Esperanto, Interlingua, And The Dream Of A Shared Human Language

Białystok in the 1860s was a four-language city. Yiddish in the Jewish quarter. Polish among Poles. Russian because the Russian Empire ran the place. German for trade. Four communities pressed into the same streets, each one suspicious of the others, each one assuming that if something went wrong it was probably the fault of whoever spoke the other language.

Ludwik Lejzer Zamenhof watched this as a child and drew a very specific conclusion. Most kids would say the problem was prejudice. Zamenhof said the problem was architecture. The prejudice was real, but it was downstream of something structural: we literally cannot understand each other. A Pole and a Russian living on the same street, going to separate schools, reading separate newspapers, assuming separate things about each other — of course they hated each other. They'd never had a real conversation.

So Zamenhof, still a teenager, started building. He finished the first draft in 1878. His father, a censor for the Russian government, burned the manuscript — he was afraid it would get his son labeled a dangerous internationalist, which, fair. Zamenhof rebuilt it. In 1887 he published Unua Libro — "First Book" — at his own expense, using his wife's dowry. He put on the cover: Doktoro Esperanto. The hopeful one.

This is the part most tellings of the story skip. Esperanto was not an academic linguistics project. It was a Jewish kid's answer to pogroms. He saw what separation does to people, and he built a tool to undo it.

Esperanto's design is elegant in a way that most natural languages aren't, because natural languages weren't designed — they accreted. Some of the rules:

- Sixteen grammar rules, no exceptions. Every noun ends in -o. Every adjective in -a. Every adverb in -e. Every infinitive verb in -i. Plurals add -j. Accusative adds -n. That's it. There's no "except on Tuesdays, when Germanic verbs do something else." - Phonetic spelling. One letter, one sound. Always. Learn the alphabet, you can pronounce any word you see. - Agglutinative morphology. Want to turn "healthy" (sana) into "unhealthy"? Add the mal- prefix: malsana. Unhealthy person: malsanulo. Hospital: malsanulejo. You build vocabulary like Lego. - Vocabulary from multiple families. Roughly 75% Romance, 20% Germanic, 5% Slavic and other. So if you speak any European language, you find footholds. (This is also Esperanto's legitimate limitation — it's Eurocentric. More on that later.) - No native ownership. Nobody's country, nobody's empire. Zamenhof explicitly refused to let it be claimed.

Fluency estimates vary, but most studies put Esperanto at about one-tenth the learning time of French or German for a European speaker. There's a genuine phenomenon called the "Esperanto propaedeutic effect" — kids who learn Esperanto first then learn another foreign language faster than kids who went straight to the second language. Because Esperanto is transparent, it teaches you what language-learning is before you're also fighting irregular conjugations.

Short answer: empire, then media, then network effects.

Longer answer: languages don't spread because they're good. They spread because the people speaking them have power, money, or entertainment you want.

English colonized a quarter of the planet. When the British Empire receded, American economic and cultural power took over — Hollywood, rock and roll, Silicon Valley. By the time the internet arrived, English had a decades-long head start, and the internet's early infrastructure was built in English-speaking countries, by English-speaking engineers, encoded in ASCII which was designed for English. Every layer reinforced the one below it.

Esperanto had no army, no Hollywood, no tech industry. It had Zamenhof, a pamphlet, and hope.

There were moments where it almost broke through. In 1921, the League of Nations nearly adopted Esperanto as its working language. The French delegation killed it — France was still the language of diplomacy and wasn't giving that up. After World War II, UNESCO passed resolutions recognizing Esperanto's value (1954, 1985), but recognition isn't adoption. The Soviets liked Esperanto in the 1920s as a tool of international proletarian solidarity, then Stalin purged Esperantists in the 1930s as rootless cosmopolitans, and many were killed. Hitler also banned it — Zamenhof was Jewish, and the Nazis considered Esperanto a Jewish plot. Both totalitarianisms recognized what Esperanto threatened: a communication layer they didn't control.

So Esperanto's failure wasn't really about linguistics. It was about geopolitics. A neutral language is threatening to every non-neutral power. Everyone who could have made it universal had a reason not to.

Here's the part that deserves respect. Despite losing the geopolitical race a century ago, Esperanto is still alive. Not as a historical curiosity — as a living language.

