Functional mapping with direct electrical stimulation (DES) is widely used during awake neurosurgery to generate causal evidence about person-specific neuroanatomical organization. According to a long-standing clinical and scientific paradigm, if the application of DES to a given brain region does not result in performance errors, that site is considered to be uninvolved in the task. Here, we show that both error rates and performance speed on correct trials are parametrically modulated by when DES starts and stops relative to the timeline of task-driven processing in stimulated brain areas and networks. We propose a framework, causal parametric mapping, which moves beyond the classic approach of binarizing the effects of DES on behavior into “positive” and “negative” mapping trials. Causal parametric mapping offers a method to functionally dissect separable processing stages in the human brain, in real time, with reversible causal evidence during invasive neurosurgical procedures.
