In 1797, during a period of illness, Samuel Taylor Coleridge composed what is now one of his most celebrated poems: the frantic and surreal Kubla Khan. As the story goes, he woke up in a creative mania, the verse having fully formed in his dream. He recorded it immediately but was interrupted by the now infamous “person on business from Porlock.” Once this stranger was dismissed, Coleridge found his creative flow disrupted, as he later described, “like the images on the surface of a stream into which a stone has been cast, but, alas! without the after restoration of the latter!”
This bit of literary apocrypha has come to embody the persistent struggle of artistic generation, and it has been referenced and lampooned in later works to illustrate the fragility of creative states and the myth-making that shadows our discussions of artistry.
Creativity is an often lauded ability central to human experience. Teachers are encouraged to foster creative thinking in their classrooms, companies tout creative solutions in their bylines, and artistic practice is becoming an increasingly popular intervention for conditions both major and mundane. It seems strange to consider that, not so long ago, the mere concept of human invention was itself controversial.
For much of Western history, creatio ex nihilo (Latin for “creation from nothing”) was not the given it is today. Early Classical models considered artistry a primarily imitative act, while Christian Europeans insisted that only God could truly create, and that our inspirations were merely divine loans. Only after many humanistic shifts in thinking has creativity become such a sacred tool in humanity’s mental repertoire.
Although our conceptions of creativity have changed, questions regarding its psychological and neurological mechanisms persist. Namely, how does the mind reconcile two diametric modes of thought—one internalized and elaborative, the other externalized and focused—to produce creative works in all their forms? In cognitive terms, these can be traced to two distinct behavioral functions and their respective neural networks:
- Incubation: a period of relative inactivity, when the mind is not attending to current states but allows for fluent memory reactivation. In neuroscientific terms, this roughly correlates with activity in a system known as the Default Mode Network.
- Execution: a state of activity in which ideas are recorded, drafted, and enacted. This requires many brain regions, but it ultimately correlates with activity in a system known as the Central Executive Network.
Huang et al. (2026) acknowledges this dilemma, noting: “Creativity possesses an inherent paradox, composed of opposing forces that must coexist in dynamic tension rather than reconcile into stable equilibrium.” It is perhaps this very paradox that gives rise to creativity—a modulation between controlling systems of the brain that allow for the cognitive dynamics necessary to produce ideas both novel and valuable. Yet, this modulation is not merely a matter of combined activation. Neuroimaging has shown that as one of the aforementioned brain networks is activated, the other is naturally inhibited. How, then, can the human brain balance its functions when they are so physiologically opposed?
Further answers may lie in our understanding of conscious states, particularly in their capacity to enable creative fecundity (as in dreams) or operational ability (as in wakefulness). Creativity is predicated on conscious tone, and thus to understand the former we must consider the impact of the latter. Unfortunately, studying the relationship between conscious states and their generative effects has proven difficult.
Consider dreaming—what some neuroscientists claim is our most natural, automatic creative state. Dreaming, which is most vivid during the rapid eye movement (REM) stage of sleep, involves desynchronized cortical activity. This activity closely mimics the activity of an awake brain, albeit with several major differences. Although cortex-wide activity remains erratic, much of the logical ordering of this activity is suspended. The aforementioned Central Executive Network is known to be inactive during REM, leaving only the diffuse activity of the Default Mode Network in brain regions related to perceptions, feelings, and internal thoughts. This disjunction of systems is thought to explain why dreaming, while generative, is often illogical, strange, and highly emotional.
This would seem like a perfect embodiment of the neural networks at the heart of creative incubation, but attempts to use REM dreaming as a creativity model are fraught with major setbacks. Most prominently, to accurately assess dream content, we generally must suspend the dreaming state. REM dreaming is therefore usually understood a posteriori, as more wakeful, attentive faculties are necessary for reporting. As we move further from the dream state, executive networks are apt to forget, alter, or mischaracterize a given dream’s subjectivity. We see then that even our means of characterizing the relationship between creative conscious periods are curtailed by these oppositional networks.
Occasionally, scientists can glean aspects of dream experiences through the observation of Dream Enactment Behaviors (DEBs). DEBs occur when subjects perform physical actions of a dream, often as a result of a sleep disorder and typically occurs during episodes of sleepwalking and REM Behavior Disorder. DEBs take many forms, ranging from ambulations, physical struggles, and, occasionally, speech. Unfortunately, the complications associated with these aberrant sleep states confound possible interpretations of native creativity. Instead, neuroscientists have turned their attention to transitional states that yield the most coherent combinations of activity underlying creativity and awareness.
N1 is the first stage of non-REM sleep and is often characterized as a “dozing off” period. Perceptions of the environment begin to diminish, but many will report an inwardness of thought akin to daydreaming. It is during N1 that the cerebral cortex exhibits a shift in firing patterns, showing a combination of Alpha, Theta, and low-amplitude mixed frequency EEG waveforms. Creative studies have found that these waveforms correlate with states of wakeful creative ideation, particularly when oscillating in the right frontal pole and right temporal lobe. However, these states are highly transient, often giving way to further desynchronization in the next non-REM sleep stage, N2, leaving only brief windows for analysis. On average, adults spend as little as five minutes in N1, and most ideas generated during this period are forgotten.
