# Amadunia — complete reference Source: https://github.com/recepcinet/amadunia-lang (CC BY-SA 4.0). Rendered at https://amadunia.com/. Summary: https://amadunia.com/llms.txt. Machine-readable: https://amadunia.com/dictionary.json, https://amadunia.com/corpus.tsv. Language tag: art-x-amadunia. Grey/fuchsia on the site marks English/Amadunia; here, Amadunia words are the ones in the left columns and in italics. # Hi! 👋 Mi ama dunia. **You just read your first Amadunia sentence.** ("Hi! I love the world.") [![License: CC BY-SA 4.0](https://img.shields.io/badge/License-CC%20BY--SA%204.0-lightgrey.svg)](https://creativecommons.org/licenses/by-sa/4.0/) [![check](https://github.com/recepcinet/amadunia-lang/actions/workflows/check.yml/badge.svg)](https://github.com/recepcinet/amadunia-lang/actions/workflows/check.yml) **Amadunia** (*ama* = love + *dunia* = world) is a constructed world auxiliary language built on one principle: **take the easiest feature from every language.** No exceptions, no conjugations, no genders, no silent letters. If a sound or rule is hard for anyone on Earth, it's out. --- ## Learn the basics in 2 minutes **Alphabet — 20 letters, each with exactly one sound:** > a b c d e f g h i k l m n o p r s t u y *c* is the sound of *chai* and *church* — never *k*, never *s*. The rest are the plain values you would guess, and [the pronunciation page](grammar/pronunciation.md) writes every one of them down. Syllables stay simple — never more than two consonants in a row, and none of the tongue-twisters other languages allow (*strengths*, *vzglyad*). If you can say "banana" and "taksi", you can say anything in Amadunia. **Stress — the second-to-last syllable, in every word, always:** a-ma-**DU**-nia, ka-**LI**-ma, **DU**-nia, **MER**-si. A vowel pair is one syllable, so *dunia* has two and *bai* has one. No exception, and no word has to be learned with its stress — in 89% of the dictionary the second-to-last syllable is also the first. [The rule](grammar/stress.md). **Numbers:** | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |---|---|---|---|---|---|---|---|---|----| | uan | du | tri | pat | fai | sis | seti | ba | nau | des | No irregular numbers, ever: 11 = *des-uan*, 20 = *du-des*, 21 = *du-des-uan*. Hundreds use *sen*: 345 = *tri-sen pat-des fai*. **First words:** *hi* (hello) · *salam* (peaceful greeting) · *ok* · *ya* (yes) · *no* · *bai* (goodbye) · *mama* · *papa* · *mi* (I) · *yu* (you) · *ta* (he/she — no gender!) · *kita* (we, including you) · *ama* (to love) · *luma* (light) · *dunia* (world) · *una* (together) **Your first conversation:** > — Hi! Ok? > — Ya, ok! Bai! Congratulations, you speak Amadunia. 🌍 --- ## The whole grammar, on one page "Grammar you can learn in a day" is easy to claim, so here is all of it. Every rule below is settled, and each has its own page in [grammar/](grammar/) giving the reasoning and the alternatives that were rejected. | | Rule | Example | English | |---|---|---|---| | **Order** | Subject – verb – object | Mi kula pan. | I eat bread. | | **Adjectives** | after the noun | dom kabir | a big house | | | and that phrase is already a sentence — no verb needed | Dom kabir. | The house is big. | | **Tense** | *suda* past, *saufa* future, before the verb; present unmarked | Ta suda lai. | She came. | | **Plural** | say the noun twice; singular after a number | anak-anak · tri anak | children · three children | | **Possession** | owner straight after the thing owned | Rafiki mi saufa lai. | My friend will come. | | **Copula** | *es* before a noun, never before an adjective | Ini es dom mi. | This is my house. | | **Existence** | *es* with no subject, first in the sentence | Es hotel in madina. | There is a hotel in the city. | | **Negation** | *no* immediately before the predicate | Mi no kula pan. | I do not eat bread. | | **Questions** | rising tone and a question mark — no other change | Yu mau ca? | Do you want tea? | | | the question word stands where the answer will stand | Itu es ke? | What is that? | | **Pronouns** | *mi · yu · ta · kita* — never change shape | Ta kula pan. | He eats bread. | | **This/that** | *ini · itu*, last in the phrase | dom kabir ini | this big house | | **Place** | *in · dari · por* before the noun; a motion verb takes a bare goal | Mi go market. | I go to the market. | | **With** | *una* + noun | Yu lai una mi? | Are you coming with me? | | **Verb chains** | verbs simply follow one another | Mi mau kula pan. | I want to eat bread. | | **Comparison** | *lebi · kurang · paling* in front of whatever they scale — an adjective or a noun — and *dari* for "than" | Dom mi lebi kabir dari dom yu. | My house is bigger than yours. | | **Clauses** | no marker at all; *porke*, *kab*, *agar* when needed | Mi fikir yu hao. | I think you are good. | | **Adverbs** | the adjective, straight after the verb | Mi kara hao libro. | I read the book well. | | **And / or** | *aur* · *o* | Mi kupi pan aur ca. | I buy bread and tea. | | **Commands** | a verb with no subject — nothing is added | Kula pan. | Eat bread. | | | and *no* denies it the same way it denies anything | No go! | Do not go! | | **Fragments** | any phrase may stand alone when the question is clear | — Yu mau ke? — Ca. | — What do you want? — Tea. | **Everything in one sentence.** The full order is subject → *no* → tense → verb → adverb → object → place → time → clause, and a sentence that uses all nine slots at once is still made of nothing but the rules above: > **Mi no saufa kula hayai pan in dom rat ini porke mi bimar.** > *I will not eat bread quickly at home tonight because I am ill.* There is nothing else to learn. No agreement, no cases, no gender, no articles, no irregular verbs, no plural exceptions, no spelling to memorise. --- ## What it sounds like From [*Uan*](texts/text-5-uan.md), the fifth text: > Mi punya kalima. > Yu punya kalima. > Kita punya uan lingua. > > Mi ama dunia. > Yu ama dunia. > Kita ama dunia una. *I have words. You have words. We have one language. I love the world. You love the world. We love the world together.* Three subjects, one verb, not a letter different. English cannot do that — *I love, you love, she loves* — and neither can most of the languages the words came from. It is what falls out of having no conjugation at all. --- ## Why another world language? Esperanto (1887) proved a constructed language can live — but its roots are ~95% European. Amadunia draws from the biggest language families of the whole world: *ba* (8) is Chinese, *pat* (4) is Indonesian–Tagalog, *ca* (tea) is at home in Chinese, Hindi, Turkish, Swahili and Russian alike, *dunia* spans Arabic–Hindi–Swahili–Indonesian–Turkish, *ama* is Latin, *kita* is Indonesian. Every continent should hear something familiar. Design rules: 1. **Zero exceptions** — every rule works everywhere 2. **Written as spoken** — one letter, one sound 3. **No conjugation, no gender, [no articles](grammar/definiteness.md)** — grammar you can learn in a day 4. **Global balance** — no language family dominates 5. **Already-global words stay** — *hi*, *ok*, *taksi*, *foto*... free vocabulary ### The balance, measured Rule 4 is the one that is easy to claim and hard to keep, so it is counted. *Origin* is where a root came from; *reach* is how many roots a speaker of that family would recognise, since a word like *dunia* is at home in five families at once. | Family | Origin | | Reach | | |---|---:|---:|---:|---:| | Austronesian | 67 | 22.3% | 115 | 38.3% | | Turkic | 23 | 7.7% | 94 | 31.3% | | Latin/Romance | 75 | 25.0% | 90 | 30.0% | | Indo-Aryan | 22 | 7.3% | 83 | 27.7% | | Semitic | 49 | 16.3% | 68 | 22.7% | | Niger-Congo | 14 | 4.7% | 64 | 21.3% | | Slavic | 5 | 1.7% | 48 | 16.0% | | Germanic | 13 | 4.3% | 46 | 15.3% | | Iranian | 9 | 3.0% | 36 | 12.0% | | Greek | 8 | 2.7% | 17 | 5.7% | | Sino-Tibetan | 11 | 3.7% | 12 | 4.0% | | Japonic | 4 | 1.3% | 8 | 2.7% | | Koreanic | 0 | 0.0% | 2 | 0.7% | | Austroasiatic | 0 | 0.0% | 1 | 0.3% | | Celtic | 0 | 0.0% | 1 | 0.3% | | Tai-Kadai | 0 | 0.0% | 1 | 0.3% | **European roots are 34% of the dictionary**, against Esperanto's ~95% — both counted the same way, by where a root came from. By reach the largest is Austronesian at 38.3%. The thin spot is real and named rather than hidden: **by reach**, Sino-Tibetan is 4.0% and Japonic 2.7%. That was recorded as structural — their words are short, and the short space is full — and the explanation turned out to be wrong: modern Chinese and Japanese words are mostly two syllables, which is exactly what the four-letter rule wants. Seven of ten ordinary words tested passed every rule. The borrowing reached for the classical monosyllables and stopped. [The full working is here](dictionary/balance.md), and what it means for the next three hundred is [in the A2 briefing](dictionary/proposal-a2.md). --- ## What a word has to survive Nine rules, then the checker. They are strict enough that most candidates die, and the ones that died are kept — [CONTRIBUTING.md](CONTRIBUTING.md) records all nine with the real words each has rejected, and running the checker as the tenth step. | Rejected | Because of | What was lost | |---|---|---| | *hijau*, *anjing*, *mvua* | letters outside the twenty | green, dog, rain | | *fenestra*, *Fenster*, *skribi* | three consonants in a row | window, to write | | *hayawan*, *sikio*, *kirei* | three vowels, or an unlisted pair | animal, ear, pretty | | *beli* | too close to *beri* (to give) | to buy | | *guru* | too close to *suru* (to do) | teacher | | *luna* | too close to *luma* (light) | moon | | *nice* | reads differently in English than it looks | pleasant | | *matar* | "to kill" in Spanish | to marry | The *l* against *r* rule outranks everything, including how widely a word travels: it is why *ruma* became *dom* after eighty-four uses. And **one root does one job** — a word is a noun or a verb, never both. Two roots out of 300 break that rule, and neither is treated as an exception: both are [open questions](grammar/verb-chains.md) waiting on a decision. --- ## Status 🚧 Amadunia was born on **September 2, 2026** and is under active construction. **300 roots**, the A1 target, reached the next day. That target is lower than the ~500 words a natural language needs at A1 because things can be described with what is already there — Esperanto launched with 917 roots, and Basic English covers daily life in 850. **A root here is simply a word.** There is no conjugation, no affix and no compound, so nothing multiplies: 300 roots are 300 words to look up, and that is the whole vocabulary. The comparison with Esperanto is not like for like — its 917 roots are multiplied by some forty affixes into tens of thousands of words, and 300 against 917 flatters this language rather than the other one. What 300 gives beyond 300 is *forms*, not words: every noun has a doubled plural, and twelve number words reach a thousand. This page used to say *compounding* does the rest. It does not: every hyphen in the repository was counted, and all 37 are either a number or a doubled plural. [No compound word has ever been formed](grammar/word-formation.md), and whether one may be is now an open question. Milestones passed: 80 for survival, then 180, then 300. Next is A2, at around 600 — and what those three hundred should be made of is measured in [a briefing](dictionary/proposal-a2.md), which makes none of the choices. **The grammar needed for A2 is complete.