By H.L. Delaney · August 24, 2026

Writing Violence Without Spectacle

Every book I have written has a massacre in it, because the history I write about has massacres in it. Putting them on the page was never the hard part. The work is deciding where to stand while they happen.

The two failures

The first failure is the one the archive already commits. Killing becomes an administrative event. Numbers, categories, and a justification that arrives in the same sentence as the count. Villages are destroyed in the passive voice. People become enemy dead, which is a category that does the work of erasing whether they were armed, whether they had done anything, and whether they were given any chance to be something other than a target.

A novel that adopts that register inherits its politics. You can write a whole massacre in the language of a dispatch and produce something a reader closes without feeling anything, which is precisely what the dispatch was engineered to do.

The second failure runs the other way. The camera goes close and stays there. Wounds get described with care. The prose gets good at it. The reader comes out the far side having had an experience, and the experience is the problem, because it is an aesthetic one attached to somebody's actual grandmother.

Where I put the camera

At the eye level of someone who has work to do. That is the single most useful rule I have found.

Somebody at a killing is almost always holding something. Horses, a lead rope, a saw, a ledger, a child's hand. Attention in a bad hour goes to the task, because the task is survivable and the rest is not. Writing from inside that attention gives you the event without turning it into a display, and it puts the reader in the only position anyone was actually in, which is too close and unable to stop it.

It also does something the wide shot cannot. It keeps the reader from getting the view. Nobody at Dokdokwas saw the whole village burn from a good vantage. They saw the part in front of them and heard the rest.

Let the paperwork carry the horror

The strongest tool I have found is the record itself, placed next to the thing it describes. A man calls out four dead. Somebody asks whether they were armed. There is a pause. Then a voice supplies enemy dead, and the officer writes it down, and the pause is the whole scene.

The same holds for what gets removed. In The Renegade Enos there is a report that says the men were wanting proper tools to bury them, and a woman with a pen who stops on that phrase, because proper tools lets the reader see the shovel. She strikes it. That edit is more violent than the burial it hides, and it does not require a single description of a body.

This is what I mean about the record being the crime scene. Reading these documents against their intent is most of the job, and I have written about working from records built to erase you and what it changes to research from the inside.

Names where the record kept them

Where a name survived, I use it. Where the count is a range because nobody troubled to count, I say that it is a range and I say why. The Modoc killed at Ben Wright's parley on the Lost River are given in the sources as somewhere between thirty and ninety people, and there is no list. That absence is a fact about the killing and it belongs on the page as one.

Naming also runs in the other direction. Ben Wright called the parley. Frémont said there was no room for prisoners. Keeping responsibility attached to a person is the opposite of what the frontier register does, which is to dissolve decisions into weather.

The test I use

When a violent scene is finished I ask whether a reader could come away from it feeling skilled. Impressed by the writing, moved in a way that flatters them, able to recommend it as a passage. If the answer is yes, the scene is working for me and not for the people in it, and I take it apart.

The other test is simpler and harder. Would I read this aloud in a room with descendants in it. That is not a hypothetical for me. These are my relatives' events, in my own basin, and the people in them have great-grandchildren who know the family versions. What fills the gap when the record is broken covers the standard I hold myself to when the paper runs out, which is reclamation rather than invention.

None of this makes the violence comfortable, and it is not supposed to. It is supposed to leave the reader where the people were, which is inside it, holding something, unable to put it down.