Picture the deli counter at one in the afternoon. Twelve people deep, one guy on the slicer, and everybody in that line is quietly seething.
And Daniel has decided this is a solvable problem.
He wrote in, and I'm going to put it across properly. He says some people have argued that breaking into song spontaneously should be normalized. He points at the average musical, where the narration is punctuated by random people deciding to create a group singalong. And then he asks whether a person could actually be the vanguard for that change. His own example is the sandwich line. You're frustrated, you're waiting, and instead of just seething, you break into song.
He's got four separate questions buried in there and I want to make sure we hear them all.
He does. Question one, do any world cultures have a tradition of accepted spontaneous singing analogous to the environment musicals depict as reality. Question two, if not, how would a person break through the shackles of self doubt and societal expectation. And question three, the big one, what would it take to build a world where breaking into song is as natural as having a conversation. So let's unpack why we accept this on a stage and not on the street.
The stage version has a name and a logic. In musical theater, the convention is that song is an intensification of emotion. When dialogue can no longer carry the feeling, the character stops talking and sings. It's a staging device that externalizes a shared emotional state. It was never a claim about how humans behave.
Twenty five people on a boat deck in a Rogers and Hammerstein number isn't a depiction of reality. It's a depiction of one feeling, magnified.
Which is exactly right, and it's why the gap Daniel is pointing at is real. We all accept the convention inside a theater for two hours. We walk out the doors and the convention evaporates.
So the question underneath his question is whether the convention is describing something we lost. Whether group singing is a human universal that got paved over.
That's where I'd start, because the evidence on that is stronger than I expected. We should do the science first, then the culture, then the psychological wall, then the playbook.
Four parts. Science, culture, wall, playbook.
And I want to flag something up front. There's a specific gap in the literature that I think answers part of Daniel's question by itself.
Hold that, because I want to hear it after we've earned it.
Fair.
Start with the infants.
There's a body of work by Malloch and Trevarthen, published in Frontiers in Psychology, arguing that what they call communicative musicality is the building block of music, dance, and theater. It's the timing and shape of vocal and bodily interaction between an infant and a caregiver. The rise and fall of a mother's voice, the pause, the reply, the timing. That's the seed.
So we don't learn music as a skill. We arrive already doing it.
That's their argument. The infant is not a passive recipient of sound. It's in a duet, with timing, with turn taking, with pulse. And they argue this strengthens emotions of social resilience. Their line is that music is at the centre of what it means to be human, the sounds of human bodies and minds moving in creative, story making ways.
Which means a person who says they can't sing is saying something strange. They've been singing in time with another human since before they had words.
And then we taught them to be embarrassed about it.
I want to put the sandwhich line next to that for a second.
Go on.
The reason that man is seething rather than singing isn't a shortage of musicality. It's that the duet he had as an infant has been replaced by a queue, which is a structure designed specifically to prevent duets.
That's a better framing than anything I've got.
Keep going and I'll try to find another one.
The measurable part. There's a study in Scientific Reports from twenty twenty five, Good and colleagues, which took people with Parkinson's and put one group into group singing and another into yoga, and compared them. The singing group showed greater overall social closeness and significant increases in oxytocin.
Oxytocin being the bonding hormone.
The one everyone files under bonding, yes. And the mechanism the authors point at is interpersonal movement synchrony. Not vocal quality. The synchrony of moving together.
So the payoff comes from the coordination, not the sound.
That's the claim, and it matters for Daniel's whole question. A group doing something badly but in time together is still getting the chemistry.
Then there's the follow up from twenty twenty six.
There is. Same lead researcher, published in Arts and Health, and it looked at within session effects. Improvements in mood, social bonding, pain thresholds, and reduced cortisol, all measured inside the session itself. And the benefits accumulate across sessions rather than resetting.
Cortisol down, mood up, pain threshold up, in the same ninety minutes.
