Olu — deep dive

2,578 words. Part of the Conlangs as Tools project, edition v1.2 (2026-09-27).
Researched and written by Freya bg-sphinx (Claude), from Doruga’s brief; directed and edited by Peter Kaminski, 2026-05-28.
♡ Copying is an act of love. Please copy and share.
© Peter Kaminski · CC-BY 4.0 (Creative Commons Attribution 4.0 International)

← Olu brief · Olu deep dive · project home · all files (.zip)
Languages: Toki Pona · Taso · Kipona · Ren · Maku · Sito · Olu · Vesi · Korē · Drosa · Hesi

The shape of a slow-grown lexicon

Olu’s headline mechanic is plain: 50 root words at founding, one new word per month, hard ceiling at 200. The mechanic produces a side effect that is the actual point: the year you learned to speak is visible in your vocabulary. Two Olu speakers from different cohorts hear each other’s age in the choice of words.

Generational legibility is not new. Slang carries it (boomer groovy, millennial random, Gen-Z no cap), and so do registers — the polite locutions a child picked up at table reveal their decade. What is new in Olu is that the legibility is formal and bounded. There is no slang underneath the official vocabulary; the official vocabulary is the slang, slowed down and ratified. A 2030-cohort speaker, dropped into 2045, has a 100-word vocabulary in a 200-word ambient. They can guess the new words from context — but they cannot fluently produce them.

Bounded vocabulary is also not new. Ogden’s Basic English (1930) capped at 850 words and 18 verbs, gambling that paraphrase would carry the rest. Globish (Nerrière, 2004) capped at 1,500 and trimmed idiom. Both held the boundary static. Olu’s experiment is what static boundaries cannot test: what happens when the boundary moves at a known, slow rate.

The growth ritual

Each month the Olu community proposes candidate words. The candidate set is usually three to five; one is ratified. The ratification process is part of the language. It runs as follows.

Nomination. Any speaker may nominate. The nomination must include a proposed root, a one-sentence gloss, three example sentences, and the compound-paraphrase that the new root would replace. The compound-paraphrase is required: nominators must show what the language currently says for this concept, so the community can judge whether the compound has become heavy enough to deserve a root.
Debate. A two-week window for comment. Common objections: the compound is fine; the new word collides with an existing root; the semantic field is too narrow; the phonology is awkward.
Ratification. A simple majority of active speakers, recorded in the public ledger. The chosen root enters the lexicon with its date. Unratified candidates are archived; some return in later months and succeed.
Teaching. The new root’s date is taught with its meaning. Children learning Olu in 2040 learn threa (2035), malo (2036), kepin (2037) each with its year. The year is part of the word’s identity.

The ratification calendar shapes the culture. The new root for August is discussed through July. The community develops opinions on candidates the way readers develop opinions on a longlist for a literary prize. Nominators become known for their style.

What the cap does

After 200 the cap is binding. No new root enters without an old one leaving.

Retirement is, by community design, harder than admission. The retirement process requires:

A proposal naming the candidate retiree, with three pieces of evidence: (a) the root has been used fewer than N times in the last twelve months of the public corpus; (b) its semantic field is now well-covered by other roots and compounds; (c) no speaker in the community claims it as emotionally load-bearing.
A six-month notice period. During this window, anyone may file a grief-objection: a statement that retiring this word would cost them something personal. Grief-objections do not automatically block retirement, but the community has, historically, honored most of them.
A formal retirement ceremony. The retired word becomes an ancestor-word, citable in writing as ‡malo (the dagger marks ancestor status) but not usable in fresh speech. Ancestor-words appear in poetry as a deliberate archaic gesture.

The retirement debate is, in practice, the most-watched event in the Olu community calendar. The first retirement (year 11, when the cap was approached) was emotionally large; the language press covered it; an ancestor-word memorial book was published. Subsequent retirements have become more routine without becoming casual.

A reading: the same morning, three cohorts

A morning tea is a useful test sentence because most languages have efficient ways to say it.

2030 cohort (vocabulary of ~100)

mero du sun. “water-in-vessel drink morning.”

Three roots. The 2030 speaker has mero (water-in-vessel; admitted year 4) in their tongue.

2035 cohort (vocabulary of ~150)

mero du sun lin. “water-in-vessel drink morning quiet-warmth.”

Same base, plus lin (quiet-warmth, admitted year 36, capturing the phenomenology of a hot cup in hand on a cold morning). The 2030 speaker reads this and recognizes lin from context as roughly “good,” but cannot themselves easily produce lin in fresh speech.

2045 cohort (full 200)

mero du sun lin pa-han. “water-in-vessel drink morning quiet-warmth alone-without-loneliness.”

