Picture one of the most respected philosophers in a country spending three straight years posting almost exclusively about his opposition to a single political vote. Not campaigning, not organizing — just posting, day after day, about the same grievance, long after the vote had already happened and nothing about it was changing. Strung end to end into a book, the record would read, in the source's own blunt words, as "the most bonkers thing I've ever seen in my life."1
The strange part isn't that he was upset. It's that the medium let the upset run for three years without ever costing him anything.
In a village of two hundred people, a person who complained about the same grievance for three straight years would burn through their social capital fast — everyone would find an excuse to be elsewhere the moment they walked into the pub. Evolution built humans to self-regulate exactly this behavior, because in a small enough group, being the person who won't let something go carries a real, felt cost: people quietly stop inviting you places.
Digital platforms remove that cost almost entirely. You can complain to an audience of thousands, interminably, from a chair, with total anonymity available if you want it, and face none of the social friction that would have shut the behavior down in any environment humans actually evolved in. The complaint becomes free to produce and free to repeat, so nothing internal or external stops it from running indefinitely.
Once the cost of complaining drops to zero, a second effect kicks in: other people join, not necessarily because they're equally upset, but because visibly joining a moral position is its own kind of signal, and once a critical mass has joined, staying silent starts to look like the position that needs explaining. Each new joiner lowers the perceived cost for the next one and raises the perceived cost of not joining — a spiral where the group's visible intensity keeps escalating independent of whether the underlying grievance actually deserves that much intensity. The mechanism is self-reinforcing in a way an in-person disagreement, bounded by actual social consequences, generally isn't.
You're deciding how seriously to take an online pile-on before you form an opinion about who's actually right. Sutherland's own diagnostic is worth running here: go look at the accounts doing the complaining. Do they post about a genuine spread of things — a cat falling off a skateboard, a holiday photo, this complaint, something else entirely next week — or is every single post from this account about the same narrow grievance? The first pattern reads as a normal, engaged person who happens to be legitimately upset about something specific. The second reads as someone the medium has let run unchecked past the point where any in-person social structure would have intervened.
You notice, too, the businesses on the receiving end of a pile-on making the same distinction. A confected outrage spiral generates enormous visible noise and very little actual cost — see this vault's page on genuine offense versus confected outrage for the parallel test applied to advertising rather than political complaint.
The core claim — that cheap digital signaling removes the natural social brakes on repetitive complaint — is well-supported by the case example and by the broader logic of costly versus cheap signaling this vault already documents elsewhere. The specific mechanism by which joining a spiral lowers others' threshold to join is plausible and consistent with documented social-proof and conformity effects, though Sutherland doesn't cite formal research for the spiral dynamic specifically — treat the spiral-escalation claim as [PLAUSIBLE — needs corroboration] while treating the underlying cheap-signal observation as well-evidenced.
The open tension: not every sustained, repeated public complaint is illegitimate — some grievances genuinely warrant years of sustained attention, and the diagnostic ("is their whole feed about one thing") risks dismissing genuinely important single-issue advocacy alongside genuinely disproportionate pile-ons. The source doesn't offer a way to distinguish a justified single-issue campaign from an unchecked spiral except by the observer's own judgment of whether the underlying grievance merits the intensity — which is exactly the judgment call the diagnostic was supposed to remove.
This converges directly with the costly-signaling literature already anchored in this vault: cheap signals are unreliable precisely because anyone can produce them without cost, and expensive signals are trustworthy precisely because only someone who really means it would pay the price. Sutherland's contribution is showing this principle operating not just in advertising or courtship display but in political and moral discourse — a domain the vault's existing costly-signaling pages haven't yet extended to. The convergence suggests the underlying principle (cost enforces honesty) is genuinely general-purpose, not specific to any one domain it's usually discussed in.
Behavioral-mechanics — Costly Signaling. That page establishes the general principle that a signal is only trustworthy in proportion to what it costs the sender to produce. This page shows the mechanism running in reverse: digital platforms drove the cost of a specific signal — public moral complaint — to near zero, and the signal's reliability collapsed in step, producing spirals of intensity disconnected from the underlying grievance's actual weight. The insight the pairing produces: you can predict where a category of signal will become unreliable simply by tracking where its production cost has fallen, without needing to evaluate any individual instance of the signal on its own merits.
Behavioral-mechanics — Confected Outrage vs. Genuine Offense. Both pages describe the same underlying phenomenon — outrage that costs nothing to perform, at scale, disconnected from proportionate cause — applied at two different scopes. The allergen-analogy page is about advertising and public communication generally; this page is about sustained individual behavior over years. Read together, they suggest a single diagnostic ("what did this cost the person expressing it") that applies whether you're evaluating a single ad backlash or a years-long social-media campaign.
Sharpest implication: the intensity of an online moral consensus tells you almost nothing about how many people genuinely hold it or how strongly, because the platform has removed the one thing — cost — that used to make intensity a reliable signal of genuine conviction.
Generative questions: