Probabilistic and distributional approaches to language acquisition

Recent computational research on natural language corpora has revealed that relatively simple statistical learning mechanisms can make an important contribution to certain aspects of language acquisition. For example, statistical and connectionist methods can provide valuable cues to word segmentation and to the acquisition of inflectional morphology, syntactic classes and aspects of word meaning. In each case, these cues are partial, and must be integrated with additional information, whether from other environmental cues or innate knowledge, to provide a complete solution to the acquisition problem. The success of these methods with real natural language corpora demonstrates their feasibility as part of the language acquisition mechanism, an area where previously most research has been limited to highly idealized artificial input or to a priori considerations regarding the feasibility of acquisition mechanisms. Exploring probabilistic learning mechanisms with natural language input provides both an empirical basis for assessing how innate constraints interact with information derived from the environment, and a source of hypotheses for experimental testing.     

The Utility of Cognitive Plausibility in Language Acquisition Modeling: Evidence From Word Segmentation

The informativity of a computational model of language acquisition is directly related to how closely it approximates the actual acquisition task, sometimes referred to as the model's cognitive plausibility. We suggest that though every computational model necessarily idealizes the modeled task, an informative language acquisition model can aim to be cognitively plausible in multiple ways. We discuss these cognitive plausibility checkpoints generally and then apply them to a case study in word segmentation, investigating a promising Bayesian segmentation strategy. We incorporate cognitive plausibility by using an age‐appropriate unit of perceptual representation, evaluating the model output in terms of its utility, and incorporating cognitive constraints into the inference process. Our more cognitively plausible model shows a beneficial effect of cognitive constraints on segmentation performance. One interpretation of this effect is as a synergy between the naive theories of language structure that infants may have and the cognitive constraints that limit the fidelity of their inference processes, where less accurate inference approximations are better when the underlying assumptions about how words are generated are less accurate. More generally, these results highlight the utility of incorporating cognitive plausibility more fully into computational models of language acquisition.     

Probabilistic models of language processing and acquisition

Probabilistic methods are providing new explanatory approaches to fundamental cognitive science questions of how humans structure, process and acquire language. This review examines probabilistic models defined over traditional symbolic structures. Language comprehension and production involve probabilistic inference in such models; and acquisition involves choosing the best model, given innate constraints and linguistic and other input. Probabilistic models can account for the learning and processing of language, while maintaining the sophistication of symbolic models. A recent burgeoning of theoretical developments and online corpus creation has enabled large models to be tested, revealing probabilistic constraints in processing, undermining acquisition arguments based on a perceived poverty of the stimulus, and suggesting fruitful links with probabilistic theories of categorization and ambiguity resolution in perception.     

Evidentiality in language and cognition

What is the relation between language and thought? Specifically, how do linguistic and conceptual representations make contact during language learning? This paper addresses these questions by investigating the acquisition of evidentiality (the linguistic encoding of information source) and its relation to children's evidential reasoning. Previous studies have hypothesized that the acquisition of evidentiality is complicated by the subtleness and abstractness of the underlying concepts; other studies have suggested that learning a language which systematically (e.g. grammatically) marks evidential categories might serve as a pacesetter for early reasoning about sources of information. We conducted experimental studies with children learning Korean (a language with evidential morphology) and English (a language without grammaticalized evidentiality) in order to test these hypotheses. Our experiments compared 3- and 4-year-old Korean children's knowledge of the semantics and discourse functions of evidential morphemes to their (non-linguistic) ability to recognize and report different types of evidential sources. They also compared Korean children's source monitoring abilities to the source monitoring abilities of English-speaking children of the same age. We found that Korean-speaking children have considerable success in producing evidential morphology but their comprehension of such morphology is very fragile. Nevertheless, young Korean speakers are able to reason successfully about sources of information in non-linguistic tasks; furthermore, their performance in these tasks is similar to that of English-speaking peers. These results support the conclusion that the acquisition of evidential expressions poses considerable problems for learners; however, these problems are not (necessarily) conceptual in nature. Our data also suggest that, contrary to relativistic expectations, children's ability to reason about sources of information proceeds along similar lines in diverse language-learning populations and is not tied to the acquisition of the linguistic markers of evidentiality in the exposure language. We discuss implications of our findings for the relationship between linguistic and conceptual representations during development.     

Sentence processing in an artificial language: Learning and using combinatorial constraints

A study combining artificial grammar and sentence comprehension methods investigated the learning and online use of probabilistic, nonadjacent combinatorial constraints. Participants learned a small artificial language describing cartoon monsters acting on objects. Self-paced reading of sentences in the artificial language revealed comprehenders’ sensitivity to nonadjacent combinatorial constraints, without explicit awareness of the probabilities embedded in the language. These results show that even newly-learned constraints have an identifiable effect on online sentence processing. The rapidity of learning in this paradigm relative to others has implications for theories of implicit learning and its role in language acquisition.     