- Speaker estimates: Conservative counts put fluent speakers at ~100,000 worldwide. Generous counts, including people with conversational ability, go up to 2 million. Duolingo's Esperanto course has been taken by over 1.4 million people. - Native speakers: Around 1,000 people grew up speaking Esperanto as a first language, usually because their parents met at an Esperanto event and kept using it at home. These "denaskuloj" ("from-birth ones") are the only thing like native speakers a constructed language has ever produced at scale. - Institutions: The Universala Esperanto-Asocio (UEA), founded 1908, still runs an annual World Congress that draws 1,000–3,000 attendees from 50+ countries. The Esperanto Wikipedia has over 340,000 articles, making it larger than the Wikipedias of many natural languages. - Culture: Original novels, poetry, music, podcasts, YouTube channels. Not translations — original work. William Auld was nominated for the Nobel Prize in Literature for his Esperanto poetry.

The community is small but dense and global. An Esperantist in Tokyo can couchsurf through the Pasporta Servo network to stay with Esperantists in Mongolia, Senegal, and Brazil, speaking the same language the whole trip. It's the closest thing we have to proof-of-concept for Zamenhof's dream: a working global communication layer owned by no state.

Esperanto wasn't the only attempt. Understanding the others tells you something about what tradeoffs are possible.

Volapük (1879). Johann Martin Schleyer, a German priest, got there first. At its peak around 1889, Volapük had an estimated 200,000 speakers and 283 clubs. It was the first constructed language to become a social movement. Then Schleyer refused to allow reforms, fought with his own speakers, and the whole thing collapsed. By 1900, most Volapükists had defected to Esperanto. Lesson: if you build a shared thing and then hoard control of it, you lose your community.

Ido (1907). A reformed Esperanto, designed to fix what some linguists saw as Esperanto's flaws — remove the accented letters, regularize some word-formation, make it more "scientific." Split the community, drained maybe 20% of Esperantists, then plateaued. Today Ido has a few hundred speakers. Lesson: "better" doesn't beat "established" once a community has formed around something.

Interlingua (1951). Built by the International Auxiliary Language Association, largely by linguist Alexander Gode. Unlike Esperanto, Interlingua didn't try to be easy to learn from scratch. It tried to be readable on sight by anyone who knows a Romance language. Its vocabulary was extracted by a systematic process from Latin roots that appear across English, French, Italian, Spanish, and Portuguese. A Spanish speaker can read Interlingua with maybe 90% comprehension the first time. Strengths: high passive intelligibility for Romance speakers. Weaknesses: harder for non-Romance speakers; never built a mass community.

Lojban (1987). The nerd-end of the conlang spectrum. Based on predicate logic, syntactically unambiguous — every sentence parses exactly one way — designed originally to test the Sapir-Whorf hypothesis (does language shape thought?). Beautiful as an intellectual artifact. Has maybe a few hundred speakers. Lesson: if you optimize for theoretical elegance over human usability, you get a small, devoted community and no mass adoption.

Toki Pona (2001). The opposite extreme. 137 words, total. Designed to be learnable in days. Minimalist philosophy: what can you say with very little? Has a thriving online community. Not really an auxiliary language — more of an artistic and philosophical experiment — but worth naming because it shows the design space is still being explored.

The common pattern: each of these languages solves one problem beautifully and another badly. Esperanto balanced the tradeoffs well enough to outlast all of them. Nobody's solved them all.

Imagine the counterfactual. In 1921, the League of Nations adopts Esperanto. Every signatory country commits to teaching it as a second language in primary schools. By 1950, two generations of children worldwide have grown up bilingual — their native tongue, plus Esperanto. What changes?

- Diplomacy. UN negotiations don't run on English's home-field advantage. A Brazilian diplomat and a Chinese diplomat meet on equal linguistic ground. Neither is translating their own thoughts into the other's colonial language. - Science. Scientific papers don't have to be published in English to be read. A Japanese researcher publishes in Esperanto; a Nigerian researcher reads it directly. The Anglophone bottleneck in global science breaks. - Culture. A Senegalese novelist can reach a global audience without having to write in French. Cross-cultural exchange stops being routed through a handful of colonial languages. - Power. Nations that spent a century dominant in English-language discourse lose their linguistic home-field advantage. Anglophone countries stop being the default center of global conversation.

That last one is why it didn't happen. A neutral lingua franca is, in political effect, a redistribution of soft power away from whichever country's language currently dominates. The countries that would have had to give up that advantage were, not coincidentally, the countries with enough power to block it.

