Chapter 6

The invented century

In August 1889, in Paris, delegates from across the world sat through a full congress conducted in a language that had not existed a decade earlier. The language was Volapük, the invention of Johann Martin Schleyer, a German Catholic priest who said the language had come to him from God in a single sleepless night (Okrent 2009). By the counts kept and repeated in its own era, Volapük then had nearly three hundred clubs, some two dozen periodicals, more than three hundred textbooks, and something near a million students (Sweet 1911; Guérard 1922). The Paris congress was its third and grandest, conducted entirely in Volapük, and the Britannica's retrospect preserved the detail everyone repeats because it deserves repeating: even the waiters spoke it. Volapük had traveled beyond the lectern and into the room, and it had done so in barely a decade.

Five years later the movement was a ruin. A small circle continued to care for Volapük, but the century that followed knew it mostly as a scatter of the faithful. Nothing about the language's grammar changed in those five years. What changed was who owned it.

Four chapters have now described what Phi is. This one describes what it is joining, because Phi enters a tradition, and the tradition's history is best told with the question a new language should ask of it: not which invented language was most beautiful, but why they died, and what the survivors did differently. The century of invented languages ran its experiments at full scale, with real people. What follows is this book's reading of the results, argued case by case: the failures were social. The grammar almost never mattered.

The emperor of words

Volapük died of its inventor. At the Paris congress the movement's leading figures pressed Schleyer to hand real authority to an elected academy, which could then settle the reforms the language's explosive growth demanded. Schleyer refused; the language, he held, was his, and he would keep a veto over every word of it. The academy proceeded without him, he denounced it, and the movement tore in half. Reform projects multiplied, clubs defected or dissolved, and the million students evaporated; the era's own encyclopedia recorded that the year of the zenith began a decline even more rapid than the rise (Sweet 1911). The language had other troubles, hard umlauts, a grammar overloaded with forms, a newer rival already circulating, but every trouble pressed on the same fracture: nobody knew who could fix anything. The rump academy eventually produced a reformed successor language that history barely remembers, and Esperanto recruited heavily from the wreckage of clubs and organizers the collapse left behind.

The autopsy is old, and a century of retellings has kept it simple. Communities sometimes forgive awkward sounds and difficult forms; they are less forgiving when every repair needs one proprietor's blessing. Sociologist Roberto Garvía, whose study of this era is the closest thing the field has to a standard account, tells the history of Volapük, Esperanto, and Ido as social history, movements shaped by who held power in the language and by how each answered the challenge to that power (Garvía 2015). The wider law, that the whole tradition's wars were never really fought over grammar, is this book's wager rather than his finding, and the rest of the chapter is the evidence offered for it.

The language that survived its own failure

The answer that lived came from Ludwik Zamenhof, an ophthalmologist raised in Bialystok, a city whose Polish, Russian, Yiddish, and German communities lived in real mutual hostility. He published Esperanto in 1887 with the opposite instinct to Schleyer's: almost immediately he renounced ownership, and at the movement's first world congress, at Boulogne-sur-Mer in 1905, the community made the renunciation constitutional: the declaration adopted there holds Esperanto to be nobody's property and its published foundation beyond anyone's right to change, its author's included (Declaration of Boulogne 1905). Zamenhof wanted more than an auxiliary code; he spoke of an inner idea, the hope that a shared second language would soften the hatreds he grew up inside. Esperantists still call each other samideano, sharer of the idea.

For a generation it looked like the idea might scale. The congresses grew from under seven hundred participants at Boulogne to nearly five thousand at Nuremberg in 1923 (Lapenna, Lins and Carlevaro 1974), and the interwar movement pressed the new League of Nations to take the language up. It came closer than one might guess. France's delegate met the first proposal, in 1920, by telling the assembly that an international language already existed and was called French. The answer was brief, if not exactly neutral. The League's secretariat went on to produce a favorable report, and the 1922 assembly adopted it as a League document (League of Nations 1922). The question of Esperanto in schools then went to the League's committee on intellectual cooperation, where the French member, instructed from Paris, helped bury it in favor of the study of national languages (Lapenna, Lins and Carlevaro 1974). The moment passed.