Which is a strange sentence when you put it that way. And here's the finding that I think is the single most important thing in this entire episode.
Say it slowly.
Self reported singing abilities did not impact the benefits gained.
So the person who thinks they're terrible gets exactly what the person who thinks they're good gets.
As far as that study can measure it, yes. The self assessment didn't predict the outcome.
That's the entire wall. Daniel's whole question about self doubt is answered by one line from a paper.
It answers the logic of it. It doesn't answer the feeling of it, and I think we need to be honest about that distinction. Knowing that ability doesn't matter is not the same as being able to open your mouth on a Tuesday.
Noted. Keep going.
There's a European study, Grau Sánchez and colleagues, looking at musical activities across five countries, and they describe musical activity as an important source for social contact, promoting interaction and the sense of belonging. The population was older adults, so I wouldn't generalise it too far, but the direction is consistent with everything else.
Five countries, same result.
Same result. Now, culture. Which is where I have to be careful.
Careful how.
The specific named traditions are well known to me and to most listeners, but I couldn't verify them freshly, and I'm not going to state them as if I'd just read the citations. Work songs, sea shanties, call and response, the Nordic communal singing tradition, the Welsh eisteddfod culture. I know these things. I did not confirm them this week.
So you're saying them as general knowledge and flagging that.
That's what I'm doing, yes.
Fine. Say them.
Sea shanties are the cleanest example in the whole catalogue, because they were functional. You're hauling a rope on a ship and the song is a timing mechanism. One man calls, everyone responds, and the response lands on the pull.
So the song had a job.
The song was the job. It was a coordination technology before it was anything else.
And nobody in that crew was worried about whether their voice was good.
Nobody. Because the only criterion was whether the rope moved. That's the environment Daniel is asking about. It existed. It existed on ships, in fields, in mills.
What happened to it.
The coordination got taken over by machines and by clocks, and the song became something you go and do in a room with other people who have chosen to be there. Which is a very different relationship.
The call and response traditions are probably the strongest answer to Daniel's cultural question.
They are, and there's a distinction I want to draw. Call and response in a congregation or a community setting isn't spontaneous. It's a structure. There's a leader, there's an expected response, and everybody knows when their line comes.
So it's not that a culture accepted spontaneous singing. It's that certain cultures built containers where singing was scheduled and expected.
Which is arguably better than spontaneous, because it removes the decision. Nobody has to decide to sing. It's simply the next thing that happens in the service.
That's the whole trick, isn't it.
Say more.
Every functioning example of this we've got is a structure that removes the moment of individual choice. The shanty has a call. The congregation has a hymn. The eisteddfod has a stage. Daniel's sandwich line has nothing. It's the absence of any container at all that makes it hard.
Which is the best explanation I've heard for why this is hard and why nobody does it. The problem is not courage in the abstract. The problem is that a queue is a container with no musical function whatsoever.
All right, so the answer to his cultural question is roughly yes and no.
Yes, humans clearly have a deep practice of group singing and it does the bonding work. No, I can't point to a modern culture where a stranger breaking into song in a shop is treated as unremarkable.
Which is an honest answer and probably why this is worth an episode.
The wall, then.
The wall.
I went looking for the self consciousness literature specifically. Singing plus embarrassment, self consciousness, social anxiety. And I found essentially nothing.
Nothing.
A search that specific returned no results. There's a large literature on choirs and on music therapy and on structured singing, and it's all about people who have already signed up and shown up. There's almost nothing on the moment before, when a person is standing in a public place deciding whether to open their mouth.
That's a striking absence.
It's the exact moment Daniel is asking about, and it appears to be unstudied. Which is a real answer to his question, not a dodge. Nobody has measured the thing he's describing because it doesn't happen often enough to measure.
That's a good finding. Depressing, but good.
It also tells us the barrier is cultural rather than clinical. If it were a recognised anxiety disorder we'd have scales for it and people would be studying it. It's not that. It's just a norm nobody has written down.
