Historians of the post-1945 world increasingly find that the decisive evidence for their questions sits in the records of intergovernmental organizations rather than national archives. Decolonization, peacekeeping, global health, development finance and refugee protection were all negotiated, administered and documented inside institutions such as the United Nations, the World Bank, the World Health Organization and the European Commission. For many questions, these archives are the only place where several national positions appear in a single file.

Yet these collections are not simply national archives on a larger scale. They belong to bodies that answer to many sovereigns at once. They enjoy legal immunities that place their records beyond the reach of national freedom-of-information laws, and they produce documentation in several languages at the same time. Their paper trails reflect diplomatic caution, bureaucratic routine and member-state sensitivities. All of this shapes what can be known long before a historian requests a box or queries a database.

Susan Pedersen's work on the League of Nations mandates system showed how much these archives can reveal about the workings of global order. The methodological challenge is to understand how they condition the evidence itself. Record-keeping cultures, declassification regimes and multilingual production are not background details. They determine which voices survive, when those voices become visible, and in what words. Treating an organization's archive as a neutral window onto world politics is the first mistake. Rigorous analysis begins with reading it as an artifact of institutional design.

Every international organization develops its own habits of documentation, and those habits are rarely neutral. The League of Nations Secretariat was shaped by Eric Drummond's ideal of an impartial international civil service. It produced dense registry files in which minutes, correspondence and internal comments were filed together by subject. The resulting archive in Geneva is now digitized through the LONTAD project, and it allows historians to follow an issue across sections and years with unusual continuity.

The United Nations inherited much of that tradition but fragmented it. Departments, specialized agencies, funds and field missions each built their own systems, with varying commitment to transferring records centrally. A Secretary-General's bulletin has set formal record-keeping obligations since 2007, yet compliance has historically been uneven. Peacekeeping missions in particular have struggled to capture records amid staff rotation and the closure of operations. What survives often depends on whether someone deliberately packed a mission's files and shipped them home.

Equally consequential is the distinction between official records and working papers. Intergovernmental bodies generate vast public documentation: resolutions, reports and meeting records, carefully negotiated and widely distributed. Behind it sits a thinner, more revealing layer of drafts, notes to file and internal memoranda. The public layer records consensus; the internal layer records how consensus was manufactured. Historians who rely mainly on the former risk reproducing the institution's preferred narrative of itself.

Even the public layer is shaped by choices about genre. United Nations bodies distinguish between verbatim records, which reproduce statements in full, and summary records, which condense debate into reported speech prepared by Secretariat précis-writers. Budgetary pressures have periodically reduced the number of meetings covered by either form. How a meeting is recorded is therefore an archival decision made at the moment of creation. That decision determines whether historians later encounter a delegate's argument or a staff member's paraphrase of it.

The digital turn has intensified these dynamics. Email, shared drives and messaging platforms have displaced the registry file. Many organizations adopted electronic records systems only after years of informal practice. The likely result is a gap in the documentary record of the late 1990s and 2000s, which is precisely the period contemporary historians are now beginning to study. Analysing how an organization recorded its work is no longer preliminary context. It is part of the historical argument itself.

Takeaway

An archive records an organization's theory of what mattered as much as what happened. Before asking what the files say, ask what the institution was built to remember.

International organizations are generally not bound by national freedom-of-information legislation. The 1946 Convention on the Privileges and Immunities of the United Nations, for instance, declares the organization's archives inviolable wherever they are located. Access is therefore governed by internal policy rather than an enforceable public right. Each organization writes its own transparency rules, and they vary strikingly. In practice, the same diplomatic episode may become visible in one institution's records decades before it emerges in another's.

The European Union follows a thirty-year rule and deposits the historical archives of its institutions with the Historical Archives of the European Union at the European University Institute in Florence. NATO likewise reviews records for public disclosure once they reach thirty years. The United Nations Secretariat reviews classified material for declassification after twenty years, though sensitive files can be withheld longer. The International Committee of the Red Cross applies a forty-year embargo to its general archives and gives longer protection to files concerning individuals.

