Estonian case inflection made simple. A case study in Word and Paradigm morphology with Linear Discriminative Learning.
Bibliographic record
Abstract
According to Word and Paradigm Morphology (Matthews, 1974; Blevins, 2016), the word is the basic cognitive unit over which paradigmatic analogy operates to predict form and meaning of novel forms. Baayen et al. (2019b, 2018) introduced a computational formalization of word and paradigm morphology which makes it possible to model the production and comprehension of complex words without requiring exponents, morphemes, inflectional classes, and separate treatment of regular and irregular morphology. This computational model, Linear Discriminative Learning (LDL), makes use of simple matrix algebra to move from words’ forms to meanings (comprehension) and from words’ meanings to their forms (production). In Baayen et al. (2018), we showed that LDL makes accurate predictions for Latin verb conjugations. The present study reports results for noun declension in Estonian. Consistent with previous findings, the model’s predictions for comprehension and production are highly accurate. Importantly, the model achieves this high accuracy without being informed about stems, exponents, and inflectional classes. The speech errors produced by the model look like errors that native speakers might make. When the model is trained on incomplete paradigms, comprehension accuracy for unseen forms is hardly affected, but production accuracy decreases, reflecting the well-known asymmetry between comprehension and production. Unseen principal parts (i.e., nominative, genitive, and partitive singulars) are particularly difficult to produce, possibly due to their more distinctive forms. Removing principal parts from training, however, does not affect accuracy for other case forms. Model performance does not degrade either when the training data includes the alternative forms that are ubiquitous in Estonian. These results are consistent with the claim of Blevins (2008) that Estonian number and case inflection is organized in a way that facilitates the deduction of full paradigms from only a small number of forms.
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) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.001 |
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".