Defining Sleep: Conceptual Boundaries and Historical Research Trajectories
For millennia, sleep was framed as a passive state of bodily repose, a temporary suspension of consciousness that required no specialized physiological explanation. This intuitive understanding, rooted in everyday lived experience, persisted until the 19th century, when advances in experimental physiology and electrophysiology first provided the methodological tools to move beyond philosophical speculation and establish sleep as a distinct, measurable biological phenomenon. To build a rigorous conceptual foundation for the field, we first must clarify the boundaries of the term “sleep” by distinguishing it from other states of reduced consciousness or inactivity, then trace the historical trajectory of sleep research from early philosophical inquiry to the modern neuroscientific paradigm, before outlining the core research questions and cross-disciplinary value of the field.
Conceptual Boundaries of Sleep: Distinguishing Sleep from Related Physiological States
While sleep is easily recognized in everyday life, its scientific definition requires precise operational criteria that differentiate it from other states of low arousal or inactivity. The American Academy of Sleep Medicine (AASM) defines sleep as a “reversible state of perceptual disengagement from and unresponsiveness to the environment, accompanied by a characteristic posture (e.g., recumbency in humans) and reduced motor activity,” with core physiological markers including specific patterns of brain activity, eye movement, and muscle tone measurable via polysomnography (PSG).<sup>[1]</sup> This definition highlights three core features of sleep: reversibility, perceptual disengagement, and measurable physiological signatures, which allow us to distinguish it from four commonly confused states: 1. Sleep vs. wakeful rest: Wakeful rest (e.g., sitting quietly with eyes closed) is a state of active wakefulness with intact consciousness and normal responsiveness to environmental stimuli. Electroencephalogram (EEG) recordings during wakeful rest show dominant alpha wave activity (8–13 Hz), the same pattern observed in relaxed wakefulness, while sleep is marked by a shift to slower theta (4–7 Hz) and delta (0.5–4 Hz) waves during non-rapid eye movement (NREM) sleep, or low-voltage mixed-frequency activity during REM sleep, alongside a 3–10-fold increase in the threshold for responding to external stimuli.<sup>[2]</sup> 2. Sleep vs. coma: Coma is a pathological state of unconsciousness caused by damage to the brainstem, thalamus, or cerebral cortex, or systemic metabolic/toxic insults. Unlike sleep, coma is characterized by the complete loss of the sleep-wake cycle, inability to be aroused even with strong stimuli, and absence of normal sleep-stage physiological patterns on EEG. In contrast, sleep is a fully reversible physiological state, with intact sleep-wake cycling and normal arousal responses to meaningful stimuli (e.g., a caregiver’s voice, a baby’s cry). 3. Sleep vs. general anesthesia: While both states involve reduced consciousness, general anesthesia is a drug-induced, reversible state marked by amnesia, analgesia, and immobility, with EEG typically showing burst suppression (alternating periods of high and low electrical activity) or dose-dependent slow-wave activity at deep anesthetic levels. Unlike natural sleep, anesthesia cannot be reversed by external sensory stimulation, and recovery requires metabolic clearance of anesthetic agents from the body. Natural sleep, by contrast, is an endogenously regulated state that can be terminated within seconds by salient sensory input, with no exogenous agent required for arousal. 4. Sleep vs. hibernation: Hibernation is an adaptive physiological state in some mammals (e.g., bears, bats, ground squirrels) characterized by a 10–30°C drop in core body temperature, a 95–99% reduction in metabolic rate, and heart rate as low as 5–10 beats per minute, allowing animals to survive periods of food scarcity and cold stress. Human sleep, by contrast, involves only a 1–2°C drop in core temperature, a 10–15% reduction in metabolic rate ($M_{sleep} \approx 0.85 \times M_{awake}$), and a heart rate reduction of 10–20% below wakeful baseline, with each sleep episode lasting only 6–9 hours rather than weeks or months. Unlike hibernation, human sleep is tightly coupled to the 24-hour circadian cycle, with sleep propensity highest during the nighttime circadian trough.
Historical Trajectory of Sleep Research: From Philosophical Speculation to Empirical Science
The scientific study of sleep has evolved through three distinct eras, each marked by paradigm shifts driven by technological and theoretical advances.
Pre-Modern Era: Philosophical and Humoral Theories (400 BCE–1700 CE)
Early conceptions of sleep were rooted in philosophical and humoral frameworks, with no empirical basis for testing hypotheses. Hippocrates (460–370 BCE), the father of Western medicine, framed sleep as a natural healing state caused by the redistribution of blood to the brain during rest, a view that persisted for nearly two millennia.<sup>[3]</sup> Galen (129–216 CE), building on Hippocratic humoral theory, argued that sleep was caused by the cooling of “animal spirits” in the brain as blood flowed to the periphery during rest, while wakefulness was driven by the heating of these spirits by incoming sensory input. This humoral framework dominated medical thinking for over 1,500 years, with no way to test the biological plausibility of these claims. The first major break from humoral theory came in 1637, when René Descartes proposed a reflex-based model of sleep in his *Treatise on Man*. Descartes argued that sleep was not caused by changes in blood flow or humors, but by a reduction in sensory input to the brain that triggered a passive withdrawal of the “animal spirits” from the sensory cortices to the brain’s interior. While still framing sleep as a passive state, Descartes’ model was the first to locate the cause of sleep in the brain rather than the body’s humors, laying the conceptual groundwork for later neurobiological models.