What came next was worse than indifference: the dictatorships understood the movement's internationalism perfectly. Hitler had named Esperanto in Mein Kampf as an instrument of Jewish world ambition, and all three of Zamenhof's children were murdered in the Holocaust; under Stalin, thousands of Soviet Esperantists were arrested and the movement's leaders shot. The standard history of this period carries the title the language earned: Dangerous Language (Lins 2016).

Esperanto never became what it set out to be. Its universal ambition put it in the path of national interest twice over, and after the war English took the role Esperanto had auditioned for. But the language did not die, and the way it did not die is the century's most instructive fact. The community that the inner idea built kept meeting, publishing, marrying, and raising children in the language: Esperanto has native speakers, about a thousand of them at the standing scholarly estimate, now two decades old (Lindstedt 2006), and no other invented language has documented anything like them. Counting everyone else is where the historian reaches for a number and discovers a definition is needed first. A statistical model of how many people would name Esperanto to a census taker lands near sixty thousand (Nielsen 2016); an extrapolation from Facebook self-identification suggests about two million internet users with some of the language (Wandel 2015). Those two figures answer different questions by different methods, not the two ends of one range, and the organized core, the membership of the world association, has always been smaller than either. Read as an auxlang, the numbers are a failure. Read as a voluntary community carried across five generations by the idea and the institutions the idea built, they have no rival among the languages in this chapter.

The improvers

Esperanto's nearest brush with death came not from governments but from its own intellectuals. In 1907 a committee in Paris, elected by a private body called the Delegation for the Adoption of an International Auxiliary Language, met to choose a definitive international language and voted for Esperanto, on condition of reforms along the lines of a project submitted anonymously under the name Ido; the anonymous author turned out to be one of the movement's own senior figures. The betrayal drama mattered less than the arithmetic that followed: by the reckoning of the movement's own standard history, the schism took about a fifth of Esperanto's leading activists and only three or four percent of its ordinary members (Lapenna, Lins and Carlevaro 1974). The leaders had followed the argument for the better design; the speakers stayed with their community. Ido then met the awkward question inside every improvement project: when is the improving finished? It kept revising, and each reform cost followers until some of its best minds left to build newer projects.

The language you could not own

The property lesson returned, a century after Schleyer, in American courtrooms. James Cooke Brown's Loglan, the logical language a later chapter weighs as a scientific instrument, was also a possession, and Brown treated it as one: he claimed copyright over the individual words of the language, asked his own volunteers for royalties, and watched them do what communities do when the founder will not share (Okrent 2009). They rebuilt the entire lexicon from scratch and continued as Lojban. Most volunteer disputes do not produce a new dictionary, but this one did. Brown's institute sued, and in 1992 a federal appeals court affirmed the narrow point that ended the fight: Loglan was a generic term, the name of a kind of language rather than of one man's product, and a generic name cannot be held as a trademark on the dictionaries and grammars that describe the language (Loglan Institute v. Logical Language Group 1992). The court ruled on a registration, not on who owns a grammar; no court has ever had to answer the larger question, because no one has brought it. Schleyer's mistake, run again under modern law, produced the modern result: the community rebuilt, the mark fell, and the language, for every practical purpose, walked away from its owner.

The language nobody came to

The saddest entry in the century's ledger is the one designed with the most scholarly care. In 1982 the linguist and novelist Suzette Haden Elgin built Láadan, a language shaped around women's perceptions, and embedded it in her novel Native Tongue as a deliberate experiment with a deadline: if women found existing languages as inadequate as her hypothesis held, they would adopt Láadan or build something better within ten years of the novel's 1984 publication. She published the verdict herself, with a scientist's plainness: the decade passed, neither happened, and the hypothesis was, in her words, proved false (Elgin 1999). The grammar was published, the dictionary existed, the novel found readers, and a small circle keeps the lessons alive to this day. What never came was the adoption the experiment was waiting for.

