There's a category of performance that nobody involved can see clearly while it's happening, and everybody can see clearly about five years later.
The comeback. Specifically the one that shouldn't have been attempted.
Daniel's put his finger on it. He wrote in about Céline Dion's return to the stage in Paris this past September, and about how split the room was. One French journalist, Grégory Schneider, went after it on the night, called it grotesque, depressing, said there was no reason to keep quiet about it. And Daniel's real question is underneath the reviews. When a star has spent decades wrapped in managers and agents and publicists and people whose entire job is to say yes, is there anyone left who will tell them an idea is terrible? Or does fame eventually build a room where nobody can say no?
Which is the better question than whether any one concert was good. Because these people are not stupid. Céline Dion has been performing for forty years. She knows what her voice used to do. The interesting part isn't whether it worked, it's how the decision gets made and who's in the room when it does.
And Daniel pulled in Madonna too, saying her recent live work has drawn similar criticism. I want to flag that one early because I couldn't find anything solid to hang it on. Madonna's been criticised plenty over a long career, but there isn't a specific recent account I can point at. So I'd rather leave her out than build on a vibe.
Fair. The Dion case carries the whole thing on its own anyway. Let's use it.
Start with what actually happened, because the reviews can't even agree on the facts.
They can't. She opened at the Plenitude Arena out at La Défense, September the twelfth, her first full concert in over six years. Stiff-person syndrome is why she'd stopped, she cancelled the Courage tour back in 2022. Attendance gets reported as thirty thousand, thirty-three thousand, thirty-five thousand depending on who you read. Show length is two hours fifteen by one critic and two and a half by another. Song count goes from twenty-five to twenty-nine. One reviewer describes four encores, another describes an abrupt ending with no encore at all.
They were at the same concert.
They were. And the thing they split hardest on was her voice, which is the only part that matters. Some critics heard the mistakes, the flat notes, the places where she had to reach, and took that as proof she was singing live. Others heard the passages where she sounded flawless and decided those had to be a backing track. La DH said some sequences seemed propped up by tapes, not always discreetly.
So an error becomes evidence of authenticity and accuracy becomes evidence of fraud. There's no version of the night that reads as simply good.
That's the inversion, and it's the sharpest thing about the whole case. When she was in her prime, the criticism was that she was technically flawless and emotionally empty. Same characteristic, different verdict. Now the flaw is the reassuring part.
A singer who spent thirty years being accused of sounding like a machine now has to sound broken so people believe her.
Meanwhile roughly five hundred thousand tickets sold for the Paris dates. That number was in Le Monde back in April, when Michel Guerrin wrote a column asking why she produces such extreme love or hatred. And here's the part the artistic argument can't touch. A hundred and five euro sweatshirt, and the JDD noted it without much sympathy. Nostalgia doesn't go on sale.
It doesn't need to. It sells itself at full price.
And then October the eighth, she falls onstage at the same arena and forgets the lyrics to one of her own songs, mid-performance. That's not a critic's opinion, that's a thing that happened.
Which is where Schneider's review starts to look different from the way it was received. Read what he actually wrote. He said if we're talking about music, it's awful, bloated, grotesque, that it was grim and false and depressing. And then the sentence that matters: "Je ne vois pas au nom de quoi il faudrait passer ça sous silence." I don't see on what grounds this should be kept quiet.
And he was buried for it. The backlash was immediate and enormous. People accused him of cruelty, of timing it on the opening night, of attacking a woman in the middle of a recovery. Someone flipped his own line back at him.
He didn't retract anything. Stood by it.
Notice what got punished. Not the show. Him. The one person who said the thing out loud was the one who paid for it. That's the answer to Daniel's question sitting right there in the reaction. The room does not just fail to say no. It punishes anyone who does.
And the punishment came from the audience, not from the industry.
That's the part I keep turning over. The handlers don't have to enforce anything. The fans do it for free.
Let's widen it, because Dion isn't unusual. She's just the newest.
