Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A language learning model for finite parameter spaces

P Niyogi1, R C Berwick

  • 1Center for Biological and Computational Learning, Massachusetts Institute of Technology, Cambridge 02142, USA. pn@ai.mit.edu, berwick@ai.mit.edu

Cognition
|October 1, 1996
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Automated vocal analysis of naturalistic recordings from children with autism, language delay, and typical development.

Proceedings of the National Academy of Sciences of the United States of America·2010
Same author

Face recognition using laplacianfaces.

IEEE transactions on pattern analysis and machine intelligence·2005
Same author

Detecting stop consonants in continuous speech.

The Journal of the Acoustical Society of America·2002
Same author

Recurrent left atrial myxoma with recurrent stroke.

Indian pediatrics·2001
Same author

The evolutionary dynamics of grammar acquisition.

Journal of theoretical biology·2001
Same author

Evolution of universal grammar.

Science (New York, N.Y.)·2001
Same journal

People make graded judgments about the inconceivable.

Cognition·2026
Same journal

The self as an image: Appearance and belief in visual representations of one's own face.

Cognition·2026
Same journal

Corrigendum to 'Consonant, vowel, and tone cues in early wordform recognition: Evidence from Cantonese-learning infants' [Cognition 275 (2026) 106624].

Cognition·2026
Same journal

Identifying distinct sources of whole number interference in children's decimal comparison: the role of numerical magnitude and inhibitory control.

Cognition·2026
Same journal

Evidence for abstract spatial concept learning in young animals.

Cognition·2026
Same journal

Blurred lines or clear boundaries? Synchrony and social dominance shape domain-specific self-other processing.

Cognition·2026
See all related articles

This study models language acquisition as a Markov structure, calculating the average number of examples needed for learning. A random step algorithm proves more efficient than local parameter adjustments for language acquisition.

Area of Science:

  • Computational linguistics
  • Formal language theory
  • Cognitive science

Background:

  • Formalizing language acquisition within finite parameter spaces is crucial for understanding linguistic development.
  • The principles-and-parameters approach offers a framework for modeling innate linguistic knowledge.

Purpose of the Study:

  • To formally characterize language learning as a Markov structure.
  • To derive explicit calculations for sample complexity in language acquisition.
  • To compare the efficiency of different learning algorithms.

Main Methods:

  • Modeling language learning as a Markov process within a finite parameter space.
  • Calculating average sample complexity based on input distributions and learning regimes.
  • Analyzing the convergence properties of random step versus local parameter adjustment algorithms.

Related Experiment Videos

Main Results:

  • Explicit calculation of sample complexity for language learning.
  • Demonstration that average convergence time for a three-parameter system is psychologically plausible (100-150 examples).
  • Identification of a random step algorithm as faster and more reliable than local parameter setting.

Conclusions:

  • Language learning can be mathematically modeled using Markov structures.
  • The efficiency of language acquisition is influenced by input distributions and learning algorithms.
  • A non-local, random step learning strategy is superior for converging to the correct target language.