Increased Activation of Indirect Semantic Associations under Psilocybin

A within-subject, double-blind, placebo-controlled design was used: eight healthy male subjects (mean age 39.4 years; seven right-handed, one left-handed; all physicians who provided written informed consent) received a capsule containing 0.2 mg/kg psilocybin or placebo on two testing days at least one week apart, in random order. Testing started at approximately the same time of day (about 11:00 AM). Institutional and national permissions for experimental psilocybin use were obtained. The behavioural task replicated an established lexical decision paradigm. Subjects decided whether a string was a real word. Word pairs of three semantic distances were presented in rapid succession (200 ms): closely related pairs (e.g., "black–white"), indirectly related pairs (e.g., "lemon–sweet"), and unrelated pairs (e.g., "cloud–cheese"). The primary dependent measures were reaction times and error rates for the second word, with semantic priming quantified as both absolute difference scores and as percentage speed gains (priming expressed as percent of baseline speed) to control for general slowing under hallucinogens. Two alternate stimulus versions were used to reduce repetition effects. Each subject performed the paradigm before drug intake and at 50, 150, and 220 minutes after intake; the main statistical comparisons reported in the extract focused on the baseline and 50-minute post-intake timepoint. Statistical analysis included a two-way analysis of variance (intervention: placebo vs psilocybin; time: before vs 50 min) on the indirect priming percentage score, and targeted t-tests assessing significance of priming effects within conditions. Error rates and general reaction-time changes were also examined.

Overall error rates were very low (0–2%) and did not vary systematically with semantic condition or drug. As expected, psilocybin produced a general slowing of reaction times across semantic conditions. Priming was reported both as raw difference scores and as percentage speed gains to account for this global slowing. A two-way ANOVA on the indirect priming percentage score (intervention by time, comparing baseline to 50 minutes) yielded a trend for an effect of time [F(1,7) = 5.09; p = .059], but no significant main effect of intervention and no significant interaction. When the authors tested indirect semantic priming within each intervention separately, they found a nonsignificant increase in indirect priming under placebo but a statistically significant increase under psilocybin [t(7) = 2.82; p = .026]. Direct (closely related) semantic priming showed nonsignificant increases after intake of either placebo or psilocybin, with numerically larger effects under psilocybin. The extract refers to a Table and a Figure illustrating mean reaction times, priming effects, and the difference between psilocybin and placebo runs, but those data are not reproduced in the provided text.

Spitzer and colleagues interpret the increased indirect semantic priming under psilocybin as evidence that the drug raises the availability of remote associations, bringing cognitive contents to mind that are normally not activated. They link this effect to the idea of a reduced signal-to-noise ratio in semantic networks under the drug, analogous to earlier interpretations of thought-disordered schizophrenia and contrary to the focusing effect attributed to dopaminergic modulation. The investigators further suggest that the observed increase in indirect priming may reflect a decreased capacity to use contextual information to focus semantic processing, which would explain why subjective reports of broadened consciousness or enhanced creativity can coincide with reductions in objective performance measures. From a methodological standpoint, the authors note that indirect semantic priming is a sensitive measure responsive to both pharmacologic manipulation and to the psychological setting. They offer a post hoc account for the tendency toward increased priming under placebo: subjects were initially anxious and tense, a state associated with stereotyped verbal associations and reduced priming, and tended to relax 30–40 minutes into testing once they inferred whether they had received placebo or psilocybin, which could facilitate broader associative access. The paper reports this as a testable hypothesis and indicates plans to use indirect semantic priming to study nonpharmacological states of anxiety and arousal. Finally, the authors present the finding as an example of linking subjective psychopathology, objective cognitive measurements, and underlying neuromodulatory physiology within a single framework.