The Language of Power

Global health speaks many languages. It still thinks in one.

Artwork by Canva

There is a quiet ritual at the heart of global health that says more about the system than any formal policy ever could. Headphones are handed out, interpretation channels flicker on, and for a moment the room sounds like the world it claims to represent. Arabic, Chinese, French, Russian, Spanish, English all coexist, layered into a carefully orchestrated performance of inclusion. It works, at least on the surface. But when the session ends and the real work begins – drafting, negotiating, refining language – almost all of it shifts to English. This is not an accident or a failure of implementation. It is how the system is structured.

Institutions like the United Nations and the World Health Organization operate with six official languages, and that matters in ways that are often underestimated. It signals balance, respect, a commitment to a plural world. Yet it coexists with a different layer of reality in which ideas are produced, shaped, and stabilized in English before they are ever translated into something else, a dynamic that has been examined in recent analyses of language and power in global health. By the time a document circulates in multiple languages, its logic is already set.

Multilingualism governs what is visible, but English governs what is made.

The system does not present this as a problem. In fact, it has developed ways of accommodating the tension. A francophone researcher can present in French, a delegate can intervene in Arabic, a minister can deliver a statement in Spanish. Access is not the issue. The issue is what happens before and after those moments. Drafts rarely begin in six languages. They begin in one. Edits are negotiated in one. Disagreements are worked through in one. By the time interpretation catches up, the space for shaping the argument has already narrowed.

Global health has tried to address this imbalance through other means. Representation is the most visible. The effort to ensure that staff reflect the diversity of member states by using country quotas has changed who is present in these spaces. The room is broader now, more reflective of the world it claims to serve. That shift matters. But it does not fully change how influence works.

Once inside, everyone operates within the same linguistic frame. A policy advisor from Dakar, a researcher from Phnom Penh, a programme lead from Lima may all sit at the same table, yet the conversation moves in English, at a pace and with a precision that rewards those who can think and respond in that language without friction. No one is excluded, but not everyone can intervene in the same way or at the same moment. Over time, this produces a pattern. Those who can draft quickly and reframe arguments fluently in English move closer to agenda-setting roles, while others remain central to implementation and coordination but less present in the moments where ideas are defined.

Representation broadens the room. Language still shapes the hierarchy within it.

What happens in those moments does not stay contained within global institutions. It travels outward through the frameworks, guidelines, and categories that structure national policy debates. Global health ideas are rarely built in multiple languages at once. They are typically assembled in English and then translated, which means that by the time they reach countries, they arrive as finished objects rather than open questions.

Concepts come with names, definitions, and implicit assumptions already in place. Translation carries these ideas across borders, but it also stabilizes them, making it harder to reopen debates at the national level. And not everything survives that journey intact. English terms like resilience, integration, or empowerment travel widely but land differently in other languages, acquiring local meanings while shedding others, sometimes becoming placeholders rather than arguments. What appears as alignment from Geneva or New York can conceal divergence on the ground.

So two conversations begin to run in parallel. One that speaks upward, aligned with global frameworks, and another that governs locally, closer to lived realities but less visible internationally. Between these layers sits a relatively small group of bi- or multilingual global health professionals who can move between these realities with ease, translating not just language but legitimacy. This group constitutes a transnational professional elite: mobile, polyglot, highly educated, and able to operate across institutional and cultural boundaries. They become indispensable. And in doing so, they also become a bottleneck through which ideas must pass. This is not an external critique. It describes a position many of us occupy.

These linguistic tensions are not unique to global health, which is what makes them harder to dismiss. The European Union offers a highly ambitious version of multilingual governance, with twenty-four official languages and full legal equivalence across all of them. Every regulation carries equal authority in every language, and the system invests heavily in making that possible. The scale is considerable: the European Parliament alone generates 552 possible language combinations, supported by more than a thousand interpreters on any given working day and nearly two thousand translators at the Commission alone. Language services account for around one percent of the EU budget. Multilingualism here is not a gesture but a legal and political commitment tied to legitimacy, sovereignty, equal representation and democratic accountability.

And yet, even within this more robust framework, a similar dynamic persists. Internal working and drafting cluster around a narrower set of languages. English dominates. French retains influence in legal and diplomatic contexts. German, despite being the most widely spoken native language in the Union, plays a more limited operational role. The rest of the linguistic system is layered on through translation. Brexit did not alter this equilibrium. The United Kingdom left the European Union, but English remained the primary working language, a reflection of how deeply embedded it has become in the functioning of the system.

The European experience matters because it removes the easy explanation. Even where multilingualism is taken seriously, institutionalized, and funded, it does not reorganize how ideas are produced.

The language in which ideas are made is not the same as the languages in which they are distributed.

The exception that tests this most directly is the Pan American Health Organization, where the balance shifts but the structure does not. PAHO’s four official languages - Spanish, English, Portuguese, and French - reflect the region it serves, and everyday interaction genuinely moves across them. Much of the written work settles into English and Spanish, shaped by institutional practice and the demands of operating across countries. The centre of gravity is more multilingual than in Geneva. Yet the distinction between languages of access and languages of production persists: more languages in the room does not change the architecture, it only changes which languages anchor it. This is not an alternative model. It is the same model with a different centre.

The issue is not that global health lacks multilingual policies or that it has failed to implement them. It is that these policies operate at a different level from the processes that shape decisions. Multilingualism ensures access and signals fairness, and those are not trivial achievements. But they do not fully address how ideas are generated, contested, and fixed.

English structures the workflow. It shapes the speed of interaction, the sequencing of arguments, and the ability to define problems early in the process. The system does not need to exclude anyone explicitly. It only needs to reward certain forms of expression consistently enough that they become the default pathway to influence.

Global health has changed who is present more than it has changed how thinking is structured. It has diversified participation and made that diversity visible, while leaving largely intact the linguistic framework through which ideas are formed.

This essay was written in English. So is almost everything that circulates in the world it describes. That is not irony. It is the point.