Development finance institutions offer a different model. The World Bank's 2010 Access to Information Policy shifted from listing what could be disclosed to listing what could not. It also set timelines for declassifying certain restricted documents after five, ten or twenty years, and the International Monetary Fund maintains its own staggered schedule for Board documents. These policies reflect how each institution understands itself. Lenders answerable to shareholders, and increasingly to civil-society critics, have reasons to perform transparency differently from security alliances.

Formal rules tell only part of the story. Many organizations allow member states to object to the release of documents they originated. Some files remain closed simply because their review was never resourced, not because anyone decided they should stay closed. Declassification is labour-intensive, and archival units are rarely priorities in budgets negotiated by governments. The effective date of access is often set less by policy than by staffing.

This creates distinctive methodological hazards. Historians may build narratives around whichever organization opened first and overstate its centrality. Cross-archival work becomes essential: a NATO file may fill a gap in UN records, and a national foreign ministry may hold the cable an organization still restricts. Mapping declassification timelines should become a standard step in research design. It should also be documented in the scholarship, so readers understand why some actors speak loudly in the sources while others remain silent.

Takeaway

Access schedules write their own historiography. The organization that opens its archives first often becomes the protagonist of the first histories, whether or not it deserves the role.

The United Nations has six official languages: Arabic, Chinese, English, French, Russian and Spanish. Its Secretariat works primarily in English and French. The European Union recognizes twenty-four official languages, while the Commission conducts much internal business in English, French and German. The League of Nations operated in English and French. These facts are not incidental. They make the records of international organizations layered linguistic artifacts in which no single version is straightforwardly the original.

Consider a Security Council meeting. A delegate speaks in Russian, interpreters render the statement simultaneously into the other languages, and verbatim reporters then prepare records in each official language. The English record of a Russian intervention is therefore mediated by interpretation and editing. In principle, the authoritative text of that speech is the Russian. Yet historians working only in English routinely cite the translation as if it captured the speaker's exact wording, including nuances that may have been lost or added.

Language also shapes outcomes. Security Council Resolution 242 of 1967 remains the classic case. The English text calls for withdrawal from territories occupied, while the French reads des territoires occupés, which many read as 'from the territories'. Both texts are equally authentic. Negotiators sometimes exploit precisely such discrepancies to secure agreement. A divergence between language versions can therefore be evidence of diplomatic strategy rather than a translation error.

Below the level of official documents, hierarchies among working languages shape the internal record. Memoranda, draft reports and field correspondence are typically written in the dominant working language, which favours staff and delegations fluent in it. Perspectives from Arabic-, Chinese- or Spanish-speaking delegations may enter the internal files only after passing through colleagues' summaries. The archive thus encodes linguistic power, quietly amplifying some voices while filtering others through translation.

Digital methods both complicate and sharpen these issues. The UN Digital Library and the Official Document System make parallel language versions searchable, which invites comparative corpus analysis of terminology across languages and decades. But text-mining tools built mostly on English corpora tend to treat the English version as the dataset, reproducing the hierarchy in computational form. Multilingual research teams and the systematic alignment of parallel texts offer a corrective. They turn the multiplicity of versions from a nuisance into a source of evidence.

Takeaway

In a multilingual archive, the gaps between versions are not noise to be cleaned away. They are traces of negotiation, mediation and power.

International organizations have become central repositories for the history of the present, but their archives carry the imprint of their peculiar institutional position. Record-keeping cultures determine what is created and kept. Declassification regimes determine when it surfaces. Multilingual production determines in what words it survives.

For contemporary historians, the implication is that archival criticism must be built into research design. That means documenting the conditions of access, triangulating across institutional and national collections, and reading parallel language versions as evidence rather than duplicates. Digital tools can scale these practices, provided they do not silently privilege one language or one repository.

The larger stakes concern the gaps of the digital era. If the email years of global governance are thinly preserved, future historians will inherit a record shaped by technical decisions made with little thought for posterity. Engaging archivists and institutions now, while those decisions remain open, may be the most consequential methodological work historians can do.