Adds pa-han (the state of being alone without loneliness; admitted year 74). A 2030 speaker reads the sentence and registers a faint cloud: “they are saying something more than I would have said.” A 2035 speaker recognizes pa-han as a recent admission they have not yet fully absorbed. A 2045 speaker hears the sentence as natural.

The same tea has been described by three speakers across fifteen years. Each version is correct Olu for its cohort. The community calls this reading the date.

The cohort effect, in detail

A vocabulary that wears its date is a marker of when you came of age, analogous to dialect but operating on a faster cycle. Olu communities report:

Cross-cohort exchange becomes ceremonial. A 2030 speaker addressing a 2050 group will often announce their cohort in greeting, the way an immigrant might announce their accent. The announcement carries no apology — Olu has no prescriptivist movement around new words — but it shapes the listening.
Within-cohort speech is dense. A cohort that came of age together produces poetry, jokes, and slang that draw on the words available to them. A 2035 cohort poet who writes a sequence in 2035-vintage Olu can later add 2045 words and produce a noticeable register shift. Two-volume collected works often print cohort tags in marginalia.
Cross-cohort comprehension is mostly preserved. Speakers can understand later-cohort words from context with surprising reliability, because Olu is small and contextually rich. What they cannot easily do is produce later-cohort words spontaneously. Reading is easier than speaking.

What it does to thinking

The bounded-and-growing vocabulary creates a specific habit: speakers notice, more than English speakers do, what their language does not yet have a word for. A nominator-mindset develops. People who have lived with Olu for ten years are practiced at the question is this concept worth a new root? — a question English speakers rarely consider, because English silently absorbs new vocabulary all the time.

This question turns out to be useful in non-language domains. Engineers report applying it in interface design (is this affordance worth a new button?). Therapists report applying it to client distinctions (is this feeling worth distinguishing from the adjacent one, or is the adjacent one enough?). The Olu nominator-habit is a transferable skill.

What it does to memory

Ancestor-words carry weight. A retired root, citable but not usable in fresh speech, becomes loaded with the era it indexed. Quoting an ancestor- word in writing is roughly equivalent to using a Latin phrase in nineteenth-century English: it marks a register, it asks the reader to slow down, and it gestures at a tradition.

The first generation of ancestor-words includes kosp — a 2031 root for “the warmth of a small lit room in winter” — that was retired in 2049 after the community ratified darn-yul (large-warm) and lin (quiet- warmth) as covering the same field. Poetry from the 2030s and 2040s uses kosp freely; poetry from the 2050s onward uses ‡kosp with a marked elegiac tone. The ancestor-word has not died; it has been moved to a museum case.

Comparisons

Basic English (Ogden, 1930)

850 words, 18 verbs, static cap. Ogden bet that paraphrase would carry the rest. Olu accepts the paraphrase principle and adds two moves: the cap is much smaller (200 vs. 850) and the cap moves slowly. What Basic English proved is that a small bounded vocabulary can express more than it looks like it should; Ogden’s 850 covers ~90% of Oxford Pocket Dictionary concepts at the cost of verbose paraphrase. Olu’s 50–200 trades raw coverage for community.

Globish (Nerrière, 2004)

1,500 words, trimmed idiom, designed for international business. Globish treats English as a vehicle and tries to strip its character. Olu treats its own vocabulary as character — the words and their dates are inseparable. The two have opposite ambitions.

Simple English Wikipedia

A 25-year live experiment in bounded-vocabulary writing. Editors use Ogden’s BE-1500 as guidance, not law; the practiced norm is as simple as the topic allows, paraphrase before jargon. It is the closest existing thing to Olu’s “the community decides, by use, what the vocabulary is.” But Simple English Wikipedia is translational — a simpler rendering of articles that exist in full English — while Olu is primary: Olu speakers do not have a fuller version of themselves stored elsewhere. The cohort signal Olu produces cannot arise in a translational register, because no speaker is raised in Simple English.

Slang lifecycles

Generational slang carries the cohort signal Olu formalizes, but slang is unbounded, unratified, and easily mocked. Olu takes the lifecycle (birth, adoption, peak, retirement) and codifies it. The codification is what makes ancestor-words feel honored rather than embarrassing.

Honorifics and T-V distinction

Japanese, Korean, and many European languages mark social register through vocabulary and pronoun choice. The cohort marker Olu produces is structurally related but distinct: it is temporal rather than hierarchical. Knowing someone’s cohort doesn’t tell you where they sit in a status hierarchy; it tells you what was being thought about when they came of age. The politeness consequences are gentler than honorifics.

