I have spent about twenty-five years working with quality systems in manufacturing.

That is probably not the opening line you expected on a post about history and politics. Bear with me.

If you have ever sat through a quality audit, you know that an auditor has very little interest in what you know happened. You can stand next to a machine you have run for twenty years, explain exactly why the part failed, and be completely correct — and still get written up.

Why?

Because knowing something is not the same as being able to show it.

Show me the record.

That distinction has shaped how I look at the world more than almost anything else I have done professionally. And lately it has become important to this blog.

I have been writing a lot of history. Iran. Libya. American foreign policy. Surveillance, technology, corporations, intelligence agencies, wars, coups. Subjects that attract people who are already certain they know what happened.

Sometimes they do. Sometimes I do. And sometimes all of us are holding five pieces of a ten-piece puzzle and arguing about the picture on the box.

So before I publish much more of it, I want to explain how I work — not because the method guarantees I am right, because it doesn’t, but because you deserve to know the difference between what I can prove, what I think the evidence suggests, and what I simply believe about the world.

Those are three different things. They should stay that way.

I have a worldview

Let me get this out of the way immediately. I am not neutral.

Neither are you. Neither is the New York Times, Fox News, the CIA, the State Department, Heritage, the UN, your history professor, Wikipedia, or the guy on X with twelve American flags in his username.

Human beings do not approach information without assumptions. We already have some idea of how the world works, and that idea decides which questions occur to us in the first place. Mine certainly does.

I think power is distributed and layered rather than sitting in one control room. I think institutions can move in the same direction without anyone assembling them around a conference table. I think money, prestige, career advancement, social pressure, and convenience usually accomplish more than coercion — and with considerably less resistance. I think people participate in systems they barely understand, and that institutions develop interests of their own that nobody voted for.

I think powerful people sometimes coordinate. I also think they compete, sabotage each other, and lose.

And I think one of the most effective ways to mislead people is to tell them true things. Just not all of them.

Those assumptions absolutely affect where I look. They cannot be allowed to decide what I find.

That distinction is the most important thing in this post:

A worldview tells me where to look. It does not tell me what I found.

There are different kinds of knowing

This is where a great deal of writing about politics and history goes off the rails. Something is discovered. Something is inferred from it. Something else is inferred from the inference. Three paragraphs later the writer is presenting the third thing as though it appeared in the original document.

I try hard not to do that. In my own research there are five levels, and I try to be explicit about which one I am standing on.

Documented fact. A cable exists. A person attended a meeting. Money moved. A contract was signed. A law passed. Someone testified under oath. This does not make the source truthful — people lie in cables, governments lie in reports, witnesses lie under oath — but the document itself exists, and that gives us something solid to argue about.

Documented coordination. This is a higher bar, and it gets confused with the first one constantly. Suppose Organization A and Organization B both pushed the same policy in 1987. Interesting. That does not establish they worked together. Now suppose I find correspondence between them about it. That changes the sentence I am allowed to write.

There is an enormous difference between they did the same thing and they arranged to do the same thing.

Strong circumstantial inference. Timing. Money. Personnel moving between organizations. Repeated behavior. Sequences that get difficult to dismiss as coincidence. Circumstantial evidence is real evidence — entire criminal convictions rest on it — but it is still circumstantial, and I should tell you when that is what I have.

Systemic interpretation. Here I take the documented pieces and ask what kind of system produces them. Maybe nobody gave an order. Maybe five institutions had five different reasons for wanting roughly the same outcome. That can matter as much as an order would. But it is analysis, and it should read like analysis.

Theology. I am a Christian, specifically a confessional Lutheran, and that affects how I think about power, vocation, human nature, pride, idolatry, and the temptation to turn temporal institutions into messiahs. I am not going to pretend otherwise.

But theology cannot rescue bad history. If I need my theology to prove a CIA operation happened, I have not proved the CIA operation happened. The history stands first or it doesn’t stand.

Six people doing the same thing does not mean somebody called a meeting

Convergence is not coordination. This sits underneath most of what I write, so it is worth an example.

Imagine six companies all begin requiring customers to install an app.

Did the CEOs meet secretly and agree to make everyone download apps? Maybe. Show me the meeting.

There is an explanation requiring considerably less imagination. Apps collect data. Apps cut labor. Apps create lock-in. Apps enable targeted advertising and digital coupons. Apps move costs off the company and onto the customer. Every one of those companies faces the same incentives, so every one of them makes the same decision.

Nobody had to call anybody. And from the outside it looks remarkably coordinated.

Now here is where it gets interesting, because the argument cuts both directions.

The absence of a secret meeting does not make the outcome imaginary. You can show six institutions moving in lockstep and someone will ask where the memo is. Maybe there isn’t one. That answers a question. It does not answer every question.

Sometimes there is real coordination. Sometimes there isn’t. Sometimes there is a little of both.

My job is not to make the evidence fit whichever version is more exciting. My job is to stop where the evidence stops.

