The Unreliable Archive
Every archive is a product of its time. This is not merely a truism about physical decay or incomplete records — about papyrus that crumbles, ink that fades, or buildings that burn — though these material facts matter. It is, more fundamentally, a claim about power. Archives are not neutral repositories of what happened; they are curated selections, shaped by the interests, assumptions, and blind spots of the people who created them. The silences within an archive — the absences, the gaps, the categories of person or experience that left no trace — are as historically significant as the documents the archive contains, and often considerably harder to interpret.
Historians working on colonial history have had to confront this problem with particular urgency. The people most profoundly affected by empire — the colonised, the enslaved, the displaced — are largely absent from the official archives that empire produced. Colonial administrations generated vast quantities of paper, but that paper recorded the perspective of administrators, merchants, soldiers, and missionaries. The experiences of those whom empire governed, exploited, and transformed survive, if at all, in fragments that require painstaking recovery: oral traditions passed down across generations, marginal annotations in documents produced for other purposes, legal records of resistance and punishment, occasional letters that somehow found their way into collections not designed to preserve them. To write history from these fragments is a different and more precarious enterprise than writing from the abundance of the imperial archive, and it requires a different kind of methodological honesty about what can and cannot be known.
Digitalisation has transformed archival practice in ways that seemed, initially, to offer solutions to some of these problems. The promise of the digital archive is universality and accessibility: if everything can be scanned and made available online, then geographical barriers dissolve, rare materials become widely available, and the marginalised voices that were previously accessible only to those who could travel to specialist collections can reach a much wider audience. There is genuine substance to this promise. Digitisation has opened up important historical materials to researchers and communities who would previously have had no access to them.
But the fundamental tensions of archival selection have not been resolved; they have been relocated. Every digitisation project involves decisions about what to scan, in what order, and at what resolution — decisions that reflect institutional priorities, available funding, and, inevitably, assumptions about what is historically important. Materials deemed peripheral, obscure, or insufficiently significant may wait decades for digitisation, if they are digitised at all. The biases of the analogue archive do not disappear when the archive goes online; they tend to migrate quietly into the digital version, now potentially amplified by the appearance of comprehensiveness that digital access can create.
Digitalisation also introduces new forms of instability that have no analogue in the physical world. A stone inscription is, barring catastrophe, durable across millennia. A handwritten manuscript on vellum may survive for centuries. A digital file, by contrast, exists only as long as the hardware capable of reading it remains functional and the software capable of interpreting its format continues to be available. Formats become obsolete — entire generations of digital content have already been rendered inaccessible by the disappearance of the applications that created them. Servers fail, institutions lose funding, links rot, and the elaborate technical infrastructure on which digital preservation depends requires constant maintenance and renewal. Archivists have coined the term 'digital dark ages' to describe the possibility that future historians may find themselves with less recoverable evidence about the early twenty-first century than about earlier periods for which physical records survive, precisely because so much of contemporary life was recorded in perishable digital form.
The question of access adds a further layer of complexity. Even where archives are comprehensive and well-preserved, they have historically been available only to those with the right credentials, the right institutional affiliations, the right languages, and the right social position. The democratisation of archival access through digitisation is genuinely significant, but it does not straightforwardly resolve these inequities; it reconfigures them, creating new distinctions between those with reliable high-speed internet connections and those without, between those fluent in the languages of dominant archive-holding institutions and those who are not.
Perhaps the most intellectually honest response to all of this is an embrace of what the philosopher of history Paul Ricoeur called the 'admitted debt' — a clear-eyed acknowledgment that historical knowledge is always partial, always perspectival, always shaped by what someone, at some point, decided was worth keeping. This is not relativism; some accounts of the past are better supported by evidence than others, and the distinction matters. But it is a form of epistemological humility that recognises the archive not as a window onto the past but as a particular kind of filtered light — illuminating some things, casting others into deeper shadow, and always shaped by the position from which it falls.