My first slept through at 8 weeks. My second destroyed everything I thought I knew.

There is a version of me that existed until about fourteen months ago — confident, mildly insufferable — who believed she understood babies.

I would like to publicly apologise to everyone who knew that version of me.

Noah was, by any objective measure, an easy baby. This is not something I said out loud at the time, because I knew enough not to. But internally — and occasionally in carefully worded things I said to close friends who I trusted not to curse me — I attributed his easiness, in part, to the things we'd done.

The routine. The schedule. The consistent bedtime. The not-picking-him-up-every-time-he-made-a-sound philosophy I'd developed partly from books and partly from a very confident friend who kept citing "sleep hygiene" like it was a science she'd personally validated.

Noah slept through at eight weeks. By four months he was doing seven-to-seven with one brief waking. By six months we'd solved it. I had solved it. The implication — unstated, but there — was that other parents who were struggling had simply not approached the problem with sufficient rigour.

I think about that person now and I feel something between embarrassment and compassion, because she was about to learn something.

Léa arrived sixteen months after Noah. She was — is — small and determined and has his exact same nose and absolutely none of his relationship with sleep.

The first week, I wasn't worried. Every baby is different in the early weeks; I knew this. The second week, I started adjusting the routine I'd used with Noah. Third week, I adjusted it again. By month two, I had run through every variation of every strategy that had worked last time and was beginning to suspect, with growing unease, that I was doing something wrong.

I was not doing anything wrong. That took me several more months to understand.

What I was doing was applying Noah's solution to Léa's problem, which is a bit like using the key to one lock and wondering why a different door won't open. Léa's sleep window was different from Noah's. Her nap needs were different. The things that settled her were different. The whole architecture of her sleep was different, and I kept arriving with Noah's blueprint and being surprised when it didn't fit.

There was a period — I'm going to say months four through six — that I do not recommend. My husband Sébastien started gently suggesting that we try something new about once a fortnight, and I resisted each time with the particular stubbornness of someone who is wrong but hasn't admitted it yet.

«What worked with Noah—» I'd start.

«Léa isn't Noah,» Sébastien would say, very carefully.

«I know she isn't Noah,» I'd say. «But the principle—»

The principle, it turned out, was not transferable.

The thing that broke my certainty was a conversation with my GP, who asked — not unkindly — whether I'd actually assessed what Léa's schedule looked like, specifically. Her wake times, her nap duration, how long she'd been awake before we started the bedtime process. Not what I thought it looked like. What it actually was.

I hadn't. I'd been working from intuition accumulated through Noah, which is not the same thing as information gathered from Léa. When I actually wrote it down — when I answered the questions the way you answer them for a specific baby, not a general idea of a baby — I saw it immediately. Her wake window before bed was forty minutes too short. She was going to bed undertired and then running out of sleep pressure at 2am and waking up. Every night, reliably, for five months.

Léa goes down at 7:50 now. She wakes once, sometimes, around 4am — feeds briefly and goes back. It's not perfect. She's nine months old and not Noah and doesn't have to be. Some nights Sébastien and I lie in bed and listen to the monitor and marvel, with full and sincere gratitude, at the silence.

The most useful thing I did — more useful than any of the things I'd tried before — was to stop treating Léa as a version of a problem I'd already solved, and start asking questions about her, specifically: this child, this age, this pattern. The answers were different from Noah's. The solution was different from Noah's. It worked.

I've stopped giving advice to new parents. When people ask me what I did with my children, I tell them what worked for Noah and then I tell them that Léa required something completely different, and that the second part of that sentence is as important as the first.

I also no longer use the phrase "sleep hygiene" like I invented it. I'm working on the humility.

*If you have a second — or third, or fourth — child and nothing from last time seems to work: the assessment starts fresh, with this specific child, this age, this pattern. Not a template. An actual look at what's happening.*

More on these topics