Archives Do Not Preserve History. They Shape It.
The archive is not a neutral repository. Every decision made in its construction — what to collect, what to discard, how to classify, who is permitted access — is an act of historical production. The archive does not wait passively for historians to come and read what is there. It determines, in advance, what histories can be written.
This is not a conspiracy. It is a structural condition. Colonial archives are the clearest case. The records kept by European colonial administrations documented what the administrations cared about — taxation, land registration, military operations, trade flows, the surveillance of populations considered threatening. The daily life of colonized people, their internal political structures, their economic arrangements, their cultural practices, appear in these records refracted through the interests of an administration that was, at minimum, indifferent to them and frequently hostile. Historians working from these archives are working with systematically distorted material. Acknowledging this is not a political gesture. It is a methodological requirement.
The problem extends well beyond the colonial case. State archives everywhere reflect the priorities of states. Corporate archives, when they exist and are accessible, reflect the priorities of corporations. Religious archives have their own systematic distortions. Even personal archives — the papers of writers, politicians, intellectuals — have been curated by their subjects or their estates with posterity in mind, which is another form of distortion. The letters that were saved are the letters someone decided were worth saving. The diaries that survive are the ones that were not burned.
Digital preservation has introduced a new version of the problem. The volume of material now produced and potentially preservable is so large that decisions about what to archive are effectively decisions about what will exist for future historians. The platforms that currently hold the largest archives of human communication and expression are private companies with no archival mandate, uncertain longevity, and every incentive to delete rather than preserve when deletion serves their operational interests. The Library of Congress famously announced a project to archive all of Twitter in 2010 and quietly discontinued it in 2017, concluding that the volume had become unmanageable. The public record of a decade of political, cultural, and social life on the platform is incomplete and inaccessible.
None of this means that archival history is impossible or that all accounts are equally distorted. It means that the archive is always a historical actor, and that reading its silences is as important as reading its records. The question a historian must ask is not only what the archive contains but who built it, for what purpose, and what it was designed not to remember.