A probabilistic computational model of cross-situational word learning

Words are the essence of communication: They are the building blocks of any language. Learning the meaning of words is thus one of the most important aspects of language acquisition: Children must first learn words before they can combine them into complex utterances. Many theories have been developed to explain the impressive efficiency of young children in acquiring the vocabulary of their language, as well as the developmental patterns observed in the course of lexical acquisition. A major source of disagreement among the different theories is whether children are equipped with special mechanisms and biases for word learning, or their general cognitive abilities are adequate for the task. We present a novel computational model of early word learning to shed light on the mechanisms that might be at work in this process. The model learns word meanings as probabilistic associations between words and semantic elements, using an incremental and probabilistic learning mechanism, and drawing only on general cognitive abilities. The results presented here demonstrate that much about word meanings can be learned from naturally occurring child-directed utterances (paired with meaning representations), without using any special biases or constraints, and without any explicit developmental changes in the underlying learning mechanism. Furthermore, our model provides explanations for the occasionally contradictory child experimental data, and offers predictions for the behavior of young word learners in novel situations.     

Maturational Constraints on Language Learning

This paper suggests that there are constraints on learning required to explain the acquisition of language, in particular, maturational constraints. First, empirical evidence for this claim is reviewed. The evidence from several studies of both first and second language acquisition suggests that normal language learning occurs only when exposure to the language begins early in life. With exposure beginning later in life, asymptotic performance in the language declines: the effects over age of first exposure are approximately linear through childhood, with a flattening of the function in adulthood. These outcomes argue that some type of constraints ensuring successful language learning exist early in life, and weaken with increasing maturation. Second, two hypotheses are considered as to the nature of these maturational changes. One hypothesis is that constraints on learning particular to language acquisition undergo maturational decay. A second hypothesis, which is considered in more detail, suggests that language learning abilities decline because of the expansion of nonlinguistic cognitive abilities.     

Constraints on Constraints: Surveying the Epigenetic Landscape

It is now relatively commonplace to advocate the need for some sorts of constraints on learning and knowledge acquisition. The critical issues to cognitive science concern the sorts of constraints that are able to best model various phenomena of learning and development. Four types of constraints on learning are proposed to be used as an interpretative framework within which to: 1. Better understand the nature of current research: 2. Allow the exploration of alternative models of learning related phenomena, and 3. See more clearly needs for further research. Superficially similar learning phenomena can be modeled by very different configurations of underlying constraints with strong implications for the sorts of representational states that are involved. Each of the five papers in this issue (Brown, Gelman, Markman, Newport, and Spelke) is considered in terms of the configuration of constraints after which each author intends to model their phenomena and in terms of alternate configurations. The papers are construed as illustrating a diverse set of models of how constraints might guide learning, and while the evidence generally favors the configurations suggested by the authors, in each case alternative models are possible and motivate quite specific future research questions. More broadly, it is suggested that asking detailed questions about the sorts of Constraints types that could potentially model complex cases of natural knowledge acquisition helps motivate fundamental questions about learning and the nature of knowledge and that the five papers in this issue are superb examples of how adopting this kind of perspective has been fruitful research orientation.     

Neural correlates of social perception on response bias

Accurate person perception is crucial in social decision-making. One of the central elements in successful social perception is the ability to understand another’s response bias; this is because the same behavior can represent different inner states depending on whether other people are yea-sayers or naysayers. In the present study, we have tried to investigate how the internal biases of others are perceived. Using a multi-trial learning paradigm, perceivers made predictions about a target’s responses to various suggested activities and then received feedback for each prediction trial-by-trial. Our hypotheses were that (1) the internal decision criterion of the targets would be realized through repeated experiences, and (2) due to positive–negative asymmetry, yea-sayers would be recognized more gradually than naysayers through the probabilistic integration of repeated experiences. To find neural evidence that tracks probabilistic integration when forming person knowledge on response biases, we employed a model-based fMRI with a State-Space Model. We discovered that person knowledge about yea-sayers modulated several brain regions, including caudate nucleus, DLPFC, hippocampus, etc. Moreover, when person knowledge was updated with incorrect performance feedback, brain regions including the caudate nucleus, DLPFC, dmPFC, and TPJ were also involved. There were overlapping regions for both processes, caudate nucleus and DLPFC, suggesting that these regions take crucial roles in forming person knowledge with repeated feedback, while reflecting acquired information up to the current prediction.     