** Phonology, pronunciation, numbers, tense, plurals, possession, questions, pronouns, the copula, existence, negation, conjunction, demonstratives, place, verb chains, comparison, subordination, adverbs, stress and the sentence types are settled — twenty-one rules, each recorded with its reasoning and with the candidates it rejected. **[Forty questions are still open](grammar/README.md)**, gathered on one page, ordered by [what the writing has actually asked for](grammar/README.md) — a word for *then* leads it with six pages, and two questions have never been reached for at all. A mark for a name scores zero there and is still the briefing that matters most, which the index says plainly: it was not found by writing but by asking how a word is said aloud. Four were not left open on purpose but were forms the material already relied on that no rule had granted: a mark for a name, how *r* is made, "want to be" followed by an adjective, and the imperative — which led the list with three pages that tried and stopped until it was granted on September 3, 2026. Each was found by writing something that was not a lesson — which is what the texts are for. Follow the commits to watch a language grow in real time. --- ## Where to start | If you want to | Go to | |---|---| | learn the language properly, from nothing | [Lesson 01](lessons/lesson-01-greetings.md), or [the syllabus](lessons/README.md) to see what each of the twenty-six teaches | | know when you can start reading | [the reading ladder](lessons/reading-ladder.md) — nearly a quarter of the texts' words after one lesson, story 1 readable at Lesson 13 | | say something today | [the phrasebook](phrasebook.md) — greeting, directions, buying, trouble | | look a word up | [the dictionary](dictionary/dictionary.md), or [the English index](dictionary/index-english.md) to write | | know which words to learn first | [the frequency list](dictionary/frequency.md) — twenty-five roots cover half of everything written | | know why a rule is the way it is | [grammar/](grammar/) — each page ends with what it rejected | | read the language, not lessons about it | [texts/](texts/) — thirty original pieces | | add a word | [CONTRIBUTING.md](CONTRIBUTING.md) | | answer one of the open questions | [grammar/README.md](grammar/README.md) — and six of the forty already have [a briefing](grammar/) that measures the evidence without making the choice | ## What's here | | | |---|---| | [lessons/](lessons/) | 26 lessons in order, teaching every one of the 300 roots | | [grammar/](grammar/) | 21 rules — what was settled, why, and what was rejected — and 8 briefings, six of them waiting on a decision | | [dictionary/](dictionary/) | 300 roots with sourced etymologies, [the English index](dictionary/index-english.md) for writing, and [where they come from](dictionary/balance.md) | | [texts/](texts/) | thirty original pieces — ten stories, two arguments, two dialogues, a poem, a letter, a recipe, a walk to school, a shared meal, a room, a life, a journey that may not happen and a notice on a school wall — each ending with what the language could not say | | [phrasebook.md](phrasebook.md) | the practical page — greeting, directions, buying, trouble | | [check.py](check.py) | every rule above, tested on every push — [and what it refuses](GUARANTEES.md), listed | | [CONTRIBUTING.md](CONTRIBUTING.md) | how to propose a word, and what has already been rejected | Each directory has its own index. Every file in `grammar/` marks what it settled and what it left open. --- ## Checking the language Amadunia holds itself to rules that are easy to state and easy to break by accident — twenty letters and no others, at most two consonants in a row, no two words a single sound apart, no *l* against *r* anywhere, every root sourced, every word in a text already in the dictionary. [`check.py`](check.py) checks all of them: ``` python3 check.py ``` It runs on every push and every pull request. A change that would put a new minimal pair into the dictionary, use an unlisted letter, leave an etymology blank, invent a word inside a text, put *es* in front of an adjective, leave a noun predicate without one, chain two verbs before Lesson 17, or quietly leave a settled question on the open list will fail it. It has earned its place several times over. It found thirteen copula errors that had accumulated across five lessons and three texts, then caught four more in the first draft of the phrasebook. It found five roots that were glossed in a word table and then never used in a single sentence — including the numbers eight and nine, so the number system had never been shown above six. And when its own hand-written verb list was replaced by one read from the dictionary, it turned out to have known 19 of the 47 roots the dictionary glosses *to something*, and three real verb chains were sitting in early lessons where the rule forbids them. The language has 50 verbs, not 47: *bisa*, *lasim* and *madad* are verbs by behaviour and not by gloss, and this sentence said *the language's 47 verbs* until September 6, 2026, which is the smaller of the two counts wearing the name of the larger. And [`test-check.py`](test-check.py) checks the checker. It breaks each guarantee in turn against a throwaway copy of the repository and demands that `check.py` rejects it, naming the specific error it expects — so a mutation caught by the wrong rule counts as a miss, and a mutation whose target has moved is reported as stale rather than as a pass. It exists because one check was added that could never fail, and a hand-run test read its empty output as success. The one thing neither can check is whether a word is a good word. That is still a decision, and the reasoning behind every one of them is in [grammar/](grammar/). --- ## License This language and all its materials are licensed under [CC BY-SA 4.0](LICENSE) — free to use, share and build upon, **forever**. Suggestions and contributions are welcome via Issues and Pull Requests; core language decisions are currently curated by the project founder to keep the language coherent while it is young. [CONTRIBUTING.md](CONTRIBUTING.md) sets out how to propose a word — the ten rules a candidate has to survive, and a sample of what they have already rejected — and how to answer one of the open questions. *Mi ama dunia. Yu ama dunia. Kita ama dunia.* 🌍 --- # Adverbs — how something is done *Taught in [Lesson 12](../lessons/lesson-12-colors-directions.md), and extended in [Lesson 18](../lessons/lesson-18-comparing-and-joining.md) for the case where the verb has an object.* *Status: settled — decided September 3, 2026.* ## The rule **An adjective placed straight after the verb describes the verb.** | | | |---|---| | Ta anda **hayai**. | She walks fast. | | Mi kara **hao**. | I read well. | | Anak kimbia **kuat**. | The child runs strongly. | | Kita rabota **cok**. | We work a lot. | No new word, no ending, no separate class of adverbs. The adjective does both jobs, and its position says which one. This is what the language has been doing since Lesson 12 — *Ta anda hayai* — without ever stating the rule. It is stated now. ## Before the object When there is an object, the adverb stays with its verb: | | | |---|---| | Mi kara **hao** libro ini. | I read this book well. | | Ta beri **hayai** ca por mi. | She quickly gives me tea. | Keeping the adverb against the verb is what makes it unambiguous. Put it after the object and it would attach to the object instead: *Mi kara libro hao* is "I read a good book". | | | |---|---| | Mi kara **hao** libro. | I read a book well. | | Mi kara libro **hao**. | I read a good book. | The word never changes; only its place does. ## The full order Everything the language has settled, in one line: > **subject → no → suda/saufa → verb → adverb → object → place → time → clause** | | | |---|---| | Mi no suda kara hao libro ini in dom rat ini porke mi sedih. | I did not read this book well at home last night because I was sad. | That sentence fills all nine slots at once, and not one word in it has changed its shape. It read "Every rule in Amadunia is in that sentence" until September 5, 2026. Nine of the twenty-one rules are in it — order, negation, tense, the adverb, the demonstrative, place, time, the clause and the invariant pronoun. The copula, existence, the plural, possession, questions, verb chains, comparison, *aur* and the command are not, and a sentence cannot hold them all: a command has no subject, and a question is a question. ## Sources Using one form for adjective and adverb is the majority pattern: Chinese, Indonesian, Malay, Turkish (*hızlı* both ways), German (*schnell* both ways), Swahili, Japanese in part. English *-ly* and Romance *-mente* are the minority, and both come with irregulars — English *good/well*, *fast/fast*, *hard/hardly*. ## Open questions - **Adverbs of degree on adverbs** — "very quickly" — needs a word for "very". *cok* is "much"; whether it covers "very" is untested. Wanted twice now: by [the phrasebook](../phrasebook.md), which lost it from "thank you very much", and on September 4, 2026 by *Mi parolas Esperanton tre flue*, which came out as *Mi sema hao Amadunia* — I speak Amadunia well, with both the degree and the manner gone. - **Whether *daima* covers "forever".** It is glossed "always", from Arabic *dāʾiman*, which carries both senses. *Natura hidup daima* would say nature lives forever, and *Mi kula daima pan* says I always eat bread — habit and endlessness in one word and one slot. Most languages separate them. This is the same shape of question as *cok* and "very" above, and it was found the same way: [text 10](../texts/text-10-mila-tahun.md) recorded that the language had no word for *forever* without checking what it did have. - **Frequency adverbs.** *daima* (always) and *kadang* (sometimes) are adverbs by meaning but not adjectives. This page used to say they sit in the adverb slot with everything else; counted, [most of them do not](proposal-frequency.md), and Lesson 19 alone uses all three positions. The split lived here as well as in the briefing until September 5, 2026, and only the briefing's copy was recounted — one word, two pages, one check. [A briefing measures it and costs the three answers](proposal-frequency.md). Until it is decided the word *kadang* cannot be written at all, because any sentence using it would settle the question by accident. --- # Comparison — bigger, biggest, as big as *Taught in [Lesson 18](../lessons/lesson-18-comparing-and-joining.md).* *Status: settled — decided September 3, 2026.* ## Three degree words, all before the adjective | | | | |---|---|---| | **lebi** | more | Dom ini **lebi** kabir. — This house is bigger. | | **kurang** | less | Dom ini **kurang** kabir. — This house is less big. | | **paling** | most | Dom ini **paling** kabir. — This house is the biggest. | They stand where *cok* stands, in front of what they scale. *lebi* has been in the language since Lesson 16, where it scaled a noun — *lebi ca*, more tea. It scales an adjective the same way. ## Than is *dari* | | | |---|---| | Dom mi lebi kabir **dari** dom yu. | My house is bigger than yours. | | Ca ini kurang garam **dari** ca itu. | This tea is less hot than that one. | | Mi lebi hayai **dari** dugu mi. | I am faster than my sibling. | **No new word.** *dari* has meant "from" since Lesson 15, and using "from" for the standard of comparison is what Arabic (*min*), Persian (*az*), Turkish (*-den*), Hindi (*se*) and Indonesian (*daripada*) all do. Roughly half the world already says "bigger *from* you". ## As ... as is *kadar* | | | |---|---| | Dom mi kabir **kadar** dom yu. | My house is as big as yours. | | Mi hayai **kadar** yu. | I am as fast as you. | *kadar* is Arabic *qadar*, spoken natively in Turkish *kadar*, Urdu and Hindi *qadar*, and Persian. No degree word is used with it — equality is not a degree. ## With verbs The same words work on a verb, in the adverb slot: | | | |---|---| | Mi kara **lebi** dari yu. | I read more than you. | | Ta rabota **paling**. | She works the most. | ## Open questions - **Superlative among a named group** — "the biggest of the three" — has no form yet. *paling kabir dari tri* is untested. - **Whether two degree words may stack.** Lesson 18's conversation read *Cok lebi hao* for "much better" until September 4, 2026. *cok*, *lebi*, *kurang* and *paling* all stand in the same slot in front of what they scale, and nothing grants putting two there at once. The line now reads *Lebi hao*, which says less. Found by reading the lesson rather than by a check, and left open rather than granted. **This page was doing it too.** The row above read *Ta rabota **paling** cok*, "she works the most", two slots deep — on the page that says nothing grants two. Corrected September 5, 2026 to *Ta rabota paling*, and a check now refuses the shape. One hit in the whole repository, and it was here: the corpus never stacked them, only the page arguing about them did. --- # Conjunction — saying "and" *Taught in [Lesson 14](../lessons/lesson-14-not-and.md).* *Status: settled — decided September 3, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## The rule ***aur* joins two things.