How to Keep a Human Record

You are, right now, the most thoroughly documented person who has ever lived, and you almost certainly could not tell me what you did on this date three years ago. There are probably forty photographs of it.

This is the central paradox of modern memory: capture has become effortless and recall has become worthless. We solved storage and accidentally destroyed the thing storage was meant to serve. This essay is about repairing that with a method, not a tool.

Storage is not memory

Human memory was never an archive. It was a process of discarding with intent — a continuous, mostly unconscious judgement about what deserved to be carried forward. Everything you remember from age nine survived a twenty-year selection process. That selection is not a limitation of the system; it is the system’s output. It is, arguably, where you are located.

A perfect archive performs no selection. It therefore contains no judgement, and a record without judgement is not a memory of a life — it is undifferentiated sediment. Twelve thousand photographs of one holiday do not remember the holiday better than four photographs. They remember it less, because remembering is choosing and no one chose.

So the first move is uncomfortable and non-optional: a record you never curate is not a record. You have to be the filter, because the filter was the part that mattered and it is now the only part not provided for you.

What actually survives being revisited

I have spent a long time looking at which personal records people return to years later, and which they never open twice. The pattern is stable and almost nobody predicts it correctly in advance.

What people never revisit:

• Photographs of scenery without people in them.
• Documentation of events that were already well documented by others — concerts, landmarks, weddings from the guest’s angle.
• Anything captured to prove attendance rather than to hold a moment.
• Long comprehensive accounts written in a formal voice.

What people return to repeatedly, often for decades:

• Voices. Any recording of a voice, especially unremarkable ones. This is the single most under-collected and most re-accessed category in existence.
• Ordinary rooms on ordinary days — the kitchen, the desk, the mess, the unremarkable light.
• Very short written fragments with a specific sensory detail and a timestamp. Two lines beat two pages, consistently.
• Handwriting, from anyone, about anything.
• Ambient sound of a place you lived. Forty-one minutes of a kitchen outperforms every photograph of that kitchen.
• Things recorded when nothing was happening.

The method

Six practices. Individually trivial; together they produce, over years, something no automatic capture system can produce.

1. Keep one line a day, not a page. One or two sentences, always including one physical detail and one thing you felt. The physical detail is what makes it re-readable later; the feeling is what makes it yours. A page invites performance. A line invites truth.
2. Record voices on purpose, badly. Once a month, ask someone to tell you a story you have already heard, and record it. Do not tidy the audio. The value is entirely in the timbre, the pauses and the specific way they say your name at the start.
3. Photograph the room, not the event. For every photograph you take of an occasion, take one of the space in its normal state. Your desk. The hallway. What is on the counter. This is the category you will want in twenty years and the one nobody collects.
4. Write things by hand, occasionally, at cost. A physical object carrying your handwriting is not sentimental clutter; it is the only form of record that is currently unforgeable and that decays at the same rate you do.
5. Curate once a year, ruthlessly. One evening annually: select twenty items from that year and place them somewhere separate. Twenty. Not two hundred. The discomfort of cutting is the entire mechanism — you are performing the selection your memory used to perform for you.
6. Keep a small list of what you refuse to document. Some things are damaged by being captured; the attention shifts from having the moment to securing it. Decide in advance which parts of your life are off the record, and honour it. A record with nothing excluded is a surveillance log.

On deliberate forgetting

The hardest part of this method is the part that involves loss, and I want to make the case for it directly rather than gesturing at it.

Forgetting used to do several jobs. It let wounds close, because the detail faded before the meaning did. It let people change, because the earlier version of them was not perpetually available for cross-examination. It made returning to a memory an event — you found something you had lost, and the finding was most of the feeling. Perfect retention removes all three. Nothing closes, nothing is outgrown, and nothing is ever found again because nothing was ever misplaced.

You cannot restore biological forgetting and should not try. But you can approximate its function architecturally: keep the curated twenty accessible and everything else genuinely inconvenient. Not deleted — buried. An unindexed folder, an external drive in a cupboard, a box. The friction is the feature. It restores the possibility of rediscovery, which is the emotional mechanism that total recall quietly abolished.

And accept a harder version too: some things should be allowed to go entirely. Not everything that happened to you needs to be retrievable for the rest of your life. The right to be forgotten includes, most importantly, by yourself.

Why bother, if it is all kept anyway

A fair question. If platforms retain everything at higher fidelity than any hand-kept record, why perform this work by hand?

Because the two objects are different in kind. An automatic archive is a record of what occurred, assembled by something with no stake in the outcome. A kept record is a record of what you decided mattered, assembled at cost, by someone who will not be here forever. The first is data about a life. The second is a life’s judgement about itself, and it is the only version anyone — including you at sixty — will find worth opening.

One further reason, less practical. Everything in the archive that reads as unbearable is unbearable for the same reason: it was kept without being chosen. A shopping list survives because storage is indiscriminate. It is devastating because nobody meant it to last. You can pre-empt that. You can be the one who chose.

If you want the reasoning underneath the method, what abundance does to meaning makes the argument in full. The attention you cannot automate covers the resource all of this spends. Or open the archive and see how the machine describes objects like yours.