Cross-modal influence of odorant stimulation on sound localization revealed by EEG and diffusion modeling

The ventriloquism effect, where sounds are mislocalized towards a spatially disparate but temporally synchronous visual stimulus, is well established. Liang et al. (2022) provided first evidence that chemosensory cues can similarly bias sound localization. This study, conducted in 2023, replicates and extends their behavioral findings, using diffusion modeling and EEG to identify underling computational and neural mechanisms. Participants localized (left vs. right) each sound in eight-item sequences, with trigeminally-potent odorants presented concurrently with the first four sounds (left, right, or both nostrils). Unbeknownst to participants, some sounds originated from the center. For central sounds, the proportion of right-ward responses increased with right-nostril stimulation but decreased with left-nostril stimulation (relative to controls). This bias was strongest in the second half of the sequence and disappeared at larger sound eccentricity. Diffusion modeling showed the bias was best explained by changes in drift rate (perceptual evidence accumulation), not in starting point (decision bias). EEG alpha-band activity lateralized towards the chosen side for central sounds, consistent with attentional orienting towards the perceived location, and was diminished at large eccentricities with incongruent chemosensory stimulation. The findings corroborate an odorant-sound spatial ventriloquism effect and provide novel insights into the cognitive and attentional processes involved. Public significance statementOur findings reveal that trigeminally potent odors can subtly influence sound localization judgments, even when those odors are irrelevant to the task at hand. This influence is strongest when sound information is uncertain (i.e., when unbeknownst to participants, a sound that had to be localized as coming from the left or right side, originated from a central location), suggesting that trigeminal stimulation can act as a spatial cue under ambiguous conditions. In addition, concurrent EEG cording revealed that the influence of odorant stimulation was also reflected in neural correlates of auditory spatial attention. Critically, diffusion modeling showed that the observed cross-modal bias occurs at a perceptual level. These findings highlight that the human brain integrates information from the chemical sense and hearing more deeply than previously assumed, revealing that even task-irrelevant odors can shape spatial perception in other senses.