** Nouns, adjectives, whole sentences — the same word for all three. | | | |---|---| | Mi kupi pan **aur** ca. | I buy bread and tea. | | Dom kabir **aur** sundar. | The house is big and beautiful. | | Mi lai **aur** mi kan yu. | I come and I see you. | | Angin **aur** yuki lai. | Wind and snow come. | ## Why the language needs a word here Because two nouns side by side are already spoken for. The [possession rule](possession.md) reads *pan ca* as "the tea's bread" — so putting nouns next to each other cannot mean "and". Without *aur*, Amadunia had no way at all to join two nouns. Writing is worse off than speech here: a comma can stand in on the page, but in the air only a pause separates *pan, ca* from *pan ca*, and a pause is not a word. ## Lists For three or more, put *aur* between each pair: > Mi kupi pan **aur** ca **aur** akua. In writing, commas may stand in for all but the last: > Mi kupi pan, ca **aur** akua. Both are correct. The full form is never wrong, which is the form to teach first. ## Sources Spanish *y*, Portuguese and Italian *e*, Russian and Polish *i*, Latin *et*, Greek *kai*, Hindi and Urdu *aur*, Arabic *wa*, Indonesian *dan*, Swahili *na*, Chinese *hé*, Turkish *ve* — every language has one, and almost none of them agree. *aur* is Hindi and Urdu, spoken by roughly eight hundred million people, and recognisable to Punjabi, Nepali and Bengali speakers besides. ## Why this and not the alternatives Almost every major word for "and" was blocked, and by the language's own rules rather than by taste. Each was checked against the whole dictionary: | Candidate | Source | Reason rejected | |---|---|---| | na | Swahili, Igbo, Zulu | minimal pair with *ba*, *ca*, *no*, *ta*, *ya* | | dan | Indonesian, Malay | minimal pair with *din*, *kan*, *pan*, *uan* | | he | Chinese *hé* 和 | minimal pair with *hi*, *ke* | | et | Latin, French | minimal pair with *es* | | kai | Greek | minimal pair with *bai*, *fai*, *kab*, *kan*, *lai* | | kon | Spanish *con* | minimal pair with *kan* | | kum | Latin *cum* | minimal pair with *kim* | | sama | Indonesian | minimal pair with *mama*, *nama*, *sasa*, *sema* | | plus | Latin, global | minimal pair with *plis* | | ve | Turkish | the alphabet has no *v* | | wa | Arabic *wa* و | the alphabet has no *w* | | i | Spanish *y*, Russian, Polish | no collision at the time, but 41 roots end in *-i* — *rafiki i mi*, *tri i du*, *beri i* would blur in speech | | e | Portuguese, Italian | one bare vowel; too light to carry a join | | hem | Turkish, Persian | clean, but reaches far fewer people | | gam | Hebrew | clean, but reaches fewer still | That so many candidates collide is not bad luck. Short common words crowd the same space, and *aur* survives partly because its shape — vowel, vowel, consonant — is one almost nothing else in the dictionary uses. ## Open questions - ~~"or"~~ — settled: **o**, Spanish/Italian *o*, Portuguese *ou*. *Ca o akua?* The language's first one-letter word; the phonology allows a bare vowel, and 13 roots end in *-o*, so it blurs less than *i* would have. Recounted September 5, 2026. It read *only four roots end in -o*, against *twenty* ending in *-i*; it is **13** against **41** today. That is the stale half. The other half is that neither figure was right when it was written: on the day *o* was settled the dictionary held **six** roots ending in *-o* — *no*, *mano*, *go*, *hao*, *o* itself and *foto* — and **22** ending in *-i*. The argument was stated as one in five and has never been better than one in three point seven; it is one in three point two now. So the comparison held all along and the margin never did, and the sentence that overstates it is the only thing on the page that ever said why *o* and not *i*. The word is settled and this changes nothing about that. A second thing the recount shows: *i* is listed above as having *no collision*, and today it has one. *o* itself is a one-letter root now, so *i* and *o* are a minimal pair — the shortest one possible. The row is true of the day it was written and false of the dictionary it sits in, which is why it now says *no collision at the time*. - ~~"with"~~ — settled in [place.md](place.md): *una* + noun. - **Joining more than two clauses** — whether long chains want a different rhythm — is a question for when there is more prose. - **There is no word for "but".** *aur* joins and *o* chooses; nothing contrasts. Found on September 4, 2026 in [Lesson 12](../lessons/lesson-12-colors-directions.md), which had been saying *Ma hao!* — an invented word — since it was written, for "but good!". The lesson now says *Hao!* and says less. [A briefing is now written](proposal-but.md): six pages have wanted one, *lakin* reaches five families with one word, and doing nothing is argued fairly. Whether the contrast is a conjunction, an adverb, or nothing at all is open; [several source languages do without one](../dictionary/README.md#words-the-writing-has-asked-for) in short speech, and Amadunia may too. --- # The copula — saying "is" *The word is given in [Lesson 06](../lessons/lesson-06-possession.md); the rule is taught in [Lesson 11](../lessons/lesson-11-being.md), and [Lesson 12](../lessons/lesson-12-colors-directions.md) drills it with a deliberately wrong sentence.* *The existential also wins the sentence-initial slot over [the command rule](sentence-types.md): a sentence beginning with* es *is "there is", never "be".* *Status: settled — decided September 2, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## The rule **Before a noun, *es*. Before anything else, nothing.** | Predicate is a… | Amadunia | English | |---|---|---| | noun | Mi **es** doktor. | I am a doctor. | | noun | Sol **es** rafiki mi. | Sol is my friend. | | adjective | Dom kabir. | The house is big. | | adjective | Ca garam. | The tea is hot. | | place word | Mi sini. | I am here. | | place word | Hotel situ. | The hotel is there. | *es* is a verb. It appears when the thing after the subject is a noun, and only then. ## Why a noun needs it and an adjective does not Because of the possession rule. Two nouns side by side already mean something: the second owns the first. | | | |---|---| | Rafiki mi doktor. | my friend's doctor — *a thing, not a sentence* | | Rafiki mi **es** doktor. | My friend is a doctor. — *a sentence* | | Nama mi Sol. | Sol's my-name — *nonsense* | | Nama mi **es** Sol. | My name is Sol. | Without *es*, *Sol doktor* can only be "Sol's doctor". With it, "Sol is a doctor". The two readings never touch. An adjective cannot own anything, so *dom kabir* has no possessive reading to collide with. It needs no *es*, and it takes none. ## Tense *es* is a verb, so it takes the tense particles like any verb: | | | |---|---| | Mi **suda es** doktor. | I was a doctor. | | Ta **saufa es** rafiki yu. | She will be your friend. | Adjectives and place words take the tense particle **directly** — in a sentence, they do the work of a verb: | | | |---|---| | Dom **suda** kabir. | The house was big. | | Ca **saufa** garam. | The tea will be hot. | | Mi **suda** sini. | I was here. | There is no *suda es kabir*. An adjective is its own verb. ## Adjective: sentence or phrase? *dom kabir* standing alone is a sentence: "the house is big". Inside a longer sentence it is a phrase: *Mi kan dom kabir* — "I see a big house". The difference is whether anything comes before or after it. This is exactly how Indonesian *rumah besar* works, and it has served three hundred million people without a copula. ## In questions The question word takes the slot of what it asks about, so *es* stays where it is: | | | |---|---| | Nama yu **es ke**? | What is your name? | | Yu **es kim**? | Who are you? | | Ta-ta **es kim**? | Who are they? | | Dom kabir? | Is the house big? — *adjective, no es* | | Hotel nali? | Where is the hotel? — *place word, no es* | ## Sources *es* is the oldest and most widely shared verb on Earth. Spanish *es*, Latin *est*, Italian *è*, French *est*, Portuguese *é*, German *ist*, English *is*, Dutch *is*, Russian *est'*, Polish *jest*, Hindi *hai*, Persian *ast*, Greek *eínai*, Sanskrit *asti* — every one from the same root, spoken today by some three billion people. The split — a copula with nouns, none with adjectives — is the Chinese rule (*wǒ shì yīshēng* but *fángzi dà*), the Swahili rule (*mimi ni daktari* but adjective agreement), the Haitian rule (*mwen se doktè* but *kay la gwo*). Zero copula with adjectives alone is the majority pattern: Russian, Arabic, Hebrew, Turkish, Hungarian, Indonesian, Japanese, Korean, Thai, Vietnamese. ## Why this and not the alternatives **Why not zero everywhere.** Arabic, Russian and Hebrew manage it, but they have case, agreement or free word order to tell "X is Y" from "X's Y". Amadunia has none of those; it has one word order and a possession rule that makes *noun noun* possessive. Zero copula with nouns was already producing ambiguous sentences in Lessons 6, 7, 9 and 10 — all corrected in this decision. **Why not *es* everywhere.** It would add a word to every adjective sentence the language has spoken since Lesson 3, for no gain: adjectives were not ambiguous when this was decided. And the tense rule for adjectives (*dom suda kabir*) falls out for free without it. That parenthesis said *never* until September 5, 2026; the fragment rule of September 3 made it untrue, and the open question below is the correction. **Why *es* and not another word:** | Candidate | Source | Reason rejected | |---|---|---| | ni | Swahili *ni* | minimal pair with *hi*, *mi*, *no* | | da | Japanese *da* | d/t pair with *ta* | | dir | Turkish *-dir* | minimal pair with *din* (day) | | si | Chinese *shì* 是 | prefix of *sis* (6) and *sini*; and "yes" in Spanish and Italian, colliding in sense with *ya* | | se | Haitian *se* | prefix of *sen*, *sema*, *seti*; recognised by few | | adala | Indonesian *adalah* | three syllables for the commonest verb in the language | | i | Tok Pisin *i* | one bare vowel; too light to hear | *es* has no collision and the widest recognition of any candidate by a factor of ten. ## Open questions - ~~**Can an adjective stand as a noun?**~~ — answered by the rule, not by a decision: **no**, and the copula is why. An adjective predicate takes no *es* and a noun predicate must have one, so the moment *genc* can be a noun, *Ta genc* stops having one reading. [Text 21](../texts/text-21-uan-umur.md) had opened seventeen sentences with *Genc* — the young one — for a day. Every word in them existed, which is why five days of checks for invented words saw nothing; the invention was a **shape**. The sentences say *insan genc*, a noun with an adjective after it, which is what the language actually has. Recorded here rather than as an open question because nothing is open: allowing it would cost a settled rule. - **A noun with an adjective is now both a sentence and a phrase.** *Dom merah* is "the house is red" by this rule, and since [the fragment rule](sentence-types.md) of September 3, 2026 granted that any constituent may stand alone as an utterance, it is also "the red house" standing on its own. Nothing distinguishes them. **154 two-word utterances in the repository have this shape**, counted everywhere but the two pages that state the figure — writing the example *Dom merah* into this very bullet made the first draft of the count wrong by one, which is the third time this repository's own prose has become its evidence. [`check.py`](../check.py) recounts it rather than trusting this page. The corpus reads only one of them the other way, and it was found by reading Lesson 12 end to end: that lesson's own table glosses *Dom merah.* as "The house is red", and eighty lines later its conversation glossed the identical string "The red house." The gloss was corrected; the ambiguity was not, because correcting it needs a decision. The two greetings *Pagi hao!* and *Aksam hao!* are the other side of it and are not a problem — an idiomatic "Good morning!" for what the words say literally is ordinary translation. So this is a formal ambiguity that context has so far carried, exactly like the third option in [proposal-names.md](proposal-names.md). It is recorded, not briefed: it may well need nothing. **Two scanners for it were written and both rejected.** Requiring one string to keep one gloss inside a lesson finds two hits, and both are the deliberate statement-and-question pairs (*Yu mau ca.* / *Yu mau ca?