Elisabeth Kubler-Ross published On Death and Dying in 1969, describing five stages she observed in terminally ill patients: denial, anger, bargaining, depression, and acceptance. The model was widely adopted in popular culture as a description of the normal grief trajectory following any significant loss. The problem is that when researchers began testing the model empirically, the evidence did not support it. Grief does not progress sequentially through discrete stages in the way the model implies, and the model has caused harm by making people who do not follow the predicted sequence feel that their grief is abnormal.
The Yale Bereavement Study
Paul Maciejewski and colleagues at Yale published a large empirical study of grief trajectories in 2007 that attempted to actually test the five-stage model. They found that acceptance was the most frequently endorsed indicator at every time point measured, including the earliest. Yearning, not depression, was the most prominent negative indicator, peaking at about four months. Disbelief peaked earliest, at about one month. Anger and depression also followed temporal patterns. The study supported a two-factor model of grief, with negative indicators resolving over time and acceptance remaining high throughout, rather than a sequential stage progression.
The Oscillation Model
Margaret Stroebe and Henk Schut developed the Dual Process Model of coping with bereavement, which has stronger empirical support than the stage model. They proposed that healthy grieving involves oscillation between loss-oriented coping, focusing on and processing the loss, and restoration-oriented coping, attending to life changes and rebuilding. Neither pole is sustainable alone: staying exclusively in loss-orientation produces rumination and complicated grief; staying exclusively in restoration-orientation produces delayed or suppressed grief. Healthy grieving oscillates adaptively between the two.
Complicated Grief: When Normal Grief Does Not Resolve
For approximately 10 to 15 per cent of bereaved people, grief does not follow the typical trajectory of acute intensity followed by gradual integration. Prolonged grief disorder, recognised in ICD-11 and DSM-5-TR, is characterised by persistent and intense yearning for the deceased, difficulty accepting the loss, emotional numbing, and significant functional impairment lasting beyond twelve months after the loss. It is distinguishable from depression and PTSD and responds to specific psychological treatments, particularly complicated grief treatment and cognitive processing therapy, that are distinct from standard depression treatment.
What Normal Grief Looks Like
The most robust finding from grief research is the enormous variability in how people grieve. George Bonanno at Columbia University has repeatedly found that resilience, broadly defined as the maintenance of stable functioning after loss, is the most common grief trajectory, seen in 35 to 65 per cent of bereaved individuals depending on the type of loss. The assumption that people who do not show visible distress are suppressing grief has not been supported: resilient grief is associated with better long-term outcomes, not with delayed processing. There is no single normal grief, and the absence of visible distress does not indicate pathology.