Thirty years of grief research says the bond continues. "Closure" didn't make it.

Pauline Boss, who spent more than forty years studying loss, puts it without softening: “Closure is a perfectly good word in real estate or when a road is closed after a flood, but it’s a hurtful word in human relationships”. “I strongly believe there is no closure after the loss of someone you love.” Funeral service says the word weekly — in arrangement conferences and in website copy — because the culture handed it over as the caring one. It isn’t a scold to notice that the science moved. It’s an edge: the professions a funeral home refers families to have already switched models, and the funeral home that catches up sounds like it.

The model behind “closure,” and when it was retired

“Closure” descends from the twentieth century’s working theory of grief — Freud’s “grief work,” in which the mourner’s task is to withdraw attachment from the dead and reinvest it, grief as managed detachment. The hinge came in 1996, when Klass, Silverman, and Nickman published Continuing Bonds, whose publisher’s framing states the old view and the reversal in two sentences: the dominant model held “that the function of grief and mourning is to cut bonds with the deceased,” while the book’s evidence showed “the health resolution of grief enables one to maintain a continuing bond with the deceased” — and that “such bonds are not denial”. Three years later, Stroebe and Schut’s dual process model (Death Studies, 1999) replaced the idea of stages entirely: grievers oscillate — sometimes confronting the loss, sometimes rebuilding the life around it — and the oscillation itself is the healthy coping, a finding the authors formally reaffirmed a decade later.

The stage model most people can still recite has two problems the field now states openly. First, it was never about the bereaved: Kübler-Ross’s five stages “originated from interviews she conducted with patients who were dying of a terminal illness” — people facing their own death, not people grieving someone else’s. Second, when the sequence was finally tested on actual grievers — Maciejewski and colleagues, JAMA, 2007, following 233 bereaved people — acceptance turned out to be the most endorsed response from the start, yearning was the dominant ache from month one to month twenty-four, and disbelief never dominated at all. Grief looked like acceptance rising on a slope, not a staircase. George Bonanno’s resilience research pushed further: the most common trajectory after loss is a stable, largely healthy course — and he judged the evidence for fixed stages “non-existent.” By 2021, the dual-process authors themselves audited how the five stages live online: 61% of the websites they reviewed presented the model, with an average critical-appraisal score of 1.9 out of 12, a portrayal they call misleading — one that can leave the bereaved “feeling that they are grieving incorrectly”. That last clause is the operational stake. A model can be wrong and harmless. This one isn’t: it hands grieving families a test they will fail.

What the science says instead

Continuing bonds, in plain terms: maintaining an inner relationship with the person who died — memories, imagined conversations, kept traditions, a sensed guidance — as a normal and often healthy part of grief. Hospice bereavement services now teach it as the working model, explicitly against “letting go” and “achieving closure”: the relationship adapts in physical absence rather than terminating. This is what grieving families already do without permission. The research’s contribution is that it stopped calling it a symptom — and then got precise about when it helps.

The precision matters, and it’s designable. Field and Filanosky’s study of 502 bereaved people (Death Studies, 2010) split continuing bonds into two forms: internalized bonds — comfort in memory, the deceased as a felt secure base — track with adjustment and even growth; externalized bonds — persistent sensed presence, illusions — track with complicated grief, particularly after violent deaths. Even the objects have a pedigree with a warning in it. Vamık Volkan, who coined “linking objects” in 1981, was describing pathological mourning — the keepsake that freezes its holder at the moment of loss. The later continuing-bonds literature recast objects-that-hold-connection as normal. Both are right: the keepsake that comforts and the relic that traps are two ends of one line, and the difference is whether the object opens memory or replaces living. (The same line, for what it’s worth, is the one to hold when a vendor offers a family an AI that talks back.)

What a funeral home designs differently

The words. Retire “closure,” “final goodbye,” “moving on,” “putting it behind you.” The replacements aren’t euphemisms — they’re the accurate model: carry, remember, honor, a place to return to. Sociologist Nancy Berns, who spent a book tracing where “closure talk” comes from, found it functions as a manufactured emotion — one many bereaved people report never finding. Some do; the word isn’t a lie. But building your language on it is building on the retired model, and a family that never finds closure will conclude something is wrong with them. Say “carry” and nobody fails.

The ritual. The funeral itself has an empirical warrant most owners have never been handed. Norton and Gino (Journal of Experimental Psychology: General, 2014) ran experiments on mourning rituals and found they measurably reduced grief — with the effect running through a restored sense of control, and working even for people who said they didn’t believe in rituals. The design implication is specific: control is the active ingredient, so make ritual participatory. A reading chosen by the family, a candle lit by the grandchildren — the family doing, not watching.

The objects and the year after. Aim keepsakes at the internalizing end of the line: a memory book that gets written in, a recorded voice that gets played on purpose, a place — physical or online — that gets visited, rather than a sealed relic on a shelf. And treat the calendar as part of the bond: an anniversary acknowledgment lands not as marketing but as ritual — in one small study, all 24 families sent a one-year bereavement card reported a positive reaction, and every one said the timing was right. A pilot, not a proof — but it points where the theory points: the door stays open, and someone should say so. (The operational case for that year of contact is its own piece; this is the science underneath it.)

The handoff. The honest boundary: for roughly one bereaved adult in ten — estimates range from about 5% to 16% depending on sampling — grief stays disabling past a year, and since March 2022 that has a clinical name: prolonged grief disorder, the newest diagnosis in the DSM-5-TR. Note the direction of the profession’s newest move: not toward helping people close the book faster, but toward naming grief that doesn’t resolve — and treating it. A funeral home’s role at that boundary is a warm referral to the hospice bereavement program or a clinician, never a diagnosis. Designing for the bond serves the nine; knowing the handoff serves the tenth.

The industry that hosts the bond

Tomorrow is National Grief Awareness Day, and the quiet playbook for marking it is its own subject. This piece is the theory underneath all of it: for thirty years the research has said the bond continues — through oscillation, through objects, through rituals of return — and the funeral is the first ritual of that continuing, not the ceremony of its ending. A funeral home is the one business in town whose whole craft is helping a family begin carrying someone well. The science finally agrees with the best directors’ instincts. The only thing left to retire is the word.