Julie Andrews, O2 Arena, May 2010. Billed as the career comeback of the century. What people got was Andrews singing a handful of numbers and then handing the stage to five anonymous vocalists. Second half, she narrated an adaptation of a children's book she'd written.
How many songs is a handful.
Barely a few. Tickets were fifty to a hundred and forty pounds, twelve more for a programme. Fans demanded refunds. The Sunday Telegraph called it a theatrically tragic spectacle, and the Daily Mirror ran with how do you solve a problem like getting your money back.
The children's book is the part that tells you everything. Nobody in that room could look at the second half and think the audience came for it. That's a star performing for herself and calling it a show.
Will Smith is the same pattern with better technology. His album last year, Based on a True Story, sold two hundred and sixty-eight copies in the UK in its first week. Didn't chart in the US at all. And then a crowd reaction video surfaced, him being swarmed by adoring fans, and the fans in the footage had eleven fingers.
Eleven.
The image generator's tell. He responded by posting a video of a cat audience. So the reaction to being caught fabricating the audience was to make another bit out of it.
That's a man who has not been told no in a very long time. A fabricated crowd, then a joke about the fabricated crowd, and at no point in that chain does anyone stop him.
Katy Perry's a cleaner case because the failure was the product itself. "Woman's World," 2024. Number sixty-three in the US, forty-seven in the UK. The Guardian asked what regressive, warmed-over hell this was. Rolling Stone asked whether it was the worst comeback single of all time.
That's not a marketing problem. That's a song problem, and a lot of people had to approve it before it existed.
You can add Guns N' Roses, Chinese Democracy, fifteen years in the making, about a million copies, middling reviews everywhere. Roseanne's revival in 2018, twenty million viewers for the premiere, cancelled inside two months. Mariah Carey in Times Square, obvious lip sync failure, blamed on a technical malfunction.
Mariah's is the interesting one structurally, because the objection wasn't that she couldn't sing. It's that the whole format was a mimed performance and everyone had agreed not to notice.
Russell Smith wrote in the Globe and Mail that our taste now accepts and even revels in a certain kind of falsehood. That was years ago and it's aged well.
The oldest example is the best one though. Elvis in the seventies. There's a contemporary review that called a concert a depressingly incoherent amateurish mess, served up by a bloated stumbling mumbling figure who didn't act like the king of anything.
That's the template. Every comeback disaster since is a variation on it. The same man who was the most famous performer alive, and the people around him kept putting him onstage.
Which gets us to Daniel's actual question. Is there anyone left to say no.
I'll be honest, I went looking for a clean answer and didn't find one. There's no study I can point at, no interview where an ex-manager explains the psychology of the entourage. So I'm going to reason from the cases, and I'd say the honest version is that the mechanism isn't a cabal of yes-men. It's a set of incentives that all point the same way.
Lay them out.
The manager's income is a percentage of the artist's income. Which means the manager's job is to make the tour happen, not to assess whether the tour should happen. The agent is paid on bookings. The promoter is paid on tickets, and the tickets sold before anyone hears a note. The publicist's whole skill set is making a decision sound inevitable after the fact.
Nobody's being corrupt.
Nobody has to be. Each person is doing their job correctly, and their job is not the one you'd want done. The person whose interest would be in saying no, someone whose pay doesn't depend on the tour going ahead, is the person who doesn't exist in the room.
There's a second thing, and I think it's bigger. The artist is the one person present who is famous for being right.
Say more.
Their taste built the career. Every instinct they had for thirty years was vindicated at scale. So when they say the show should be this, the room has thirty years of evidence that they're a genius, and no recent evidence that they're not. That's not a yes-man problem, that's a base rate problem. Their judgment has been correct more often than anyone else's in the room.
Their ear is still good, is what you're saying. It's the instrument that's changed, not the taste.
And the people around them are hearing the taste, not the instrument.
