has focused on adjacent dependency learning, which has shown that learners are capable of extracting adjacent statistics from continuous language streams. In contrast, studies on non-adjacent dependency learning have mixed results, with some showing success and others failure. We review the literature on non-adjacent dependency learning and examine various theories proposed to account for these results, including the proposed necessity of the presence of pauses in the learning stream, or proposals regarding competition between adjacent and non-adjacent dependency learning such that high variability of middle elements is beneficial to learning. Here we challenge those accounts by showing successful learning of non-adjacent dependencies under conditions that are inconsistent with predictions of previous theories. We show that non-adjacent dependencies are learnable without pauses at dependency edges in a variety of artificial language designs. Moreover, we find no evidence of a relationship between non-adjacent dependency learning and the robustness of adjacent statistics. We demonstrate that our two-step statistical learning model can account for all of our non-adjacent dependency learning results, and provides a unified learning account of adjacent and non-adjacent dependency learning. Finally, we discussed the theoretical implications of our findings for natural language acquisition, and argue that the dependency learning process can be a precursor to other language acquisition tasks that are vital to natural language acquisition.

Learning nonadjacent dependencies embedded in sentences of an artificial language: When learning breaks down.

The structure of natural languages give rise to many dependencies in the linear sequences of words, and within words themselves. Detecting these dependencies is arguably critical for young children in learning the underlying structure of their language. There is considerable evidence that human adults and infants are sensitive to the statistical properties of sequentially adjacent items. However, the conditions under which learners detect nonadjacent dependencies (NADs) appears to be much more limited. This has resulted in proposals that the kinds of learning mechanisms learners deploy in processing adjacent dependencies are fundamentally different from those deployed in learning NADs. Here we challenge this view. In 4 experiments, we show that learning both kinds of dependencies is hindered in conditions when they are embedded in longer sequences of words, and facilitated when they are isolated by silences. We argue that the findings from the present study and prior research is consistent with a theory that similar mechanisms are deployed for adjacent and nonadjacent dependency learning, but that NAD learning is simply computationally more complex. Hence, in some situations NAD learning is only successful when constraining information is provided, but critically, that additional information benefits adjacent dependency learning in similar ways. (PsycINFO Database Record

The role of reference in cross-situational word learning.

Word learning involves massive ambiguity, since in a particular encounter with a novel word, there are an unlimited number of potential referents. One proposal for how learners surmount the problem of ambiguity is that learners use cross-situational statistics to constrain the ambiguity: When a word and its referent co-occur across multiple situations, learners will associate the word with the correct referent. Yu and Smith (2007) propose that these co-occurrence statistics are sufficient for word-to-referent mapping. Alternative accounts hold that co-occurrence statistics alone are insufficient to support learning, and that learners are further guided by knowledge that words are referential (e.g., Waxman & Gelman, 2009). However, no behavioral word learning studies we are aware of explicitly manipulate subjects' prior assumptions about the role of the words in the experiments in order to test the influence of these assumptions. In this study, we directly test whether, when faced with referential ambiguity, co-occurrence statistics are sufficient for word-to-referent mappings in adult word-learners. Across a series of cross-situational learning experiments, we varied the degree to which there was support for the notion that the words were referential. At the same time, the statistical information about the words' meanings was held constant. When we overrode support for the notion that words were referential, subjects failed to learn the word-to-referent mappings, but otherwise they succeeded. Thus, cross-situational statistics were useful only when learners had the goal of discovering mappings between words and referents. We discuss the implications of these results for theories of word learning in children's language acquisition.

Infants' sensitivity to vowel harmony and its role in segmenting speech.

