Statistical Language Learning

  title={Statistical Language Learning},
  author={Jenny R. Saffran},
  journal={Current Directions in Psychological Science},
  pages={110 - 114}
  • J. Saffran
  • Published 2003
  • Psychology
  • Current Directions in Psychological Science
What types of mechanisms underlie the acquisition of human language? Recent evidence suggests that learners, including infants, can use statistical properties of linguistic input to discover structure, including sound patterns, words, and the beginnings of grammar. These abilities appear to be both powerful and constrained, such that some statistical patterns are more readily detected and used than others. Implications for the structure of human languages are discussed. 

Figures from this paper

Language Learning and Language Universals
The role of learning in generative grammar is explored, highlighting interactions between distributional patterns in the environment and the innate structure of the language faculty. Expand
Statistical Learning of Language
The fundamental challenge of language acquisition is outlined and recent findings are reviewed which suggest that infants (and adults) are able to track statistical information in linguistic input and that they may be able to use that information for phonetic acquisition, speech segmentation, word learning, and acquisition of simple grammatical structure. Expand
Domain general constraints on statistical learning.
Infant's learning of visual patterns shows the same constraints as infants' learning of phonotactic patterns, consistent with theories suggesting that constraints arise from domain-general sources and, as such, should operate over many kinds of stimuli in addition to linguistic stimuli. Expand
Perceptual and memory constraints on language acquisition
It is shown that two basic mechanisms can explain many results in artificial grammar learning experiments, and different linguistic regularities ranging from stress assignment to interfaces between different components of grammar. Expand
Using mathematical models of language experimentally
This work suggests that the problem of not knowing what to test in animals and children lacking the expressive abilities of normal, mature adults can be mitigated by tapping equally rich, but more formal mathematical approaches to language. Expand
A Role for Chunk Formation in Statistical Learning of Second Language Syntax.
Humans are remarkably sensitive to the statistical structure of language. However, different mechanisms have been proposed to account for such statistical sensitivities. The present study comparedExpand
Statistical learning and developmental language impairments
Important studies of statistical learning are reviewed and how these perspectives may be used to improve the study of developmental language impairments and their remedial treatment and a particular focus is on the prospects of designing training tasks in a non-linguistic domain of behavior. Expand
Learning and Long-Term Retention of Large-Scale Artificial Languages
A large-scale learning experiment is reported that demonstrates that adults can learn words from unsegmented input in much larger languages than previously documented and that they retain the words they learn for years, suggesting that statistical word segmentation could be scalable to the challenges of lexical acquisition in natural language learning. Expand
Exclusion Constraints Facilitate Statistical Word Learning
In environments in which performance could benefit from exclusion, a learning advantage for speech over nonspeech is found, revealing an interaction between statistical and exclusion processes in associative word learning. Expand
Two Distinct Sequence Learning Mechanisms for Syntax Acquisition and Word Learning
It is demonstrated that the learning of repeating sequences is related to vocabulary development in these children, suggesting that there may be at least two relatively distinct domain-general sequential processing skills, with each supporting a different aspect of language acquisition. Expand


The Use of Predictive Dependencies in Language Learning
To what extent is linguistic structure learnable from statistical information in the input? This research investigated the role played by statistical learning in the acquisition of rudimentary phraseExpand
Constraints on Statistical Language Learning
Abstract How do learners discover the structure in linguistic input? One set of cues which learners might use to acquire phrase structure are the dependencies, or predictive relationships, which linkExpand
Pattern induction by infant language learners.
9-month-old infants were given the opportunity to induce specific phonological patterns in 3 experiments in which syllable structure, consonant voicing position, and segmental position were manipulated and revealed that infants rapidly extracted new phonological regularities. Expand
Statistical Learning by 8-Month-Old Infants
The present study shows that a fundamental task of language acquisition, segmentation of words from fluent speech, can be accomplished by 8-month-old infants based solely on the statistical relationships between neighboring speech sounds. Expand
Does Grammar Start Where Statistics Stop?
How do we acquire language? Do we identify words in a stream of speech by identifying statistical regularities or by seeking grammatical structure? A new study of adults ([PeA±a][1] et al .) suggestsExpand
Infant artificial language learning and language acquisition
This article reviews studies using the domain-general, associative learning paradigm that have led to a number of exciting discoveries regarding the learning mechanisms available during infancy and raises important issues with respect to whether such mechanisms are general or specific to language. Expand
Structural packaging in the input to language learning: Contributions of prosodic and morphological marking of phrases to the acquisition of language
It is suggested that the finding that phrase structure cues are a necessary aspect of language input reflects the limited capacities of human language learners; languages may incorporate structural cues in part to circumvent such limitations and ensure successful acquisition. Expand
Visual statistical learning in infancy: evidence for a domain general learning mechanism
The results provide support for the likelihood of domain general statistical learning in infancy, and imply that mechanisms designed to detect structure inherent in the environment may play an important role in cognitive development. Expand
From Syllables to Syntax: Multilevel Statistical Learning by 12-Month-Old Infants
To successfully acquire language, infants must be able to track multiple levels of regularities in the input. In many cases, regularities only emerge after some learning has already occurred. ForExpand
Signal-Driven Computations in Speech Processing
It is shown that both statistical computations to identify words in speech and algebraic-like computation to discover higher level (grammatical) structure can be influenced by subtle cues in the speech signal. Expand