Nobody measured the letter to someone who died

Writing a letter to someone who died is a practice older than any study of it. Pennebaker and Beall (1986) — the paper most often cited to back it up — measured something else: how writing about a hard event, addressed to no one, changed the health of 46 students six months later. That gap between what got measured and what gets promised is the subject of this piece.

What Pennebaker and Beall actually measured

They measured whether writing about a difficult event, privately and with no addressee, changed health outcomes over six months. Forty-six introductory psychology students at Southern Methodist University, recruited for course credit, wrote for fifteen minutes a night on four consecutive nights — November 14 to 17, 1983 — about one of four kinds of topic.

The design had four groups, not two. Twelve wrote about trivial topics — their living room one night, the shoes they had on the next — as the control group. Twelve wrote only about how they felt about a personal trauma, without saying what happened. Eleven wrote only the facts, with no mention of feelings. Eleven wrote both, fact and feeling together. That last group — facts and feelings combined — is the one the line “writing about hard things works” almost always points to.

Nobody, in any group, addressed another person. Campus health center staff, blind to which group each student belonged to, counted how many times each one had visited before the experiment (late August to mid-November) and after it (mid-November to mid-May, six months).

How large the effect was, and how sure

An effect right at the edge of statistical significance, in a group of eleven people. The group that wrote facts and feelings together was the only one whose health center visits didn’t rise over the following six months; the other three groups’ visits went up, by roughly one to one and a quarter visits on average. The difference across the four groups landed at p = .055 — on the “not significant” side of the conventional .05 threshold, not inside it.

Pennebaker and Beall said as much themselves: the results “should be viewed as promising rather than definitive.” They published no effect size with a confidence interval, only the test statistic and that borderline p. The sample was small. Nobody was selected for carrying an unresolved trauma. And the authors themselves note they can’t rule out that some participants adjusted both their health center visits and their follow-up questionnaire answers once they knew, after the debrief, what the study was looking for. The visit records were kept by staff blind to each student’s condition, but a blind record cannot stop someone from deciding not to go: the authors put the main measure inside their own caveat.

What the study never tested

Nobody wrote a letter. The format in all four conditions was a first-person essay about something hard, not a communication addressed to anyone, let alone to someone no longer alive.

Of the 127 essays written by the three trauma groups combined, 27% dealt with the death of a close friend, family member, or pet. Loss shows up in the raw material of the study. Just not like this: nobody was asked to say something to the person who died, nobody signed anything, nobody wrote with a “you” pointed at one specific person. These were essays on a topic, not letters to a recipient. And the sample wasn’t people in recent grief, either — healthy students, not selected for a loss or for how long ago it happened.

What the evidence says about grief specifically

That there’s no evidence emotional disclosure in writing helps people adjust to a loss in normal bereavement. Stroebe, Schut, and Stroebe (2005) reviewed four research domains — social support, emotional disclosure, experimentally induced emotional disclosure (the family of studies Pennebaker belongs to), and grief intervention — and found that evidence in none of the four.

It’s the most direct review that exists on whether the mechanism behind expressive writing — telling what happened and what it felt like — transfers to grief. Their answer, across all four domains, is that the data doesn’t support it. That doesn’t mean writing does harm; the review didn’t measure that either. It says something less comfortable, and less visible behind the practice’s popularity: the intuition that “getting it out” helps has less backing than its reputation suggests.

Between those two papers — the one measuring writing about hard things in general, and the one reviewing whether disclosure helps in grief — we did not find, in preparing this piece, an experiment that tested writing a letter addressed to someone who died specifically, against another condition, with an outcome measured afterward. It may exist and we missed it. What’s certain is that it isn’t the study usually cited to defend the practice.

What writing it is for, then

For putting on record what went unsaid, with a date. There’s no evidence the letter speeds up grief or softens it. There’s no evidence of the opposite either — nobody measured it in either direction.

What it produces, verifiably, is a document. Your version of what was left unfinished, written while you still remember it in detail. Memory of the first months after a loss edits itself — the order shifts, some sentences soften, others go missing; what you write this week keeps words you won’t have exactly a year from now. A journal that keeps that letter with its date doesn’t improve it. It keeps it.

How to write it, if you decide to

• Give it a time limit. Fifteen or twenty minutes, with a timer. The original design was short and finite; there’s no evidence more time produces more effect.
• Write the fact and the feeling together. It’s the only condition in the 1986 study that showed any difference, even though that study wasn’t about letters.
• Address the person, by name, in the second person. That’s what makes it a letter instead of an essay. It’s a change of format, not a formula with studies behind it.
• Decide before you write what you’ll do with it. Keep it, tear it up, or just close the notebook. As far as we could verify, no option has more backing than the others.
• If what surfaces overwhelms you — you can’t function, can’t sleep, or the thoughts that show up scare you — that letter isn’t yours to resolve alone: seek professional help; findahelpline.com lists services by country and language.

Frequently asked questions

Does the relationship have to have been good for this to work? No study, not the ones cited here or any other we found, has tested that condition. What the letter asks for is honesty, not a resolved relationship; some people write it precisely because something was left unresolved, not despite that.

Do I have to send it anywhere? There’s no recipient who could receive it, so no. Keeping it in your journal leaves the dated record; tearing it up is a physical close that works just as well for some people. Neither one has, based on what we reviewed for this piece, more backing than the other.

How many times should I write it? The science on this specific practice doesn’t answer that — there’s no study comparing writing it once against writing it several times — so there’s no rule to hand you with an author and a year attached. What you can measure is how you leave the page each time: lighter once is different from more sunk each night.

Does this replace grief, or speed it up? No, based on the evidence available. Stroebe, Schut, and Stroebe (2005) found no evidence that emotional disclosure in writing helps people adjust to a loss in normal bereavement, and their review didn’t measure it slowing things down either. What the letter offers is more modest than an accelerant: a dated document of what’s true for you today.

What can be said today: writing about something hard, in general, showed a small difference right at the edge of significance, in a study of 46 people who had not recently lost anyone. What can’t be said is that finding is about this letter — nobody has measured writing to someone who died, in this exact format, against another condition. There’s more on this territory in why haven’t I cried since my dad died and in grieving your home country, on loss with no funeral to mark it. If this is your first time keeping a record of something hard, starting a journal is the simplest place to begin.