Perceived versus Target-Scored Empathic Accuracy under Psychedelics: A Measurement Gap and a Proposed Test
Vasileios Sampalis
Zenodo (CERN European Organization for Nuclear Research) October 26, 2026 preprint DOI: 10.5281/zenodo.23667350 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Theoretical or philosophical paper (proposal with a structured search) Randomized Placebo-controlled Double-blind |
|---|---|
| Measures | Empathic Accuracy Task, clip-level confidence ratings |
| Key points | Argues that claims psychedelics deepen understanding of others rest on an untested assumption, since no published study has measured empathic accuracy under a psychedelic. Proposes that perceived empathic accuracy will rise while actual empathic accuracy does not, with perceivers becoming less able to distinguish accurate from inaccurate readings, and outlines a randomized, double-blind, placebo-controlled crossover to test enhancement, impairment, and no-change hypotheses. |
Abstract
Psychedelics are increasingly described as deepening empathy and understanding of others, a claim drawn from increases in emotional empathy and self-reported connectedness. These measures index how people feel. None establishes whether the inferences people draw about others are accurate. Empathic accuracy (EA) is accurately inferring the specific content of another's thoughts and feelings, scored against the target's own report. A structured search of 15 platforms to October 2026 reached no published study measuring EA in anyone who has taken a psychedelic. Nor can perceived empathic accuracy (perceived EA), a perceiver's belief that they have read another person correctly, be assumed to coincide with accurate understanding: where the two have been compared, they were weakly related or unrelated. Six constructs differ in what each is scored against: emotional empathy, felt connection, self-reported empathy, and perceived EA against the participant's own account, cognitive empathy against a preset key, EA alone against the target's own report. Psychedelic studies have scored accuracy about others only for emotion recognition and cognitive empathy, against a preset key rather than the target's own report. LSD, psilocybin, and MDMA raise emotional empathy while acutely reducing recognition of negative emotions, though later crossovers reproduced neither the psilocybin rise nor MDMA's recognition effect. I set out three hypotheses for EA under a psychedelic: enhancement, impairment, or no change. I predict the third: perceived EA rises while EA does not, and each perceiver grows less able to tell accurate readings from inaccurate ones. To test all three I propose a randomized, double-blind, placebo-controlled crossover in healthy volunteers. Each completes two dosing sessions, one with a classical psychedelic, one with inert placebo. Each session carries an acute assessment on the declining phase, clock-matched under placebo, and a seven-day follow-up: four assessments. The primary instrument is the Empathic Accuracy Task, with clip-level confidence ratings measuring perceived EA; a dyadic interaction task at each follow-up is the ecological check. Until EA is measured under a psychedelic, the claim that these drugs deepen understanding of other people rests on an untested assumption.