Distinct audio and visual accumulators co-activate motor preparation for multisensory detection

Detecting targets in multisensory environments is an elemental brain function, but it is not yet known whether information from different sensory modalities is accumulated by distinct processes, and, if so, whether the processes are subject to separate decision criteria. Here we address this in two experiments (n = 22, n = 21) using a paradigm design that enables neural evidence accumulation to be traced through a centro-parietal positivity and modelled alongside response time distributions. Through analysis of both redundant (respond-to-either-modality) and conjunctive (respond-only-to-both) audio-visual detection data, joint neural-behavioural modelling, and a follow-up onset-asynchrony experiment, we found that auditory and visual evidence is accumulated in distinct processes during multisensory detection, and cumulative evidence in the two modalities sub-additively co-activates a single, thresholded motor process during redundant detection. These findings answer long-standing questions about information integration and accumulation in multisensory conditions.