If you read police blotters enough, you start to notice the colors: Pink bike. A yellow color backpack. Silver four-door sedan. Heather-blue uniform pants. A black male wearing a black hat with white lettering, a navy blue coat, and blue jeans.

You notice the colors mostly because there aren’t many other details. Place, time, incident description—color of suspect’s clothing or car. Often the color of his or her skin.

I read police blotters because I write about crime, but I also enjoy getting lost in them. I can spend hours scrolling through outdated lists of events that happened places I’ve never visited: Rutherford, New Jersey; Killeen, Texas; Eliot, Maine. Someone told me once—and I’m not sure how he would have known, but I chose to believe him—that police blotters are the most-read pages on local news websites. So perhaps I’m not alone.

Blotters vary in content by region. In Lamoille County, Vermont, there were five incidents involving deer between November 13 and 19. Car vs. deer, the log read the first four times, noting the injuries to drivers and deer (the driver was always unharmed). Then, finally, the recording officer became overwhelmed by the casualties: Nov. 17, 6:38 pm, a bad night for deer. The blotters of cities tend to list fewer animals and more violent incidents. Sometimes, though, the suburbs are brutal, too: Wellesley, Massachusetts: Knife flashed in road incident, 4:59 p.m.

The style tends to be more or less consistent across the board. Spare syntax, short sentences that sometimes aren’t quite sentences, sparse details. Small worlds are compressed into bite-size stories that are often without endings.

When entries don’t end badly, the stories they tell can be charming. Some recent incidents, pulled from blotters across the country:

Nov. 18 at 9:19 a.m., a woman in Johnson lost her glasses; she thought she’d placed them on a nightstand before bed. Police were able to locate them in the house.

8:06 p.m. — A caller from Highway 49 reported a man showed up at their residence asking for an ambulance. [He] smelled like a skunk.

A resident said Nov. 20 someone rang her doorbell three times and tried to open her front door. Responding officers found the front door ajar and a package sticking out. It was found to be an apple pie that was left by a neighbor.

11:17 p.m. — A caller from Clear Creek Place reported their husband was being really emotional and crying.

Read in succession, they might almost be a series of haikus, or examples of that now-clichéd writing exercise: tell a story in six words. The Fresno police, in the final sentence of their description of suspect #1, who may or may not have lit a fire on school property, tell such a story: Last seen riding a grey bike.

Who was he and where did he go? I have imagined him with so many different faces, going in so many different directions.

Of course, the police blotter is not actually a series of haikus. More often than not, it’s a laundry list of violence, theft, drunk driving, and accidents. It’s also not a benign document, or an apolitical one. Read closely in another way, it’s not just a record of what has occurred: it’s a record of who controls information.

Police get to say what happened and how. Police get to decide which codes are used to describe certain types of crime.

Take this call for service:

On Tuesday police responded to an Old 255 Road residence for a reported domestic disturbance. Those involved were separated prior to police arrival, and the situation was handled without incident.

“Domestic disturbance.” We understand something from that phrase, but what? A shouting match? Assault? We are to understand from the word “domestic” that the expected order of the home was upset. We are to understand that everything has gone back to some version of normal—“the situation was handled without incident.” I read this entry with unease. What was left out? Is everything okay? How does this story really end?

This is not to say that more information necessarily belongs in this entry. It’s also not to say that the decision to leave out details or include them is nefarious in intent. But these decisions matter.

We learn these patterns. We borrow this language, and it becomes, for us, a domestic disturbance rather than something vaster and more complex, a major upset in two people’s lives. We think therefore that we understand these crimes, rewritten occasionally as short stories in the newspaper, and we come to forget how much we never knew.

Here’s an entry that’s both more specific and more vague.

Strangulation, petit larceny, domestic battery, 5:30 p.m. Sunday, 1300 block of Huntington Avenue.

It is spare, and it is chilling.

V was asleep when he was awakened by a noise. V opened his eyes and saw S’s silhouette. V muttered something. S fled scene.

This is one of the most mysterious incidents I have found, and possibly my favorite. Maybe because nothing really happened. But so much happens, in those tiny, tight sentences.

I like the verbs: “awakened,” “muttered,” “fled.” The use of “silhouette.” These words feel like they belong in a Victorian novel, heightening the drama of the nonincident in the incident report. The threat of violence lurks and builds—then fails to materialize. All that really happens here is watching.

Who watches whom? S and V, shorthand for suspectand victim, but perhaps they could also be Sophie and Victor, or Sylvia and Valentine. S watches sleeping V, and yet V is watching S, really, for it is V who retold the story to officer P, who recorded it in the log for you and me.

This tripleness of watching becomes distilled in the police blotter, for us to rewatch from afar. We have a view into a room at night where nothing has occurred, or onto a street corner in broad daylight where one human has committed a violent act against another, or into a house where a woman has lost her glasses. Our view is slanted, partial, manipulated, and yet we keep looking.

Police get to decide when the race of a suspect makes it into the public record—a practice for which there seems to be little standard protocol and arguably little purpose, most of the time.

Race is entirely absent from some blotters. In some, it appears occasionally. In others, it’s always there. I started wondering, after awhile, why it ever appears at all.

On Nov. 24 an employee at Layyah in Greenport phoned police to report that a black male wearing NY Yankees sunglasses, a black wool jacket and a gray sweatshirt stole a bag of candy valued at $1.99. Officers were unable to locate a suspect.

How much more do we know about the crime, or the suspect, because we know he was a black male? I imagine police release the information because it’s more information. Someone might be able to identify the suspect on the basis of a description and call in to say, I know him. But who would know him from his sunglasses and the color of his skin? Who could call?

The effect, I think, is often a kind of deadening confirmation of biases. Massive structures of inequality are compressed and stripped away. The context is gone and we are left to make what we will of race and gender.

We are left on alert for details too ubiquitous to be of any use; the suspect could be anyone and anywhere, a black male wearing NY Yankees sunglasses, a black wool jacket and a gray sweatshirt.A man with a bag of candy in his pocket, a bag of candy that will soon be eaten, and gone.

Sophie Haigney is a writer and journalist based in San Francisco.