Whose Archive, and What Is Missing

An archive is not a neutral residue of the past. It is what an institution chose to keep, catalogued under the categories that institution used, and every stage of that involved a decision by somebody. What got written down. What was retained rather than discarded. What was catalogued, and under what heading, and therefore what a later researcher can find at all.

The consequence is a record systematically richer about the literate, the propertied and the prosecuted, because those are the people institutions generate paper about. Everybody else appears when they come to official attention, which usually means a court, a workhouse or a parish register: moments of trouble rather than moments of life.

Historians call the resulting absences archival silences, and the word is precise. A silence is a systematic absence, not a random gap, and it is not a problem that more searching will fix, because the documents were never created. Recognising this changes what can honestly be claimed: the difference between saying we do not know and saying it did not happen is the whole of the discipline's integrity on such questions.

The technique that gets around it is reading against the grain, using a document for something other than its purpose. A slave ledger was made to record property and can be read for names, ages, family groupings and the fact of resistance. A trial transcript was made to convict and preserves how ordinary people actually spoke. A probate inventory was made for tax and shows what was in a house, room by room. The method works precisely because the document's author did not care about the thing you are after and therefore had no reason to shape it.

The migrated archives are the clearest modern demonstration that a collection is a set of decisions rather than a survival. Britain acknowledged in 2011 that a substantial body of colonial administrative records had been removed at independence rather than handed over, with others destroyed outright. The consequence is that successor states cannot fully reconstruct their own administrative history, and historians of those countries work with a record that was deliberately shaped by its previous custodians.

Digital records have not solved any of this, whatever the volume suggests. They have changed the shape of the problem and arguably made it worse: retention is decided by whoever owns the platform, under terms that can change without notice; formats become unreadable within decades; and what survives skews heavily towards whoever was online and using the dominant service. A historian in 2200 will face exactly the same question (who decided what survived, and for what purpose) with a considerably less accountable answer than a nineteenth-century registry provides.

Which closes a loop this shelf opened with its first book. Method taught that a source has a purpose and must be read for it. The archive is that same insight one level up: the collection has a purpose too, and the person who decided what went into it is usually not named on the folder.