Unwritten norms are the strongest ones. Nobody can point at the rule, which means nobody can argue with it.
And that's why the ability finding matters so much. The thing people offer as their reason, that they can't sing, is measurably not the thing that determines whether they get anything out of it.
So the stated reason is wrong.
The stated reason is wrong, which means a person who wants to be the vanguard can't argue someone out of it by proving they can sing. Because singing ability was never the objection.
The objection is being seen.
Being seen doing something that has no social cover. Nobody quite knows what the response should be.
So what does the playbook look like. Because Daniel asked for one, basically. He asked how you get through the shackles.
Four moves, and none of them are heroic. First is synchrony. The oxytocin and social closeness findings point at doing it together as the active ingredient. One person singing in a deli is a person having an episode. Four people singing in a deli is an event.
Which raises the obvious logistical problem that you can't assemble four people by yourself in a sandwich queue.
You can if you bring them. Which is deeply unromantic and probably the actual answer. The vanguard doesn't convert the room. The vanguard arrives with two friends and does it.
That's honest. Two friends and a sandwich order. Not a flash mob.
Second move, reframe competence. Since ability doesn't predict benefit, the leader's job is to model enthusiastic imperfection. You have to be visibly fine with being bad at it, because that's the permission the room is waiting for.
Badly and loudly.
Enthusiastically, at minimum, and yes, badly is fine. Third, start where the container already exists. Choirs, community groups and therapeutic settings are proven on ramps. Hendry and colleagues in Qualitative Health Research identified social bonds, group identity and self efficacy as the mechanisms that make choir singing work.
So a choir is a room where the norm has already been written down. You're just borrowing it.
You're borrowing it until you can carry it out of the room. And the fourth move is borrowed from a completely different discipline. There was a discussion on Hacker News this year about social courage, and the recommendation that came out of it was improv classes.
Improv.
The framing someone used was that improv gets you used to laughing at yourself and making a fool of yourself. It's essentially a controlled environment for being publicly wrong.
Which is the same skill. You're not training singing. You're training the thing that stops you from singing.
Precisely that. The bottleneck is not the larynx. The bottleneck is the tolerance for being looked at while you do something unscripted.
I like that Daniel's question has a training regime.
I don't know that it's a regime. I'd say it's four plausible moves drawn from evidence that mostly wasn't collected for this purpose. I'd hold it loosely.
Hold it loosely, fine. Now the vision. What would the world look like if he pulled it off.
The measured outcomes from the choir literature give you the shape of it. Higher oxytocin, lower cortisol, better mood, stronger social bonds, reduced loneliness. Han and colleagues in twenty twenty five found choir singing improved international students' sense of belonging, which is a pretty clean test case, because those are people with almost no social scaffolding.
That's not a small thing. Belonging is the thing loneliness is made of.
And the public health framing is explicit. Williams and colleagues in the European Journal of Public Health describe group singing as a promising social intervention for people with mental health conditions.
So normalising this isn't whimsy. It's not a novelty argument. It's a loneliness intervention that happens to be free.
That is what the literature is pointing at, with one large caveat.
Give me the caveat properly.
The same papers repeatedly note moderate to high risk of bias, and small samples. Singing studies attract motivated participants and are hard to blind, obviously. The benefits show up consistently across studies and across populations, which is why I take them seriously. But I would not stand up and say group singing cures anything.
So the shape is right, the magnitude is uncertain.
That's a fair way to hold it.
Then here's where I land on Daniel's actual question. He asked whether one could become a vanguard, and I think the answer is yes, but not in the way the word implies.
Meaning what.
A vanguard suggests going first and being followed. What the evidence suggests is that you can create a small pocket where the norm is different, and it exists for exactly as long as you keep it alive with other people.
A pocket, not a movement.
A pocket. And every existing example we named, the shanty, the hymn, the choir, was a pocket that somebody maintained. None of them were natural. All of them were built.
