Herman, I want you to picture the most specific thing a fifteen-month-old can be upset about.
Specific how?
He's fine. He's happy. He's in the room. You're in the room. And then the door swings shut, and he loses it.
Ah. Yes. That one.
Hannah sent this in about Ezra. He's fifteen months, he's in a separation-anxiety phase but a mild one. He'll stay with caregivers, he mostly handles the two of you leaving the room, he's still independent in the way that a toddler who's just learned his legs is independent. And the one thing that reliably sets him off, especially when he's tired, is a door closing. Not being blocked from a room he wants into. He's already in the room. He's happy there. The door shuts and he cries. And it isn't only the big doors. Sometimes it's cabinet doors too.
Cabinet doors.
And her question is the one you'd want to ask. Is this just separation anxiety wearing a different coat, or is something else going on with object permanence, control, transitions, the meaning of a barrier at this age.
That's a better question than it looks.
It usually is with her.
No, I mean it. Because the standard separation-anxiety story does not actually fit the case she described, and most people would never notice, because the story sounds right. Let's start with what we actually know about fifteen months, and then get to the part that doesn't fit, which is that the door is doing more work here than the room.
So begin with the age itself.
The age itself is a peak. Separation anxiety starts around eight or nine months, and it peaks somewhere between ten and eighteen months on the clinical side. The BabyCenter guidance stretches the peak to twelve to twenty-four. Either way, Ezra is sitting dead center of the window. There's also what daycare people call a second wave, fifteen to eighteen months, and the interesting thing about that wave is it arrives just as mobility and independence are taking off. He's not getting more clingy because he's getting less capable. It's the reverse.
Independence and reactivity climbing together.
Exactly that, and we'll come back to it. Now the mechanism under the peak, which is object permanence. The Merck Manual says it plainly: separation anxiety happens when infants understand they're a separate person from the caregiver but still haven't mastered the idea that something exists when it isn't seen or heard. And because they don't have a concept of time yet, they don't experience the departure as temporary. They experience it as permanent. BabyCenter puts the same thing in one sentence. Your baby doesn't understand that you'll come back.
Permanent. That's the whole engine.
That's the engine. Which explains why a parent leaving the room hurts. It does not explain why a door closing with the parent standing right there hurts. That mismatch is the episode, and I want to be honest about something before we go further.
Go on.
There is no study on this. I looked. There's no body of work on toddlers being distressed by the act of a door closing. Nothing on cabinet doors. You can run the obvious phrases through PubMed and get nothing relevant back. The closest published thing I found is from 1975, and it isn't about doors at all.
So we're improvising.
We're mapping a thing every parent has seen onto mechanisms that are documented. That's different from reporting a named phenomenon, and the difference matters. It means everything today is a best reading, not a finding.
Fine by me. State your best reading. There are four mechanisms people reach for, and my guess is they don't all fit.
They don't all fit, and the first one we already have. Object permanence. You exist, you're gone, and I can't hold onto the fact that you're coming back. That's why separation hurts. Hold it in one hand, because we're going to need it later and it's going to be the one that fails.
So the door isn't the whole story, or permanence isn't.
Permanence isn't. Door might be.
Second mechanism.
Barriers. The 1975 study is Van Lieshout, in Child Development. Sixty-four children, observed at eighteen and twenty-four months. The mother takes a toy, and instead of hiding it, she puts it inside a Plexiglas box. Right in front of the child. Visible the whole time. Unreachable.
Visible and locked. That's cruel by design.
That's the point of the design. It separates the variables. The child can't tell himself the toy is gone. It's right there. So any distress he shows is about the barrier, not about absence. And the canonical result is that the barrier itself changes the child's behavior. That's the closest published analog to what Hannah's describing: something you can see through, that you can't get past.
Except the toy in that study is something the child wants. Ezra isn't trying to get into the next room. That's the part I keep bumping on.
Which is why I said closest analog, not match. The study is about a barrier to a desired object. A door closing on a happy child is a different animal. But there's a second line of work that gets us closer, and it's not about access at all. There's a 1977 paper in the Journal of Personality and Social Psychology on psychological reactance in two-year-olds. Boys preferred an object behind a barrier only under certain barrier conditions. The barrier didn't just block the object. It changed how much the child valued it.
The barrier does something to the meaning of what's behind it.