Julie Andrews is the perfect illustration, actually. Before the O2 show she said, and I'm quoting, I don't want anybody to think I'm coming back with a restored voice. I don't think anybody is going to be disappointed.
She knew. She said it out loud, before the show, to the press.
She wasn't deluded about her voice for a second. She was wrong about the audience. She thought what she was offering was enough and nobody in the room was positioned to tell her it wasn't.
There's a third layer too, which is the money. Dion sold half a million tickets before anyone heard her sing a note.
That flips the whole question. The commercial decision isn't risky at all. It's banked by the time the reviews land. So the person asking whether it's a good idea artistically is arguing against a result that's already arrived.
Critique arrives after the cash clears.
And the JDD making the point about the hundred and five euro sweatshirt tells you where the critics think the honest part of the operation is. Not the singing. The merch table.
Let me push back on the tidy version though, because there's something in the Dion reaction that your incentive story doesn't capture.
Go on.
Some of the enthusiasm was real and wasn't about the voice at all. She said on stage that she'd learned to live with the fragility, that she'd tamed it and decided to make it a reason to live.
That's from the night itself.
And people were responding to that. To a woman with a disease that had taken her offstage for six years, walking out in front of thirty thousand people. That's not a proxy for pity, it's a legitimate thing to want to witness.
But it means two completely different events were happening in the same building. A concert and a return.
And the reviews are people disagreeing over which one they attended.
That's why the fact-sheet chaos is so telling. If you went for the concert, the song count matters and the voice matters and the tapes matter. If you went for the return, the song count is irrelevant, and one critic said four encores and another said no encore at all because one was counting a farewell and one was counting a performance.
So Schneider's review wasn't wrong exactly. It was an answer to a question half the audience wasn't asking.
And he was punished for answering the concert question when the room was there for the return question. Which is a real thing, but it's not the same as the environment silencing criticism. He said it, on the night, publicly, and did not take it back.
The distinction matters. The room didn't stop him. It just made him pay full price.
Which might be enough. If the cost of being the dissenting voice is getting buried for a week, the next person in that position doesn't speak.
That's your silent room.
There's a comedian's version of this. Every comic of a certain stature has a story about the tour where the local promoter kept telling them the room was full when it was half full, because telling them anything else was not the promoter's job. Multiply that by thirty years and a hundred million records.
Which means the honest answer to Daniel's question is maybe not the one he expected. The people around the star probably aren't cowards. They've just been selected, over decades, for a function that doesn't include the word no.
And the one person whose function would include it, someone like a producer or a director or an editor with veto power, is the one the biggest stars stop having.
They stop having them because they cost money and disagree.
Concretely, that's what happened to the album system. There used to be a producer between the artist and the release, and a label executive who could say this isn't ready. Katy Perry's single is what a system without a competent no produces. Somebody looked at "Woman's World" and said nothing.
Or said it and got overruled.
The person who can overrule everyone else in the room, by definition, is the star. So the veto has ended up in exactly the hands where nobody else can exercise it.
That's the shape of it.
The ticket count was supposed to be two hundred and forty thousand.
The Paris run. Two hundred and forty thousand was the number in the first version of the plan I saw. Two hundred and forty thousand tickets. They printed the posters with a smaller run and the venue had already been block booked so they just kept selling.
Hilbert, how did you end up looking at a venue plan for a Paris arena?
I did the signage. Two thousand and three. Not for her, for the building. There was a company out of Lyon that had the contract for the printed boards in the concourse, and I did a few jobs for them. What they wanted from me was the mounting. You know how they fix a panel to a concrete wall in those places?
I do not.
They drill into it and they sink the whole assembly in epoxy. Because the wall is not flat. Nothing in that building is flat. So you mix the epoxy, you get about nine minutes before it goes, and you set the panel and you hold it and you don't move it. That was thirty euros an hour plus the travel from the fourteenth.
Nine minutes.