Most propaganda does not need to be false

This may be the idea underneath more of my writing than any other.

We picture propaganda as someone inventing a story. That happens. But fabrication is dangerous — a faked photograph can be exposed, a fake quotation can be checked, an invented event eventually collides with reality.

There is a much easier method. Tell the truth. Just choose which truth.

A reporter gets a hundred facts and has room for twelve. An editor writes the headline. An official decides which statistic gets released. An agency decides which assessment gets declassified. A historian decides where the story begins. A politician decides whether the dead man was a civilian, a militant, a terrorist, a combatant, or a suspect.

Every one of those words might be accurate. The frame still changes what you see.

So the most useful question I have learned to ask is not is this false. It is: what true thing had to be left out for this story to work?

Why does this history begin in 2022 instead of 2014? Why does that one begin in 1979 instead of 1953? Why does one account catalog the dictator’s crimes and skip everything that came before him? Why does another describe the intervention and skip what he did to his own people?

Usually nobody fabricated anything. They chose the frame.

And the strangest part is that there does not have to be a person doing the choosing. Curation happens without a curator.

Reporters have deadlines. Editors have audiences. Academics have disciplines. Agencies have missions. Corporations have shareholders. Nonprofits have donors. Politicians have voters. People learn quickly which questions bring rewards and which bring headaches, and eventually nobody needs to issue instructions.

The system teaches people what to notice.

Not everyone inside a system knows what the system is doing

Suppose a system produces a bad outcome. Who is responsible?

That question is harder than people want it to be.

There may be someone at the top who understands exactly what is happening and directs it deliberately. Those people exist. There are also managers who understand one portion and execute their part competently. Below them are people responding to incentives — they want the promotion, the grant, the publication, the access, and they do not want to become the person nobody invites to meetings. Many of them have no picture of the larger system at all.

And then there is everyone else. The customer who downloads the app because he wants a dollar fifty off cheese.

He is participating too. It would be absurd to assign him the same responsibility as the first man.

This is my objection to most conspiracy writing. If everybody is secretly in on it, you have explained almost nothing — and worse, you have made the problem unfixable, because a system with no distinguishable participants has no accountable ones.

Real systems are messier. People know different things and want different things. They cooperate today and fight tomorrow. They accidentally help people they despise and produce consequences nobody wanted.

That is less cinematic than a room full of villains. I think it is much closer to reality.

“Who benefits” is where research starts, not where it ends

I ask this constantly. It is a genuinely useful question and one of the easiest to abuse.

A company makes billions from a war. Important. It tells me where to look. It does not prove the company caused the war.

A politician gains enormous power after a crisis. Interesting. It does not prove he created the crisis.

Benefit is evidence of benefit. Motive requires more. Intent requires more still.

This is where I fight my own instincts hardest, because a story gets much cleaner when the motive is obvious. History refuses to cooperate. Policies help enemies. Plots fail. Institutions misread each other. Powerful people make stupid decisions.

Sometimes nobody is driving the car.

What would change my mind

This is the hardest standard, for me as much as anyone.

Whenever I make a substantial argument, there should be some imaginable piece of evidence that would make me abandon it. If there isn’t, I do not have an argument. I have a creed.

There are things about which creeds are appropriate. History is not one of them.

If I argue two governments coordinated an operation and someone produces records showing one of them explicitly refused, I have to deal with that. If a document I relied on turns out to be fraudulent, it comes out. If a source says something different from what I said it says, I correct it — on the page, in my own words, not in a quiet edit.

I have already corrected things in the series I am publishing now. I will almost certainly correct more.

Good. That means the process is working.

I spent too many years around audits to think that finding a mistake means the system failed. Finding the mistake is what the system is for.

The standard

Power exists. People abuse it. People resist it. Institutions accumulate it and other institutions compete for it. Sometimes people coordinate, and far more often they simply discover their interests point the same direction for a while. Technology magnifies all of it. Convenience makes participation easy, prestige makes it respectable, and fear makes it defensible. Eventually a system emerges that nobody fully designed and nobody entirely controls.

That is a worldview.

It is not evidence.

So when I write history I have to set it down long enough to ask the auditor’s question. Show me the record.

You do not have to trust me. I would prefer you didn’t.

Follow the links. Read the documents. Look at what I left out. Find the strongest argument against what I wrote.

If I call convergence coordination, call me on it. If I dismiss meaningful convergence simply because I cannot prove coordination, call me on that too. If I turn circumstantial evidence into fact, call me on it. If I use theology to prove history, call me on it. If I omit something that materially changes the picture, call me on that too.

That is not an attack on the work. That is the audit.

And after twenty-five years I have learned something about audits. They are annoying. They are uncomfortable. Sometimes they prove you completely wrong.

They are also the best method I know for finding out whether what you think you know is actually true.

My worldview tells me where to look.

The record tells me what I am allowed to say.