The standard advice, repeated everywhere from GP waiting rooms to airline magazines, is this: adults need roughly seven to nine hours of sleep, taken in one continuous block, ideally at the same time each night. It is good advice in the way "eat vegetables" is good advice — broadly correct, lightly evidenced at the edges, and quietly shaped by the culture that issued it.
We want to take that prescription seriously and also take it apart. Sleep culture — the inherited assumptions about when, how long, and in how many pieces a person should sleep — varies more across history and geography than the one-block rule suggests. That doesn't mean the rule is wrong. It means it's narrower than it sounds.
Is one long block of sleep actually natural for humans?
The honest answer is: probably yes, under certain conditions, and probably not under others. When humans live with stable temperatures, dim evenings, and no artificial light past sunset, they tend to consolidate sleep into a single nocturnal bout of roughly six and a half to seven hours. When they don't — when nights are hot, lit, noisy, or socially active — they often split sleep into two or more episodes. Monophasic sleep is a real human pattern. It is not the only one.
The most-cited field data on this comes from Gandhi Yetish and colleagues, whose 2015 paper in Current Biology tracked sleep in three pre-industrial groups (the Hadza, San, and Tsimane) using actigraphy. Across 94 adults, average sleep duration was 5.7 to 7.1 hours per night, sleep was largely consolidated, and — notably — none of the groups napped routinely outside the hottest summer months. That finding cut against the assumption that "natural" human sleep is biphasic by default.
Where the standard advice is roughly right
Two parts of the conventional prescription are well-supported.
The first is total duration. Large epidemiological work, including Shawn Youngstedt's analyses and the consensus statement from the American Academy of Sleep Medicine and Sleep Research Society (Watson et al., 2015), converges on seven hours or more for most adults, with risks of cardiometabolic and cognitive harm rising below about six. The number is fuzzy at the edges, but the central tendency is real.
The second is regularity. A 2023 analysis by Windred and colleagues in Sleep, using accelerometer data from over 60,000 UK Biobank participants, found that sleep regularity predicted all-cause mortality more strongly than sleep duration did. Going to bed and waking at consistent times appears to matter at least as much as how long the night lasts. Whatever your schedule is, the body seems to prefer that you keep it.
Where the standard advice breaks down
Three places, mainly.
Historical segmented sleep. The historian A. Roger Ekirch, in his 2005 book At Day's Close and a series of journal articles beginning in 2001, assembled hundreds of references in pre-industrial European texts to a "first sleep" and "second sleep," separated by a quiet wakeful hour around midnight used for prayer, conversation, sex, or chores. Thomas Wehr's 1992 photoperiod experiment at the NIMH gave the pattern a possible physiological basis: when 15 healthy men were placed on 14-hour dark nights for a month, their sleep spontaneously split into two bouts with a calm, prolactin-elevated interval between. The sample was tiny, the conditions artificial, and the result has never been replicated at scale. Plausible but thin.
The siesta belt. Across parts of Spain, Greece, southern Italy, much of Latin America, and historically across the Mediterranean and West Africa, an afternoon rest of 20 to 90 minutes was — and in places still is — built into the day. Whether the siesta is a thermoregulatory adaptation, an agricultural artifact, or a cultural preference is genuinely unsettled. What the data does suggest is that brief afternoon naps (under 30 minutes) reliably improve alertness and don't appear to harm nocturnal sleep in habitual nappers. Longer naps are more ambiguous, with some cohort studies (notably Naska et al., 2007, Archives of Internal Medicine) initially linking them to lower cardiac mortality and later work finding the opposite for naps over an hour.
Hot, crowded, or socially dense environments. Carol Worthman's anthropological work in the 2000s documented sleep in settings where people share bedding, sleep in shifts, nurse infants through the night, or take their main rest during the heat of the day. In these contexts, the single-block night isn't an aspiration. It's a category error.
How a consolidated night actually assembles itself
It's worth walking through what the standard "eight hours" is doing internally, because that's what the advice is really protecting.
In the first 20 to 40 minutes after sleep onset, slow-wave (deep) sleep dominates, driven by adenosine pressure accumulated during the day. The first third of the night is when most of that deep sleep occurs, which is why a truncated early night costs more than a truncated late one. REM episodes lengthen as the night progresses, with the longest bouts in the final two to three hours — so a 5 a.m. alarm preferentially cuts REM, not slow-wave sleep. Core body temperature reaches its minimum roughly two hours before habitual wake time, and cortisol begins climbing about an hour before that.
A biphasic schedule can, in principle, deliver similar architecture if both bouts are long enough. The problem is that most modern split-sleep experiments — particularly the polyphasic schedules popular in optimization forums — sacrifice slow-wave sleep on the assumption it can be compressed. It cannot, reliably.
An honest version of the rule
The eight-hour-one-block prescription is best understood as a reasonable default for adults living under electric light, working fixed daytime hours, in temperate climates, who don't share a bed with a nursing infant. For that population, regularity and duration matter more than block count, and most people will land on monophasic sleep when both are protected.
For everyone else — shift workers, parents of young children, people in hot climates, people whose evenings are socially structured around 10 p.m. dinners — the rule is a useful baseline to deviate from intelligently, not a biological mandate.
One small thing to try this week
Hold your wake time constant within a 30-minute window for seven consecutive days, including the weekend, and let bedtime float. This is the lowest-cost way to test whether your dissatisfaction is about quantity, timing, or regularity — and it's the variable the strongest recent evidence points to.
The question worth asking isn't whether your sleep is normal. It's whether it's regular, sufficient, and yours.