*) that teach rising tone — keep the punctuation and it finds nothing, including the Lesson 12 fault, which lives in a conversation's paragraph translation rather than a table row. Aligning a paragraph translation to the lines above it is guesswork. Neither was added; a check that cannot catch the fault that prompted it is worse than no check, because it reads as cover. - **"want to be" plus an adjective has no form.** *es* is a verb and can be chained — *mau es doktor*, "want to be a doctor". An adjective is its own predicate and takes no verb, so *mau* has nothing to attach to: "wants to be the easiest" cannot be built. Both [text 3](../texts/text-3-lingua-ini.md) and Lesson 21 hit this and both were rewritten around it — text 3 by supplying a noun (*mau es lingua paling asan*), Lesson 21 by dropping the modal. Whether *es* is allowed before an adjective inside a verb chain, or something else is needed, is undecided. Found September 3, 2026 by checking every sentence in the repository against the copula rule. - ~~Verb chaining~~ — settled in [verb-chains.md](verb-chains.md): the verbs simply follow one another, *Mi mau kula pan*. - ~~Negation~~ — settled in [negation.md](negation.md): *Mi no es doktor*. *no* goes before *es*, and before the tense particle when there is one. - ~~Existence~~ — settled September 3, 2026: **es at the front of the sentence, with no subject, means "there is".** | | | |---|---| | Es hotel in madina. | There is a hotel in the city. | | No es ca. | There is no tea. | | Es dom kabir sini? | Is there a big house here? | | Suda es skola in negara ini. | There was a school in this country. | No new word. Amadunia never drops a subject, so a sentence beginning with *es* can only be existential. Spanish *hay*, French *il y a*, Portuguese *tem*, Turkish *var* and Chinese *yǒu* all press an existing verb into this job; *punya* (to have) was the alternative and was rejected because it needs an owner, and existence has none. --- # Definiteness *Taught in [Lesson 02](../lessons/lesson-02-food-and-home.md), where the first noun is glossed both ways.* **Settled by design rule 3: there are no articles. A bare noun is neither definite nor indefinite, and context decides.** *dom* is *a house* or *the house*. Nothing is added either way, and there is nothing to leave out. Not one of the 300 roots is an article, and none ever will be. ## The material has relied on this from the beginning **Forty-one nouns** are translated both ways in the lessons, the grammar files and the texts, and no page has ever said why. Seven of them: | Root | Translated *the* | Translated *a* | |---|---|---| | din | Din finis. — The day finishes. | Din baru lai. — A new day comes. | | farasi | Farasi kuat. — The horse is strong. | Farasi lebi kabir dari kanis. — A horse is bigger than a dog. | | kanis | Kanis kimbia in hutan. — The dog runs in the forest. | Kanis keci lai una ta-ta. — A small dog comes with them. | | kalima | Kalima-kalima lai dari dunia. — The words come from the world. | Kalima no sakil in lingua ini. — A word is not heavy in this language. | | ponte | Ponte in sungai. — The bridge is at the river. | — Ponte? — A bridge? | | rafiki | Rafiki-rafiki suda lai. — The friends came. | Rafiki eski hao. — An old friend is good. | | insan | Insan-insan hao sini. — The people are good here. | Insan hao. — A good person. | A learner reading those pages meets the same noun under two English words and is told nothing. That is what this page is for. The count read **seven** until September 5, 2026 — the number of rows in the table, mistaken for the number in the corpus. Forty is the figure, and it makes the page's case rather than weakening it: the material has leaned on articlelessness five times harder than the page claimed. [`check.py`](../check.py) recounts it, and it counts a noun whose dictionary gloss is a single word and which some gloss renders "the X" and some "a X". ## Where the world's languages sit | | | |---|---| | No article at all | Russian, Turkish (definite), Hindi, Indonesian, Japanese, Chinese, Swahili | | Definite only | Arabic *al-*, Hebrew *ha-* | | Both | English, French, German, Spanish | More people learn to speak without articles than with them, and the ones who have them do not agree on how many. Every learner of English has to be taught that it is *the sun* and *a house*, and no rule explains it. Leaving them out costs nothing a sentence cannot recover. ## Saying it precisely when it matters Nothing was added for this. Three settled rules already do the work: | | | | |---|---|---| | dom **ini** | this house | [demonstratives](demonstratives.md) | | dom **mi** | my house | [possession](possession.md) | | **uan** dom | one house | [numbers](numbers.md) | ## Open questions - **Is anything needed when context does not settle it?** *Mi kan dom* is "I see a house" and "I see the house" and there is no way to force one. Three workarounds exist — *ini*, an owner, a number — and each says more than "definite". Whether that is a gap or the point is undecided. Recorded September 3, 2026, after finding seven nouns translated both ways with no page saying why. --- # Demonstratives — this and that *Taught in [Lesson 15](../lessons/lesson-15-pointing-placing.md).* *Status: settled — decided September 3, 2026.* ## The rule ***ini* is this, *itu* is that. They follow the noun, and they come last.** | | | |---|---| | dom ini | this house | | anak itu | that child | | ca ini garam | this tea is hot | | dom mi kabir ini | this big house of mine | The order inside a noun phrase is now complete: **noun → owner → adjective → this/that.** Each piece is optional; the order never changes. ## Standing alone *ini* and *itu* can stand by themselves wherever a noun phrase stands. Before a noun they need *es*, like any subject: | | | |---|---| | Ini es dom mi. | This is my house. | | Itu es ke? | What is that? | | Ini hao, itu mal. | This is good, that is bad. | | Mi mau ini. Mi no mau itu. | I want this. I do not want that. | | Mi mau kupi ini. | I want to buy this. | This section said **the subject** until September 5, 2026, and three sentences were already using them as objects — two of them in [the phrasebook](../phrasebook.md), where wanting this and not that is one of the lines a traveller actually needs. Nothing was wrong with the sentences: a rule page had described half of what its own rule does. The last two rows are those sentences, so the page now shows both slots. ## The pattern you already knew | this | here | that | there | |---|---|---|---| | **ini** | s**ini** | **itu** | s**itu** | *sini* and *situ* have been in the language since Lessons 8 and 10. *ini* and *itu* are what is left when the *s-* comes off. Indonesian built them the same way, and Amadunia inherits the set whole. ## Today, tonight *din ini* is "this day" — today. *rat ini* is tonight. *din itu* is that day. No new words for any of them. ## Sources Indonesian and Malay *ini / itu*; Tagalog *ito*. Austronesian, some three hundred million speakers, and the same family that gave the language *kita*, *sini*, *situ*, *anak* and its plural rule. ## Why this and not the alternatives **Why after the noun.** Every other modifier in the language follows its noun — the owner, the adjective. A demonstrative before the noun would be the one modifier out of place. **Why last in the phrase.** Because it points at the whole thing. *dom mi kabir ini* is "this [big house of mine]"; the pointing wraps everything else. | Candidate | Source | Reason rejected | |---|---|---| | bu / su | Turkish | *bu* is a minimal pair with *du*, *ba*, *yu* | | to | Russian *to*, Japanese *sore* | *to* is a minimal pair with *ta* | | na | Chinese *nà* 那 | minimal pair with *ba*, *ca*, *no*, *ta*, *ya* | | ili | Russian *ili* (or), Swahili *ile* | *ili* is a minimal pair with *ini* | | este / ese | Spanish | one family; and *ese* would sit one sound from *es* | | yah / vah | Hindi | the alphabet has no *v* | | kore / sore | Japanese | clean — neither collides with anything, and *sore* is two sounds from *sol*, not one — but one family | ## A note on checking *dom mi kabir ini* is a noun phrase and *ini* is last in it. But *Dom ini kabir* — "this house is big" — has *kabir* **outside** the phrase, as the sentence's predicate. The two look identical to a machine, so [`check.py`](../check.py) cannot test the rule in general; it tests only the case that has no predicate reading, an owner after *ini* or *itu*. An audit on September 3, 2026 found no violation of either kind. ## Open questions - A third distance (*that over there*, Indonesian *sana*, Japanese *are*) is not provided. *situ* covers it for now. --- # Negation — saying "not" *First met in [Lesson 01](../lessons/lesson-01-greetings.md) as an answer; taught as a rule in [Lesson 14](../lessons/lesson-14-not-and.md).* *Status: settled — decided September 3, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## The rule **Put *no* in front of what you are denying.** It is the same *no* you already answer questions with. Nothing else changes. | | | |---|---| | Mi kula pan. | I eat bread. | | Mi **no** kula pan. | I do not eat bread. | | Mi **no** es doktor. | I am not a doctor. | | Dom **no** kabir. | The house is not big. | | Mi **no** sini. | I am not here. | *no* goes immediately before the predicate — the verb, the adjective, the place word, or *es*. ## With tense *no* comes first, then the tense particle, then the verb. Negation covers the whole thing: | | | |---|---| | Mi **no suda** kula. | I did not eat. | | Yu **no saufa** lai. | You will not come. | | Dom **no suda** kabir. | The house was not big. | | Ta **no saufa es** doktor. | She will not be a doctor. | Never *suda no*. The order is always **no → tense → verb**. ## Answering is unchanged *no* keeps its first job exactly as it was: > — Yu mau ca? > — No. > — Yu mau ca? > — No, mi no mau ca. Mi mau akua. One word, two jobs, no new form for either. ## Sources *no* comes from the deepest and most widely shared negation on Earth — the root behind Spanish and Italian *no*, English *no* and *not*, Portuguese *não*, Latin *non* and *ne*, Russian *ne*, Polish *nie*, Serbian *ne*, Greek, Sanskrit *na*, Hindi and Urdu *na* and *nahī̃*, Bengali *na*, Persian *na-*, Nepali *na*. Well over three billion people already carry some form of it. Using the answer word as the negator is what Spanish, Italian and Portuguese do: *No.* / *No vengo.* Nothing is borrowed that the language did not already have. ## Why this and not the alternatives **Why not *na*.** *na* is the same root and was the first candidate. It is a minimal pair with five settled words — *no*, *ta*, *ya*, *ba*, *ca*. The collision with *no* is the fatal one: a negator one sound away from the word for "no", meaning almost the same thing, would be misheard constantly and the mistake would never be visible. The collision with *ta* is nearly as bad, since negation stands right beside pronouns. **Why reuse *no* instead of adding a word.** It costs nothing, it is already in the founder's first word list, and it is the same root every rejected candidate was reaching for. A language whose whole claim is "no exceptions" should not carry two words for one idea. **Why *no* before the tense particle.** Negation is about the whole statement, not about the time. Putting it outermost means the rule can be stated once — *no* goes in front of what you deny — and never needs a second clause. | Candidate | Source | Reason rejected | |---|---|---| | na | Sanskrit, Hindi, Persian | minimal pair with *no*, *ta*, *ya*, *ba*, *ca* | | ne | Latin, Russian, Polish, French | minimal pair with *ke* and *no* | | la | Arabic *lā* | minimal pair with *ta*, *ya*, *ba*, *ca* | | bu | Chinese *bù* 不 | minimal pair with *du*, *ba*, *yu* | | nai | Japanese *nai* ない | minimal pair with *bai*, *fai*, *lai*, *nau* | | non | Latin, French, Italian | no collision, but a second word for a job *no* already does | | tidak | Indonesian | one family, and two syllables for the commonest word in any language | That row said **three** syllables until September 6, 2026. *tidak* has two by the rule this language counts by — [one syllable per vowel group](stress.md) — and the overstatement was on the half of the reason that is a measurement rather than a judgement. The verdict does not move: one family is the other half, and it is why *no* was chosen. But a rejection is a record, and a record that overstates the cost of what it turned away is worth less than one that does not. ## Open questions - **Contrastive ellipsis — "X, not Y" — has no form.