Byron's designation of the dream as “The mystical Usurper of the mind” (Don Juan 4.30.4) captures two aspects of its perennial fascination. Dreams are mysterious, and they take possession of our minds as if they had an irresistible power of their own. After more than two millennia of research and theorizing, even defining dreaming remains nontrivial and controversial. By adopting “the most broad, general, and indisputable definition of dreaming: mental activity occurring in sleep” (Hobson 2002, 7), one excludes phenomena often not distinguished from dreams in the past— waking visions. In accounts from antiquity and the Middle Ages one sometimes cannot tell whether uisio refers to a waking vision or a dream, and sometimes waking visions are included in classifications of dreams. In 1999, a task force of the Associated Professional Sleep Societies failed to agree upon a definition of dreaming, and some contemporary scientists argue that dreams from REM sleep should be distinguished from the “dream-like mentation” of NREM sleep (Wamsley and Stickgold 2013, 132).

The mystery of dreaming extends beyond problems of definition. The question that has preoccupied people since Homer— “What do dreams mean?”— has yet to be answered to general satisfaction. Since the discovery of REM sleep by Aserinsky and Kleitman (1953), scientists have learned a good deal about the physiology of sleep— for example, which areas of the brain are active in different stages of sleep, or which neurotransmitters are performing which functions— but no one has come up with a persuasive theory of the function or significance of dreams. Hobson states a widely held suspicion: “dreaming itself could be an epiphenomenon without any direct effect on normal or abnormal cognition” (2009, 805). Solms concludes a recent survey of the neurobiology and neurology of dreaming: “An adaptive function for dreaming has, however, not been empirically demonstrated” (2011, 540). In fact, today there is less of a consensus on the relation of REM sleep to dreaming than a couple of decades ago because dreams also occur in NREM sleep, although not as frequently and, more controversially, are not of the same character as in REM. Consequently, a history of conceptions of dreams cannot be written as a series of approaches to and deviations from an accepted theory. The debate about the significance or insignificance of dreams continues, often in terms familiar to one who knows the history of dream theory.