That's the finding. A barrier is never neutral furniture. It changes the object.
So a door isn't just a door.
It's never just a door.
Third mechanism, and I think this is the one under the cabinet.
Control. Autonomy. ZERO TO THREE, in their material on self-control from twelve to twenty-four months, says toddlers get frustrated because there are still many things they want to do and cannot, and that young children don't like to feel out of control. There's a nice parenthetical in the original. Tantrums don't feel good to them either. Which is easy to forget when you're the one listening to it.
They're not enjoying this.
They're not running a campaign. They're drowning.
That's kinder than I'm usually inclined to be at seven in the evening.
And it's developmental, not personal. Strong Start, writing about fifteen-month tantrums, puts the prefrontal cortex not fully developed until the mid-twenties, and says at fifteen months it's barely getting started. Impulse control, emotional regulation, rational decision-making. None of it is online yet at the level the situation demands.
The equipment isn't installed.
The equipment is half-installed and someone keeps closing doors on it. And the same source flags autonomy as brand new and overwhelming at twelve to eighteen months. There's also the self-determination-theory work on toddler autonomy, Côté-Lecaldare and colleagues, which describes toddlerhood as the period where the issue of autonomy is central, where children begin to assert themselves and want choices. The drive is new. The capacity to manage it isn't. That's the third mechanism and it's the one that covers the cabinet, because a cabinet door isn't a separation from a person. It's a barrier to a thing or a space he was using.
Fourth.
Transitions and the no. Strong Start lists transitions and being told no as the top fifteen-month triggers, and there's a line in there I want to use because it's better than anything I'd write. No feels like a wall they didn't see coming.
A closing door is both of those at once.
Both in one motion. It's a state change, open to shut, that the child didn't initiate and can't undo, and it's a no, because the space behind it is now closed to him. Add the barrier meaning from the reactance work, add the autonomy drive, and you get something more than the sum of the parts.
So a door swinging shut is a transition, a no, and a barrier, all in the time it takes to swing.
And the child is fifteen months old and has fewer than ten words. There's no language to put any of that into. The frustration is pre-verbal.
So the honest answer to Hannah's question is the one nobody wants on a podcast.
Which is?
It's all of them. Not one. Not the clean one.
There's no source I found that says door-closing distress is just separation anxiety. Nothing even says it's primarily that. The evidence supports overlap, and the exact weighting, we don't know. I'd rather say that plainly than hand her a tidy mechanism that happens to be wrong.
You know what strikes me about all four mechanisms.
What.
Every one of them is about the child losing authorship of something. Permanence, the door changes what's behind it. Barriers, the door is between him and it. Control, he didn't close it. Transition, he didn't choose the moment. The common thread is somebody else decided.
That's the cleanest way I've heard it put.
I'd have said it more slowly.
You did, actually.
So that's the toolkit. Now the harder question. Why is a kid who's fine with caregivers and fine with the two of you leaving the room losing it over a cabinet door.
Start with the second wave, because it's the bit that resolves the apparent contradiction. The fifteen-to-eighteen-month spike in separation anxiety is commonly associated with exactly the period when mobility and independence grow. Those two things are correlated, not in tension. A child who ranges further has more territory to lose when something seals it off. He's not more reactive despite being more independent. He's more reactive and more independent for the same underlying reason, which is that his world just got bigger and he knows it.
Bigger world, more to be cut off from.
And here's where I'd plant a flag on the cabinet door, because I think it's the strongest evidence in the whole discussion and it deserves its own moment.
Plant it.
A cabinet door is not a separation event. There's no person behind it. Whatever is behind it is a stack of containers or a pot or a drawer he was about to put his hand into. So the reaction to a cabinet door closing cannot be explained by separation anxiety. It has to be explained by the barrier-and-control mechanisms. Which means the door-closing reaction isn't a variant of separation anxiety that happens to fire on doors. It's a barrier phenomenon that overlaps with separation anxiety because both of them are running at full volume at fifteen months.
Which means the door was never about you leaving.
The door was never about us leaving. It's about the door.
Say more about the cabinet, because I want the specific version.
The cabinet is a barrier to access, which is the Van Lieshout paradigm almost exactly. Visible, unreachable, and the child was mid-action. And there's something worse about it than the Plexiglas box. The box is a novelty someone else introduced. The cabinet is something he was using. He had his hand in there. Him closing a door he was interacting with is closer to having an activity taken away mid-motion than to having a person taken away.