** *no* goes before the predicate, and the predicate is a verb, an adjective, a place word or *es*. It never goes before a noun. Five sentences were writing *X, no Y* anyway, dropping the verb out of the second half the way English does: | Where | Was | Now | |---|---|---| | [Lesson 19](../lessons/lesson-19-home-and-nature.md) | *…korku dari animal kabir, no dari kucing* | a second clause, with its verb | | [Lesson 20](../lessons/lesson-20-colours-and-health.md) | *Mi mau kamisa asul, no kamisa merah* | *Mi mau kamisa asul. Mi no mau kamisa merah.* | | [Lesson 21](../lessons/lesson-21-things-and-ideas.md) | *Anahtar mi in kamra, no in bage* | two sentences | | [the phrasebook](../phrasebook.md) | *Mi mau ini, no itu* | *Mi mau ini. Mi no mau itu.* | | [demonstratives.md](demonstratives.md) | the same sentence | the same | Every one of them is longer now, and a phrasebook line that a traveller actually says is longer than it needs to be. That is the argument **for** deciding this, and it is the founder's to make: a rule that lets *no* stand in front of a repeated constituent would give all five back at no cost in words. Nothing here decides it. A sixth sentence was a different fault and is simply fixed: [Lesson 22](../lessons/lesson-22-doing-and-feeling.md) said *No korku* for "no fear", which needs the existential — *No es korku*, there is no fear. Found September 5, 2026. The last of them had been promoted **into a rule page the day before**, when *Mi mau ini, no itu* was cited on [demonstratives.md](demonstratives.md) as evidence that a demonstrative can be an object. It is evidence of that. Nobody had looked at its *no*. - ~~Answering a negative question~~ — settled: *ya* and *no* answer the **fact**, never the wording. *Yu no mau ca? — No* means I do not want it; *Ya* means I do. This is the English, Spanish, Hindi, Arabic, Turkish and Indonesian system, by far the majority. When it matters, answer with the verb: *Mi mau* / *Mi no mau*. - **Nothing, nobody, never.** Not yet designed. Whether they are built from *no* plus an existing word, or given their own roots, is open. --- # Numbers *Taught on [the front page](../README.md) rather than in a lesson, because there is nothing to teach — twelve words and no exception. [Lesson 05](../lessons/lesson-05-plural.md) is where they are first used and now says where they came from. That arrangement is why six, eight and nine sat in no sentence in the whole course until [Lesson 21](../lessons/lesson-21-things-and-ideas.md) and [Lesson 24](../lessons/lesson-24-the-table-and-the-city.md) were given sentences for them.* *Status: settled for 1–10, 100, 1000.* ## Digits | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |---|---|---|---|---|---|---|---|---|----| | uan | du | tri | pat | fai | sis | seti | ba | nau | des | ## Units above ten | Value | Word | |---|---| | 10 | des | | 100 | sen | | 1000 | mila | **Twelve, not eleven.** This page opened with *eleven words and no exception* until September 5, 2026, and so did [Lesson 05](../lessons/lesson-05-plural.md); [the front page](../README.md) said eleven number words *reach a thousand*, which eleven of them cannot — *mila* is the twelfth and it is the one that reaches it. Eleven was right until *mila* was settled, and the sentence that carried it was never the sentence that changed. The [dictionary](../dictionary/dictionary.md) still says eleven and is still correct: it is counting the words that reach **a hundred**, and that ceiling did not move. One number, two ceilings, and only the pages that named the higher one were wrong — which is why finding it took reading all four rather than grepping for the figure. [`check.py`](../check.py) counts both from the glosses now. ## Building numbers There are no irregular numbers. Larger numbers are built from *des*, *sen* and *mila* by two positions, and the position is what carries the arithmetic: - **A digit before a base multiplies it.** *du-des* is twenty, *tri-sen* is three hundred. - **A digit after a base adds to it.** *des-uan* is eleven, *du-des-uan* is twenty-one. So *du-des* and *des-du* are different numbers — twenty and twelve — and the only thing telling them apart is which side of *des* the *du* stands on. A multiplier of one is never written: *uan-des* would be a second spelling of *des*, and there are no second spellings here. This page used to say "largest unit first", which contradicts its own table: *du-des* puts the smaller unit first, because two is multiplying ten rather than being added to it. Corrected September 3, 2026, when [`check.py`](../check.py) was taught to parse these forms and the rule it needed turned out not to be the rule that was written down. **Teens** — *des* followed by the digit: | 11 | 12 | 19 | |---|---|---| | des-uan | des-du | des-nau | **Tens** — the digit followed by *des*: | 20 | 30 | 40 | 50 | 60 | 70 | 80 | 90 | |---|---|---|---|---|---|---|---| | du-des | tri-des | pat-des | fai-des | sis-des | seti-des | ba-des | nau-des | **Tens plus a digit** — join with a hyphen: - 21 = *du-des-uan* - 45 = *pat-des fai* **Hundreds** — the digit followed by *sen*, then the remainder as a separate word: - 345 = *tri-sen pat-des fai* ## Open questions - **A number cannot count a pronoun.** *The three of us* has no form. Numbers go before what they count — *tri anak* — and a pronoun takes neither that shape nor the reverse: *tri kita* is not a construction, and *kita tri* would read as possession if it read as anything. Found on September 5, 2026 in [text 19](../texts/text-19-kamra-mi.md), which had written *Kita tri sini* since the day it was written and had it read back as "the three of us are here" by everyone including its own line-by-line table. The line is *Kita sini* now. The invention was a **shape**, not a word, which is why nothing caught it: every check the repository had was looking for words that do not exist, and every word in that sentence exists. - **Ordinal numbers have no form.** *First, second, eighth* cannot be built. [Text 6](../texts/text-6-seti-din.md) is the clearest case: a story that runs on days — ill seven days, medicine eight days — wanted *on the eighth day* and could only say *for eight days*. Nothing in the cardinal system suggests a derivation, and inventing one would be the first affix in the language. Listed in the index since it was first noticed and recorded here, where the decision would live, on September 3, 2026. **It blocks more than a sentence.** The language has no name for any weekday and none for any month — two closed sets missing entire. A language built on economy would number them rather than name them, as Chinese does with *week-one* for Monday, and that would cost no roots at all. It cannot, because numbering the second day is naming an ordinal. [Measured with the other closed sets](../dictionary/proposal-a2.md) on September 4, 2026: seven weekday names and twelve month names is nineteen roots that numbering would make unnecessary, which no other open question on the list can say. [The briefing is written](proposal-ordinals.md). - ~~Words for 1000~~ — settled: *mila*. Above a thousand is still open. - Whether the separator between groups is a hyphen or a space is used inconsistently above (*du-des-uan* vs *pat-des fai*, both inherited from the README) and needs a single rule. --- # Phonology *Taught on [the front page](../README.md) and in [Lesson 01](../lessons/lesson-01-greetings.md).* *Status: settled.* ## Alphabet Twenty letters, each with exactly one sound. No digraphs, no diacritics. | | Letters | |---|---| | **Vowels** (5) | a e i o u | | **Consonants** (15) | b c d f g h k l m n p r s t y | `c` is always the sound of *chai*, *church* — as in *ca* (tea), *cok* (many), *keci* (small). Never *k*, never *s*. This follows Indonesian, Malay, Hausa and Somali, which all write that sound with a plain *c*. (Until September 2, 2026 it was written *ch*; the digraph was dropped because it was the alphabet's one exception to "one letter, one sound".) `y` is always a consonant glide (*ya*, *yu*), never a vowel. ## One letter, one sound There are no silent letters, no letters that change value by position, and no letter that can be read two ways. A word is pronounced exactly as it is written, and written exactly as it is pronounced. ## Syllable shape Syllables stay simple: never more than two consonants in a row, and none of the tongue-twisters other languages allow (*strengths*, *vzglyad*). If you can say "banana" and "taksi", you can say anything in Amadunia. Two-consonant sequences that occur in current vocabulary: - `tr` — *tri* (3) - `ks` — *taksi* (taxi), across the syllable break *tak-si* - `kt` — *doktor*, across the syllable break *dok-tor* - `rb` — *corba* (soup) - `rg` — *harga* (price) - `rk` — *porke* (why), *market*, across the syllable break *por-ke* - `rs` — *mersi* (thank you), across the syllable break *mer-si* - `rd` — *pardon*, across the syllable break *par-don* - `ns` — *insan* (person), across the syllable break *in-san* - `pl` — *plis* (please), word-initial like *tr* - `fr` — *fruta* (fruit), word-initial like *tr* - `mb` — *ambil* (to take), across the syllable break *am-bil* - `mr` — *kamra* (room), across the syllable break *kam-ra* - `nd` — *anda* (to walk), across the syllable break *an-da* - `ng` — *angin* (wind), *cang* (long) - `bl` — *problema* - `br` — *libro* (book), across the syllable break *li-bro* - `gr` — *grupo* (group), word-initial - `hr` — *cehra* (face) - `ht` — *anahtar* (key) - `lc` — *dulce* (sweet) - `fl` — *flor* (flower), word-initial - `ft` — *hafta* (week) - `lb` — *kalb* (heart) - `lt` — *kultura* - `mp` — *tempo* - `nc` — *pencere* (window) - `nk` — *banka* - `nl` — *yanlis* (wrong) - `nt` — *kanta* (to sing) - `ny` — *nyama* (meat), *tanya* (to ask), *punya* (to have) - `pr` — *problema* - `rp` — *korpo* (body) - `rt` — *start* - `sf` — *asfar* (yellow) - `sk` — *eski* (old), *skola* (school), across the syllable break and word-initial - `sm` — *asman* (sky) - `sp` — *espera* (to wait), across the syllable break *es-pera* - `st` — *studi* (to learn), word-initial like *tr* and *pl* ## No two words a sound apart **Settled: no two roots of the same length may differ in exactly one letter. Between *l* and *r* the ban is absolute and outranks everything.** One letter is one sound here, so a minimal pair is a pair of words a single sound apart. *beli* and *beri* would be such a pair, and telling them apart is the hardest contrast on Earth for well over a billion speakers — Japanese, Korean, and much of southern China and East Africa. A world language cannot rest a distinction on it. This is the rule that kills most candidates. *beli* (to buy) fell to *beri*, *guru* to *suru*, *luna* to *luma*, *kota* to *kita*, *buku* to *buka*. It also outranks how widely a word travels: *ruma*, a house, was in use eighty-four times before it was replaced by *dom*, because *ruma* stood against *luma*. The eighty-four was recounted on September 5, 2026 and it is exact. The repository as it stood at the commit before the rename holds eighty-four occurrences of *ruma*, spread across twenty-one files — ten of them in Lesson 06 alone. ### The twenty-four that were already here The rule was adopted after the language had started, so twenty-four pairs predate it and are exempt. They are listed in [`check.py`](../check.py), which rejects any pair that is not among them. What the exempt pairs have in common is not where they came from but how short they are: **23 of the 24 are under four letters** — *ba/ca*, *go/no*, *hi/mi*, *kan/pan*, *pat/rat* — and *mama/nama* is the only exception. Short words cannot avoid each other; there are not enough of them. That is the same fact that closed the space below four letters in [CONTRIBUTING rule 2](../CONTRIBUTING.md), and the two rules are really one observation seen from two sides. Not one of the twenty-four is *l* against *r*, and check.py tests the exempt list for that too — an exemption may excuse a pair, never that pair. ## Vowel sequences **Settled: these five sequences and no others, and never three vowels in a row.