Recognising faces and reading words : investigations into visual perceptual expertise
Bibliographic record
Abstract
Faces and words are ‘objects of expertise’. Both have many parts, yet are processed by expert mechanisms which emphasise the whole. The following behavioural studies investigated holistic integration in faces, and parallel-letter, lexical processing in words. People normally only read upright words, so inverting words may reveal markers of perceptual expertise. \n \nStudy 1 explored the impact of word inversion on potential markers. The ‘word-length effect’ was found to be most suitable as it was only exacerbated by inversion in normal word formats. \n \nStudy 2 inverted paragraphs of text to reveal further markers of perceptual expertise. 15 hours of training in reading inverted novels partially reversed many of the deleterious effects of inversion. We saw a trend to the reduction of the word-length effect, which may reflect increased use of expert mechanisms. \n \nStudy 3 investigated whether expert word and expert face perception networks overlap. Subjects with prosopagnosia due to unilateral right lesions showed normal word-length effects, but struggled to differentiate visual text styles. Therefore, the expert face network in the right hemisphere may not overlap with the expert word network, but it may contribute to the perception of visual text style. \n \nStudy 4 asked whether internal features contribute more than external features to mental representations of faces. Isolated internal features produced stronger identity aftereffects, supporting this idea. However, when placed in a whole-face context, the contribution of the internal features was weakened. Holistic integration therefore reduces the saliency of the internal features. This occurs in both familiar and unfamiliar faces. \n \nOverall we find that word perceptual expertise is well characterised by the word length effect and may be acquired relatively quickly. However, it may not be served by the face recognition network in the right hemisphere. We also confirm that internal and external parts of faces, regardless of familiarity, are represented holistically.
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 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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| 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 teacher head, 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".