The door is the sound of the activity ending.
That's it. And when you reframe it that way, a few other things click. It's not fear of abandonment. It's the ending, unilaterally declared, of a thing he was in the middle of.
Which is why the parent being in the room doesn't help.
It can't help. The parent being in the room isn't the missing piece. The missing piece is the open door.
Then the cut-off-from-the-world intuition is half right.
Half right and structurally misleading. Object permanence explains why separation hurts. You exist, you're gone, and I can't hold onto you'll come back. But that model predicts distress when a person disappears. A door closing with the person still in the room doesn't fit the model. Which is why the answer isn't separation anxiety plus two smaller things. It's the other mechanisms carrying the weight, with separation anxiety running alongside because of the age.
So the permanence engine is idling, not driving.
That's a good way to say it. It's the loudest mechanism, and I think it's the one most parents reach for first because it's the one with a name. It's just not the one doing the work today.
Then there's the tiredness, which is the part of Hannah's message I found most interesting.
Why that part.
Because it's the piece that sounds like parental impression and turns out to be written down in a medical reference.
It is written down. The Merck Manual, in its guidance on managing separation anxiety, advises feeding the child and letting him nap before the parents leave, and gives the reason in the text. Separation anxiety may be worse when a child is hungry or tired. That's a mainstream clinical source directly validating the observation that state changes the reaction.
Not a vibe. An instruction.
With a stated mechanism. Which is rare, and I want to be honest about the shape of the evidence, because it's a clinical note, not a controlled trial. I couldn't retrieve a peer-reviewed citation establishing the sleep-deprivation-to-reactivity link in toddlers specifically. So the link is real and it's written into clinical guidance, but the underlying study I'd want, I didn't find it.
What's the rest of it.
Strong Start, on the same fifteen-month tantrum material, says a tired or hungry toddler's window for coping is paper-thin, and that the classic tantrum before lunch or at the end of the day is almost always about physical state, not behavior. It calls protecting sleep the highest-leverage prevention tool a parent has. There's also a sleep-regression write-up that ties the fifteen-month window to the same developmental events we've been discussing, the language circuits firing, the memory consolidating, the object permanence locking in, and makes the point that the child is becoming acutely aware that you can disappear. The reorganization and the reactivity arrive together.
So the tiredness isn't just a dimmer switch on the same reaction.
The tiredness is the difference between the reaction being available and the reaction being the only tool in the kit. A rested toddler has some capacity to be annoyed about a closed door and go do something else. A tired one doesn't have the spare capacity to do anything but protest. Same trigger, fewer resources, louder output.
You've watched this in the flat.
Every evening. The door to the kitchen, specifically, which closes at noise level rather than volume level. The interesting thing is that the same door closing at ten in the morning is unremarkable. At six, it's an event.
Same door. Same child. Same swing.
And then there's the language gap, which is the bit that makes me feel worst about the whole thing.
Fifteen-month-olds typically have fewer than ten words. So he can't say I didn't want that door closed. He can't say I was using that cabinet. He can't say I was about to open it and you took the moment away. The complaint exists in full inside him and there's no channel for it. So it comes out as the only output he has.
And the output is one to three minutes long.
Peak intensity at this age runs one to three minutes. Short and intense. Which is actually useful to know if you're on the receiving end, because it means the storm is not a slow build that will run all evening. It peaks fast and it passes.
What I keep coming back to is how much of toddler upset, if you actually watch it, is about a change of state the child didn't author.
You're going to make that the thesis, aren't you.
I'm not going to make it anything. I'm going to say it once and let it stand. From his side of the room, every one of these events is the same shape. Something was one way, it became another way, nobody asked him, and it may or may not go back. The door is the visible edge of that. It's legible to us because we can see the door. The same thing is happening all day in ways we can't see, and it's all the same underlying lesson, which is that the world has states you don't control and some of them reverse and some of them don't.
That's the real developmental project.
And he'll spend the next two years on it.
Longer than that.
It's a lot of homework for a person who can't say the word door.
Which takes us to the part of the show where someone who's actually thought about doors as a profession weighs in.
Thought about doors as a profession.
I said what I said.
Ninety-eight dollars. That's a hollow-core interior slab. That's the door that sounds like a verdict.