** Both halves are rules, not observations — the page listed them as "observed in current vocabulary" while [`check.py`](../check.py) rejected anything else, and until September 3, 2026 the checker enforced only the first half. *kuai* is legal as two pairs, *ua* then *ai*, and illegal as a run of three; it is the example [CONTRIBUTING rule 4](../CONTRIBUTING.md) names. | Sequence | Example | |---|---| | ai | *bai*, *fai*, *kaifa* | | ao | *hao* | | au | *nau*, *mau* | | ia | *dunia* | | ua | *uan*, *akua* | The `ua` row read *uan*, *una* until September 5, 2026. *una* has no `ua` in it — u-n-a — so the table that settles which vowel sequences exist illustrated one of them with a word that does not contain it. It was the only one of the nine in this table that failed, and it had been there since the table was written. Every example on this page — in this table and in the consonant list above — is checked against the sequence it illustrates now. ## Open questions - ~~Whether a vowel pair is one syllable or two~~ — settled in [stress.md](stress.md) on September 3, 2026, together with the stress rule that needed it: **a vowel pair is one nucleus**, so *dunia* is two syllables and *familia* three. Counting a syllable means counting vowel groups. It affected 28 of the 300 roots. - ~~Whether stress is fixed (and if so, on which syllable)~~ — settled in [stress.md](stress.md) on September 3, 2026: **the second-to-last syllable of every word**. This bullet outlived the decision by a day, and the day is the point: [the briefing it points at](proposal-stress.md) had said *Decided* at the top of it the whole time. - Whether the two-consonant sequences above are the complete permitted set, or merely the ones used so far, is not yet decided. --- # Place and role — in, dari, por *Taught in [Lesson 15](../lessons/lesson-15-pointing-placing.md), after [Lesson 13](../lessons/lesson-13-weather-directions.md) found it missing.* *Status: settled — decided September 3, 2026.* ## Three small words | | | | |---|---|---| | **in** | at, in, on | Mi **in** dom. — I am at home. | | **dari** | from | Mi lai **dari** market. — I come from the market. | | **por** | to, for | Mi beri pan **por** yu. — I give bread to you. | Each goes **before** its noun. *in* turns a noun into a place; *dari* into a source; *por* into a recipient or a purpose. ## Going somewhere needs nothing After a verb of motion — *go, lai, anda, kimbia* — the place that follows is the destination. No word is added: | | | |---|---| | Mi go market. | I go to the market. | | Ta lai dom. | She comes home. | | Anak kimbia market. | The child runs to the market. | This has been true since Lesson 2. It is now the rule: **a motion verb's object is where it is going.** Chinese does exactly this — *wǒ qù shìchǎng*, "I go market". ## In | | | |---|---| | Anak lala in dom. | The child sleeps in the house. | | Kita kula in market. | We eat at the market. | | Sol in dunia. | The sun is in the world. | Place words — *sini, situ, upar, sub, kiri, yamin* — are already places and take no *in*. *Mi sini*, never *mi in sini*. ## Dari | | | |---|---| | Ca ini dari dugu mi. | This tea is from my sibling. | | Yu dari nali? | Where are you from? | | Akua dari yuki. | Water from snow. | ## Por | | | |---|---| | Ca ini por kim? — Por mi! | Who is this tea for? — For me! | | Ta rabota por familia ta. | She works for her family. | | Mi kupi pan por anak-anak. | I buy bread for the children. | *por* was already in the language, folded inside *porke* — "for what", why. Now it stands on its own. **Why giving needs *por*.** *Mi beri pan dugu mi* would read, by the possession rule, as "I give my sibling's bread". *por* keeps the recipient apart from the owner: *Mi beri pan por dugu mi*. ## With *una* means "together". Put a noun after it and it means "with": | | | |---|---| | Kita kula una. | We eat together. | | Kita kula una familia. | We eat with the family. | | Mi go market una rafiki mi. | I go to the market with my friend. | No new word. *una* is Latin *ūnā* — the founder's word — and in Latin it already meant both "together" and, with *cum*, "together with". ## Where these phrases go **Place first, then time, both at the end of the sentence.** | | | |---|---| | Mi kula pan in dom. | I eat bread at home. | | Mi kula pan din ini. | I eat bread today. | | Mi kula pan in dom din ini. | I eat bread at home today. | | Kita saufa lai dari market rat ini. | We will come from the market tonight. | This settles a question every lesson since 7 had left open: *sini, situ, sasa, kab, porke, kaifa* all go last, and now the rule says so. ## Sources | Word | Source | |---|---| | in | Latin *in*, English *in*, German *in*, Italian *in*, Dutch *in* — one of the few words shared unchanged across the whole Indo-European west | | dari | Indonesian and Malay *dari* | | por | Spanish and Portuguese *por*, *para*; Italian *per*; French *pour*; Latin *pro* | Three words for at, from and to is the floor. Indonesian has exactly these three (*di, dari, ke*); so does Turkish (*-da, -dan, -a*), Arabic (*fī, min, ilā*), and Toki Pona (*lon, tan, tawa*). Fewer than three and a language leans on its verbs; Amadunia does that for "to" and no further. ## Why prepositions, not suffixes Amadunia is subject-verb-object with modifiers after the noun. Languages built that way — Indonesian, Swahili, Thai, Vietnamese, Arabic, Spanish — put their place words *before* the noun, and languages built the other way — Turkish, Japanese, Hindi — put them after. The language follows its own shape. ## Why this and not the alternatives Nearly every short candidate collided: | Candidate | For | Reason rejected | |---|---|---| | di | at (Indonesian) | minimal pair with *du*, *hi*, *mi* | | fi | at (Arabic) | minimal pair with *mi*, *hi* | | ni | at (Japanese) | minimal pair with *hi*, *mi*, *no* | | da | at (Turkish) | minimal pair with *ba*, *ca*, *du*, *ta*, *ya* | | en | at (Spanish) | minimal pair with *es*, *in* | | me | in (Hindi) | minimal pair with *mi*, *ke* | | min | from (Arabic) | minimal pair with *ain*, *din* | | de | from (Spanish) | minimal pair with *du*, *ke* | | se | from (Hindi) | minimal pair with *ke* | | ot | from (Russian) | minimal pair with *ok* | | a | to (Spanish) | 69 roots end in *-a*; *ca a mama* blurs | | ko | to (Hindi) | minimal pair with *go*, *ke*, *no* | | li | to (Arabic) | minimal pair with *hi*, *mi* | | pada | to (Indonesian) | minimal pair with *papa* | | kwa | with (Swahili) | the alphabet has no *w* | | kon / kum | with (Spanish, Latin) | minimal pairs with *kan*, *kim* | | to | with (Japanese) | minimal pair with *go*, *no*, *ta* | | sat | with (Hindi) | minimal pair with *pat*, *rat*, *sar* | ## Open questions - **Whether a time word may open a sentence.** The order this page settles puts place first and time after it, both at the end, and grants fronting to one thing only: a subordinate clause, with a comma. **Eight sentences front a time word instead**, and two of them are lessons: | Where | Sentence | | |---|---|---| | [Lesson 20](../lessons/lesson-20-colours-and-health.md) | *sasa ta hao* | now she is well | | [Lesson 22](../lessons/lesson-22-doing-and-feeling.md) | *sasa kita go* | now we go | | [text 16](../texts/text-16-berapa-harga.md) | *Besok mi saufa tabak samaki una mama mi* | tomorrow I will cook fish with my mother | | [text 17](../texts/text-17-tren-aur-farasi.md) | *Besok mi saufa go dom mi* | tomorrow I will go to my house | | [text 20](../texts/text-20-in-skola.md) | *Sasa mi faham* | now I understand | | [text 20](../texts/text-20-in-skola.md) | *Besok mi saufa go skola tena* | tomorrow I will go to school again | | [text 21](../texts/text-21-uan-umur.md) | *Sasa anak faham* | now the child understands | | [text 26](../texts/text-26-dom-eski.md) | *Sasa dom itu no es dom mi* | now that house is not my house | Every one of them is a contrast — *now*, as against before; *tomorrow*, as against today — and English, Chinese, Indonesian and Turkish all front a time word for exactly that. Nothing here grants it and nothing here forbids it either: the rule says where time *goes*, not that it may go nowhere else. That is the same silence that covered the imperative until September 3 and [a tense marker before an adjective](tense.md) until yesterday. Found September 6, 2026 by reducing every sentence in the material to its parts of speech and reading the shapes that occur once — the method that found the four sentences below. The most recent of the eight is [text 26](../texts/text-26-dom-eski.md), written two days ago, which used the shape without noticing it was not granted. [`check.py`](../check.py) holds the count at eight while the question is open. - **A place word cannot attach to a noun, and cannot take one.** *sini, situ, upar, sub, kiri, yamin* are whole place expressions and go last, so there is no way to say **the people here**, **the soup there**, or **below the house**. Four sentences had been writing those anyway, and this rule's own page did not notice because the sentences are made entirely of real words: | Where | Was | Wanted | Now | |---|---|---|---| | [definiteness.md](definiteness.md) | *Insan-insan sini hao* | the people **here** are good | *Insan-insan hao sini* — the people are good here | | [text 7](../texts/text-7-surat-por-mama.md) | *Insan-insan sini hao* | the same | the same | | [text 4](../texts/text-4-in-madina-baru.md) | *Corba situ paling hao in madina* | the soup **there** is the best | *Corba paling hao in madina* | | [text 21](../texts/text-21-uan-umur.md) | *Es bage sub dom* | a garden **below the house** | *Es bage sub* — a garden below | The first three put a place word in front of the predicate, where the rule above says place goes last. The fourth is the other half of the same gap: it used *sub* as a preposition, and the language has three prepositions — *in*, *dari*, *por* — none of which means above or below. **The gloss moved with it.** *Ponte in sungai* was translated "the bridge is **over** the river" on this rule's own illustration and in Lesson 24. *in* is at or in; over has no word. Both now say "at the river". Found September 5, 2026 by counting every adjacent pair of word classes in the repository and reading the rare ones. Nothing in the language was invented to close the gap. **A fifth sentence, and a different shape.** [Text 24](../texts/text-24-mama-mama-mi.md) tried to say that a house stood on a road near a river, with *dekat* taking the river after it, and the checker refused the line. *dekat* is not a place word at all; it is an adjective, so a noun in front of it is described by it and a noun after it has nothing to attach to. The first four sentences wanted a place word to take a noun; this one wanted an **adjective** to, which the language allows even less. The text now states the road and the river in two sentences and leaves the relation between them to the reader. Found September 6, 2026. Every gap this repository has found by writing has been a missing word. This one is a missing **slot**: *in*, *dari* and *por* are the whole set, and a language that can say a road and a river cannot say one is beside the other. Whether the set should grow is the founder's call and nothing here proposes a candidate — [the ones already rejected are in the table above](#why-this-and-not-the-alternatives). - ~~Existence~~ — settled in [copula.md](copula.md) on September 3, 2026: *es* at the front of a sentence, with no subject, means "there is". - **Does a time expression take *in*?** The rule above says place first, then time, and stops there — it never says whether "when" carries a preposition. The material has settled into two forms without anyone deciding on one: | Form | Count | Examples | |---|---|---| | *in* + a bare time-of-day noun | 6 | *Mi kula **in pagi*** · *Mi lai **in aksam*** · *Mama mi sema masal por mi **in rat*** | | the noun bare, carrying *ini*, *besok* or *kemarin* | 20+ | *rat ini* · *din ini* · *besok pagi* · *kemarin aksam* | Read together they look like one pattern: a time word that points at a particular time stands alone, and a plain one needs *in* to become "when". That is a rule nobody wrote, and it may simply be the shape English gave the hand that wrote these sentences. One case does not fit either form — *Kita saufa kula una rat*, we will eat together at night, where *una* may be doing the work instead. Deciding this either blesses *in* for time or takes it out of six sentences. - **With an object, the two jobs of *una* collide.