People who are asleep and having a lucid dream—aware that they are dreaming—can perceive questions from an experimenter and answer them in real time using eye movements and facial muscle contractions. In a study of 36 individuals during REM sleep, including frequent lucid dreamers, a novice, and a patient with narcolepsy, participants performed perceptual analysis of new information, held information in working memory, computed simple answers, and gave volitional replies. Correct answers occurred on 29 occasions across 6 individuals, documented by four independent laboratories. This two-way communication channel allows real-time interrogation of dream cognition and characteristics.
Lucid dreaming—being aware that one is dreaming while still asleep—is rare but can be trained by regularly questioning whether current experience is real or a dream. Virtual reality (VR) scenarios containing dream-like elements enhanced this training. Over four weeks, volunteers who received VR-assisted lucid dreaming training showed significantly greater increases in lucid dreaming than those who received no training. Eye-signal-verified lucid dreams during polysomnography supported these behavioral results. Potential mechanisms include synthetic dream-like experiences, incorporation of VR content into dream imagery as memory cues, and dissociative effects of VR that may amplify lucid dreaming training during wakefulness.