How to Read a File Responsibly

The docket is the only document in this file you can get without anyone's permission.

Everything before it was internal. The responding officer's report, the detective's spiral notes, the autopsy protocol, the tip log, the transcript, the review memo, the lab result: most of those pages sit behind a records exemption for as long as a case is open, and often after. The docket is different. It belongs to the court, and in most American jurisdictions the court's index of a criminal case is a public record. You can usually search it from a home computer. For federal cases you can use PACER, the federal judiciary's electronic records system, which charges by the page in a way that has not changed much since the dial-up era. For state cases you can use the state or county court portal, or you can visit the clerk's office in person.

A docket does not tell a story. It is a list of entries in date order, each one a line or two long. A charging document filed. An initial appearance. Counsel appointed. A motion to continue, granted. Another motion to continue, granted. A change of plea hearing. A plea entered. A sentencing date set. A sentencing held. Judgment entered.

In April Tinsley's case, the docket for the prosecution of John D. Miller is short, because he pleaded guilty. There was no trial and no parade of witnesses, so none of what the lab report set in motion appears there in narrative form. The docket records that the case moved, and how quickly. It cannot hold an eight-year-old who had gone looking for her lost umbrella. It cannot hold thirty years.

You know how to ask the three questions by now, because you have been asking them since the first page. Ask them once more of this one.

It says that the State of Indiana brought a case, that the case proceeded through hearings, and that it ended in a plea of guilty and a sentence. It says the names of the court, the parties, and the judge. It records dates, and the dates are facts you can cite.

It misses almost everything you have learned to care about. It does not include the swab, the genealogy, the family trees built from strangers' consumer tests, the trash collected for a confirmatory sample, or the detectives who kept the file from being thinned. It contains no explanation. A docket is designed to record procedure, and it does that well.

It forces a question about what a public record is for. The docket exists so that the public can see that the state did what it said it did, in the open, under rules. It does not exist so that the public can retry the case, and it was never built to tell you why anything happened.

That distinction is the hinge of this chapter. Everything that follows is a method for reading public records in the same disciplined way the earlier chapters read the private ones, and for knowing where the reading has to stop.

Start with a resolved case. This is the first rule of practice, and it is the one most people skip. If you want to learn to read a file, do not begin with the unsolved case that keeps you awake. Begin with one that has ended in a conviction or a documented exoneration, where the record is complete enough to check your reasoning against the outcome. The National Registry of Exonerations, maintained by researchers at the University of California Irvine, the University of Michigan, and Michigan State University, publishes case summaries with sources. Appellate opinions are public, and they often contain the most detailed account of the evidence that a court has ever written down. Those are your training files.

Then assemble the record in the order this book has followed.

First, the earliest public account. Usually this is a newspaper story from the first day or two, which you can find in a library's microfilm or a newspaper archive. Read it as if it were the first-responder report. It says what the police released. It misses whatever they held back, and early reporting is often wrong about times, ages, and locations. It forces a question you should write down: which of these details will later turn out to be true?

Second, everything about the child that the family chose to make public: an obituary, a memorial page, a school statement. Read it for the person. Do not mine it for clues. You are reading it to remember whom the file is about.

Third, whatever official documents your state releases. Some states treat autopsy reports as public records and many restrict them, often more tightly for children. Police records of a closed case may be available through a state public records request, which is usually a written letter citing the statute. Expect redactions, delays, and occasional fees. Expect, at least once, to receive forty pages of black rectangles with a cover letter thanking you for your interest.

Fourth, the court record: docket, charging documents, any published opinions, and transcripts if you are willing to pay a court reporter for them.

Fifth, build a timeline, and give every entry a source in a column beside it. If a fact has no source, it does not go on the timeline. If two sources disagree, both go on, marked as a conflict. For each document, write three short lines: says, misses, forces.

Now the hard rules. There are only two, and they do not bend.

You do not name, in public or in private correspondence with anyone other than investigators, a living person who has not been charged. This holds even if the name appeared in an old newspaper as someone "questioned," and even if a podcast has already said it. Chapter 6 showed how a person can be moved from "witness" to "suspect" with a pen stroke and no initials. You do not get to make that stroke. A name repeated online can cost a person a job, a marriage, or more, and no later correction reaches all the places the original went.