** Alone after a verb it is "together" — *Kita kula una* — and before a noun it is "with". Put an object in the sentence and the reading flips: *Kita kula una pan* is "we eat with bread", never "we eat the bread together", so **there is no way to say that at all**. Found on September 4, 2026 by [text 13](../texts/text-13-kula-una.md), which was written for *una* and had to keep every *together* sentence free of an object. Whether the answer is a position rule, a second word, or nothing, is open. - *Through, without, about, before, after, **near*** — the second tier of prepositions — are for later. **near** was asked for again on September 4 by [text 12](../texts/text-12-tarik-por-skola.md), which could not write *the school is near the tree* and said *the school is on this road* instead; *dekat* is an adjective and putting it in front of a noun would be inventing a preposition. It was found on September 3, 2026: three pages were writing *dekat ponte*, near the bridge, as though *dekat* were a preposition. It is an adjective — *ponte dekat* is "the near bridge" — and the prepositions are *in*, *dari* and *por*, so "near the bridge" has no form. All three were rewritten around it. [Text 9](../texts/text-9-pagi-in-madina.md) is the first page to want one: *Mi anda por tarik*, I walk along the road, borrows *por* for a route, and nothing here grants that. The rule settles direction and location and says nothing about a path. --- # Plural *Taught in [Lesson 05](../lessons/lesson-05-plural.md).* *Status: settled — decided September 2, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## The rule **To make a noun plural, say it twice, joined by a hyphen.** Nothing is added, nothing changes inside the word. | Singular | Plural | English | |---|---|---| | anak | anak-anak | children | | dom | dom-dom | houses | | rafiki | rafiki-rafiki | friends | | din | din-din | days | That works on every noun in the language, with no exceptions, because there is nothing that could go wrong: no ending to choose, no sound to change, no list of irregulars. ## After a number, the noun stays single If the sentence already says how many, the noun is not doubled. | | | |---|---| | tri anak | three children | | du dom | two houses | | cok rafiki | many friends | The number has done the work; repeating it would be saying the same thing twice. This is the same principle as tense: once the information is there, it is not restated. (*Anak-anak* on its own means "children, more than one, count unspecified".) ## Where plurals are not needed The unmarked noun is neutral, not singular. *Mi kan anak* means "I see a child" **or** "I see children" — whichever the situation makes obvious. Doubling is used when the plural actually matters and nothing else in the sentence shows it. ## Sources Doubling a word to mean "more than one" is one of the most widespread devices on Earth: | Family | Example | |---|---| | Austronesian | Indonesian/Malay *anak-anak* (children), *rumah-rumah* (houses); Tagalog | | Sino-Tibetan | Chinese *rén-rén* 人人 (everyone), *tiān-tiān* 天天 (every day) | | Japonic | Japanese *hito-bito* 人々 (people), *yama-yama* 山々 (mountains) | | Dravidian | Tamil, Telugu — distributive and plural reduplication | | Afro-Asiatic | Somali, Hausa — plural by partial reduplication | | Turkic | Turkish keeps nouns single after a number (*üç çocuk*, never *üç çocuklar*) — the second half of this rule | Keeping the noun single after a numeral is itself shared by Turkish, Hungarian, Persian, Chinese, Japanese, Korean and Indonesian. ## Why this and not the alternatives **Why not a suffix (-s, -ler).** A suffix is easy while it is regular, and in every natural language that has one it stops being regular: English *child/children*, Spanish *lápiz/lápices*, Arabic broken plurals. Doubling cannot drift because there is no form to drift. **Why not a particle (like *suda* for tense).** A plural particle placed before the noun would behave exactly like an article, and design rule 3 says no articles. Placed after the noun it would be one more word to learn and one more position rule; doubling needs neither. **Why not "no plural at all" (the Chinese model).** Tempting, and it is why the unmarked noun is neutral here. But learners coming from languages with obligatory plurals need *some* way to say "more than one" when it matters. Doubling gives them one without making it compulsory. **Candidates rejected, and why:** | Candidate | Source | Reason rejected | |---|---|---| | -s | English, Spanish, French, Portuguese | Indo-European only; breaks rule 4 for a feature this basic | | -ler | Turkish | one family; and as a separate particle it has no meaning to a non-Turkish speaker | | men | Chinese *-men* 们 | in the source it works only on people and pronouns — borrowing it as a general plural would misrepresent Chinese | | tachi | Japanese *-tachi* たち | same problem: people only | | duo (many) | Chinese *duō* 多 | introduces the vowel sequence *uo*, not attested in the language | ## Open questions - **All, some, none.** *cok* covers "many" and stands before the noun — *cok rafiki*. There is nothing for *all* or *some*. **"None" may already be covered**: the existential negation settled in [copula.md](copula.md) gives *No es ca*, there is no tea, which is what "none" usually means in a sentence — so the question is narrower than its name, and is really about *all*, *some* and *every*. Recorded here, where the decision would live, on September 3, 2026. **One page has since tried.** This entry read "nothing written so far has tried to say either" until September 5, 2026, and [text 19](../texts/text-19-kamra-mi.md) had already recorded the want — *Mi beri akua por flor in pagi*, which says "in the morning" where somebody waters plants **every** morning — and cited *this page* while doing it. The page it pointed at said no page had pointed. It is also the same want as the last question on this list: *din-din* for "every day" is the distributive sense of doubling, and if doubling carries it, a doubled *pagi* is already the answer to text 19 and *every* needs no word at all. Two questions, one sentence, and neither knew about the other. - ~~Pronouns~~ — settled: *mi-mi*, *yu-yu*, *ta-ta* are confirmed; see [pronouns.md](pronouns.md). - Whether doubling may also carry the distributive sense some source languages give it (*din-din* = "every day"), or stays strictly plural. --- # Possession *Taught in [Lesson 06](../lessons/lesson-06-possession.md).* *Status: settled — decided September 2, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## The rule **The owner comes right after the thing owned. Nothing is added.** | Amadunia | English | |---|---| | dom mi | my house | | nama yu | your name | | anak ta | her child, his child | | rafiki kita | our friend | | dom rafiki | the friend's house | | nama anak | the child's name | No possessive pronouns to learn (*my, your, his*), no suffix, no "of". The ordinary pronoun does the job, and so does any ordinary noun. ## Where it sits The owner takes the slot immediately after the noun — the same slot an adjective takes. When both are present, the owner comes first, then the adjective: | Amadunia | English | |---|---| | dom mi kabir | my big house | | anak ta keci | her small child | | rafiki kita hao | our good friend | *dom mi* is a unit: "my house". Whatever follows describes that unit. ## Chains An owner can itself have an owner. Each one follows what it owns, and the chain can go as long as it needs to: | Amadunia | English | |---|---| | dom rafiki mi | my friend's house — *house (of) friend (of) me* | | nama anak rafiki yu | your friend's child's name | Read left to right, each word belongs to the one before it. No exception at any depth. ## In a sentence | Amadunia | English | |---|---| | Mi kan dom yu. | I see your house. | | Anak ta suda nomu akua. | Her child drank water. | | Rafiki mi saufa lai. | My friend will come. | | Kita bil nama doktor. | We know the doctor's name. | ## Sources Placing the owner straight after the noun, with no marker, is how a large share of the world already does it: | Family | Example | |---|---| | Austronesian | Indonesian/Malay *rumah saya* (house I), *nama dia* (name he) | | Semitic | Arabic *bayt-u r-rajul* (house the-man), Hebrew construct state | | Iranian | Persian *khāne-ye man* (house-of me) — with a linking vowel, same order | | Niger-Congo | Swahili *nyumba yangu* (house my) — owner after noun | | Celtic | Welsh *tŷ Siôn* (house John), Irish | | Tai-Kadai, Austroasiatic | Thai *bâan khǎw* (house he), Vietnamese *nhà tôi* (house I) | | Romance | *la casa de Juan* — noun first, with *de* | It also matches what Amadunia already does with adjectives: modifiers follow the noun. ## Why this and not the alternatives **Why no marker.** Every marker candidate with real reach collides with a word already settled: Chinese *de* 的 is a minimal pair with *du* (2) and *ke* (what); Japanese *no* の is already "no"; Swahili *ya* is already "yes". That three independent candidates all collide is a sign the language does not want a marker here — and juxtaposition needs none. **Why owner before adjective.** The owner is the most identifying information ("whose house"); adjectives describe. Arabic and Indonesian, the two largest noun-first models, both bind the owner tighter to the noun than the adjective. **Why not separate possessive pronouns (*my, your*).** A second pronoun set doubles what a learner must memorise and, in every language that has one, drifts away from the base set (English *I/my*, Spanish *yo/mi*). Reusing the ordinary pronoun means there is nothing extra to learn and nothing that can drift. **Candidates rejected, and why:** | Candidate | Source | Reason rejected | |---|---|---| | de (particle) | Chinese *de* 的, Romance *de* | minimal pair with *du* and *ke* — and two letters, where the space is closed | | no (particle) | Japanese *no* の | already "no" | | ya (particle) | Swahili *-a* class agreement | already "yes" | | -im / -in (suffix) | Turkish, Hungarian, Arabic | inflection on the noun — against the spirit of rule 3 | | owner first (*mi dom*) | Chinese, Japanese, Turkish, Hindi, English | would be the only modifier placed before the noun — one exception to an otherwise clean pattern | ## Open questions - ~~Attributive vs predicate~~ — settled in [copula.md](copula.md): a noun predicate takes *es* (*rafiki mi es doktor*), which keeps it apart from possession (*rafiki mi doktor* = my friend's doctor); adjectives take nothing. - **Alienable vs inalienable.** Some languages distinguish "my hand" from "my house". Amadunia currently does not; whether it ever should is open. --- # Pronouns *Taught in [Lesson 09](../lessons/lesson-09-pronouns.md).* *Status: settled — decided September 2, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## The set | | one | more than one | |---|---|---| | **I** | mi | **mi-mi** — we, not you | | | | **kita** — we, with you | | **you** | yu | **yu-yu** — you all | | **he, she, it** | ta | **ta-ta** — they | Six forms. Four of them you already had; the other two are the plural rule doing its ordinary job. ## Four things a pronoun never does here **It never changes for case.** *Mi kan ta. Ta kan mi.* — I see her; she sees me. The same word whether it acts or is acted on. (English *I/me*, *she/her*; Amadunia *mi*, *ta*.) **It never changes for gender.