"Waking at three means something's wrong"

Before we get to the thinking, it's worth looking at the waking itself, because many people add a layer of alarm at this point. You surface at three and the first thought is, "Oh no, not again. Why can't I sleep like a normal person?" That thought alone can be enough to start the shift.

Waking in the night is normal. Sleep doesn't run as one unbroken stretch from bedtime to morning. It moves through lighter and deeper stages in cycles, and brief surfacings between them are part of how it works. Most are so short that you don't remember them. That's the standard account in sleep science, and I'm giving it from general knowledge without a particular study to cite. It's also well established that wakings become more frequent and more noticeable as people get older.

There's some intriguing evidence that a longer wakeful spell in the middle of the night may be natural to us. In the early 1990s the sleep researcher Thomas Wehr ran an experiment in which healthy volunteers spent fourteen hours of every twenty-four in darkness, roughly a winter night without artificial light. Their sleep changed shape. It spread out, and it usually divided into two blocks of several hours, with one to three hours of wakefulness in between.

About a decade later the historian Roger Ekirch published a study of sleep in Britain before the industrial age. He'd found references in diaries, court records and literature to a "first sleep" and a "second sleep". He argued that people commonly woke for an hour or so in between, and that they regarded it as quite unremarkable. According to his account they used the interval to pray, talk, tend the fire or lie quietly. I should tell you that this is disputed. Another writer, Niall Boyce, has reassessed the evidence and argued that Ekirch overstated how general the pattern was. Ekirch has replied, standing by his case. I'm not qualified to referee it.

I'm not suggesting you adopt a fourteen-hour night, or that you should want to be awake at three. But both pieces of work weaken the idea that a waking in the night means something has gone wrong with you. People have probably always had them. If Ekirch is even partly right, our ancestors lay awake in the small hours as a matter of routine, and it didn't occur to them that it was a disorder.

This matters because of how the shift begins. There's a gap between the waking and the first problem, often only a few seconds long. During that gap you're simply a person who's awake in a dark room, comfortable, with nothing required of you. What happens next isn't fixed. If your first response is alarm, then you've supplied the jolt of alertness that the night shift needs to get going. If your first response is closer to "Ah, I'm awake, that happens", you've at least left the door to sleep open.

It's also worth knowing that lying awake and resting isn't worthless. You aren't asleep, and it isn't as good as sleep. But a body lying still in the dark with a quiet mind is getting some rest, and is well placed to slide back under. Lying still in the dark while doing your household accounts and rehearsing an argument with your sister is different. That's work, and it's being done at a time when you should be resting.

So the waking isn't the problem, and you can stop treating it as one. That's the first small change, and it costs nothing. Some wakings do deserve a doctor's attention, as the preface said, and chapter 15 goes through them. For the ordinary kind, the trouble starts with what you do next.