The five stages of grief, and why grief does not arrive in stages
Denial, anger, bargaining, depression, acceptance. Where the five stages came from, what researchers found when they tested the order, and the model that fits real grief better.
Denial, anger, bargaining, depression, acceptance. Almost everyone can name them, and a great many people going through a loss quietly conclude that they are doing it wrong, because what they actually feel does not look like that list.
They are not doing it wrong. The five stages are the most famous idea in grief and one of the least supported, and the mismatch between the model and the experience causes real harm: people wait for stages that never come, worry about ones they skipped, and try to force themselves into an acceptance they were told comes last.
Where the stages actually came from
Elisabeth Kubler-Ross published On Death and Dying in 1969. Her subject was not bereavement. She was writing about people who were dying, drawn from interviews with terminally ill patients in a Chicago hospital, and her argument was largely a humane one: that dying people were being managed rather than listened to.
Two things happened afterwards that she did not do. The stages were transferred from dying to grieving, which are different experiences facing in opposite directions. And they were hardened into a sequence, a set of boxes to be ticked in order, when the original work described them as common responses that could recur, overlap or be absent altogether.
The model spread anyway, because it is memorable, because it offers a map at the moment people most want one, and because there was very little else in general circulation.
What happened when someone tested it
For most of its life the model was never checked against data. When it finally was, in a study by Maciejewski and colleagues that followed bereaved people over two years and measured each of the five responses at intervals, two findings broke the sequence.
Yearning dominated, not denial or anger. The strongest negative response at every stage of the study was missing the person: longing, searching, the reflex to tell them something before remembering. Denial and anger were present but comparatively minor. The emotion at the centre of grief is simply absence, and it is the one the famous list does not contain.
Acceptance was highest from the start. Rather than arriving at the end, acceptance was the most endorsed item from the earliest measurement onward, and it rose gradually. People know almost immediately that the person has died. What changes over time is not the knowing but how much room it takes up.
Later work has largely confirmed the shape: the emotions in the model are real and common, the ordering is not.
The model that fits better
Most bereavement researchers now work from something called the dual process model, put forward by Margaret Stroebe and Henk Schut. It is less quotable and much more recognisable.
It describes two kinds of coping that a grieving person moves between:
| Loss-oriented | Restoration-oriented | |
|---|---|---|
| What it involves | Facing the loss: crying, remembering, looking at photographs, missing them | Facing the life that continues: the paperwork, the job, the new routines, other people |
| What it feels like | Painful and necessary | Sometimes a relief, sometimes like betrayal |
| Common worry | “I am stuck in this” | “I am avoiding it” |
The claim of the model is that healthy grieving oscillates between the two, and that the oscillation is not a distraction from grief but the mechanism of it. You cannot face a loss continuously; nobody can. Nor can you outrun it with activity. What people actually do is alternate, sometimes within a single hour.
This explains the thing that most confuses people about their own grief: laughing at something four days after a funeral, then being taken apart by a song in a supermarket the following week. That is not instability. On this model it is exactly what the process looks like.
Two other findings are worth knowing because they contradict advice people are still given.
Continuing bonds are not a problem. Older grief theory treated the goal as detachment, letting go, cutting the tie. The work of Klass, Silverman and Nickman turned that around: most people maintain a relationship with the person who died, through talking to them, keeping objects, marking dates, and this is associated with adjusting well rather than badly. You do not have to let go of anyone to recover.
Most people are more resilient than they expect. George Bonanno’s long-running trajectory research finds that the single largest group after a bereavement shows relatively low, stable distress that eases over months, without a period of intense grief. A smaller group has intense grief that resolves, and a smaller group again has grief that stays severe. If you feel less shattered than you assumed you would be, you are in the largest group, not the coldest one.
What this changes in practice
The practical value of dropping the stage model is that it removes two false questions. Not “which stage am I in”, and not “why have I not reached acceptance yet”.
What replaces them is more useful and considerably less tidy.
- Expect oscillation, and stop grading it. A good day is not progress and a terrible day is not relapse. Over months the ratio shifts, and the ratio is the only trend worth watching.
- Let both sides happen. Deliberately avoiding all reminders and deliberately staying inside the grief are both single-sided, and both tend to stall. The people who do best move between the two, badly and unevenly.
- Anniversaries reload. The first birthday, the date itself, the season the light changes. These are predictable, and knowing a hard week is coming is most of what makes it survivable.
- The practical load is part of the grief. Estates, accounts, the hundred small notifications that a person has died. This is restoration-oriented work, not an interruption of mourning, and it is exhausting in its own right.