You do not contact a victim's family unless you have a reason that serves them, and when you do, you contact them once, in writing, through a channel they have chosen to make public. Say who you are, what you are doing, and that you will not write again unless they reply. Then keep that promise. Families of murdered children have described being called by strangers on anniversaries, sent theories, and confronted with photographs. Some have spent years with a question to detectives that no one has answered. A letter from you will not answer it.

Here is where the method meets pressure. The following scenario is a hypothetical, built to show a decision, and it is not a real case.

You have been reading an unsolved case from the 1990s. A girl of nine, a town of twenty thousand. In the archive you find a story from the second week, which reports that police "spoke with a neighbor" and gives the neighbor's first name and street. Ten minutes in a people-search database tells you the neighbor is still alive, still in that town, and has a criminal record from the 2000s for something unrelated. In an online forum, three people have already connected the same dots. One has posted the address.

The timeline you built is on your desk. Your hand is on the keyboard.

Ask the three questions of the newspaper story, the way you would ask them of any page.

It says that police spoke with a neighbor. That is all it says. It does not say "suspect," "person of interest," or "cleared," and the reporter may not have known which, if any, applied.

It misses nearly everything that would give the line meaning. You do not know why police spoke with him, whether every house on the street was canvassed, or what he said. You do not know what the tip log holds, what the lab found, or whether the detective's notes from that week ever reached a typed report. Chapter 5 showed how a tip log fills with neighbors, ex-partners, and anyone seen near the scene, and how most of those names are noise. The later record is a separate fact about a separate period of the man's life.

It forces one legitimate question: does the agency currently investigating this case know about this article? They almost certainly do, since it came from their own release. The question you are actually tempted by, whether he did it, is one the newspaper cannot answer. Neither can you, and neither can the forum.

So you close the forum tab without posting. You do not save the address. On your timeline you write the line from the story with its source, and in the "forces" column you write: Canvass scope? Unknown. Then you turn to the parts of the record you can actually affect. Is the biological evidence preserved? Has the case been reviewed for genetic genealogy? Those are questions with paper answers, and the last chapter showed you how much depends on them.

Run the lens honestly and the result is a discipline more than a theory. A line that seems to point at a person turns out, when you ask what it misses, to point at a gap in your own knowledge. The gap is real, and it belongs to the investigators. They will fill it only if the right information reaches them in a form they can use.

That is the last practical skill: passing something on.

Sometimes your reading will turn up something that may matter, such as a document that contradicts a released timeline, a witness account in an old article that never reappeared, or a factual error that has hardened into accepted fact. If the information concerns a living person, it goes to investigators and nowhere else. If it concerns the record, it goes to investigators first.

Find the agency that holds the case now. That may not be the department that handled it originally. Cases move to state police, county cold-case units, or regional task forces. Look for a published tip line or a cold-case unit address. For cases with a federal connection, the FBI accepts tips at tips.fbi.gov. For a missing child, the National Center for Missing & Exploited Children runs a tip line. For unidentified remains, NamUs, the federally funded National Missing and Unidentified Persons System, accepts information through its website.

Write the way a good report is written. Give the case name and date. In one sentence, say what you found. Name the source precisely, with publication, date, and page, or the record number. Keep fact separate from inference, and label inference as inference. Leave out adjectives and theories. Give your name and how to reach you. Keep a copy, and note the date you sent it.

Then expect silence. Silence does not mean you were ignored. An investigator who tells a member of the public what a tip did or did not confirm can damage a future prosecution, alert a suspect, or hurt a family. You have read enough of this file to know that the people who say least are often protecting the case. You will probably never learn what happened to your letter, and you send it anyway.

The tools you have now are modest ones. You know the order of the documents and which of them you can lawfully see. You can build a timeline in which every line has a source. You know the two rules you will not break. You can ask what a page says, what it misses, and what it forces, and you can see that the third answer usually points back at your own ignorance. And you know how to send what you find into the one channel where it can do any good.

Choose one resolved case this week, ideally from your own state, and request its docket. Read it alongside one contemporary news story. Write the three lines for each. Then compare what you thought the record would hold with what it actually holds, and note the difference. Every serious reader of these files started with that exercise. It takes an afternoon and a records fee of a few dollars, and it sends no strangers to anyone's door.