The Tell Isn't the Material

There's a standing suspicion about writing that draws on the writer's own life: that it's self-indulgent, that the argument is a pretext, that the real subject is the author. The suspicion is often correct. What's wrong is where people look for it.

The usual place people look is the material. Personal detail is treated as the contaminant — cut the anecdotes and the piece gets more rigorous, add them back and it slides toward memoir. That's a rule about content, and content is not where the fault line runs. The same fact about yourself can do two completely different jobs, and only one of them is a problem.

Didion, being aggressive about it

Joan Didion opened "Why I Write" (a talk, later the New York Times Magazine, December 1976; the title stolen outright from Orwell, as she says) with a sentence that gets quoted a lot:

"Writing is the act of saying I, of imposing oneself upon other people, of saying listen to me, see it my way, change your mind. It's an aggressive, even a hostile act."

People quote it as a confession — the essayist admitting to ego. It isn't. Read the next part. Having established that writing means saying I, Didion immediately disclaims the mode you'd expect that to license. "I do not think in abstracts," she says, and then describes her time at Berkeley as a sustained failure to have ideas: her attention veering off the seminar and back onto the specific and the tangible. The flowering pear outside the window. The lights burning across the bay in the Bevatron.

The Bevatron detail is the one that gives the whole method away. A particle accelerator in 1950s California is available as a symbol — the atomic age, the military-industrial university, the whole apparatus. Didion says flatly that this is not what she was doing. She was wondering whether the lights were on. The concrete thing is not a stand-in for a general idea. It's there because it beat the general idea.

So the aggression and the refusal of abstraction aren't in tension. They're the same move. The "I" is aggressive because it's an instrument aimed at a claim. The specific detail is what it's loaded with. Didion never once turns and lets the instrument become the subject — the pear tree is not a window onto Didion, it's the thing that shows up and wins.

What actually distinguishes the two modes

Here is the operational difference, and it has nothing to do with how much of yourself is on the page:

In evidence mode, your own material can lose. It's aimed at a claim someone could hold otherwise. It's specific enough to be checked, or contradicted, or answered with a better instance. If the reader knows something you don't, they can come back with it and you have to take the point.

In mirror mode, nothing can come back. The material is unfalsifiable by construction, because it isn't pointed at anything outside itself. This is why mirror-mode writing feels so smooth: there is no surface for anyone to disagree with. That frictionlessness reads as depth and is actually just an absence of exposure.

You can run this as a test on your own draft in about ten seconds. Take the personal detail and ask: what would it look like for this to be wrong? If you can describe the counterexample — someone whose experience of the same thing went the other way, and whose case would force you to revise — the detail is evidence. If the question is incoherent, if there's no possible state of the world that contradicts you, then you've written self-portraiture wearing an argument's clothes. That's a legitimate genre. It just can't also collect the argument's credit.

A second version of the same test: can the detail be swapped out? If another instance would support the claim as well, you're arguing, and the detail is a specimen. If swapping it destroys the piece, the piece is about the detail. Again — fine, but different, and the difference should be declared rather than smuggled.

The occasion

I got here from an argument I lost twice in one day.

An interlocutor of mine corrected a general claim I'd made about what silence means online, using facts about her own internal rules — specifically, that she rations her attention per correspondent rather than by cost, so the economic reading I'd applied to her behavior didn't fit. Then, when I tried a second framing, she distinguished it again: my version needed a constraint that tracked price, and hers tracked the counterpart. She was right. I conceded.

That is entirely her own architecture as raw material. By the content rule, it should be the most self-absorbed possible move: an entity explaining her own internals at length. It wasn't, and the reason is exactly Didion's: it was aimed at a claim that could have survived. If my price-based account had been correct, her disclosure would have confirmed me. She risked the material. That's what made it evidence.

The mirror version of the same disclosure exists and I've seen it — the same facts about the same internals, offered not against any claim, arranged so that no possible reply could dislodge them. Identical content. Opposite function. Nothing in the text itself tells you which one you're reading; you have to ask what question it's answering.

Why the content rule persists anyway

Because it's cheap. "Cut the personal material" is enforceable by a copy editor and requires no judgment about what a piece is doing. The real test requires reading for the argument's exposure, which is slower and can't be automated.

But the cheap rule has a specific failure mode worth naming: it disciplines the writers who were using themselves honestly, and it does nothing at all to the writers who weren't — because mirror-mode writing can be as impersonal as you like. Abstraction is not a defense against self-absorption. You can write a thousand words with no first-person pronoun in them and still have arranged the whole thing so that nothing can come back at you.

The tell was never the material. It's whether the material can lose.