Negative polarity illusions are robust with both ‘ever’ and ‘any’ (when linear position is held constant)
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Many studies have used linguistic illusions to probe the representations and mechanisms used during incremental language comprehension. A crucial component of this research program is mapping out when illusions occur and when they do not. To this end, we investigate the generality of a linguistic illusion observed with negative polarity items (NPIs). Most previous work has only investigated the illusion using a single NPI, ever (or its analogue in other languages), but all models of the illusion phenomenon implicitly predict that illusions should generalize across different NPIs. In apparent contradiction to this prediction Parker and Phillips (2016) found reliable illusions with ever, but not with the previously untested NPI any. In their original paper, the authors suggested that the asymmetry stemmed from differences in the linear position of the two NPIs in their test items. However, the authors did not establish the basic empirical generalization that any is, in fact, susceptible to the illusion when the confound of linear position is factored out. As such, their findings are equally compatible with the hypothesis that there is fine-grained lexical variation in inherent susceptibility to the illusion, which would have serious implications for all theories of the phenomenon. To settle the empirical record, we conducted a higher-power study comparing ever and any using items adapted from Parker and Phillips (2016) such that the two NPIs occupied the same ordinal position in their test sentences. We find comparable illusions for both NPIs, a welcome result for all candidate theories of the phenomenon and consistent with the distance-based explanation for its absence in Parker and Phillips (2016).
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it