Can statistical learning bootstrap early language acquisition? A modeling investigation
Bibliographic record
Abstract
Before they even produce their first word, infants start developing a language-specific perception, recognize the auditory form of frequent words, and develop a rudimentary knowledge of grammatical categories. A major question in language development is: what mechanisms are responsible for the effortless learning infants demonstrate? In-laboratory experiments have shown that young infants are exquisitely sensitive to fine-grained statistical regularities of their speech input. This has led researchers to propose statistical learning as the cornerstone mechanism of early language acquisition. While the statistical learning account has been influential, the extent to which it can explain early language acquisition is still controversial. Recent computational studies provide evidence in favour of the statistical learning hypothesis for sound learning, but can this result be extended to higher level linguistic categories? Here, we introduce STELA, a developmental and psycholinguistic-inspired computational model that simulates how infants might learn at multiple linguistic levels simultaneously based on statistical analysis of raw audio signals. Our algorithm uses only the raw input without any human annotation, and it is trained to predict future segments of speech based on past ones. It reproduces the pattern of parallel learning across sound and word levels reported in infants: it learns to discriminate sounds, recognizes the auditory form of words, and organizes sounds and words along linguistic dimensions. This suggests that statistical learning from raw speech is sufficient to bootstrap early language acquisition at the sound and word levels.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.007 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".