Ask someone what has been going on and they will not answer with topics. They will answer with people. Work is fine, but my manager has been strange. The move was easy, but my mother has opinions about it. Even the answers that sound like subjects turn out, one question later, to be about someone.
Software almost never files a life this way. It files by session, by date, by tag, by project, by mood. All of these are real organising principles and all of them cut across the grain of how a life is actually structured, which is: a small number of people, recurring, for years.
The number is small, and that is the point
The set of people who matter to any of us is not large. Robin Dunbar's work on social group size proposed a cognitive ceiling on stable relationships, with an inner structure of much smaller layers — roughly five people in the closest circle, around fifteen in the next.1 The precise numbers are debated. The shape is not seriously in question: intimacy is steeply tiered, and the top tier is tiny.
This matters for a practical reason. A record organised by topic faces an unbounded space — there is no limit to the number of subjects a life can touch. A record organised by person faces a small, slow-changing set. Most of what you will ever write about will concern perhaps a dozen people, and the composition of that dozen changes over years rather than weeks.
That is what makes the person a usable unit. It is stable enough to accumulate against.
What a topic loses
Say you file by subject. Work goes in one place, family in another, health in a third. It is clean.
Now: your father is ill, your work has become unbearable because you are exhausted from hospital visits, and you have stopped seeing anyone. Three folders. One situation. The filing has separated the parts of a single thing and stored them where their relationship is invisible.
File by person instead and the same material arranges itself correctly without effort. Your father: ill since March, and here is everything downstream of that. The work entries and the health entries and the vanished-friendships entries sit together because they are, in fact, together — they are all about your father, and you knew that, but the folders did not.
Topics are a taxonomy imposed from outside. People are the structure already there.
Absence becomes visible
There is a property of person-filing that nothing else gives you, and it may be the most valuable one.
If the record is organised by person, then a person can be missing. Someone who appeared every week for two years and has not appeared in eleven weeks is a visible fact — a gap with a shape and a length.
Organise by session and that same gap is nothing at all. It is not stored anywhere. It is the absence of entries among other entries, which is indistinguishable from you having been busy. Nothing can point at it because there is no it — no row, no record, nothing that says this used to be here.
Yet the friendship that thinned is one of the most common quiet losses in adult life, and it is made entirely of absence. No argument, no event, nothing to write about on any particular day. You do not notice it happening; you notice, eventually, that it has happened. A record filed by person is one of the few instruments that can catch it while it is still in progress.
The same person is several people over time
Filing by person invites an error worth naming, because a naive version of this idea is worse than no version.
A person in your life is not a fixed entity with a fixed valence. Your relationship with your mother at twenty-four and at thirty-four are different relationships involving the same woman. If the record collapses those into one profile — your mother, difficult — it has built a caricature and will keep serving it back to you long after it stopped being true. Worse, it will make it harder to see that it stopped being true, because the summary is right there, confident, every time.
This is where the person and the timeline have to work together rather than separately. A person is not a label. It is a thread, with a direction, along which things changed. What she was in March can be superseded by what she was in October without March being erased — because the difference between them is exactly what you would want to see.
Martin Conway's account of autobiographical memory describes something similar in human terms: memories are organised into lifetime periods and general events, and the periods are frequently defined by relationships — the years with someone, the years after.2 We already store our lives in person-shaped chapters. The chapters have boundaries because the relationships changed.
Two people, one record
The last thing person-filing gives you is the connection between people, which no other organising principle surfaces at all.
You write about your partner. You write about your sister. Separately, over months, both records grow. Filed by session, they are interleaved noise. Filed by person, something becomes visible that neither entry contains on its own: the weeks when your sister appears are disproportionately the weeks when your partner appears unhappily. Neither of you has ever said the two are connected. You may not think they are.
We would not tell you they are. That is exactly the kind of claim that is not ours to make — it might be a coincidence, and even if it is not, the reason is yours to know and not ours to guess. But the co-occurrence is a fact about the record, and a fact about the record is something we can honestly show you. What it means is a different question, and it belongs to you.
If any of this resonated, you can read more about the people you stopped mentioning or why your relationship patterns keep repeating.
- Dunbar, R. I. M. (1993). Coevolution of neocortical size, group size and language in humans. Behavioral and Brain Sciences, 16(4), 681–694. The layered structure of the inner circles is developed in Dunbar, R. I. M. (1998). The social brain hypothesis. Evolutionary Anthropology, 6(5), 178–190.
- Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107(2), 261–288, on the hierarchy of lifetime periods, general events and event-specific knowledge.
- Wegner, D. M. (1987). Transactive memory: A contemporary analysis of the group mind. In B. Mullen & G. R. Goethals (Eds.), Theories of Group Behavior. New York: Springer.
- Bartlett, F. C. (1932). Remembering: A Study in Experimental and Social Psychology. Cambridge: Cambridge University Press, on the way stored impressions simplify toward a stable and often inaccurate summary.