A placebo-controlled investigation of synaesthesia-like experiences under LSD

M. and colleagues used a within-groups, placebo-controlled design in which each participant completed sessions under LSD and under placebo on separate days. The study probed colour experiences evoked by standardised graphemes and sounds, and assessed both spontaneous synaesthesia-like experiences (via post-task questionnaires) and two commonly accepted diagnostic criteria for synaesthesia: stimulus–concurrent consistency (consistency of inducer–colour pairings across repetitions) and inducer specificity (the degree to which particular stimuli reliably evoke particular concurrents). The authors note that full methodological details are reported elsewhere. The extracted text does not clearly report the sample size, the LSD dose(s), participant inclusion/exclusion criteria, the order or counterbalancing of sessions, nor details of any psychological support provided during sessions. Outcomes were measured behaviourally (responses to graphemes and sounds) and with self-report questionnaires including measures of spontaneous synaesthesia-like experiences and a psychometric measure of absorption (the MODTAS is mentioned). Data analysis was performed using MATLAB (2014a) and SPSS (version not fully reported). The analysis approach (for example, whether analyses were intention-to-treat, use of mixed models, or specific statistical tests) is not clearly described in the extracted text.

The authors report that participants endorsed more spontaneous synaesthesia-like experiences under LSD than under placebo on post-task questionnaires. By contrast, there were no reliable differences between LSD and placebo in the extent to which participants experienced colours in direct response to the experimental inducers (standardised graphemes and sounds). Likewise, inducer–concurrent consistency and inducer specificity—two widely used diagnostic markers of congenital synaesthesia—did not differ between drug conditions. In fact, stimulus–colour consistency was numerically lower under LSD than under placebo. The extracted text indicates heterogeneity in participant responses: some individuals showed pronounced drug-related effects while others showed minimal change. Preliminary correlational analyses suggested that individual differences in absorption moderated several effects: higher trait absorption related to a pattern in which participants who experienced LSD-induced colours exhibited reduced stimulus–colour consistency. The blind was ineffective — all participants correctly identified the drug condition — and self-reported drug intensity in the placebo condition was uniformly zero at task completion. The results reported here do not include specific numerical effect sizes, p-values, confidence intervals, or the sample size in the extracted text.

M. and colleagues interpret their findings as indicating that LSD can increase spontaneous synaesthesia-like experiences but does not produce stimulus-locked, consistent, or inducer-specific synaesthesia-like responses in the controlled paradigm used here. They emphasise that the absence of greater inducer–concurrent consistency and inducer specificity under LSD argues against classifying these drug-evoked reports as genuine synaesthesia according to standard definitions. The authors note that this conclusion is preliminary given sample-size limitations. Several explanations for the null effects on consistency and specificity are discussed. The authors consider and mostly reject inadequate dosing and expectancy effects as primary explanations, noting that participants reported strong subjective drug effects and did not report believing they had taken LSD during placebo sessions. They propose that drug-induced synaesthesia-like phenomena may be phenomenologically or mechanistically distinct from congenital synaesthesia; transient, spontaneous perceptual states induced by psychedelics may not produce the stable inducer–concurrent pairings that characterise congenital forms. The authors also note that the simple stimuli used here may be less likely to evoke drug-related concurrents than richer, more complex stimuli (for example, music), and that selective attention or distractor suppression under LSD might impair performance on consistency measures. The authors highlight heterogeneity across participants and present preliminary evidence that trait absorption moderates the relationship between experiencing drug-induced colours and measures of stimulus–colour consistency. They caution that correlations with absorption may be confounded by the content of the MODTAS measure, which may include items overlapping with crossmodal correspondences or synaesthesia-like experiences. Key methodological limitations are acknowledged: the likely under-powered sample, and an ineffective blind in which participants correctly identified drug condition. The authors argue that the ineffective blind probably would inflate demand characteristics in the LSD condition, so the absence of consistency and specificity effects despite this suggests that such effects were genuinely absent in this paradigm. They recommend further research with larger samples, counterbalanced designs, variable doses, stimuli of differing complexity, experiential-sampling methods to capture spontaneous occurrences, and combined neuroimaging to investigate neural correlates. The authors also discuss the possibility that consistency may require consolidation over time and thus be an inappropriate criterion for transient or early-stage synaesthesia-like phenomena, suggesting that training paradigms combined with pharmacological manipulation might be informative.