The autopsy here is different from Volapük's, and the difference is the point. Elgin wrote the materials, put them in the public's hands, and asked a real question. The diagnosis that follows is this book's, not hers: what Láadan lacked was a felt need. The audience it was built for did not experience their languages as broken, and a language offered as a remedy for an undiagnosed illness recruits no patients. Busy sympathizers adopt words, not grammars. If that reading is right, this is the failure in the century that governance could least have touched, which makes it the one every values-built language, Phi included, has to sit with longest.

The language of good

Toki Pona is the tradition's clearest modern success, and it is instructively small. In 2001 a Canadian translator and polyglot named Sonja Lang, working through a depression (Lang 2021), began building a language of about a hundred and twenty words as a private instrument for simplifying her own thought, "like mapping my brain onto paper," as she later put it (Lang 2024). Toki Pona spread quietly among hobbyists for a decade, and then Lang did the thing the century's ghosts kept whispering: she finished it. The official book, published in 2014, gave the language a stable, complete form a learner could hold in two hands (Lang 2014); in 2026 she went further and dedicated that edition to the public domain (Lang 2026). The community did the rest. A free video course by one YouTuber became the most common first resource, named by about half of the two thousand respondents to the community's own 2024 census (Toki Pona census 2024). Chat servers gave the language a place to be spoken daily, and the whole lexicon fits on one page, so a learner can at least see the edge of the lexical task. In 2022, on the third application, Toki Pona received an ISO language code, tok (ISO 639-3 Registration Authority). The code is an identifier, not a verdict; the registry certifies nothing about a language's health. The fact is still worth reading twice: a hundred-word art project built for one person's peace of mind, filed in the world's catalogues beside the natural tongues.

The community's own censuses keep the success honest, and the census keepers are honest about the census: recruitment is online and self-selected, turnover between rounds is high, and the organizers say plainly that no single census can count how many people have learned the language (Toki Pona census 2024). Each round draws about two thousand respondents. In the 2024 round about a fifth said they did not yet know the language, just over half placed themselves at conversation level or better, and about one in eighteen claimed the top rung. Toki Pona did not conquer anything. It also never tried: its official pages describe a community, kulupu, and decline the word movement, and the language never claimed to be for everyone. Read against the century's failures, the success looks like four choices. Small scope made fluency reachable; a finished book made the target stable; a warm and light-touch authority left forks nothing to feed on; modest goals kept it out of everyone's crossfire. Every one of those choices reads as a negative image of some earlier grave.

The funnel and the epitaph

One more set of numbers calibrates the whole century. When a television fantasy put an invented language in millions of ears, more than 1.2 million people signed up for the High Valyrian course on Duolingo inside two years, by the company's count as the press reported it (Pereira 2019), and no one has measured how many of them ever reached a conversation. Enrolling in a course is quick; reaching conversation requires the learner to return. Klingon, the most famous invented language on earth, had twenty or thirty people who could carry on a spontaneous conversation, at Arika Okrent's estimate from inside the community (Okrent 2009). Curiosity outnumbers learners by orders of magnitude, learners outnumber speakers the same way again, and nothing in this chapter's record beats that funnel. Communities are not recruited. They condense, slowly, out of use.

Where does a language like Phi stand among these failed ambitions and surviving communities? It enters with choices that read like a close study of both. The artifact is finished and machine-checked rather than perpetually reformed. Authority is explicit and personal, but the design is given away: the project's license declares the language itself, as a system, free for anyone to use, teach, and build on, forever, which is the claim Brown spent his last decades refusing to make. The universal ambition that put Esperanto in the path of governments is refused outright in Phi's founding documents, which claim a practice for whoever wants one and a world language for no one. And where Láadan arrived as a diagnosis its audience had not asked for, Phi arrives as a literature first, a shelf of finished texts a reader can weigh without committing to anything. None of that earns it a community; nothing earns anyone a community. The century only says what the price of one is.

