Text Verification and Verb Factivity: An ERP Investigation
Bibliographic record
Abstract
Text Verification and Verb Factivity: An ERP Investigation Todd R. Ferretti Murray Singer Department of Psychology, Wilfrid Laurier University Waterloo, Ontario, Canada Department of Psychology, University of Manitoba Winnipeg, Manitoba, Canada Courtney Patterson Department of Psychology, Wilfrid Laurier University Waterloo, Ontario, Canada epochs that extended 100 ms before the critical word in the Keywords: verb meaning; discourse processing; event- target sentence (i.e., truck) to 1000 ms post stimulus onset. related brain potentials (ERP). Introduction Results and Discussion Singer (2006; JML) has recently investigated text verification processes when people read passages similar to (1). These passages included sentences that varied in their truth with reference to antecedent text (i.e., truck (true) / bus (false)), factivity of the main verb (comprehended (factive) / implied (nonfactive)), and negation (was (affirmative) / wasn’t (negative)). The results demonstrated that in the Late Positivity Complex (LPC) region (600-1000 ms post stimulus onset), amplitudes were more positive for true, factive sentences than for false, factive sentences. However, truth had no influence on nonfactive verbs. Similarly, in the P2 region (200-300 ms), amplitudes varied in the same way as a function of truth and factivity. Despite these clear differences in early and late components, in the N400 region (300-500 ms), amplitudes varied only as a function of truth (see Figure 1). These findings are consistent with Singer’s (2006) reading time data, and provide insight into how the brain processes information about truth of discourse constituents in conjunction with the factivity associated with verbs. In particular, the interaction in the P2 results suggest the brain is most prepared to process the visual features of the targets words in true, factive sentences, and the least prepared for the targets in false, factive sentences. The same pattern of findings in the LPC data also suggests that people had the least difficulty integrating the targets into the discourse in the true, factive condition, and the most difficulty for the targets in the false, factive condition. Dan had been driving all night in order to get home for Thanksgiving. Before long, Dan drove past a truck/bus which was stopped with a flat tire. He couldn’t help but laugh because its spare tire must have been underneath everything and suitcases and boxes were strewn everywhere. Later, while Dan was sitting in a diner, drinking some coffee, a policeman came in and started a conversation with him. He implied/comprehended that the vehicle with the flat was/wasn’t a truck. Singer found that reading times varied systematically with truth, factivity, and negation. For factive, but not nonfactive verbs, reading times for false, affirmative sentences were read more slowly than true, affirmative sentences. Alternatively, false, negative sentences were read more slowly than true, negative sentences, but this was only true for nonfactive verbs. These results are consistent with Singer’s proposal that readers verify discourse constituents against the referents that they passively cue during reading. In the present research, we extend these results by providing converging neurocognitive evidence for these reading processes by employing ERP methodology. The main focus of this research was on affirmative sentences. Method µV 0.0 Materials Stimuli consisted of 32 target passages and 21 filler passages. The target passages were identical to (1) with the exception that we only used affirmative target sentences. Words in the target sentences were presented one a time for a duration of 300 ms and an SOA of 500 ms. ms Figure 1: Results at a central - parietal electrode located on the midline. Solid = true / factive, dots = false / factive, dash = true / nonfactive, dash + dot = false /nonfactive. Acknowledgments EEG Recording Parameters This research was supported by a CFI grant to the first author, and by separate NSERC discovery grants awarded to the first and second authors. EEG was recorded from 64 electrodes from 48 participants. Impedances were kept below 5KΩ. ERPs were computed in
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 | 0.000 | 0.003 |
| 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 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 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".