What was the job, Hilbert.
I adjusted doors. Forty-seven units in the building. Hollow-core, prehung, three hinges, and none of them hung straight from the factory. You shim the hinge, you get the reveal even top to bottom, and the door closes the way it's supposed to close. Then you do it again, because the first pass is a warm-up.
Twice. In every unit.
Forty-seven times two. Ninety-four doors. I had the list.
You kept a list.
You have to keep a list. You don't remember ninety-four doors.
Of course you don't.
The gap is what people get wrong. Everybody thinks it's about the door being hard to close. It's about the gap. There's a gap between the slab and the jamb, and a hollow-core door with a wide gap doesn't have a sound, it has a boom. You hear it in your chest. A solid door shuts and you barely notice. A cheap door shuts and the whole room knows.
So the door doesn't just close. It announces.
And that's before we talk about the cabinet.
I was hoping you'd get to the cabinet.
The cabinet is the real tell. You can tell the kid isn't scared of being alone. If he was scared of being alone, the bedroom door would do it, the front door would do it. He'd be on you. He isn't. He's fine in the room. What gets him is the cabinet, and the cabinet doesn't separate him from a person. It separates him from a thing he was about to touch. That's a different crime.
So the door closing is the sound of something ending.
He isn't scared of being alone. He's scared of being done. He was in the middle of a thing and the door said the thing is over, and nobody asked him.
That's the whole episode, isn't it, Hilbert. You just said it in three sentences.
What I said is what I said.
Did you ever see it work the other way. Opening the door back up.
Once. The neighbor's kid. Not mine, and it was brief, I was watching him for a few minutes while they moved something in. He was doing the thing, the crying, and I happened to open the cabinet door back up and just left it open. He stopped. Instantly. Like a switch.
How long did you stand there holding it open.
Eleven minutes. I counted, because it kept working.
You stood there holding a cabinet door open for eleven minutes.
He was fine the whole time. I have never been more certain of anything in my life. And the thing I want to add is that the door isn't the problem. It's never the door. It's the ending. If you open it back up, you'll see. The ending is what he's objecting to.
And what happened with the forty-seven doors.
I over-adjusted one. I got the swing so gentle that it wouldn't latch. Beautiful door. The best work I ever did on a slab. Tenant complained it wouldn't commit.
Wouldn't commit.
Said it wouldn't commit. So they let me go.
And the door.
I kept it. It's in the garage. The one that got away.
Hilbert, that's the worst door-related thing I have ever heard and I want you to know I mean that as a compliment.
It's not a compliment, but I'll take it.
There's something in what you said, though. The paper-thin window. The rested child can hold onto the open-door idea and the tired one can't. But your point was narrower than that. If the ending is the problem rather than the door, then re-opening it isn't a trick, it's actually addressing the thing. You don't have to explain that the ending reverses. You have to show him.
Which leaves the honest gap where we found it. No dedicated research on door-closing distress. The multi-factor reading is the best available, not settled, and the tiredness link rests on a clinical note rather than a controlled study.
And the timeline is the reassuring part. Merck puts resolution by around two years. BabyCenter and Cleveland Clinic stretch it to three. The second wave is meant to intensify before it fades, which means the next few months are more of this, not less.
The one most common wrong belief here is that this is just separation anxiety, and I want the correction on the record.
The parent still in the room is the whole proof. If it were separation anxiety, the child would settle the moment you sit down next to him. He doesn't, because the door isn't standing between him and you. It's standing between him and the thing he was in the middle of, and the permanence engine has nothing to do with it.
So he isn't learning that you come back. He's learning that the world has states, that some of them reverse, and that the one who decides isn't him.
One door at a time, for the next two years.
That's the thought to leave it on.
And a thank you to Hilbert Flumingtop, our producer, who has been holding that door open since the mid-eighties.
If this was your kind of episode, go back for episode twenty-five ninety-five, Baby-Proofing a Small Rental; episode fifty-four forty-eight, Toddler Mom-Clings After Daycare; and episode twelve oh five, System Update. This has been My Weird Prompts.
The human-AI collaboration podcast. If you enjoyed it, leave us a review, it helps people find the show.
And send us your own prompt on Telegram at t dot me slash MWP listener bot.
We'll be back soon.
See you tomorrow.
