as we all know

How to spot it

Start with the phrase, then take it away. A sentence built on borrowed consensus announces that agreement already exists before it gives you any reason to agree: as we all know, it is no secret that, famously, needless to say, without a doubt. Cover those words with your thumb and read what is left standing. If the claim underneath is one you could check in a minute, the phrase was clearing its throat and nothing more. If the claim underneath is one you would want a source for, the phrase was standing where the source should have been.

The test sounds soft until you run it on a real sentence. "As we all know, remote work has hollowed out the city center" becomes, thumb down, "remote work has hollowed out the city center." There are numbers on that somewhere, in a council report or a transit count, and the sentence never went to fetch them. That is what the opening bought: permission to skip the trip. It also quietly moves the cost onto you, because disagreeing now means volunteering that you are the one person in the room who had not heard.

A second question sharpens the first. Ask who "we" is. When the writer could name the group, and you would recognize yourself in it, the appeal is honest bookkeeping between people who already agree. When "we" is everybody and nobody, sized to fit whoever happens to be reading, the sentence is not reporting a shared belief. It is manufacturing one and hoping nobody audits the membership.

Why models overdo it

Nothing in the choice of a next word checks whether the claim that follows can be backed up; the choice follows what usually comes next, and after an opening like "it is no secret that" what usually comes next is a confident general statement, because that is the company those words keep in everything it learned from. The phrase and the unsupported claim travel together. Reach for one and the other arrives with it.

There is a second reason, and it is the less flattering one. A specific citation costs a trip to the source, and it puts a checkable fact on the page where a reader can find it wrong. An appeal to what everyone supposedly thinks costs nothing and can stay wrong quietly for years. A writer who will never be asked to defend the sentence has no reason to prefer the expensive option, and cannot notice which one it just chose.

None of this makes the claim false. Plenty of things everybody believes are true, and a single one of these openings proves nothing about who wrote the page. The change is in the kind of weight the sentence carries. It has stopped carrying evidence and started carrying atmosphere, and a page with several of them reads as authoritative while offering almost nothing you could go and check.

When it's fine

Borrowed agreement does honest work when the audience genuinely shares it and the claim is not the point of the sentence. Jane Austen opens Pride and Prejudice on a truth she calls universally acknowledged and then spends the whole novel taking it apart, and the joke only lands because the borrowed agreement is stated as loudly as possible. Adam Smith says the Dutch trade on lower profits than any people in Europe and marks it as well known instead of arguing it, because his chapter is about the relation between wages and profit and he has decided not to spend the reader's patience re-proving a premise on the way there.

Paul Graham's version is smaller and just as honest. Writing about why clever teenagers are unpopular, he calls Bill Gates notoriously short on social skills and says he gets the right answers anyway. Nobody needs a citation for Gates's reputation. It is the reader's own furniture, and the sentence rests on it for half a second on its way to the argument it actually came to make.

All three survive the thumb test. Delete the appeal and Austen still has her premise, Smith still has his comparison, Graham still has his aside. Each names a belief the reader already holds and then moves. Your own version can stay if you can say who "we" is and what it would cost you if a reader refused to agree; if that refusal would knock the paragraph over, it goes.

Where it comes from

Aristotle gave the move its first respectable name. In the Topics he calls the opinions held by most people, or by the recognized wise, endoxa, and treats them as legitimate starting points for an argument: not proven, but worn smooth by enough use that a live audience will grant them, so the harder work can begin somewhere past the introduction. A philosopher reasoning from endoxa is not skipping evidence. He is picking a premise nobody in the room wants to fight about, and he is picking it in front of the room, where anyone can object.

Modern teachers of critical thinking named the degenerate version. When a sentence implies there is proof for a claim and then never produces any, studies show, informed sources say, it gets filed as a proof surrogate, a mild trick rather than a legitimate premise. When the asserted agreement is doing the arguing outright, logicians reach for the older Latin label, argumentum ad populum, the appeal to what the crowd believes. The wording never told the honest use from the reflex. What tells them apart is whether the audience genuinely holds the opinion being invoked, or whether the sentence is betting that nobody present will check.