Why Certain Languages Keep Their Grip on Diplomacy Long After the Empires Behind Them Collapse
You're in a reading room somewhere, turning through a formal protest note or the founding charter of some multilateral body, and the phrasing stops you. It feels almost architectural: precise, layered, built to outlast whoever signed it. Often, that phrasing is French. Not because France is the dominant world power, which it hasn't been, by most measures, since the mid-twentieth century. The language persists anyway, embedded in institutions, legal traditions, and professional expectations that were poured like concrete decades or centuries ago and have simply never been repoured.
This is the central puzzle of diplomatic language: prestige outlasts power. Dramatically, sometimes by centuries. Understanding why requires looking not at geopolitics but at the mechanics of institutional inertia, the economics of coordination, and the surprisingly human tendency to mistake the vessel for the water.
The Lock-In That Treaties Built
Consider what happens when two states negotiate a treaty. Both sides need identical legal meaning from identical words. Ambiguity is not a stylistic flaw here. It is a diplomatic failure that can, in the worst cases, justify a war. So the parties select a language with an established tradition of legal precision, one whose diplomatic vocabulary has been tested across hundreds of prior agreements and whose terms carry shared interpretations that courts and foreign ministries already recognize.
French became that language during the seventeenth century, formalized by the Peace of Westphalia in 1648 and cemented through the Treaty of Utrecht in 1713, where it displaced Latin as the working tongue of European statecraft. Once enough treaties were written in French, all subsequent diplomats had to learn French to read, interpret, and contest those treaties. Every new agreement that cited or built on older ones reinforced the network. The cost of switching grew with each decade.
Economists call this a coordination equilibrium. It is the same reason QWERTY keyboards persist even though alternative layouts demonstrably reduce finger travel: the moment enough people are trained on a system, the collective cost of retraining exceeds the individual benefit of a better alternative. Replace "keyboard layout" with "a language in which five hundred years of binding legal precedent is written" and the stickiness becomes almost gravitational. That is not a metaphor. It is a balance sheet problem, and the liability column keeps growing.
The Prestige Loop and How It Feeds Itself
There is a second mechanism layered on top of the coordination problem, and it is more psychological. Diplomatic language accumulates what linguists sometimes call prestige by association: the belief that a language is inherently suited to precision, elegance, or authority. A belief that is self-reinforcing once widely held.
French diplomats and scholars spent roughly two centuries actively promoting this association. The Académie française, founded in 1635, was partly a geopolitical project: standardize the language, and you standardize the tool through which France projected influence. It worked. By the eighteenth century, Frederick the Great of Prussia wrote his personal correspondence in French and regarded his native German as a coarse instrument. Courts from St. Petersburg to Vienna conducted sophisticated discourse in a language they had learned as adults specifically because fluency signaled membership in a civilized international order.
The signal and the substance became confused. This is where most accounts go soft, and they shouldn't. French was not precise because it was French. It was precise because generations of legal scholars had developed a precise vocabulary within it. But once enough people believed the language itself carried that quality, they treated it as a precondition of seriousness. Ambassadors who wrote in anything else risked appearing provincial.
This loop is not unique to French. Latin held it for a millennium before French. English holds a version of it now, not because English is structurally superior for legal drafting (it demonstrably isn't: its synonyms are vast, its verb-aspect system ambiguous in ways that cause real interpretive headaches in contract law) but because the institutions built after 1945 were predominantly Anglophone in leadership, and the technical literature of economics, science, and international law has since been produced overwhelmingly in English.
What People Get Wrong About Language and Power
The common assumption is that diplomatic language tracks power in something close to real time: the dominant state imposes its tongue, and when it weakens, the tongue retreats. History is considerably messier than that.
Latin is the bluntest counterexample. The Western Roman Empire collapsed in 476 CE. Latin remained the language of European law, theology, science, and formal diplomacy for roughly a thousand years afterward. It was nobody's native tongue and hadn't been for centuries, which, paradoxically, made it more useful: no country owned it. Two medieval Italian city-states negotiating a trade agreement could both write in Latin without either conceding the other's cultural dominance.
