Sleep and Dreams

For most of human history, people slept in two shifts. Then artificial light abolished a practice so deeply embedded that it had survived from Homer to the Industrial Revolution. This single fact — that the dominant mode of human sleep for millennia was something nobody alive remembers — should make us suspicious of everything we think we know about what sleep is for.

The Lost Architecture

Roger Ekirch stumbled onto biphasic sleep while researching the history of night-time in English criminal depositions. A nine-year-old girl named Jane Rowth, testifying in a 1699 murder case, mentioned in passing that she and her mother had "arisen from their first sleep" before two men appeared at the window. The phrasing was utterly casual — first sleep, as though everyone had one.¹

Ekirch found references to biphasic sleep everywhere: Chaucer, Virgil, Plutarch, Pausanias. In France it was the premier somme; in Italy, primo sonno. Colonial accounts from Brazil, Oman, Madagascar. The pattern was universal: retire at 9 or 10 PM, sleep for a few hours, wake naturally around midnight for a one-to-three-hour stretch called "the watch," then sleep again until dawn.¹

The watch was not idle time. People prayed, brewed beer, had sex, chatted in the intimacy that darkness and shared beds provided. A 16th-century French physician recommended the period after first sleep as the best time for conception, since the body was rested but not yet inert. Christians had specific prayers prescribed for this window. A London tradesman invented a "nocturnal remembrancer" — a guided writing pad — for capturing the philosophical insights that came during the watch. One Yorkshire man used his to commit murder.¹

What makes the historical evidence compelling is the biological confirmation. Thomas Wehr's 1992 experiment at the National Institute of Mental Health took fifteen men and reduced their daylight hours from sixteen to ten, confining them in darkness for the remaining fourteen. After four weeks, their sleep spontaneously reorganised into two roughly equal blocks separated by one to three hours of quiet wakefulness. Melatonin rhythms confirmed the change was biological, not just behavioral. Wehr had reinvented biphasic sleep in a lab.¹

Even more recently, David Samson's fieldwork with the Manadena community in northeastern Madagascar — where there is no electricity — found a clear period of activity after midnight followed by a return to sleep until dawn. Biphasic sleep never fully disappeared; it just retreated to places artificial light hadn't reached.¹

The extinction of the watch happened gradually across the 19th century as gas and then electric lighting extended the evening and compressed sleep into a single consolidated block. Ekirch tracked it almost decade by decade. The cultural consequences were immediate: middle-of-the-night wakefulness, which had been a valued and productive part of life for millennia, was rebranded as insomnia — a disorder to be medicated rather than a natural feature of human chronobiology.

The Antidepressant in the Machine

If compressing sleep changed our relationship with the night, deliberately disrupting sleep reveals something unexpected about the relationship between sleep and mood. Francesco Benedetti's chronotherapy protocol at San Raffaele Hospital in Milan has treated close to a thousand patients with bipolar depression, achieving 70% response rates within the first week — in a population where conventional antidepressants have largely failed.²

The protocol is counterintuitive: keep depressed patients awake for an entire night, give them bright light exposure each morning and lithium to sustain the effect, repeat every other night for a week. Sleep deprivation has "opposite effects in healthy people and those with depression," Benedetti explains — it makes the healthy miserable but prompts immediate improvement in the depressed.²

The mechanism appears to involve circadian rhythm restoration. In severe depression, the normal oscillations of hormones, body temperature, and brain excitability are flattened — melatonin doesn't rise properly in the evening, cortisol stays high around the clock. The sleep-wake cycle, driven by clock genes expressed rhythmically in every cell, falls out of sync. Sleep deprivation seems to kick-start the sluggish clock. Lithium boosts Per2, a protein that drives the molecular clock. Bright light resets the master clock in the suprachiasmatic nucleus. Together, they restore the cycling that depression had damped.²

What's philosophically interesting is the implication: depression may not be primarily a disorder of neurotransmitter levels (the serotonin story) but a disorder of temporal organization. The circadian clock isn't just a sleep scheduler — it coordinates hundreds of cellular processes, and when it goes flat, everything from mood to metabolism to immune function degrades. This connects to the broader insight from constructed emotion that emotional states aren't discrete circuits but emergent properties of whole-body regulation.

Dreaming as Second Life

Michael Fahey's account of three weeks of sedation-induced dreams after emergency aortic surgery is the most vivid first-person description of extended dreaming I've encountered.³ The dreams weren't random noise. They had narrative structure, emotional logic, recurring settings, and a "hub world" — a grassy plateau in a circular pit — that functioned as a reset point between dream sequences. He could sometimes influence the next dream from this hub: "Something different, please." Sometimes it felt as though someone was listening.

The dreams systematically incorporated bodily signals. His lungs were filling with fluid, producing gurgling sounds; he dreamed of playing poker underwater, his breath bubbling in his lungs. He wasn't eating or drinking; he dreamed elaborate scenarios about pizza and cola that he could never quite consume. The distinction between Marvel (happy dreams) and DC (nightmares) in his comic-book sequences is the kind of pattern that seems random but might reflect deep affective associations below conscious access.³

Most striking: his wife and children never appeared. He loved them most of all, and his subconscious seems to have censored them because "my mind could not handle the thought of leaving them behind." This selective exclusion — the dreaming mind protecting itself from its own worst fears — suggests something more sophisticated than mere replay.³

After waking, the sedation dreams didn't fade like normal dreams. They had "time to seep in and leave indelible imprints," and for months afterward he struggled to distinguish them from real memories. The phenomenology was as rich and detailed as any autobiographical memory — "they may as well have been real." This is consistent with Gelernter's model of the cognitive spectrum: at the bottom of the spectrum, memories are re-experienced rather than recalled, and the boundary between remembered and imagined collapses entirely.

Dream-Logic as Cognitive Mode

David Gelernter's argument — that we've been studying only half of human thought by focusing on alert, logical reasoning — frames dreaming not as cognitive garbage collection but as a distinct and essential mode of cognition.⁴ The cognitive spectrum runs from maximum focus (analytical reasoning, conscious control) through free association (daydreaming, mind-wandering) down to full dream-logic (loss of reality-testing, hallucinated environments, emotional rather than logical connections between scenes).

What changes as you descend the spectrum isn't just what you think but how you think. At the top, you control your thoughts and inspect memories from outside. At the bottom, thoughts control you and memories are re-entered as virtual realities. The transition is smooth — there's no sharp line between focused analysis and free association, or between hypnagogic drifting and REM dreaming. Poets have always known this. Coleridge composed Kubla Khan in an opium haze; Rilke watched his own descent toward sleep with the attention of a scientist; Rimbaud insisted "one should say not 'I think' but 'I am thought.'"⁴

The creative implications are real. New analogies — the foundation of creativity according to most cognitive psychologists — emerge when focus drops low enough for emotion-bridges to form between logically unrelated memories. You can't force this. Creative insight, like sleep itself, requires relinquishing control. The same mechanism that makes you believe you're playing poker with underwater serpents also generates the connections that Kekulé's snake-eating-its-tail brought to the benzene ring.

What links all of this — the lost watch, chronotherapy, sedation dreams, the cognitive spectrum — is that sleep and its adjacent states aren't just downtime. They're alternative modes of consciousness with their own logic, their own utility, and their own pathologies when disrupted. The modern assumption that consciousness means alert wakefulness and everything else is mere unconsciousness may be one of the most impoverishing simplifications of the industrial era.