This semilucidity, known as “hypnagogia,” has been reportedly used by major thinkers such as Albert Einstein and Salvador Dali. Although their individual approaches varied, most utilized a hypnagogic trigger to abruptly end the N1 phase and restore complete wakefulness: the subject would typically sit comfortably in a chair while holding a heavy object in one hand. As they entered the N1 state, their brain would begin the usual shift into Theta oscillations, raising sensory thresholds and reducing muscle tone. This slackening of limbs caused the subject to drop the held object, and the resulting noise would rouse them. Whatever imagined scenarios persisted in their short-term memory could then be recorded or elaborated in workings of their choosing. Charles Dickens supposedly came to rely on this method for many of his literary inspirations, including Oliver Twist and A Christmas Carol.
More recent studies have linked both dreaming and hypnagogic states with increased creative thinking when awake. These unfocused, desynchronized periods are naturally suitable for a cognitive effect known as “spreading activations.” This occurs when neuronal assemblies responsible for specific feelings, memories, etc., are activated and, in turn, activate other assemblies without strict oversight from executive systems. Because of this, the assemblies activated may be more distantly related, thus leading to novel links often called “remote associations.” The effects described in these REM/N1 studies may represent a lingering potential for remote associations in the wakeful state. Unfortunately, data for the permanence of these associative abilities are scarce and require further study.
But while historic and modern examples cite hypnagogia as a rectification of the creativity paradox, many of these same examples further delineate incubating and executing states rather than illustrate their synergy.
A good example of this is the story of Caedmon, who medievalists cite as the first named English-language poet. His story is related in Bede’s Historia ecclesiastica gentis Anglorum and focuses on the creation of his most notable verse, the eponymous “Caedmon’s Hymn.” Bede intimates that Caedmon, a lay brother in a North Yorkshire abbey, lacked the knowledge (or, more likely, the confidence) to compose poetry, and so would isolate himself in the evenings when the other brothers recited verse. One night, he fell asleep and was visited by an unknown, but likely divine, entity that urged him to sing about the world’s creation. Caedmon supposedly launched into perfect verse, which was later reconstructed by Bede. Curiously, Bede’s report of the verse is followed by an admission that the “translated” verse is a pale imitation of Caedmon’s original dreamed-up song. There is something intrinsically lost in committing this virtual creation to something concrete and tangible.
This is further reflected in our initial example, Kubla Khan. Although much blame is placed on the “person on business from Porlock,” it is unlikely that Coleridge’s verse would be perfectly remembered, as modern researchers now understand that recall of dream events is extremely flawed. Dreaming and hypnagogia can certainly enhance the mind’s creative associations, but these are still a far cry from the effortless generation in Default Mode Network-dominated periods. Many scholars likewise assume Coleridge’s “slumber” was actually an opium-induced stupor, which would involve neural activations distinct from those of normal sleep.
This introduces yet another dimension to our discussion: how other altered states can enhance or diminish creative thinking. The relationship between atypical neural states, whether the result of psychoactive substances or neurological disorder, and creative fecundity has been hotly debated for centuries. In fact, these cases may represent the most ardent evidence for the delineation between incubative and executing states. Many recreational drugs (such as opium) highly stimulate or disinhibit neurons that release the neurotransmitter dopamine. High dopamine levels have been linked to a broad range of cognitive effects, including greater flexibility of thinking. Excess dopamine, however, is also implicated in a range of psychotic disorders, including schizophrenia. While the hallucinations that typify schizophrenia might reflect the automatic perceptual generation of dreams/imagination, it should be noted that these states are, again, fraught with distress and dysfunction and cannot be considered exact correlates.
The archetype of the “tortured artist” has endured well into our social media age, but any notion that this represents an ideal creative state is sorely mistaken. Earlier studies proposing a link between mental illness and creativity have been soundly criticized for romanticizing an apparent link between suffering and artistic output. Indeed, many mentally ill artists, such as Vincent Van Gogh and Robert Schumann, are exceptions rather than indicators. As Kaufman and Paul note in their 2014 editorial, “there is very little evidence suggesting that clinical, debilitating mental illness is conducive to productivity and innovation.” Rather, the neurology underpinning many of these cases is more likely to inhibit artistic output, as the distress associated with this experience is far more likely to arrest the attentive, executive, emotional, and logistical efforts required.
It is clear that creativity represents not one conscious state but the dynamic tension of many distinct activations. The paradox of these opposing systems is illustrated in the stark differences between incubated and executed states, both of which are required for artistic practice. While there exists some overlap between these networks, the manner in which they coactivate continues to thwart simple explanations. The study of sleep may yield more insights into these topics, but current models resist simple categorizations of mental productions. For the average creative, this may mean very little. Artistic production has flourished for centuries without a clearly elucidated mechanism, but this does not diminish this explanation’s potential benefits. Further investigations into these cognitive networks will expand our understanding of conscious states, well-being, and the specifics of human ingenuity.
