Somewhere around the third month, people begin asking a version of the same question. Not out loud, usually, and not unkindly. It arrives as a pause in a conversation, or as a gentle suggestion about getting back out there. The question underneath it is whether you are on schedule. Almost everyone who has lost someone has felt it land, and almost everyone has quietly asked it of themselves: should I be further along than this by now?

Where the Stages Came From

The schedule people have in mind is nearly always the same one. Denial, anger, bargaining, depression, acceptance. It is one of the few pieces of psychological theory an ordinary person can recite from memory without ever having read a word of the source, which is a strange sort of success for a book.

The book is On Death and Dying, published in 1969 by the Swiss-American psychiatrist Elisabeth Kübler-Ross. It grew out of hundreds of interviews conducted in Chicago hospitals with people who were terminally ill, at a time when doctors routinely did not tell patients they were dying and nobody thought to ask them how that felt. It was an act of real moral courage, and it changed how hospitals treat the dying.

Notice what it was about. The five stages describe how people met their own approaching death. They were not a study of the bereaved. The move from one to the other happened afterward, in the culture rather than in the literature, and for a long time nobody checked whether it held.

What Happened When Somebody Tested It

The first serious empirical test arrived in 2007, when a team led by Paul Maciejewski published the Yale Bereavement Study in JAMA. They followed bereaved people after a death from natural causes and tracked five responses, disbelief, yearning, anger, depression and acceptance, from one month to two years out.

The result was not the clean debunking anyone expected. The five responses did reach their peaks roughly in the predicted order, which is a genuine point in the model's favor and is usually left out when the study gets cited. But two other findings pull the ground out from under the way the stages are actually used.

The first is that acceptance was the most strongly endorsed response at every single point of measurement, including the first month. It did not arrive at the end of a journey. It was there from the beginning and it grew. The second is that the dominant painful feeling, throughout, was not denial and not anger. It was yearning. Plain missing. Anger and depression were present, but they were never the main event.

That reframes the whole thing. Grief, measured rather than theorized, looks less like a sequence of psychological positions to be worked through and more like a long ache that slowly makes room around itself.

Nearly two decades earlier, Camille Wortman and Roxane Silver had made a related point in a paper on the myths of coping with loss. A substantial proportion of bereaved people never show the intense distress the model predicts, and when researchers followed them, that turned out not to be delayed grief or repressed grief. It was just their grief. The trouble with a stage model is that it tells everyone what they ought to be feeling, and then leaves those who feel otherwise convinced something is wrong with them.

What It Actually Looks Like From Inside

The description that tends to make bereaved people sit up is not a stage model at all. In 1999 Margaret Stroebe and Henk Schut proposed what they called the dual process model, which says that grieving people move back and forth between two kinds of attention.

One is loss-oriented: the crying, the photographs, the going over of the last conversation. The other is restoration-oriented: the paperwork, the new routines, the business of building a life that has a hole in it. People oscillate between the two, sometimes across months, often across a single afternoon.

The oscillation is not a failure of focus. In this model it is the actual mechanism of grieving, and it explains two things that otherwise seem shameful. It explains how a person can laugh properly at a wake, because attention has swung to the room and the living. And it explains why the crash so often lands on an unremarkable Tuesday, weeks later, over something trivial like the sound of a particular car door.

You Do Not Have to Break the Bond

There is an older assumption buried under the popular idea of grief, which is that the work of mourning is to detach: to withdraw the attachment, close the account, and reinvest elsewhere. Most people have absorbed some version of it, which is why they feel faintly foolish talking to someone who has died.

In 1996 Dennis Klass, Phyllis Silverman and Steven Nickman published a collection that argued the opposite from the evidence in front of them. Bereaved people, they observed, do not generally sever the relationship. It changes form and continues. They talk to the person. They keep the coat in the closet for years. They ask, silently, what she would have made of this. They named this continuing bonds, and it has been one of the more useful ideas in the field.

Honesty requires a caveat, because this one has hardened into a slogan. A systematic review published in 2023 found that the relationship between continuing bonds and how well people do is genuinely mixed, and appears to depend on the kind of bond. A connection that comforts, and that coexists with knowing the person is dead, tends to sit alongside good adjustment. A connection that requires the death not to be quite real does not. That distinction is worth more than the reassurance.

When It Is More Than Grief

In 2022 the American Psychiatric Association added prolonged grief disorder to the DSM-5-TR, placing it among the trauma and stressor-related conditions rather than with the depressive disorders. The criteria require that the death was at least twelve months ago for an adult, six months for a child or adolescent, alongside persistent intense yearning or preoccupation and a cluster of symptoms that can include identity disruption, disbelief, avoidance of reminders, emotional numbness, intense loneliness and an inability to re-engage with life. The World Health Organization's ICD-11 recognizes the same condition at six months.

How common is it? A pooled analysis by Marie Lundorff and colleagues in 2017 put it at close to ten percent of adults bereaved by non-violent causes, while estimates drawn from representative general population samples run lower, in the range of three to four percent. The number matters less than what it is for.

It is worth being precise about what a diagnosis of this kind is claiming, because the announcement of it in 2022 drew a fair amount of alarm about grief being medicalized. It is not a ruling that someone has grieved wrongly or for too long. It describes a state in which grief has stopped moving at all and has become the thing standing between a person and their own life. That state has treatments, grief-focused cognitive behavioural approaches chief among them, with reasonable evidence behind them. Naming it is what makes it treatable. It is not a verdict on the size of anyone's love.

What Helps

Say the name. The most common complaint of the bereaved is not that people said the wrong thing. It is that people stopped mentioning the person at all, out of a fear of causing pain that is already there. Saying the name gives the grieving person permission to speak.

Offer the specific thing. Anything shaped like let me know if you need anything puts the work of asking onto the person least able to do it. I am bringing dinner Thursday, is six alright, does not.

Expect the anniversary. Bodies keep dates. The week before a birthday, a death day or a wedding anniversary is often harder than the day itself, and knowing that in advance takes some of its power away.

Let it come in waves. Grief is not a slope. It arrives in swells that get further apart rather than smaller, which is why year two can catch people badly: the swells are rarer, so there is less warning and less patience available from everyone else.

Distrust the timetable, including your own. The stages were never a schedule and were never about the bereaved. If a private deadline is making you feel like a failure at grieving, it is worth knowing it came from a misread of a book about dying patients.

Sit with other people who have lost someone. Bereavement groups do something that even good friends cannot, which is to stop being surprised by you.

The Part Worth Keeping

Kübler-Ross spent her later years pointing out that the stages were never meant as stops on a line, that not everyone passes through all of them, and that they do not come in a set order. That correction has had a fraction of the reach of the original, which is how these things go.

What survives from her work is the part that mattered in the first place: that a person facing loss deserves to be asked how it is, and listened to without being managed. There is no schedule to fall behind. There is only a relationship that has changed its shape and has to be carried differently, and a community whose one real job is to keep letting you talk about the person as though they were someone who actually lived here.