FEEDFORWARD AND FEEDBACK AS DETERMINERS OF KNOWLEDGE AND POLICY: NOTES ON A NEGLECTED ISSUE

B jörkman M. Feedforward and feedback as determiners of knowledge and policy: notes on a neglected issue. Scand. J. Psychol . 1972, 13 , 152–158.—Studies of cognitive learning should distinguish between acquisition of knowledge and policy formation. Two learning operators, feedforward, that is, task information by instructions, and feedback must also be taken into account. This double aim requires experimenters to analyze and relate two sets of data: subjective estimates about the task learned and predictions of outcomes. A broader approach is thus needed with a focus on how policy is related to knowledge, how these processes depend on feedback and feedfonvard and how the latter two interact in producing knowledge and policy.     

What does it mean to ‘live’ and ‘die’? A cross‐linguistic analysis of parent–child conversations in English and Indonesian

Previous work on children's intuitive knowledge about the natural world has documented their difficulty in acquiring an overarching concept of biological life that includes plants as well as humans and non‐human animals. It has also suggested that the acquisition of fundamental biological concepts like alive and die may be influenced by the language used to describe them, as evidenced by differences between English‐ and Indonesian‐speaking children's performance in tasks involving these concepts. Here, we examine one particularly important source of linguistic information available to children during this acquisition process: everyday conversations with their parents. We take a cross‐linguistic approach in analysing the evidence available to English‐ and Indonesian‐speaking children as they acquire meanings for words corresponding to the concepts alive and die . Our analysis illustrates that young children acquiring English and Indonesian are faced with distinct problems, but that parental input in both languages does little to support the acquisition of broad, inclusive biological concepts.     

Relational knowledge: the foundation of higher cognition

Accumulating evidence on the nature, function and acquisition of relational knowledge indicates a crucial role of such knowledge in higher cognitive processes. In this review, we specify the essential properties of relational knowledge, together with the role it plays in reasoning, categorisation, planning, quantification and language. Furthermore, we discuss the processes involved in its acquisition and how these processes have been implemented in contemporary neural network models. We present evidence demonstrating that relational knowledge integrates heuristic and analytic cognition, is important for symbolic processes and the creation of novelty, activates specific regions of the prefrontal cortex, and is the most recently evolved and slowest-developing cognitive process. Arguably, relational knowledge represents the core of higher cognition.     

The knowledge-learning-instruction framework: bridging the science-practice chasm to enhance robust student learning

Despite the accumulation of substantial cognitive science research relevant to education, there remains confusion and controversy in the application of research to educational practice. In support of a more systematic approach, we describe the Knowledge-Learning-Instruction (KLI) framework. KLI promotes the emergence of instructional principles of high potential for generality, while explicitly identifying constraints of and opportunities for detailed analysis of the knowledge students may acquire in courses. Drawing on research across domains of science, math, and language learning, we illustrate the analyses of knowledge, learning, and instructional events that the KLI framework affords. We present a set of three coordinated taxonomies of knowledge, learning, and instruction. For example, we identify three broad classes of learning events (LEs): (a) memory and fluency processes, (b) induction and refinement processes, and (c) understanding and sense-making processes, and we show how these can lead to different knowledge changes and constraints on optimal instructional choices.     

Kinematic information feedback for learning a rapid arm movement

The experiments reported examine the notion that knowledge of results (KR) about the outcome of a response does not provide the necessary information for optimizing performance in many skilled activities. The effect of traditional KR was contrasted with various kinematic feedback parameters in the acquisition of a single degree of freedom response requiring the minimization of movement time. Experiment 1 showed that the presentation of discrete kinematic information feedback (peak accelaration, time to peak accelaration, and velocity at the target location) did not facilitate performance over movement-time KR. Experiment 2 revealed that presentation of a computer generated velocity-time representation of the movement as terminal information feedback improved performance over movement-time KR. This facilitation occured even without knowledge of the kinematics for optimal performance. The findings suggest that the task criterion specifies the appropriate information feedback for skill learning in that the information feedback must match the constraints imposed upon response output.     

Analogical Processes in Language Learning

ABSTRACT— The acquisition of language has long stood as a challenge to general learning accounts, leading many theorists to propose domain-specific knowledge and processes to explain language acquisition. Here we review evidence that analogical comparison is instrumental in language learning, suggesting a larger role for general learning processes in the acquisition of language.     

How does a word become a message? An illustration on a developmental time-scale

Recent changes in views on cognition underscore its embodied, situated and distributed character. These changes are compatible with the conceptual framework of ecological psychology. However for ecological psychology to propose explanations for a broad range of cognitive phenomena, including language, it needs an account of how to link the dynamics of coupling between the organism and the environment with the apparent symbolicity of informational structures. In this paper it is proposed that a theory of information in biological systems, advocated by Howard Pattee, may help forge this link. By treating informational structures as constraints on dynamics this approach helps to identify which processes, in which systems and on what time-scales are needed for structures to ‘become messages’. I will illustrate how these processes might work on developmental time-scale in the domain of language, building on the work by Edward Reed (1995, 1996) and extending it using the view of linguistic structures as constraints.