A critical part of infants' ability to acquire any language involves segmenting continuous speech input into discrete word forms. Certain properties of words could provide infants with reliable cues to word boundaries. Here we investigate the potential utility of vowel harmony (VH), a phonological property whereby vowels within a word systematically exhibit similarity ("harmony") for some aspect of the way they are pronounced. We present evidence that infants with no experience of VH in their native language nevertheless actively use these patterns to generate hypotheses about where words begin and end in the speech stream. In two sets of experiments, we exposed infants learning English, a language without VH, to a continuous speech stream in which the only systematic patterns available to be used as cues to word boundaries came from syllable sequences that showed VH or those that showed vowel disharmony (dissimilarity). After hearing less than one minute of the streams, infants showed evidence of sensitivity to VH cues. These results suggest that infants have an experience-independent sensitivity to VH, and are predisposed to segment speech according to harmony patterns. We also found that when the VH patterns were more subtle (Experiment 2), infants required more exposure to the speech stream before they segmented based on VH, consistent with previous work on infants' preferences relating to processing load. Our findings evidence a previously unknown mechanism by which infants could discover the words of their language, and they shed light on the perceptual mechanisms that might be responsible for the emergence of vowel harmony as an organizing principle for the sound structure of words in many languages.

Top-down structure influences learning of nonadjacent dependencies in an artificial language.

Because of the hierarchical organization of natural languages, words that are syntactically related are not always linearly adjacent. For example, the subject and verb in the child always runs agree in person and number, although they are not adjacent in the sequences of words. Since such dependencies are indicative of abstract linguist structure, it is of significant theoretical interest how these relationships are acquired by language learners. Most experiments that investigate nonadjacent dependency (NAD) learning have used artificial languages in which the to-be-learned dependencies are isolated, by presenting the minimal sequences that contain the dependent elements. However, dependencies in natural language are not typically isolated in this way. We report the first demonstration to our knowledge of successful learning of embedded NADs, in which silences do not mark dependency boundaries. Subjects heard passages of English with a predictable structure, interspersed with passages of the artificial language. The English sentences were designed to induce boundaries in the artificial languages. In Experiment 1 & 3 the artificial NADs were contained within the induced boundaries and subjects learned them, whereas in Experiment 2 & 4, the NADs crossed the induced boundaries and subjects did not learn them. We take this as evidence that sentential structure was "carried over" from the English sentences and used to organize the artificial language. This approach provides several new insights into the basic mechanisms of NAD learning in particular and statistical learning in general. (PsycINFO Database Record

Prosodic differences between declaratives and interrogatives in infant-directed speech.

In many languages, declaratives and interrogatives differ in word order properties, and in syntactic organization more broadly. Thus, in order to learn the distinct syntactic properties of the two sentence types, learners must first be able to distinguish them using non-syntactic information. Prosodic information is often assumed to be a useful basis for this type of discrimination, although no systematic studies of the prosodic cues available to infants have been reported. Analysis of maternal speech in three Standard American English-speaking mother-infant dyads found that polar interrogatives differed from declaratives on the patterning of pitch and duration on the final two syllables, but wh-questions did not. Thus, while prosody is unlikely to aid discrimination of declaratives from wh-questions, infant-directed speech provides prosodic information that infants could use to distinguish declaratives and polar interrogatives. We discuss how learners could leverage this information to identify all question forms, in the context of syntax acquisition.

Language acquisition is model-based rather than model-free.

Christiansen & Chater (C&C) propose that learning language is learning to process language. However, we believe that the general-purpose prediction mechanism they propose is insufficient to account for many phenomena in language acquisition. We argue from theoretical considerations and empirical evidence that many acquisition tasks are model-based, and that different acquisition tasks require different, specialized models.

Word categorization from distributional information: frames confer more than the sum of their (Bigram) parts.

