What the Body Can and Cannot Say

The protocol arrives as eleven pages, stapled at the corner, with a case number on every sheet so that no page can wander away from the others. The first page holds a diagram: two outlines of a human figure, front and back, the kind printed by the thousand and filled in by hand. Below it are boxes for height, weight, and the clothing as received. On the third day after a child disappears into a file, this is the document everyone has been waiting for. Detectives have been told it will tell them what happened.

It will tell them some of what happened. Learning which part is most of the work.

The file we have been building is a composite, assembled from the ordinary documents of real cases and labeled as such from the first chapter, so that we can read each page without trespassing on any one family. Its forms, its gaps, and its language are drawn from the record. The girl at its center has had a usual route, a perimeter, a canvass sheet, and one sentence in a detective's hand. Now she has a forensic pathologist, a physician trained to determine why and how a person died by examining the body. The pathologist has never met her family, and the report reflects this by design. It is meant to be the one document in the file that knows nothing about what anyone hopes is true.

Turn to page two. The protocol begins with the external examination, then the internal, then a list of specimens retained. Near the end comes the part people skip to, a short block of text headed Opinion. Inside it are two lines that look like the same line twice.

Cause of death: a specific injury or disease.

Manner of death: Homicide.

Cause is the medical mechanism, the thing that stopped this particular body from living. Manner is a classification of the circumstances, and most American jurisdictions allow only five options: natural, accident, suicide, homicide, and undetermined. The two lines answer different questions. Cause asks what happened inside her. Manner asks what kind of event this was in the world. A pathologist can be certain of the first and still hesitate over the second, because manner depends partly on facts that never reach the autopsy table: what a scene looked like, what a witness said, whether a fall was a fall.

In our file, both lines are filled in, and they agree with what the detectives believed on the first night. That agreement feels like confirmation. Hold that feeling at a distance for a while.

Look at what the opinion block says, and then at what it does not say. It does not say who. No line on this form has ever said who. Television has taught a generation of viewers to expect the autopsy scene to end with a pathologist straightening up and describing the killer's height, handedness, and temper. Real protocols contain observations, measurements, and opinions about mechanism. When they speak about a person other than the deceased, it is usually to note that a sample has been taken that might, later, be compared to someone.

The detective in our file reads the opinion block twice. Then she turns back to page six, because the line she actually needs is somewhere else.

Page six is headed Postmortem Changes. It records three observations that pathologists have relied on for more than a century.

The first is lividity, the settling of blood to the lowest parts of the body after the heart stops, which shows as discoloration and becomes fixed in place after a number of hours. Lividity can show whether a body was moved after death, since fixed discoloration on a surface that was not facing down tells the examiner the body lay in another position for a time. The second is rigor mortis, the temporary stiffening of muscles that typically begins within a few hours of death, peaks, and then passes over the following day or two. The third is body temperature, which falls toward the temperature of the surroundings at a rate that can be modeled. Pathologists in many countries use a chart called the Henssge nomogram to turn a temperature reading into an estimate.

Every one of these is a process, and every process runs at a speed set by conditions. Heat speeds rigor along and cold slows it. A small body cools faster than a large one, so a child loses heat more quickly than an adult. Clothing, wind, damp ground, a blanket, the time a body lay in sun and then in shade, all of it shifts the curve. The nomogram itself is published with confidence bands that widen as the hours pass. The method is honest about its own uncertainty, and so is the pathologist who uses it well.

So the line on page six in our file does not read Time of death: 4:15 p.m. It reads like this:

Postmortem interval estimated at approximately 12 to 24 hours prior to examination of the body at the scene, based on the totality of findings.

Twelve hours wide. Read in the morning, that window opens the previous afternoon and closes in the night.

Return to the margin you made at the end of the last chapter, the pencil hours beside your child's name. Try to fit a window that wide over any single one of them. It covers several. It covers most of an evening. A body cannot report the moment of a death. It reports a span of time during which the death is consistent with what the body shows, and the honest width of that span is the most important number on the page.

Now go back to the line about the usual route. The detective wrote, on the first night, that the girl usually walked home a particular way and would have reached a particular corner around a particular time. That sentence was a reasonable starting place. It was also an assumption, and it was written in ink before any of this arrived.

Here is where the file splits, and the split happens at a desk.

The detective lays the protocol beside her own timeline. Her timeline puts the girl on the route in late afternoon, out of sight of witnesses somewhere along it, and dead not long after. The protocol's window includes late afternoon. It also includes eight o'clock, and ten, and the first hours after midnight. Nothing in the pathology contradicts her theory. Nothing in it chooses her theory either.

She has two ways to write the next supplemental. In the first, she records that the autopsy findings are consistent with the established timeline. Every word of that sentence would be true. A supervisor skimming the file would read it as confirmation, and so would every detective who opened the folder in ten years. In the second, she records the window as the pathologist gave it and lists the hours inside it that her timeline does not yet account for.

