Grief comes in waves, not stages

Grief comes in waves — that’s the phrase people actually reach for, not “which stage am I in.” The five stages most people learned come from a different source entirely: interviews with patients who were dying, not with anyone who had lost someone else. When a study finally tested the stage model on people who had actually lost someone, the tidy order didn’t hold up the way the checklist implies.

Three weeks in, it’s easy to end up scrolling a list, wondering why anger hasn’t shown up yet, or why you skipped bargaining entirely. The list you’re checking yourself against isn’t yours. It belongs to someone else, in a different situation, from more than fifty years ago.

What are the five stages of grief, in the version everyone knows

The version that circulates names five stages, in this order: denial, anger, bargaining, depression, and acceptance. It’s usually told as a route: you enter through denial, you exit through acceptance. The person who proposed it, though, included a diagram where the stages overlapped. The version that survived in popular culture is the orderly one, not the overlapping one.

Where the five stages actually came from

Elisabeth Kübler-Ross proposed them in On Death and Dying (1969), after a project of interdisciplinary seminars at the University of Chicago. Four divinity students had approached her with a simple question: what dying patients actually needed. Kübler-Ross ended up interviewing more than two hundred of them, about facing their own death. Not a parent’s death, a partner’s, or a child’s. Their own.

The book presents the model as a qualitative study, built from conversations, not as a law you could predict in advance. In the original text, working through anger helped a patient reach a temporary bargain, then a depression described as a step toward final acceptance. It was a fairly ordered sequence, despite the overlap diagram the same book included. Kübler-Ross later broadened the model herself, from facing one’s own death to grief in general, across later talks and books. Applying it to losing someone else was never the starting point. It was an extension that came after, and one she pushed for herself.

What happened when someone tested it on people who were actually grieving

Maciejewski and colleagues (2007) ran the most-cited attempt: 233 bereaved people in Connecticut, mostly widowed, average age 63, tracked from one to twenty-four months after the loss. They measured five indicators, not quite Kübler-Ross’s original five — they swapped denial for disbelief, added yearning, and dropped bargaining entirely.

The central finding undoes the popular use of the model. Acceptance was the most frequently reported response from the first month, not the endpoint you finally reach. Disbelief, the closest stand-in for the original denial, was never the dominant response at any point in the follow-up. When the authors rescaled the data, the peaks of each indicator landed in an order close to what the theory predicts. Disbelief at one month, yearning at four, anger at five, depression at six, acceptance climbing the whole time. That’s a statistical probability the authors themselves highlight. But it measures the relative shape of each curve, not how many people were actually feeling each thing at each point. At the level of raw frequency, the order people assume didn’t hold.

The study deliberately excluded traumatic deaths and complicated grief: cases where the loss keeps occupying the center of someone’s life months later, with sleep, work, or relationships stalled. That situation isn’t described in this data, and a list of phases isn’t the tool for it. That’s where talking to someone trained helps, and findahelpline.com lists lines by country and language.

Why “what stage am I in” isn’t the question that helps

The question assumes two things that neither the original source nor the one large follow-up study backs up. That grief has a fixed shape, and that the shape is walked through in order until it’s done. Neither the study of 233 people nor the book that started the model describes anything like that.

Researchers who study grief today tend to reach for a different image. Stroebe and Schut (1999) proposed a dual process model. A bereaved person oscillates between confronting the loss directly and stepping back from it, to attend to the life that continues. Not one or the other, steadily. Back and forth, and the authors themselves present it as a proposed model, not a settled finding. It’s a closer match to what people actually describe — not a stretch you cross once, but something that keeps returning.

Grief also takes shapes no five-name list covers: not crying is one of them, and leaving a country leaves a loss with no funeral that’s just as hard to name. It isn’t the only popular piece of advice that drifted from its source, either: the mandate to forgive has a similar genealogy.

You’re still not sure what stage you’re in because the question never had an answer to begin with. The list you checked at three weeks was built for a different question: how someone faces their own death, not how someone keeps living after someone else’s. What’s worth writing down is simpler: what you felt today, without needing it to fit a name or an order.

Frequently asked questions

What are the five stages of grief? The most-cited model names five: denial, anger, bargaining, depression, and acceptance. Kübler-Ross (1969) proposed them from interviews with patients facing their own death, not with people who had lost someone else. The one large follow-up study in bereaved people didn’t find the fixed order the name implies.

Is it normal to skip some stages of grief? Yes. Maciejewski and colleagues (2007) found that disbelief — the closest match to the original denial — was never the dominant response among the 233 people they followed. Bargaining wasn’t even part of what they measured. Skipping a stage, or not recognizing it, doesn’t mean grief is going wrong.

How long does each stage of grief last? The original book doesn’t set durations, because it described people facing their own death, in circumstances that varied widely. The 2007 follow-up found peaks between the first and sixth month. But that was in one specific population — mostly widowed, no traumatic deaths — that doesn’t represent everyone who goes through a loss.