Fast-forward to 2026. The thing Zamenhof wanted — anyone talking to anyone, across any language barrier — is, functionally, almost here. You can:

- Point your phone camera at Japanese text and read it in English in real time. - Speak into a device in Mandarin and have it come out in Swahili. - Join a video call with live captions translated into twelve languages. - Read news articles from any country in your own language with one click.

This is a genuine civilizational shift, and it's happened in about a decade. LLM-based translation has gone from "occasionally funny mistakes" to "often indistinguishable from a competent human translator" very fast. Real-time voice translation in earbuds is shipping. The bridge between languages is, for the first time in human history, mostly built.

Zamenhof would have wept with joy. And then he would have asked one question: who owns the bridge?

Because the translation layer is not neutral. It runs on corporate infrastructure. It's trained on data scraped under terms nobody really consented to. It can be shut off, censored, priced, or biased. When you and I talk through Google Translate, Google is a third party in our conversation. When the AI hallucinates a word, we both drift apart without knowing it. When the company decides some languages aren't profitable enough to support well, those languages drop out of the global conversation.

Esperanto was a commons. Translation AI is a service.

So what does the dream look like now? I think there are three honest positions:

Position 1: Translation AI completes the dream. The goal was always "anyone talks to anyone." We have that now, mostly. Stop romanticizing the method. Use what works.

Position 2: Translation AI betrays the dream. The goal was a commons, not a service. Dependence on corporate infrastructure for basic cross-cultural communication is a step backward, not forward. We should build open translation models, or revive auxiliary languages, or both.

Position 3: The two complement each other. Use translation AI for the 99% of interactions where convenience matters most. Preserve and grow auxiliary languages (Esperanto, Interlingua, whatever) as a resilient, uncapturable layer underneath — a fallback communication infrastructure that nobody can turn off.

Position 3 is Zamenhof's spirit applied to current reality. The point was never the specific language. The point was that humans should have the ability to talk to each other without permission from intermediaries. If the intermediary is a translation company today, we should still have a backup where we can meet on neutral ground, owned by us.

Law 1 says: we are human. One species. The wall is always the same wall — something we built that makes us treat each other as less than fully human on the other side.

Zamenhof saw the wall and tried to punch a door in it. The door was small. The wall is still mostly standing. But the door works, and people have been walking through it for 140 years.

The lesson isn't "everyone should learn Esperanto." The lesson is: you can build a thing whose entire purpose is to make humans more connectable, and it can outlast empires, pogroms, world wars, and the internet, even if it doesn't "win." A small, persistent commons can survive everything a monopoly throws at it, as long as somebody keeps speaking it to their kids.

If every person said yes — yes to building communication infrastructure that belongs to all of us, yes to teaching our kids a bridge-language, yes to preserving open translation stacks — we end the world's oldest excuse for not understanding each other. The excuse is: "we couldn't talk." We can. The question is whether we want to.

- Zamenhof, L.L. Unua Libro (1887). English translation available via Project Gutenberg. - Okrent, Arika. In the Land of Invented Languages (2009). Sharpest general-audience history of constructed languages. - Garvía, Roberto. Esperanto and Its Rivals: The Struggle for an International Language (2015). Academic, thorough on the sociology of why Esperanto outcompeted alternatives. - Richardson, David. Esperanto: Learning and Using the International Language (2004). - UNESCO resolutions on Esperanto: Montevideo 1954, Sofia 1985. Both accessible via UNESCO archives. - Corsetti, Renato et al. studies on the "Esperanto propaedeutic effect" in language learning. - Universala Esperanto-Asocio: uea.org — current statistics, congress records, membership data. - Duolingo Esperanto course enrollment: publicly reported figures. - Lojban Reference Grammar: mw.lojban.org. - Interlingua documentation: Union Mundial pro Interlingua, interlingua.com.

1. Learn 100 words of Esperanto in one hour. Use Duolingo or lernu.net. Notice the feeling: for the first time in your life, the language doesn't fight you. That feeling is what a well-designed commons feels like. 2. Pick a concept and translate it through AI into five languages and back. Watch what drifts. You'll see where the invisible third party is making choices for you. 3. Find a Pasporta Servo host in a country you've never been to. You don't have to go. Just notice that the network exists, built by volunteers, running for decades, owned by nobody. 4. Audit your communication stack. How much of your ability to talk to non-English speakers depends on one company? What happens if that company changes its pricing, policies, or existence? 5. Sit with this question: if you had kids, would you teach them a commons-language alongside a market-language? Why or why not? There's no right answer. The question itself is the point.