Almost every new parent arrives with an expectation of how a baby sleeps, assembled from books, from advice, and from other people's accounts. Almost none of it survives the first fortnight.
The distress that follows is largely about the gap between expectation and reality rather than about the sleep itself.
What newborn sleep actually looks like
Newborns sleep a lot in total, distributed across the whole twenty-four hours in short stretches. There is no day-night distinction initially because the circadian system is immature.
Circadian rhythm develops over the first months, with melatonin production establishing gradually. Before that, expecting a baby to sleep at night is expecting something the biology doesn't yet support.
Sleep cycles are also much shorter than adult ones — around fifty to sixty minutes rather than ninety — with a higher proportion of active sleep, during which babies twitch, make noises, and appear to be waking.
That last point causes a specific problem. Parents frequently intervene during active sleep, believing the baby is waking, and the intervention wakes them.
The night-waking question
Waking at night is normal well beyond the point most guidance implies.
Studies using objective measurement rather than parental report have consistently found that a substantial proportion of infants continue to wake at night through the first year and beyond.
Crucially, the same research finds that many babies described as sleeping through are in fact waking and resettling without signalling — the change is in whether they call out, not in whether they wake.
So "sleeping through the night" describes a behaviour, not the absence of arousals, and the developmental milestone is self-settling rather than continuous sleep.
There's also considerable individual variation, and it appears to be substantially constitutional rather than a consequence of parenting.
Where the books go wrong
Several recurring problems.
Presenting averages as targets. Sleep duration figures are averages across wide distributions. A baby sleeping considerably more or less than the stated figure may be entirely normal.
Implying that a method produces a result. Books frequently describe a routine and the outcome that follows, without acknowledging that babies vary enormously in how they respond.
Selective reporting. Advice books are written by people whose methods worked for the babies they encountered. The babies for whom nothing worked don't appear.
Confident schedules. Rigid feeding and sleeping schedules for very young infants are not supported by evidence and can interfere with feeding, particularly breastfeeding, which depends on frequent demand-led feeding to establish supply.
What has reasonable support
A short list of things with some evidence behind them.
Light exposure. Bright light during the day and dim light in the evening supports circadian development. Cheap, easy, and probably genuinely helpful.
A consistent bedtime routine. Studies of bedtime routines have found associations with better sleep outcomes. The specific content matters less than consistency.
Putting a baby down drowsy but awake, where feasible, on the reasoning that falling asleep independently supports resettling. This is easier to say than to do and doesn't work for every baby.
Responding to hunger cues rather than a clock in the early weeks.
Sleep training, briefly and carefully
A contested area where the evidence is more limited than either advocates or opponents suggest.
Trials of behavioural sleep interventions in infants over about six months have generally found short-term improvements in parent-reported sleep and in parental mood, with follow-up studies not finding evidence of harm to attachment or emotional outcomes.
Critics point to methodological limitations, reliance on parent report, and the difficulty of studying long-term effects.
What's reasonably clear: these approaches are not recommended for young infants, they don't work for every child, and the decision is a family one made in a specific context rather than a universal prescription.
What's also clear is that severe parental sleep deprivation carries genuine risks — for mental health, for driving safety, for the capacity to care responsively — and those risks are real considerations rather than reasons for guilt.
The thing that helps most
Reducing the expectation gap.
Parents who know that frequent waking is normal report less distress than those who believe something has gone wrong, at the same actual level of sleep disruption.
The practical consequence is that accurate information is itself an intervention. Not because it produces more sleep, but because a difficult period you understand is easier to sustain than one you think you're failing at.
And it does change. Not on any schedule anybody can promise, and it does.
When to seek advice
Some situations warrant a conversation with a health professional rather than more reading.
Poor weight gain or feeding difficulties. Persistent breathing irregularity, snoring or pauses. Extreme irritability that doesn't settle. Any concern about a baby's development.
And for the parents: if you are not coping, if your mood is persistently low, or if you are having frightening thoughts, that warrants support urgently. Postnatal mental health difficulties are common, treatable, and frequently unreported because of the assumption that struggling means failing.
General information only. Any concerns about your baby's health or sleep should be discussed with a qualified healthcare professional.