There is a common late-night behaviour that most people who have lost someone will recognise. You open a phone. You scroll back through a thread with someone who is dead, or estranged, or simply gone from your life in the way people sometimes go. You reread the messages you have already read many times. You linger on the ordinary ones. You close the phone. You do not tell anyone you did this.
The behaviour is often described, when it is described at all, in the language of being stuck. Unable to move on. Ruminating. Not doing the work. The vocabulary around grief in modern secular life tilts heavily toward forward motion, toward closure, toward the milestone of finally deleting the thread.
The people who actually work with the bereaved tell a different story.
What older cultures did with the same impulse
Many human cultures have built a scaffolding around continued contact with the dead. Not metaphorical contact. Actual, ritualised, calendared contact. Anniversary meals. Photographs kept on a shelf with candles. Letters written and read aloud. Set days each year when the living speak to the dead as though the dead can hear, because within the ritual, they can. These practices were not fringe. They were structural. They gave a person a container for the fact that love does not obey the timetable of a funeral.
Modern secular life has quietly retired most of this. The default framework most people inherited was closure: grieve, integrate, move on, and if you are still doing something that looks like contact after a set period, something has gone wrong with you.
That idea has been slowly abandoned inside the bereavement field itself. Kenneth Doka, writing in Psychology Today, describes the older clinical model as one where recovery meant withdrawing emotional energy from the person you lost and reinvesting it elsewhere. The newer model, which he credits in part to the work of Dennis Klass on parents of deceased children, treats the bond as something that continues. The people we have loved become part of our biography. They keep leaving marks. The connection is not the problem. It is the point.
That shift is now the mainstream position rather than a minority one. A January 2026 essay by psychotherapist Mark Shelvock describes the early-twentieth-century framing of grief as detachment, letting go, severing ties, and sets against it the continuing-bonds understanding the field now works from: bonds are transformed rather than cut.
What the phone actually holds
A message thread is not a diary and not a shrine, but it does something both of those older objects used to do. It preserves the specific voice of a specific person on a specific ordinary day. Not their eulogy voice. Not the curated voice of a memorial post. The voice that asked whether you had eaten, that sent a photograph of a dog, that argued with you about a restaurant, that said goodnight.
This is different from most of what social media has done to grief. Digital mourning practices often involve memorial pages, hashtags, comment threads, and public tribute posts, all of which are essentially performances of loss for an audience. Rereading a private thread is the opposite. It is unwitnessed. Nobody hearts it. Nobody comments. It is closer to the older private rituals than to the newer public ones.
Which may be part of why it feels shameful. It has no external structure to legitimise it.
The mechanism, described carefully
A caveat is worth stating plainly. Silicon Canals is a publication, not a clinic, and what follows is a reading of the available material, not psychological advice.
What the continuing-bonds literature suggests is that the mind of a bereaved person does not treat the deceased as gone in the way the physical world treats them as gone. The bond kept updating. The person kept, in a certain sense, existing. This was not denial. It was how the mind organises love that has nowhere obvious to go.
Rereading messages sits inside this pattern comfortably. It is a way of letting the inner representation refresh against the actual voice, the actual phrasing, the actual jokes. The thread is evidence. It confirms that the person was real, that the relationship was real, that you have not invented the version of them you carry.
An August 2026 piece in the Index-Journal, summarising practitioner views on continued conversation with the dead, makes a related point. Speaking to a deceased person, silently or aloud, is not something most therapists treat as a problem in itself; they look instead at whether it helps the person carry the loss and keep going. What rereading messages does is give that conversation a script the other side once actually wrote.
Why it looks like being stuck when it isn’t
The behaviour draws suspicion because it does not produce a visible outcome. There is no post, no photo album, no trip, no tattoo. Nothing to point at. In a culture that reads productive grief as grief that generates artefacts, the person rereading a thread at midnight looks like a person who has failed to convert their sadness into anything legible.
But the same culture is comfortable with other private returns. Silicon Canals has covered the pattern of adults rereading the same handful of novels over decades, using the books as a way of checking in with earlier versions of themselves. The mechanism there is not so different. A reread is a form of return. The self who first read the book, and the person you were with when they first sent the message, are both being visited.
There is also the quieter version of loss that does not involve death at all. Silicon Canals has written elsewhere about the grief inside long marriages, where the person you married slowly becomes someone only you remember the earlier version of. People reread threads from living people for the same reason they reread threads from the dead. The person on the other end of that older conversation is, in a real sense, no longer available.
What the behaviour is not
None of this is an argument that rereading is always benign. There is a version of it that takes over, that displaces sleep, that substitutes for the harder work of speaking to people who are still here. The continuing-bonds framing, as Doka is careful to point out, is not a case for perpetual pain. The bond is meant to keep updating, not to become a chain.
Doka makes a further observation in the same essay that is worth holding onto here: grief tends to return in surges attached to milestones, a birth, a wedding, an anniversary, and those surges are not evidence that someone has slipped backwards. They are evidence that the relationship is still doing what relationships do.
The line between rereading that helps and rereading that hurts is not one Silicon Canals is qualified to draw for anyone. What can be said is that the behaviour, considered on its own, is not a sign that something has gone wrong. It is a small private ritual in a culture that no longer supplies public ones.
The wider pattern
Across the reporting on grief in the digital age, the same theme keeps surfacing: people are building their own scaffolding in the absence of shared ones. Some of it is public, in online memorials, group threads, and tribute videos. Some of it is deliberate and structured, like the grief travel Forbes examined in August 2025, where bereaved people take trips built around reflection and ritual rather than distraction. And some of it is entirely private, unwitnessed, and looks from the outside like doing nothing.
If you have watched someone do it, or done it yourself, the thing worth knowing is that it belongs to a very old family of behaviours. The ritual is not missing. The framework around it is.
If the pattern is displacing sleep, work, or the people still in your life, that is worth talking to a GP or a bereavement counsellor about. Cruse Bereavement Support in the UK and similar organisations across Europe exist for exactly this.
The rest of it, the quiet scroll, the thread reread on an ordinary Tuesday, is closer to lighting a candle than to being stuck.