Bootstrapping language acquisition

Abstract

The semantic bootstrapping hypothesis proposes that children acquire their native language through exposure to sentences of the language paired with structured representations of their meaning, whose component substructures can be associated with words and syntactic structures used to express these concepts. The child's task is then to learn a language-specific grammar and lexicon based on (probably contextually ambiguous, possibly somewhat noisy) pairs of sentences and their meaning representations (logical forms). Starting from these assumptions, we develop a Bayesian probabilistic account of semantically bootstrapped first-language acquisition in the child, based on techniques from computational parsing and interpretation of unrestricted text. Our learner jointly models (a) word learning: the mapping between components of the given sentential meaning and lexical words (or phrases) of the language, and (b) syntax learning: the projection of lexical elements onto sentences by universal construction-free syntactic rules. Using an incremental learning algorithm, we apply the model to a dataset of real syntactically complex child-directed utterances and (pseudo) logical forms, the latter including contextually plausible but irrelevant distractors. Taking the Eve section of the CHILDES corpus as input, the model simulates several well-documented phenomena from the developmental literature. In particular, the model exhibits syntactic bootstrapping effects (in which previously learned constructions facilitate the learning of novel words), sudden jumps in learning without explicit parameter setting, acceleration of word-learning (the "vocabulary spurt"), an initial bias favoring the learning of nouns over verbs, and one-shot learning of words and their meanings. The learner thus demonstrates how statistical learning over structured representations can provide a unified account for these seemingly disparate phenomena.

Distinctions in the Acquisition of Vocabulary and Grammar: An Individual Differences Approach

“…(submitted). While this study did not investigate the mechanisms underlying the learning, our findings are consistent with the use of a combination of statistical learning mechanisms (Yu & Smith, 2007), syntactic and semantic bootstrapping (Abend, Kwiatkowski, Smith, Goldwater, & Steedman, 2017; Gleitman, 1990), and a propose‐but‐verify procedure (Trueswell, Medina, Hafri, & Gleitman, 2013), with initial learning occurring implicitly (consistent with associative learning mechanisms) and then top‐down knowledge interacting more explicitly and working in unison with unconscious learning (consistent with strategic approaches to learning). Replicating the findings of Rebuschat et al.…”

The role of feedback and instruction on the cross-situational learning of vocabulary and morphosyntax: Mixed effects models reveal local and global effects on acquisition

“…Indeed, learning is extremely rapid under these conditions: after just a few dozen exposures to varying utterances appearing with scenes to which they refer, participants are better than chance at knowing the meaning of the words and identifying the syntax within those utterances. Hence, the chicken-and-egg problem of vocabulary and grammar acquisition is solvable as a consequence of sensitivity to cross-situational statistics (see also Abend, Kwiatkowski, Smith, Goldwater, & Steedman, 2017, for a computational demonstration of this ability).…”

Computational and Robotic Models of Early Language Development

“…Importantly, the studies reviewed so far primarily address learning the semantics of words, largely ignoring syntax. However, a variety of models have been proposed to address the problem of learning syntax from experiencing sentences (e.g., Chang, Dell, & Bock, 2006;Freudenthal et al, 2007;Thomas and Knowland, 2014), as well as for simultaneously learning semantics and syntax (e.g., a Bayesian model: Abend et al, 2017; a connectionist model: Li, Farkas, & MacWhinney, 2004). Due to space limitations we will not cover syntax learning in-depth, but many of the issues we identify for cross-situational word learning models also apply to syntax learning models.…”