Related Experiment Video
Updated: Jun 22, 2026

Measuring Statistical Learning Across Modalities and Domains in School-Aged Children Via an Online Platform and Neuroimaging Techniques
Published on: June 30, 2020
The metamorphosis of the statistical segmentation output: lexicalization during artificial language learning
Tânia Fernandes1, Régine Kolinsky, Paulo Ventura
1Faculdade de Psicologia e de Ciências da Educação da Universidade de Lisboa, Alameda da Universidade, Lisbon, Portugal.
Artificial language learning (ALL) outputs are integrated into the mental lexicon only when speech segmentation cues are consistent. This process, crucial for language acquisition, is independent of word frequency.
Area of Science:
- Cognitive Psychology
- Psycholinguistics
- Computational Linguistics
Background:
- Understanding how infants learn spoken language is a key challenge in psycholinguistics.
- Statistical learning mechanisms are proposed to play a role in speech segmentation and word acquisition.
- The integration of learned word forms into the mental lexicon remains incompletely understood.
Purpose of the Study:
- To investigate whether statistical speech segmentation outputs from artificial language learning (ALL) are integrated into adult listeners' mental lexicon.
- To determine the role of congruent segmentation cues (transitional probabilities, wordlikeness) in the lexicalization of ALL outputs.
- To examine whether lexicalization is influenced by stimulus frequency.
Main Methods:
- Employed artificial language learning (ALL) coupled with experimental psycholinguistic techniques.
- Assessed lexicalization by measuring inhibitory effects of novel ALL-derived neighbors on auditory lexical decision tasks.
- Manipulated segmentation cues (transitional probabilities, wordlikeness) and stimulus frequency during familiarization.
Main Results:
- ALL outputs were lexicalized only when segmentation cues were congruent, both immediately and one week post-familiarization.
- No lexicalization effect was observed when segmentation cues were incongruent.
- ALL outputs showed effects beyond chance even with incongruent cues, indicating a dissociation between item knowledge and full lexicalization.
- Equating stimulus frequency did not alter the pattern of results.
Conclusions:
- Lexicalization of ALL outputs depends critically on the congruence of segmentation cues.
- The findings suggest that statistical learning contributes to mental lexicon formation under specific cue conditions.
- Lexicalization is not solely driven by raw frequency of occurrence but by the integration of segmented forms within a consistent linguistic structure.
Related Concept Videos
Language Development
The critical period for language acquisition suggests that the ability to acquire language is at its peak early in life. As people age, this proficiency decreases. Language development begins very...
Language and Cognition
Components of Language
Automatic Processing and Automatic Social Behavior