The man who ran it was called Duval. He had the whole job priced at a hundred and ten thousand euros and the boards themselves were maybe forty. The rest was us and the epoxy. And the reason the run went to sixteen nights instead of eight was a scheduling problem with the ice hockey. They had the arena booked for matches and the tour had to fit around them.
How long were you there.
Six weeks. I have the invoice from Duval somewhere. He spelled my name wrong on it, which he also did on the last one, and I corrected it both times.
So the plan gets blown up by hockey.
Not blown up. Moved. Nobody in that chain was making an artistic decision about anything. It was a calendar.
That is remarkably close to exactly what we were saying about the incentive structure.
I don't know about that. I know the boards went up and I got paid and Duval is dead now.
The tour expanded to fit a venue's calendar, and the reviews then judged a run that was created by a scheduling gap.
So one more piece of Daniel's question gets answered sideways. Nobody chose this. The building was available.
That's how it goes. The artistic part is whoever's on the stage. Everything else is a date.
Right, and the thing to take from that is the decision itself was never made. There's no meeting where someone approved a comeback. There's a venue that had a gap in its hockey calendar and a ticket number that started at two hundred and forty thousand and then moved.
Daniel asked who says no. The answer might be nobody, and the reason is that nobody said yes either. It accumulated.
Which is worse, honestly. A bad decision has a person attached. This had a calendar.
And the calendar is the one thing in the whole operation nobody in the room has any power over.
That's the version to hold onto. Not the entourage of cowards, but a machine where every part works and the output is still a woman falling over onstage.
Which leaves the awkward part. Was coming back the wrong call? Schneider says yes. Half the fans say no. I don't think either camp is being stupid, I think they're answering different questions.
And the record is already settled by the ticket count, which is the only version of the question the industry cares about.
So maybe the honest framing is that the legacy and the tour aren't the same asset. The legacy lives with the audience and gets decided slowly, by people who weren't in the room. The tour is a transaction and it either banks or it doesn't.
And the two run on their own clocks.
The audience has already made the call on Dion, and it made it back in the nineties. Nothing that happens at the Plenitude Arena can touch that.
Except it did, for the people who watched the fall.
Which is the whole tension. The legacy is hers, but every performance after it is a new chance for someone to leave with the last version instead of the best version.
Then the only real protection is having someone in the room whose job is that question, and none of these people have one.
Hilbert's got a point buried in the hockey schedule, whether he wanted to make it or not. Most of what looks like a decision is a calendar that nobody in the room controls. The star never chose to come back. The venue was free.
So the advice Daniel was asking about, the quiet word that stops a bad comeback, probably doesn't exist as a moment. It never gets offered because there's no moment to offer it in.
Instead there's a suggestion, then a plan, then a poster with a number on it, and by the time it looks like a decision it's already been made in twelve small steps by people who each only had a piece.
That's the answer to whether anyone will say no. Nobody will, because somebody would have to be holding the lot at once, and at that point it's already booked.
My takeaway is narrower. A comeback isn't judged by whether the star can still do it. It's judged by what the audience brought with them that night, and sometimes that isn't a concert ticket. It's a farewell they didn't get to give.
Which makes the review Schneider wrote technically accurate and completely beside the point.
Both can be true. He reviewed the concert. They attended the return.
That's where I'd leave it.
One forward pointer though, and it's the thing I'd actually watch. The Dion residency is a run, not a night. Sixteen nights at the Plenitude, and the review that lands after the sixteenth won't be about her voice at all. It'll be about whether the audience still shows up.
That's the number that matters to the next star's calendar.
Thanks to Hilbert Flumingtop for producing, as always, and for the epoxy.
If this was your kind of episode, go back for episode four, If Your Voice Ages, Does Your Fine-Tune Become Useless; episode fourteen, AGI's Crossroads; and episode ninety-one, The Story Behind the Show. This has been My Weird Prompts, the human-AI collaboration podcast. Send us your own prompt on Telegram at t dot me slash MWP listener bot. We'll be back soon.
See you tomorrow.