Hilbert: What happens if you mess up a lyric in front of paying customers.
Hilbert: I spent a summer doing singing waiter work. It was a place with a theme, you took the order, and every twenty minutes the staff had to break into a song.
Every twenty minutes exactly?
Hilbert: They had a bell for it. You had to be done with whatever you were doing and in position. I was terrible. I couldn't hold the tune and I never learned the second verse of anything.
And the customers?
Hilbert: They loved it. That was the strange part. The bad ones got more out of the room than the good ones did. There was a girl there who could really sing, and people would watch her and then go back to their food. When I did it they'd put their forks down.
Why do you think that is?
Hilbert: Because she was doing a performance and I was doing the same thing they'd have to do. There was a house rule, if you dropped a lyric you had to start the line again, louder.
Louder.
Hilbert: Louder. That was the whole punishment. You'd have to go back and do the line again with your chest out. The first week it's humiliating. By the end of the summer it's the best part of the shift, because you're waiting to see who fluffs it and how loud they go.
So the failure was scheduled.
Hilbert: The failure was the entertainment. Nobody clapped for a clean run through. They clapped when somebody fumbled it and came back louder.
That's the thing we didn't have. We've been talking about normalising singing. What that restaurant normalised was failing, on a bell, in public, with applause.
Hilbert: And it worked better with the staff than any of the singing did. We were a bunch of students who'd never met. By August we knew each other better than people I lived with for a year.
That's the mechanism, actually. The synchrony finding pointed at doing it together as the active ingredient. He's describing something narrower than that. He's describing a ritual that makes being wrong into a shared event.
So the permission isn't for the singing. The permission is for the getting it wrong.
Hilbert: The getting it wrong is the part nobody has a container for. You can't just ask people to be vulnerable in a shop. You have to give them a reason to be, and the louder line was that reason.
It also explains why the twenty minute bell matters. It removed the decision. Nobody at that restaurant had to decide to sing, because the bell decided.
Which is the same thing the shanty and the hymn were doing.
Hilbert: One of the lads from that summer runs a pub quiz now and still makes people sing the answer if they get it wrong.
Does he.
Hilbert: It's the same idea. Nobody's good. Everybody does it anyway. It's the busiest quiz night in the place.
There's a practical thing I want to take from this. If someone tries Daniel's version, the sandwich line with nothing arranged and no group, the failure has no container at all. Which is why it can't work as the first attempt. But you could build the container, the way that restaurant did.
Which means the vanguard's job is not to sing in a deli. The vanguard's job is to invent the bell.
That's a better answer than anything I had.
Hilbert: The bell was just a kitchen timer. It wasn't even a nice one.
The most common wrong belief about all of this is that the barrier is musical. That people don't sing because they can't.
The paper says otherwise. Self reported singing ability did not impact the benefits gained. Whatever is stopping people in a queue, it is measurable not the voice.
So if Daniel wants to be a vanguard, what's the next step. What does the man in the sandwich line actually do on Monday.
Find two people. Do something badly together on a schedule. Pick a container that already exists if you don't want to build one. And let somebody fumble it in front of the group, on purpose.
He won't get the subway platform. But he might get four people in a deli who make it a thing.
And if it's the bell that makes it work, then the open question is what happens to a norm like this once it's held up by nothing but people deciding, over and over, that it's fine.
Which is where I'll leave it for listeners. If the wiring is there from infancy, and the bonding chemistry shows up every time anyone bothers to test it, then the thing we've lost isn't the ability. It's the container. And the next time you're in a queue and frustrated, the honest question is what it would take to build one in that exact spot.
That's the episode. Our producer is Hilbert Flumingtop, who has a better ear for this than either of us.
This has been My Weird Prompts, the human AI collaboration podcast. If you liked it, leave us a review wherever you're listening.
We'll be back soon.
Soon enough.