Tracking without turning grief into a project
There is a narrow, genuine use for writing any of this down, and it is not to monitor your recovery.
It is that grief distorts memory of itself. In the middle of a bad week, the previous good fortnight is unavailable to you; it does not feel like it happened. A record is the only thing that can tell you the ratio has shifted, because your memory in a bad week will insist it has not. It also finds the things that reliably help and the things that reliably do not, which is otherwise almost impossible to see from the inside.
Three things are enough: the date, roughly how heavy the day was, and one line on what happened around it. Anything more elaborate turns into a chore you abandon in week two, which is a worse outcome than not starting.
GriefOS keeps that record without asking you to perform progress: heavy days and light ones side by side, the dates that are coming, and space for the letters to the person that most people write and never send. It runs offline in your browser and stores nothing off your device, which matters more here than in most subjects. A notebook does the same job. So does a plain mood record, if the grief is sitting on top of something you were already managing.
If you would rather have prompts than a blank page, grief journal prompts covers what to actually write, and what the research does and does not show about whether writing helps.
The part that is genuinely a medical question
Grief is not a disorder. A minority of it turns into one, and the line is now formally drawn.
Prolonged grief disorder entered the DSM-5-TR in 2022. For adults it requires at least twelve months since the death, intense yearning or preoccupation with the person on most days, and impairment that persists rather than gradually loosening. Around one in ten bereaved adults meets the criteria, with higher rates after sudden, violent or unexpected deaths and after the loss of a child.
The twelve month threshold exists to avoid pathologising an ordinary first year, and it is worth being clear that it cuts both ways: it is not a target and it is not a permission slip to wait. If at any point there are thoughts of harming yourself, if you cannot eat or sleep for extended stretches, or if daily functioning has stopped rather than slowed, that is a reason to talk to a doctor now.
What you do not need a doctor for is failing to progress through five stages in the right order. That was never the map, and there is nothing wrong with the way you are doing this.
This is general information about grief, not medical advice or therapy. If you are struggling, a GP or a bereavement service is the right next step, and if you are in crisis, contact your local emergency number or a suicide prevention line now.
Frequently asked questions
What are the five stages of grief?
Denial, anger, bargaining, depression and acceptance, described by Elisabeth Kubler-Ross in her 1969 book On Death and Dying. She based them on conversations with people who were dying, not with people who had been bereaved, and she did not present them as a fixed sequence everyone passes through. The application to grief came later, largely from other people.
Do the five stages of grief happen in order?
No. When researchers actually tested the sequence in bereaved people, the strongest response at every point was yearning rather than denial or anger, and acceptance was the most endorsed item from the earliest measurement onward rather than the final destination. The emotions in the model are real and common. The order is not.
How long does grief last?
There is no schedule, and anyone who gives you one is guessing. What research does show is that intensity usually falls over the first year or two for most people, while the grief itself does not end so much as change shape. Waves that arrive years later, especially at anniversaries and birthdays, are normal and are not evidence that you have gone backwards.
Is it normal not to feel the stages at all?
Yes, and it is more common than most people expect. Studies that follow bereaved people over time find that the largest single group shows relatively stable, moderate distress that lifts over months, without an intense grief phase. Feeling less devastated than you thought you would is not coldness and it is not denial.
What is the dual process model of grief?
It is the model most bereavement researchers now work from, proposed by Margaret Stroebe and Henk Schut. It describes grieving as oscillation between two kinds of coping: loss-oriented, which is facing the loss itself, and restoration-oriented, which is dealing with the practical and social life that continues. Healthy grief moves back and forth between the two, and the movement is the mechanism rather than a distraction from it.
When does grief become something a doctor should look at?
Prolonged grief disorder was added to the DSM-5-TR in 2022. It requires at least twelve months after the death for adults, intense yearning or preoccupation on most days, and significant impairment in daily life. Roughly one in ten bereaved adults meets that threshold. Separately, any thoughts of harming yourself are a reason to get help now, not at twelve months.
Should I try to reach acceptance?
Acceptance is not a finish line you cross and then feel differently afterwards. In the research it looks less like a stage and more like a gradual, uneven shift in how much of your attention the loss takes. Aiming at it directly tends to produce a performance of being fine. Living the ordinary week, and letting grief interrupt it, gets you there more reliably.
Medical disclaimer: This guide is for general information and education only. It is not medical advice, diagnosis or treatment, and it is not written by a clinician. A planner is a place to keep your own notes, including notes you may want to show your doctor. Always talk to your doctor or midwife about your health and any decisions.
Ecuato builds interactive dashboard planners as single offline HTML apps. About Ecuato, browse all planners or visit the Etsy shop.