Mechanisms of spontaneous confabulations: a strategic retrieval account
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
The 'temporality' hypothesis of confabulation posits that confabulations are true memories displaced in time, while the 'strategic retrieval' hypothesis suggests a general retrieval failure of which temporal confusion is a common symptom. Four confabulating patients with rupture of an anterior communicating artery (ACoA) aneurysm, eight non-confabulating ACoA controls and 16 normal controls participated in three experiments designed to test the two hypotheses. In Experiment 1, participants were tested on two continuous recognition tasks, one requiring temporal context distinctions, previously shown to be sensitive to confabulation and another that only requires content distinctions. Both manipulations were sensitive to confabulation, but not specific to it. Temporal context and content confusions (TCCs and CCs) can be explained as failures to make fine-grained distinctions within memory. In Experiment 2, free recall of semantic narratives that require strategic retrieval but are independent of temporal context was used to induce confabulations associated with remote memory, acquired before the onset of amnesia. Confabulators produced significantly more errors. Thus, when retrieval demands are equated, confabulations can be induced in the absence of temporal confusions. Only confabulators conflated semantic content from different remote semantic narratives and introduced idiosyncratic content, suggesting that qualitatively different mechanisms are responsible for distortions due to normal memory failure and for confabulation. Lesion analyses revealed that damage to ventromedial prefrontal cortex is sufficient for temporal context errors to occur, but additional orbitofrontal damage is crucial for spontaneous confabulation. In Experiment 3, we tested whether failure in memory monitoring is crucial for confabulation. Recognition of details from semantic and autobiographical narratives was used to minimize the initiation and search components of strategic retrieval. Only confabulators made more false alarms on both tasks, endorsed even highly implausible lures related to autobiographical events and were indiscriminately confident about their choices. These findings support a strategic retrieval account of confabulation of which monitoring is a critical component. Post-retrieval monitoring has at least two components: one is early, rapid and pre-conscious and the other is conscious and elaborate. Failure of at least the former is necessary and sufficient for confabulation. Other deficits, including TCC and CC, may be required for spontaneous confabulations to arise. The confluence of different sub-components of strategic retrieval would determine the content of confabulation and exacerbate its occurrence.
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