Who Keeps the Digital Past? — Part 7

When people first bring us a family archive, they often ask which files are worth saving. That sounds like a technical question, but it is usually a question about scale. A folder containing every near-identical holiday photograph can make the genuinely revealing images harder to find. I encourage families to retain a smaller number of representative versions, together with enough information to explain where they came from and why someone kept them. A voice message without its date, sender or surrounding circumstances may remain playable for decades while losing most of its meaning. This is why I resist the idea that preservation equals accumulation. An archive is not improved merely by becoming larger; it is improved when a future reader can understand the relationship between an item and the life in which it mattered. The best arrangement therefore preserves enough structure for later interpretation without drowning it in duplicates, stray exports and files no one can plausibly identify.

The most difficult conversations do not concern passwords. They concern responsibility. After a death, relatives may discover that one person has access to thousands of photographs, private messages and online accounts but no shared understanding of what should happen next. A legal document can nominate an executor, which is useful, but it cannot decide whether a sibling should read old correspondence or whether a child ought to inherit a public profile. I advise families to name a memory steward while everyone is still able to discuss preferences without pressure. The title is deliberately informal. Its purpose is not to create another bureaucratic role, but to make clear that someone will need judgement, restraint and the confidence to ask other people what they are comfortable with. Those conversations can be awkward because they acknowledge competing loyalties, yet avoiding them merely leaves the conflict to grief and hurried decisions.

Our local history project receives images from people who can say very little about them. A photograph may arrive labelled only “Grandad’s shop”, although no one can agree which street it stood on or who took the picture. Early on, we treated such uncertainty as an obstacle to be eliminated. Now we display it. Visitors can see competing captions, dates that remain provisional and notes explaining why one interpretation is stronger than another. This has made the archive less tidy but more honest. Ambiguity is not always a defect in a historical record; sometimes it is evidence of how stories travel through families and change with repetition. The important thing is not to turn a reasonable guess into a fact simply because a database prefers one clear field. A transparent uncertainty note also gives later contributors somewhere to add evidence, rather than inviting them to overwrite a previous claim with confidence.

People tend to imagine a digital archive as something created by one person for their own future use. In reality, a single inbox may contain the lives of dozens of others: colleagues, former partners, children and people who never expected their messages to be kept. Consent therefore cannot be treated as a box ticked at the moment a file is created. It may need to be reconsidered when material is organised, transferred or made accessible. In some cases, delayed access is more respectful than deletion. A collection can be sealed for ten years, for example, allowing distance without pretending that the material never existed. Users dislike such restrictions at first because they want a simple yes-or-no decision. They usually understand their value when they imagine being the person described in the records. The point is to create room for relationships to change over time, not to treat privacy as a one-off technical preference.

As a novelist, I keep fewer drafts than people assume. I save the versions that show a genuine turn in the work: a scene moved to another voice, a structure abandoned, a sentence that suddenly made the book possible. The rest I delete. This is not because I think early attempts are embarrassing; it is because a complete trail can become a substitute for deciding what the work was trying to do. I do keep a notebook of false starts, but it is selective and deliberately rough. The point is to preserve evidence of uncertainty without pretending that every hesitation deserves equal status. Readers sometimes ask for access to all my files, as though quantity would reveal the secret of writing. I suspect it would reveal mostly the ordinary mess that any process produces. Selective preservation can make the surviving material more legible because it records decisions about value as well as the mess of production.