If the hundred and forty years have a single epitaph, this book would phrase it the way its case histories keep pointing: nobody adopts a language. People adopt a community, and the language comes along inside it.

References

Brown, James Cooke. 1960. Loglan. Scientific American 202(6): 53-63. https://doi.org/10.1038/scientificamerican0660-53.

Declaration of Boulogne. 1905. Adopted at the first World Esperanto Congress, Boulogne-sur-Mer.

Elgin, Suzette Haden. 1999. Láadan, the Constructed Language in Native Tongue. Essay first published on the author's SFWA member site, updated 2 April 1999; now hosted, under a slightly expanded title, at https://laadanlanguage.com/articles/articles-by-suzette/laadan-constructed-language/.

Garvía, Roberto. 2015. Esperanto and Its Rivals: The Struggle for an International Language. Philadelphia: University of Pennsylvania Press.

Guérard, Albert Léon. 1922. A Short History of the International Language Movement. New York: Boni and Liveright.

ISO 639-3 Registration Authority. Change requests 2007-011 and 2017-035 (rejected) and 2021-043 (adopted; the code tok took effect January 20, 2022). iso639-3.sil.org.

Lang, Sonja. 2014. Toki Pona: The Language of Good. Self-published; official site https://tokipona.org/.

Lang, Sonja. 2021. Le Toki Pona: une langue universelle à découvrir. Interview with Radio-Canada, July 31; English translation at tokipona.org.

Lang, Sonja. 2024. Toki Pona: from personal art project to small world language. Roundtable talk, Department of Linguistics, University of Colorado Boulder, April 8.

Lang, Sonja. 2026. Public-domain dedication of the 2014 English edition of Toki Pona: The Language of Good, April 14; recorded at Wikisource.

Lapenna, Ivo, Ulrich Lins, and Tazio Carlevaro. 1974. Esperanto en perspektivo: faktoj kaj analizoj pri la internacia lingvo. London and Rotterdam: Universala Esperanto-Asocio, Centre for Research and Documentation.

League of Nations. 1922. Esperanto as an International Auxiliary Language: Report of the General Secretariat Adopted by the Third Assembly. Geneva: League of Nations.

Lindstedt, Jouko. 2006. Native Esperanto as a test case for natural language. In A Man of Measure: Festschrift in Honour of Fred Karlsson, special supplement to SKY Journal of Linguistics 19: 47-55.

Lins, Ulrich. 2016. Dangerous Language: Esperanto under Hitler and Stalin. Translated by Humphrey Tonkin. London: Palgrave Macmillan.

Loglan Institute v. Logical Language Group. 1992. 962 F.2d 1038. United States Court of Appeals for the Federal Circuit.

Nielsen, Svend. 2016. Per-country rates of Esperanto speakers. Self-published statistical model, svendvnielsen.wordpress.com, December 2016.

Okrent, Arika. 2009. In the Land of Invented Languages. New York: Spiegel and Grau.

Pereira, Alyssa. 2019. 1.2 million Game of Thrones fans have signed up to learn High Valyrian, created by Cal grad. SFGATE, April 12.

Sweet, Henry. 1911. Volapük. Encyclopaedia Britannica, 11th edition, volume 28.

Toki Pona census. 2024. Results of the 2024 Toki Pona census. Updated February 27, 2025. https://tokiponacensus.github.io/results2024/.

Wandel, Amri. 2015. How many people speak Esperanto? Esperanto on the web. Interdisciplinary Description of Complex Systems 13(2): 318-321. https://doi.org/10.7906/indecs.13.2.9.

Phi sources

The Phi material in this chapter comes from LICENSE, the opening of project/development_protocol.md, and the catalog in texts/README.md.