Anecdotal reports on scientific discovery, inventive originality, and artistic productivity suggest that creativity can be triggered or enhanced by sleeping and dreaming. Several studies confirm these anecdotes, showing that sleep promotes creative problem-solving compared to wakefulness. For example, when subjects performed a cognitive task that could be solved much faster through applying a hidden rule, after a night of sleep more than twice as many subjects gained insight into the hidden rule as in a control group staying awake (Wagner et al. 2004). Similarly, subjects benefited in a creativity task from an afternoon nap but not from staying awake (Cai et al. 2009; Beijamini et al. 2014), and the likelihood of solving a problem encountered before sleep can be increased by cued reactivations during sleep (Ritter et al. 2012).
According to the classical stage model of creativity, creative insights may be described by a process consisting of several stages, of which the incubation phase appears to be most intimately associated with sleep and dreaming (Dresler 2011, 2012; Ritter & Dijksterhuis 2014). The most common psychological approaches support this view: psychoanalytical models of creativity emphasize the primary process concept, which denotes free-associative and dream-like thinking, compared to the more rational and analytical secondary-process thinking (Kris 1952). Cognitive models propose that a state of defocused attention facilitates creativity (Mendelsohn 1976)—creative individuals seem to have less narrowly-focused attention than uncreative ones, which leads to unorthodox connections of remote ideas that might eventually lead to creative cognitions. In a similar vein, creative individuals are thought to have relatively flat association hierarchies (i.e., more, yet weaker associations between cognitive elements), which accounts for the ability to make remote associations; whereas uncreative individuals are thought to have relatively steep association hierarchies (Mednick 1962). Physiological models emphasize the level of cortical arousal as an important variable influencing creativity: both a lower level of cortical arousal—particularly in the prefrontal cortex—and a higher variability in cortical arousal levels are expected in creative compared to uncreative individuals, depending on specific phases of the creative process (Martindale 1999). In addition, low levels of norepinephrine are thought to facilitate creativity, shifting the brain toward intrinsic neuronal activation with an increase in the size of distributed concept representations and co-activation across modular networks (Heilman et al. 2003). The prefrontal cortex seems to be of particular importance for creative processes; however there is evidence that both prefrontal activation and prefrontal deactivation facilitate creativity—maybe depending on the specific phase of the creative process. Brain areas showing selective activation for insight events are—besides the prefrontal cortex—the visual cortices, the hippocampus, and in particular the anterior cingulated cortex, which is thought to be involved in breaking the impasse that marks the critical step of insight into a problem (Dietrich & Kanso 2010).
Both theoretical models and empirical neuroscience of creativity suggest that sleep and dreaming provide an ideal environment for creative incubation: primary-process thinking is explicitly conceptualized as dream-like, and the hyper-associative nature of dreams can be considered a prime example of a flat associative hierarchy. Defocused attention is a phenomenal feature of most dreams, physiologically probably caused by prefrontal cortex deactivation. And daydreaming has the potential to increase creativity (Lewin 1989), while the level of engagement in such mind-wandering in contrast to explicitly directed thoughts is associated with creative performance (Baird et al. 2012). The sleep cycle provides the brain with highly alternating arousal levels, and the chaotic activation of the cortex in REM sleep through brain stem regions in absence of external sense data leads to a much more radical renunciation of unsuccessful problem solving attempts, leading to co-activations of cognitive data that are highly remote in waking life (Kahn et al. 2002a). These co-activations, woven into a dream narrative in a self-organizing manner, repeatedly receive further innervations by the brainstem, leading to bizarre sequences of loosely associated dream topics that might eventually activate particular problem-relevant cognitions or creative cognitions in general (Hobson & Wohl 2005). In addition, in REM sleep, which is characterized by low levels of norepinephrine, visual cortices, the hippocampus, and the anterior cingulate cortex have all been shown to be strongly activated, potentially facilitating insight events. In conclusion, the phenomenological and neural correlates of sleeping and dreaming provide ideal conditions for the genesis of creative ideas and insights.
In summary, a third important function of sleep and dreaming is the association of remote cognitive elements in order to facilitate creativity and problem solving.