Grammatical categories, such as noun and verb, are the building blocks of syntactic structure and the components that govern the grammatical patterns of language. However, in many languages words are not explicitly marked with their category information, hence a critical part of acquiring a language is categorizing the words. Computational analyses of child-directed speech have shown that distributional information-information about how words pattern with one another in sentences-could be a useful source of initial category information. Yet questions remain as to whether learners use this kind of information, and if so, what kinds of distributional patterns facilitate categorization. In this paper we investigated how adults exposed to an artificial language use distributional information to categorize words. We compared training situations in which target words occurred in frames (i.e., surrounded by two words that frequently co-occur) against situations in which target words occurred in simpler bigram contexts (where an immediately adjacent word provides the context for categorization). We found that learners categorized words together when they occurred in similar frame contexts, but not when they occurred in similar bigram contexts. These findings are particularly relevant because they accord with computational investigations showing that frame contexts provide accurate category information cross-linguistically. We discuss these findings in the context of prior research on distribution-based categorization and the broader implications for the role of distributional categorization in language acquisition.

Greater sensitivity to nonaccidental than metric shape properties in preschool children.

Nonaccidental properties (NAPs) are image properties that are invariant over orientation in depth and allow facile recognition of objects at varied orientations. NAPs are distinguished from metric properties (MPs) that generally vary continuously with changes in orientation in depth. While a number of studies have demonstrated greater sensitivity to NAPs in human adults, pigeons, and macaque IT cells, the few studies that investigated sensitivities in preschool children did not find significantly greater sensitivity to NAPs. However, these studies did not provide a principled measure of the physical image differences for the MP and NAP variations. We assessed sensitivity to NAP vs. MP differences in a nonmatch-to-sample task in which 14 preschool children were instructed to choose which of two shapes was different from a sample shape in a triangular display. Importantly, we scaled the shape differences so that MP and NAP differences were roughly equal (although the MP differences were slightly larger), using the Gabor-Jet model of V1 similarity (Lades & et al., 1993). Mean reaction times (RTs) for every child were shorter when the target shape differed from the sample in a NAP than an MP. The results suggest that preschoolers, like adults, are more sensitive to NAPs, which could explain their ability to rapidly learn new objects, even without observing them from every possible orientation.

The segmentation of sub-lexical morphemes in english-learning 15-month-olds.

In most human languages, important components of linguistic structure are carried by affixes, also called bound morphemes. The affixes in a language comprise a relatively small but frequently occurring set of forms that surface as parts of words, but never occur without a stem. They combine productively with word stems and other grammatical entities in systematic and predictable ways. For example, the English suffix -ing occurs on verb stems, and in combination with a form of the auxiliary verb be, marks the verb with progressive aspect (e.g., was walking). In acquiring a language, learners must acquire rules of combination for affixes. However, prior to learning these combinatorial rules, learners are faced with discovering what the sub-lexical forms are over which the rules operate. That is, they have to discover the bound morphemes themselves. It is not known when English-learners begin to analyze words into morphological units. Previous research with learners of English found evidence that 18-month-olds have started to learn the combinatorial rules involving bound morphemes, and that 15-month-olds have not. However, it is not known whether 15-month-olds nevertheless represent the morphemes as distinct entities. This present study demonstrates that when 15-month-olds process words that end in -ing, they segment the suffix from the word, but they do not do so with endings that are not morphemes. Eight-month olds do not show this capacity. Thus, 15-month-olds have already started to identify bound morphemes and actively use them in processing speech.

Categorizing words using 'frequent frames': what cross-linguistic analyses reveal about distributional acquisition strategies.

Mintz (2003) described a distributional environment called a frame, defined as the co-occurrence of two context words with one intervening target word. Analyses of English child-directed speech showed that words that fell within any frequently occurring frame consistently belonged to the same grammatical category (e.g. noun, verb, adjective, etc.). In this paper, we first generalize this result to French, a language in which the function word system allows patterns that are potentially detrimental to a frame-based analysis procedure. Second, we show that the discontinuity of the chosen environments (i.e. the fact that target words are framed by the context words) is crucial for the mechanism to be efficient. This property might be relevant for any computational approach to grammatical categorization. Finally, we investigate a recursive application of the procedure and observe that the categorization is paradoxically worse when context elements are categories rather than actual lexical items. Item-specificity is thus also a core computational principle for this type of algorithm. Our analysis, along with results from behavioural studies (Gómez, 2002; Gómez and Maye, 2005; Mintz, 2006), provides strong support for frames as a basis for the acquisition of grammatical categories by infants. Discontinuity and item-specificity appear to be crucial features.