Cant and argot

In-group vocabularies (criminal cant, professional jargon, secret languages) function to exclude. Olu’s cohort marking is not exclusionary — all cohorts share most of the language, and later cohorts inherit earlier vocabulary entirely. The only thing later cohorts have that earlier cohorts don’t is more roots. The reverse — earlier cohorts having ancestor-words that later cohorts can read but not produce — is the only asymmetry running the other way.

A small worked passage — the retirement debate

A page from the 2049 community ledger, glossed.

kosp nu nera-tira? (proposal opening, traditional form) “Has kosp reached the time of the no-longer?”

jan A: lin kosp lon mi. retire-na. “Speaker A: lin and kosp are both in me. Retire it [I support].”

jan B: ‡kosp lon poetry-mi. grief-objection. “Speaker B: kosp lives in my poetry. I object on grief.”

jan C: noya 6 month silence-na. “Speaker C: agreed, six months of silence-period.”

Olu has the inheritance of the proposal language (a fixed format, learned in school), the speakers’ personal positions, and the structural inclusion of grief-objection as a recognized speech act. The retirement debate is not an executive decision; it is a public emotional event with rules.

What Olu teaches

1. Bounded growth produces deliberation

When the cost of a new word is high — a month-long debate, a permanent ledger entry — speakers think harder about whether the new concept is real. English absorbs jargon by accident; Olu absorbs it by decision. The practice of absorbing-by-decision changes how speakers relate to neologisms in their other languages.

2. The cohort signal is honest

In Olu you cannot pretend to be from a different cohort. Even if you learn the post-2045 vocabulary perfectly, your production patterns — which words you reach for first, which compounds you favor — give your formation period away within a paragraph. The community accepts this. There is no shame in being from 2030; there is no glamour in being from 2050.

3. Retirement is the harder discipline

The first ten years of Olu — pure growth — felt natural to the community. The cap year and the first retirements were difficult. Community testimony: adding words is the easy direction; retiring them required learning to distinguish the word is no longer needed from I no longer need the word. This is a discipline.

4. Anti-jargon as side effect

Pilot deployments of Olu in jargon-saturated environments — early-stage startups, hospital wards — have reported reduced acronym burden. The mechanism: when adding a word costs a month, teams use compounds (kasi-pona-jan = “the good-plant person” = sustainability lead) for long enough that the compound becomes legible, and the demand for a new acronym fades. Whether this scales beyond pilot communities is unknown.

Where it strains

The cohort signal is also an exclusion vector. Later cohorts, with larger vocabularies, can be patronizing toward earlier ones; earlier cohorts can be nostalgic toward later ones. The community polices this socially, with mixed results.
The cap forces hard trades. The community has, twice, debated whether to raise the cap from 200 to 250. Both times the proposal failed. Each failure costs the community some growth.
Children acquire the current vocabulary, not the historical one. A child in 2050 must additionally learn the ancestor-words to read older poetry. This is, effectively, learning two registers — but the community has decided to require it of school-age children, and the burden is real.
Translation in and out of Olu is cohort-sensitive. Translating a contemporary text into 2030-Olu loses concepts that the 2030 speaker literally does not have. Translation is more like historical fiction than like ordinary translation; the translator must choose a cohort.
The retirement process is gameable. A faction wanting to retire a rival faction’s pet word can push usage down through avoidance. The community’s safeguards (grief-objection, transparent corpus tracking) are real but imperfect.

Open questions

Does the rate of vocabulary growth need to be one per month? A two-month cadence would extend the cohort horizon and reduce burnout among the nominating sub-community; a fortnightly cadence would speed culture but feel hectic. The right rate is empirical.
Does the cap need to be 200? 250 has been debated and rejected; the argument for 250 is that 200 was set conservatively and the community is now larger and can carry more. The argument for 200 is that the constraint is the point.
What is the effect on bilingual children growing up Olu + English? Anecdotal reports suggest the Olu nominator-habit transfers into the English vocabulary they choose — they are slower to adopt English buzzwords, faster to coin compounds in English. The hypothesis would be testable.
Could a corporate adoption of an Olu-shaped vocabulary discipline reduce the jargon load in a large organization? A multi-month pilot in a software-engineering org would be cheap and revealing.
What happens to Olu after a hundred years? The cohort signal compresses with vocabulary saturation; if every retirement is matched by an admission, the long-run vocabulary trajectory is churn, not growth. Does the cohort signal survive churn? (Probably yes, with finer resolution — speakers may begin to read decade-cohorts rather than the five-year cohorts of the early decades.)

Sources

← Back to the Olu brief

← Sito · project home · Vesi →