Multiple-choice reading

In the basement of a municipal archive, where the air tastes faintly of vinegar and dust, I once watched an archivist lift a strip of early colour film as though it were a relic from a drowned civilisation. The images—street parades, women in cloche hats, a child craning towards a balloon—were still legible, but their hues had migrated, the reds bleeding into a spectral orange. “It’s not decaying,” she said, with a kind of professional tenderness. “It’s transforming.” Her remark has stayed with me because it contains, in miniature, the moral ambiguity of preservation. To keep something is never simply to arrest time; it is to choose a manner of change, and thereby to impose a story on what survives.

We have grown accustomed to thinking of archives as neutral storehouses: a democracy’s memory bank, an insurance policy against forgetting. Yet the archive is also an instrument of power—less in the melodramatic sense of secret dossiers than in the quieter fact that the past must be sorted, described, and made retrievable. The catalogue is a form of rhetoric. It decides which lives become “correspondence” and which become “ephemera”; which events are “disturbances” and which are “uprisings”. Even the most conscientious archivist cannot avoid the violence of classification, because categories are never merely technical. They are theories about what matters.

The digital age was supposed to liberate us from this predicament. Storage became cheap, duplication effortless, access seemingly universal. Why fret about selection when you can keep everything? The promise, in its most utopian form, was that history would no longer be written by the victors or curated by institutions, but by the vast, impartial accretion of data. And yet the paradox of abundance is that it reintroduces scarcity by other means. When everything is saved, attention becomes the limiting resource, and the gatekeepers simply change their names: search algorithms, recommendation systems, proprietary formats. The question is no longer what is preserved, but what is findable, and under what terms.

Consider the way a photograph circulates now: uploaded, compressed, stripped of context, re-captioned, embedded in an argument it was never meant to make. The philosopher Walter Benjamin wrote of the “aura” of the original artwork, that fragile emanation of time and place, which mechanical reproduction erodes. Digital reproduction goes further, because it doesn’t merely multiply copies; it dissolves the notion of a stable original. The “same” image exists in a dozen resolutions, with metadata altered or deleted, with colours adjusted by automatic filters whose aesthetic assumptions are coded into software. Authenticity becomes probabilistic rather than absolute, a matter of forensic inference. In this sense, our new archives resemble ecosystems more than libraries: they are dynamic, adaptive, and full of invasive species.

This ecological analogy is not merely decorative. The internet’s memory is often described as ethereal, but it is built on a very material infrastructure—server farms that drink electricity, undersea cables that thread the ocean floor, rare minerals extracted at social and environmental cost. The cloud has a carbon footprint and, like any industrial system, it externalises its waste. One of the stranger ironies of contemporary culture is that we produce an ever-thickening layer of digital traces—messages, photos, location pings—while simultaneously accelerating the physical processes that threaten the archives of the future. Floods do not respect reading rooms. Heat warps magnetic tape. Political instability turns museums into targets. The more we rely on preservation, the more we live in conditions hostile to it.

There is, too, a psychological dimension that our discourse about memory tends to neglect. We congratulate ourselves on our capacity to record, as though documentation were a moral achievement. But the human mind is not a hard drive; forgetting is not simply failure, but a function. Trauma researchers have long noted that memory is reconstructive: we do not retrieve the past intact so much as reassemble it, guided by present needs and narratives. Societies behave similarly. A nation that claims to “face its history” is often, on closer inspection, negotiating which version of that history can be tolerated without shattering its self-image. Archives are enlisted in this negotiation. They can be sites of reckoning, but they can also be mausoleums—places where uncomfortable truths are entombed under the weight of procedure.

None of this is to indulge in cynicism about institutions or technology. The archive, for all its compromises, remains one of the few mechanisms by which a society can argue with itself across time. The existence of a record makes denial harder, even if it never makes it impossible. When activists demand the release of police footage or the declassification of state documents, they are appealing to an idea older than any database: that power should leave traces it cannot wholly control. Similarly, when artists work with found photographs or bureaucratic files, they are not merely scavenging; they are reanimating the past, exposing the seams of official narratives. The most interesting art of the last decade has often been archival in this sense—less obsessed with originality than with montage, with the ethics of quotation, with the question of who gets to speak from the record.

Yet we should resist the comforting belief that more memory automatically yields more justice. Sometimes the proliferation of evidence produces not clarity but fatigue, a numbing equivalence in which atrocities compete for attention like streaming shows. Sometimes visibility becomes a trap, especially for those already surveilled. The same technologies that allow marginalised communities to document abuse can also be used to catalogue them, to predict them, to police them. The archive is not just a mirror; it is a map, and maps have always been tools of governance.

What, then, does responsible preservation look like in a world where both forgetting and remembering can be forms of harm? Perhaps it begins with humility: an acknowledgement that the past is not a possession, and that our custodianship is provisional. It requires, paradoxically, a willingness to let some things fade—not through neglect, but through an ethics of restraint. Not every life needs to be rendered into data; not every grief needs to be searchable. At the same time, it demands vigilance about the infrastructures that silently shape what endures: the formats that become obsolete, the companies that collapse, the laws that decide who may access what. The archivist in the basement was right: transformation is inevitable. The question is whether we will pretend it isn’t happening, or whether we will learn to read the colour-shifts, the gaps, the misfilings, as part of the story we are always, unavoidably, telling.