Stress changes the representational landscape: evidence from word segmentation.

Over the past couple of decades, research has established that infants are sensitive to the predominant stress pattern of their native language. However, the degree to which the stress pattern shapes infants' language development has yet to be fully determined. Whether stress is merely a cue to help organize the patterns of speech or whether it is an important part of the representation of speech sound sequences has still to be explored. Building on research in the areas of infant speech perception and segmentation, we asked how several months of exposure to the target language shapes infants' speech processing biases with respect to lexical stress. We hypothesize that infants represent stressed and unstressed syllables differently, and employed analyses of child-directed speech to show how this change to the representational landscape results in better distribution-based word segmentation as well as an advantage for stress-initial syllable sequences. A series of experiments then tested 9- and 7-month-old infants on their ability to use lexical stress without any other cues present to parse sequences from an artificial language. We found that infants adopted a stress-initial syllable strategy and that they appear to encode stress information as part of their proto-lexical representations. Together, the results of these studies suggest that stress information in the ambient language not only shapes how statistics are calculated over the speech input, but that it is also encoded in the representations of parsed speech sequences.

Linguistic and conceptual influences on adjective acquisition in 24- and 36-month-olds.

Two hundred forty English-speaking toddlers (24- and 36-month-olds) heard novel adjectives applied to familiar objects (Experiment 1) and novel objects (Experiment 2). Children were successful in mapping adjectives to target properties only when information provided by the noun, in conjunction with participants' knowledge of the objects, provided coherent category information: when basic-level nouns or superordinate-level nouns were used with familiar objects, when novel basic-level nouns were used with novel objects, and--for 36-month-olds--when the nouns were underspecified with respect to category (thing or one) but participants could nonetheless infer a category from pragmatic and conceptual knowledge. These results provide evidence concerning how nouns influence adjective learning, and they support the notion that toddlers consider pragmatic factors when learning new words.

Frequent frames as a cue for grammatical categories in child directed speech.

This paper introduces the notion of frequent frames, distributional patterns based on co-occurrence patterns of words in sentences, then investigates the usefulness of this information in grammatical categorization. A frame is defined as two jointly occurring words with one word intervening. Qualitative and quantitative results from distributional analyses of six different corpora of child directed speech are presented in two experiments. In the analyses, words that were surrounded by the same frequent frame were categorized together. The results show that frequent frames yield very accurate categories. Furthermore, evidence from behavioral studies suggests that infants and adults are sensitive to frame-like units, and that adults use them to categorize words. This evidence, along with the success of frames in categorizing words, provides support for frames as a basis for the acquisition of grammatical categories.

Category induction from distributional cues in an artificial language.

The ability to identify the grammatical category of a word (e.g., noun, verb, adjective) is a fundamental aspect of competence in a natural language. Children show evidence of categorization by as early as 18 months, and in some cases younger. However, the mechanisms that underlie this ability are not well understood. The lexical co-occurrence patterns of words in sentences could provide information about word categories--for example, words that follow the in English often belong to the same category. As a step in understanding the role distributional mechanisms might play in language learning, the present study investigated the ability of adults to categorize words on the basis of distributional information. Forty participants listened for approximately 6 min to sentences in an artificial language and were told that they would later be tested on their memory for what they had heard. Participants were next tested on an additional set of sentences and asked to report which sentences they recognized from the first 6 min. The results suggested that learners performed a distributional analysis on the initial set of sentences and recognized sentences on the basis of their memory of sequences of categories of words. Thus, mechanisms that would be useful in natural language learning were shown to be active in adults in an artificial language learning task.