She calls the medical examiner's office first. Her question is whether the window can be narrowed. The pathologist does not narrow it. He explains that stomach contents were recorded and retained, but that digestion varies too widely between people, and even within one child on different days, to pin an hour; he will not put a time on it in writing that he could not defend under oath. He tells her what he would need to say more: a reliable last-seen time from someone other than a family member, a receipt, a timestamped image. Evidence, in other words, from the world outside the body.

She hangs up and writes the second version. Then she adds a line requesting that the hours from six in the evening to two in the morning be canvassed with the same care as the afternoon.

The decision costs her. It reopens a canvass the department considered finished, and it puts in writing that the first night's picture may be incomplete. It is the kind of sentence that a defense attorney will one day read aloud. She writes it anyway.

The protocol's last pages are the least dramatic and, over the long life of a file, often the most important. They are a list.

Retained: fingernail scrapings and clippings, bagged separately by hand. Swabs from specified areas, each in its own sealed container. Hair standards. A blood card, a small square of filter paper on which a few drops of blood are dried for later testing. Clothing as received, each garment packaged in paper rather than plastic, because plastic traps moisture and moisture lets biological material degrade. Each item is given a number. Each number is written on the container, on the seal, and in the protocol.

This is evidence collection, and its purpose is narrow. The pathologist is not testing these items to find a killer. Most of them will not be tested at all during the autopsy. He is preserving them so that someone, later, can ask a question of them that no one has thought of yet.

Then the chain of custody begins, the written record of every person who has held an item of evidence, when, and why, from the moment it is collected until it is presented in court. In our file, it takes the form of a property sheet attached to each package. The first line bears the pathologist's initials. The second bears the initials of the evidence technician who carried the sealed packages from the medical examiner's office. The third bears the clerk who received them at the department's property room and logged them into a numbered location.

The sheets look like bureaucracy because they are bureaucracy. A form for every handoff, a signature for every signature. It would be easy to see this as the system protecting itself from paperwork errors. It is protecting something more fragile. A swab is only evidence if a court can be shown that it is the same swab, taken from the same child, untouched by anyone not listed, unaltered in a way no one recorded. A gap in the chain, a missing date or an unsigned transfer, invites the argument that something could have been switched, contaminated, or mishandled. Sometimes that argument is right. The forms exist so that it can be answered.

The last line on the property sheet is a location code: a room, a shelf, a box. The evidence locker is not a dramatic place. It is a climate-controlled room with metal shelving and boxes labeled in marker, some of them with handwriting that belongs to people long since retired. When the packages from our file go onto their shelf, the investigation outside keeps moving. The detective canvasses the evening hours. Tips come in. Seasons change. Inside the box, almost nothing changes at all. The paper bags stay folded. The seals stay sealed. The swabs hold whatever they held on the third day, and the room holds them.

Put the protocol through the three questions.

It says how the girl died, in medical terms, and that the circumstances were a homicide. It says the death fell within a span of hours, and it says, with care, that the span is wide. It says what was kept, and where.

It misses the moment of death, which no protocol truly records. It misses who was there. It misses everything the body could not carry: the route she took, the people she saw, the reasons. It also misses, by design, every theory the detectives brought with them, and this absence is easy to mistake for agreement.

The question it forces is the one the detective answered at her desk: what, specifically, would this document have to show to prove the working theory wrong? If the answer is nothing, if every possible finding can be described as consistent, then the protocol has not tested the theory at all. It has been used to decorate it. Forensic science is most powerful in exactly the opposite role, as an instrument for eliminating. A window that excludes the hours a suspect can account for clears that suspect. Lividity that shows a body was moved rules out the idea that it fell where it was found. A profile from a fingernail scraping that does not match a man the neighborhood has already convicted in its mind can give him his life back. The science rarely names anyone. It takes possibilities off the table, and it waits for someone to bring it a question sharp enough to answer.

Those of you who listen to podcasts about cases like this have heard hosts announce what an autopsy proves. Listen more carefully next time. Ask whether they are quoting a cause, a manner, or a window, and whether they are reporting its width. If a time of death is given to the minute, someone has trimmed the edges off a range, and the hours they trimmed may be the ones that matter.

Go back to your margin. Beside each pencil hour, you wrote what could turn it to pen and how long that proof would last. Now add a third mark. Put a small circle beside the hours a body could speak to, and be strict: a body can speak to a span, never to a minute, and never to a name. Most of your circles will cover several cells at once, and none will tell you who was nearby. Everything left without a circle depends on a camera, a neighbor, a receipt, a person who remembers, and you have already written down how quickly those disappear.

That is the trade a homicide file makes on the third day. The living evidence keeps aging. The preserved evidence goes onto a shelf, under a location code, behind a signature, and stops. Whoever opens the next document, the tip log that has been filling up since the first news broadcast, will be working against the first clock while the second one stands still in a numbered box.