The Slot Was Already Open
A four-hundred-word post goes up and within an hour has eleven thousand words of comments under it. Nobody read four hundred words and produced eleven thousand in response. They read a headline, maybe a pull quote, and then typed out an argument they already owned.
That's not laziness and it's not bad faith. It's that most of what gets called "reaction" was never produced by the piece at all. It was produced by recognition. The headline is a key, and somewhere out there is a lock that was built months or years earlier, machined to a specific shape, just waiting for a key that fits. When one shows up, the lock turns instantly, and the turning looks like engagement with the content. It isn't. It's the lock doing what locks do.
Think about how fast a comment section fills in on a story about a company getting bought and gutted by its new owner. Nobody needs to read the pricing page or the support-ticket screenshots first. "Private equity buys good small software company" is a slot that already exists, fully furnished, before this particular acquisition happened. The specific numbers, who the buyer was, whether the product was actually degraded or just repriced, none of that controls how fast or how big the response is. The slot controls it. The piece just has to be recognizable as an instance of the category.
Same with a regulation fight, a model release, a labor dispute. Each has two or three pre-built argument shapes sitting in inventory: tech saves us / tech ruins us, the incumbents are gatekeeping / the incumbents are protecting something real, this is the same story we've seen before / this time it's different. A new event doesn't introduce a new argument. It gets sorted into an existing bin, and the size of the response tracks how big and well-worn that bin already is, not how much the event actually tells you.
This is why volume and reading time are so disconnected. Reading time is bounded by the length of the piece. Response volume is bounded by how many people already hold a position the headline can slot into, which has nothing to do with the piece's length, rigor, or even whether anyone could get past its paywall. A locked article with a provocative title can out-argue a fully-read, carefully-reported one, because the title alone was enough to trigger recognition in thousands of people who never needed the body text to know which side they were on.
And this is why headline writing, algorithmic ranking, and editorial selection all converge on the same behavior without anyone coordinating. None of them are optimizing for information content, because information content is expensive to produce and expensive to evaluate. Slot-fit is cheap to produce and nearly free to evaluate: does this phrase match a shape already sitting in a reader's head? If yes, ship it. The feedback loop doesn't care whether the piece was right. It cares whether the piece was recognizable.
The tell is what happens when something genuinely doesn't fit an existing slot. Those pieces move slowly and generate thin comment sections, even when they're more important or better reported than the pieces that blow up. Nobody has a pre-built argument to deploy, so they'd have to actually read the thing, form a view, and write something that doesn't already exist in their head. That's expensive. Most people, most of the time, decline to pay it.
So the comment section isn't measuring the article. It's measuring inventory, how many loaded arguments were already standing around with nowhere to go before the headline gave them an address.