The neutrality argument is real and underappreciated. French survived partly because, by the nineteenth century, it was a second language for almost everyone using it diplomatically, which meant no single power could claim it as a home-field advantage. When the League of Nations was founded after the First World War, French and English were both designated working languages precisely because neither could be granted sole status without appearing to hand a permanent edge to France or Britain.
English now sits in a structurally different position: it is the native language of two historically powerful states and a working language absorbed by most of the world's professional class. That breadth is its strength. It is also a latent vulnerability. As more non-native speakers use English in international settings, the language drifts from its Anglo-American anchors. Scholars of international law note that the English of UN documents is already a distinct register from the English of London or Washington, shaped as much by translation conventions and legal formulas from other traditions as by any Anglophone original. The language is, slowly, becoming nobody's in particular. Which might be exactly what keeps it.
A Tale of Two Diplomats
Imagine two foreign ministry trainees who enter service around the same time. Call them Marta and Jonas. Marta joins a mid-sized European foreign ministry in the 1970s and invests heavily in French: the Quai d'Orsay's formulations, the vocabulary of the Vienna Convention, the specific phrasing conventions of UN Security Council resolutions. Jonas, at a different ministry, bets on English and only English, judging French a fading asset.
By the 1990s, Jonas looks prescient. English is everywhere. But Marta is the one who gets assigned to the International Court of Justice proceedings, because the older briefs they are contesting are in French, and the court's founding documents carry French-language provisions that have no settled English translation. Her investment paid a return Jonas couldn't collect because he'd never made it.
The point isn't that French was the better bet. It's that the value of a diplomatic language isn't just current-use frequency. It's access to the archive. Whoever holds the ability to read, interpret, and contest the foundational documents possesses a form of legal advantage that doesn't appear in any power index, and doesn't depreciate with a change of government.
The Archive Is the Argument
So ask yourself: when does a language actually lose its authority? Not when the empire falls. Not even when the last native speaker of a diplomatic generation retires. A language loses authority only when the documents it produced stop being cited, and treaties, it turns out, have an extraordinarily long half-life.
The Congress of Vienna's territorial arrangements were still being invoked in European border disputes well into the twentieth century. The Ottoman capitulations, written in a mix of Ottoman Turkish, French, and Latin, created legal ambiguities that successor states argued over for decades after the empire ceased to exist. Every time a state wants to contest, invoke, or build on a prior agreement, it must engage the language in which that agreement was written. There is no clean translation that carries full legal weight. Translation is interpretation, and interpretation is political.
This creates a professional class with a direct interest in maintaining the prestige of specific languages: international lawyers, treaty archivists, diplomatic historians, specialist translators whose careers depend on the continued relevance of their particular linguistic expertise. They are not a conspiracy. They don't need to be. Their incentives align naturally with the preservation of existing linguistic hierarchies, and they occupy the institutions where those hierarchies are reproduced.
What Survives When the Power Doesn't
Find the right archive and you will find French-language dispatches from states that no longer exist, in handwriting that describes a world rearranged many times over since. The language is still there. Still technically operative in some clause somewhere.
That persistence should prompt a useful skepticism the next time someone declares that English's global dominance is simply natural, inevitable, a reflection of some inherent communicative fitness. It isn't, and the historical record is unambiguous on this point. It reflects a particular historical moment when the institutions that matter most were built by Anglophone powers with Anglophone assumptions, and those institutions have since compounded interest on that founding advantage.
Language, in the end, is infrastructure, and nobody replaces infrastructure because a better design came along. They replace it when the cost of keeping the old system finally, visibly, exceeds the cost of the transition. For any language embedded deeply enough in the architecture of international law, that threshold is much further away than it looks, and the tab for reaching it keeps running.