** *ta* is he, she and it. Whether the tea, the doctor or the child — *ta*. The English translations cannot say that, and every one of them has to choose. **159 glossed sentences in the repository contain *ta*, and 144 of them assign a gender the sentence does not have** — 85 she, 56 he, 3 both, 12 *it*. That is the shape of a language with no gender being read in one that has it, and it is not a fault in any single line. The split was 69 she to 41 he until September 5, 2026, when twelve glosses were changed for the reason below. It was a fault in seven pages. On September 5, 2026 the material was counted and [Lesson 06](../lessons/lesson-06-possession.md), [Lesson 17](../lessons/lesson-17-wanting-and-able.md), [Lesson 18](../lessons/lesson-18-comparing-and-joining.md), [Lesson 22](../lessons/lesson-22-doing-and-feeling.md), [Lesson 25](../lessons/lesson-25-doing-and-being.md), [subordination.md](subordination.md) and [verb-chains.md](verb-chains.md) each glossed *ta* three to five times and **only ever as a woman**. A learner reading any of them in order met *ta* as feminine and nothing else — which is the one thing this rule says it is not. [possession.md](possession.md) had been doing it correctly all along, glossing *anak ta* as "her child, his child"; Lesson 06, which teaches the same phrase, said "her child" and then said her four more times. Every one of those pages shows both now, and [`check.py`](../check.py) holds them to it. The texts are exempt and stay as they are: a character in a story keeps one gender, and [text 18](../texts/text-18-gusa-in-tarik.md) has a driver, a police officer and a child taking *he*, *she* and *he* in nine lines, which is the point rather than an inconsistency. Whether the glosses should instead use singular *they* everywhere is a real question and nobody has decided it. It would say what the language says, at the cost of 112 rewrites and of English prose that reads oddly in a beginner's lesson. Recorded, not decided. **It never changes for politeness.** One *yu* for a friend, a stranger, a child, a president. (Spanish *tú/usted*, Hindi *tu/tum/aap*, Indonesian *kamu/anda* — Amadunia *yu*.) **It pluralises like everything else.** Say it twice: *yu-yu*, *ta-ta*, *mi-mi*. The rule from [plural.md](plural.md), applied without exception. ## Two ways to say "we" This is the one place the pronoun set has more than the bare minimum, and it is there because the founder's first word list already contained it: *kita — we, including you*. | | means | built from | |---|---|---| | mi-mi | me and others — **not** you | *mi* doubled: several of me | | kita | me and others — **and** you | its own word | The form tells you the meaning. *mi-mi* is made only of *mi*, so it cannot contain *yu*. *kita* is the word for the group that has both of us in it. | | | |---|---| | Mi-mi go market. Yu sini? | We're going to the market. You staying here? | | Kita go market! | Let's go to the market — all of us. | | Mi-mi suda kula. Yu-yu? | We've eaten. Have you all? | | Kita suda kula. | We've eaten — you and I both. | ## In use | Amadunia | English | |---|---| | Ta-ta suda lai. | They came. | | Yu-yu mau ca? | Do you all want tea? | | Dom ta-ta kabir. | Their house is big. | | Ca garam. Mi mau ta. | The tea is hot. I want it. | | Rafiki mi-mi es doktor. | Our friend is a doctor. | | Kita saufa kan sol. | We will see the sun. | With possession, nothing new: *dom mi-mi* (our house), *anak yu-yu* (your children — all of yours), *nama ta-ta* (their names). ## Sources | Form | Source | |---|---| | mi | Spanish/Italian *mi*, English *me*, Swahili *mimi* | | yu | English *you* | | ta | Chinese *tā* 他 / 她 / 它 — one sound for he, she and it | | kita | Indonesian/Malay/Tagalog *kita* — we, including the listener | | doubling for plural | colloquial Indonesian *kamu-kamu* (you all), *kita-kita* (us lot) | The inclusive / exclusive split (*kita* / *mi-mi*) is shared by Indonesian, Tagalog, Tamil (*nām / nāṅkaḷ*), Vietnamese (*chúng ta / chúng tôi*), northern Mandarin (*zánmen / wǒmen*), Hausa, Quechua and most of the Pacific. A single genderless third person (*ta*) is how Turkish, Persian, Hindi, Finnish, Hungarian, Yoruba, Swahili and spoken Chinese all work. ## Why this and not the alternatives **Why doubled pronouns, not new words.** Every big language with a *regular* pronoun plural does it with a marker on the singular: Chinese *wǒ → wǒ-men*, Japanese *watashi → watashi-tachi*, Vietnamese *tôi → chúng tôi*. Amadunia's plural marker is doubling, so doubling it is. It costs no new words, and a learner who knows *ta* and the plural rule already knows *ta-ta*. **Why keep the two "we"s.** Because *kita* was glossed "we, including you" from the first day; a contrast that is named must have a partner. And the partner comes free: *mi-mi* is what the plural rule produces anyway. Dropping the distinction would have meant either changing the founder's gloss or leaving *mi-mi* meaning the same as *kita* — a synonym with no job. **Why *ta* covers "it".** One third person for everything is the majority pattern by speakers and the only one with no exception. Adding a separate "it" would reintroduce a human / non-human split — a gender by another name. **Why no formal "you".** Politeness levels are the single most common thing learners get wrong in languages that have them, and the most-learned language on Earth does without them entirely. **Candidates rejected, and why:** | Candidate | For | Reason rejected | |---|---|---| | kami | we (exclusive) | not needed — *mi-mi* already fills the slot with no new word | | hum | they | is "we" in Hindi/Urdu — a pronoun with the opposite reference for 600M people | | nos | we | three letters, where the space is closed at 49 roots; it is a near miss for *no* rather than a minimal pair, which needs equal length | | oni, mereka, kalian, antum | they / you all | each adds a word only one family recognises, to do a job the plural rule already does | | a separate "it" | it | would reintroduce a human / non-human distinction | | a formal "you" | you | politeness levels are an exception generator | ## Open questions - ~~Demonstratives and *there*~~ — settled: *ini* and *itu* in [demonstratives.md](demonstratives.md), *situ* in Lesson 10. - **Reflexives** (*myself*) and **reciprocals** (*each other*) are not yet designed. - Whether *mi-mi* and *ta-ta* may ever be shortened in fast speech is a question for later, once there is speech. --- # Pronunciation *Taught on [the front page](../README.md) and in [Lesson 01](../lessons/lesson-01-greetings.md).* *Status: written down September 3, 2026. Most of it was already decided and had never been said.* ## What this file fixes Amadunia's second design rule is **one letter, one sound**. Until now the language had named that sound for exactly two of its twenty letters — `c` and `y` — and left the other eighteen, including all five vowels, to be guessed. Nothing here is a new decision. Three hundred roots with sourced etymologies already commit the language to these values: *ca* comes from Chinese *chá*, *dunia* from Arabic and Swahili, *kita* from Indonesian, *dom* from Russian. Every one of those languages reads these letters the same way, and the README's own test — "if you can say *banana*" — pins the vowels. This file reads the values off the vocabulary rather than choosing them. The one thing the evidence does not settle is at the bottom. ## Vowels The plain five. This is the most common vowel system on Earth — Spanish, Japanese, Swahili, Indonesian, Greek, Hausa and Turkish all use it, and a speaker of any of them already knows these five sounds. | | Sound | As in | Amadunia | |---|---|---|---| | **a** | /a/ | Spanish *casa*, Japanese *aka*, Swahili *baba* | *ama* | | **e** | /e/ | Spanish *peso*, Japanese *eki*, Indonesian *desa* | *beri* | | **i** | /i/ | Spanish *sí*, Japanese *ichi*, Swahili *nini* | *mi* | | **o** | /o/ | Spanish *como*, Japanese *oto*, Indonesian *toko* | *dom* | | **u** | /u/ | Spanish *luna*, Japanese *uma*, Swahili *kuku* | *yu* | They never change. There is no long and short, and **no reduced variant away from stress**: the *a* in *ama* and the *a* in *salam* are the same sound, and so are both unstressed *a*'s in a-ma-**DU**-nia. [Stress](stress.md) makes a syllable louder, not different — the way it works in Indonesian, Swahili and Japanese, and not the way it works in English, Russian or Portuguese, where an unstressed vowel collapses toward a schwa. If it collapsed here, *written as spoken* would stop being true the moment a word was said aloud. That rule read "no reduced variant under stress" until September 5, 2026, which names the wrong condition: a vowel that reduces reduces when it is *un*stressed. As written it promised something no language on Earth breaks, while every example under it — and the whole paragraph it belongs to — is about the unstressed syllable. [stress.md](stress.md), [the phrasebook](../phrasebook.md) and this page's own next sentence all had it the right way round. The page that defines the twenty sounds is the one that had it backwards, in the six bold words that state the rule. ## Consonants | | Sound | Note | Amadunia | |---|---|---|---| | **b** | /b/ | | *bai* | | **c** | /tʃ/ | as in *chai*, *church*. Never /k/, never /s/ | *ca* | | **d** | /d/ | | *din* | | **f** | /f/ | | *fai* | | **g** | /ɡ/ | **always hard** — as in *go*, never as in *gem* | *garam* | | **h** | /h/ | always pronounced, never silent | *hao* | | **k** | /k/ | | *kan* | | **l** | /l/ | | *lai* | | **m** | /m/ | | *mi* | | **n** | /n/ | | *no* | | **p** | /p/ | | *pan* | | **r** | — | see below | *rat* | | **s** | /s/ | **always voiceless** — as in *sun*, never as in *rose* | *sol* | | **t** | /t/ | | *ta* | | **y** | /j/ | a consonant, as in *yes*. Never a vowel | *yu* | ## Two that need saying out loud **`g` never softens.** In Italian, Spanish and French a *g* changes value before *e* and *i*. In Amadunia it does not, because no letter changes value by position — that is the rule. *genc* begins like *go*, not like *gem*. **`s` never voices.** English turns *s* into /z/ between vowels — *rose*, *music*. Amadunia does not. *pesa*, *risi* and *musik* all have the *s* of *sun*. Both follow from "one letter, one sound" and neither is obvious to a speaker whose own language does the opposite, which is exactly why they are written here. ## Aspiration, and other things that do not matter Whether *p*, *t* and *k* are said with a puff of air is not distinguished. *pan* said the English way and *pan* said the Spanish way are the same word. The language does not use aspiration to tell any two words apart, and it never will, because that would break the no-minimal-pairs rule. The same goes for whether *t* and *d* are said against the teeth or the ridge behind them. Say them the way your own language does. ## Open questions - **How `r` is made.** A tap (Spanish *pero*, Japanese, Turkish, Indonesian), a trill (Spanish *perro*, Italian), or an approximant (English)? The language has never said, and the source words point in several directions at once — *rat* is Indo-Aryan, *rabota* Slavic, *risi* Mediterranean, *rafiki* Swahili. The tap is the safest of the three: it is what Spanish, Japanese, Turkish, Indonesian, Swahili and Hindi speakers already do, and it is easier for everyone else than a trill. But it is a decision and it has not been made. Until it is, any *r* is understood — nothing in the language distinguishes two words by their *r*. - ~~Where the stress falls~~ — settled in [stress.md](stress.md) on September 3, 2026: the second-to-last syllable of every word. It does not change any sound on this page. --- # Proposal: a word for "but" **Nothing here is adopted.** The dictionary is closed at 300 roots and this page does not open it. It sets out the evidence, the candidates and the costs for a decision that is the founder's. *aur* joins and *o* chooses. Nothing contrasts, and [conjunction.md](conjunction.md) records that as open, with the honest possibility that the language may not need one. ## The evidence is six pages, and it is the strongest on the open list Every one of these is a sentence that stopped, not a line on a wordlist. | Page | The sentence | What it wanted | |---|---|---| | [Lesson 12](../lessons/lesson-12-colors-directions.md) | *— No, eski. Hao!* | no, old, **but** good | | [Text 14](../texts/text-14-yamur-aur-ca.md) | *Ca garam.* | the tea is hot **but** the night is cold | | [Text 16](../texts/text-16-berapa-harga.md) | *Nyama lebi eski.* | the meat is cheaper **but** older | | [Text 17](../texts/text-17-tren-aur-farasi.md) | *Tarik duan aur safari no muskil.* | the road is short **but** the journey is not easy | | [Text 18](../texts/text-18-gusa-in-tarik.md) | *Tarik baru duan.* | the new road is short **but** slow | | [Text 21](../texts/text-21-uan-umur.md) | *Pesa keci.* | the money is little **but** the bread is good — written as *lakin* twice in draft, by reflex | Lesson 12 is the one that matters most. It did not work around the gap — it **invented a word**, *ma*, and that word sat in the course unnoticed until September 4, 2026. A gap that makes a careful writer invent is a different kind of gap from one that makes a writer rephrase. Four of the six are a bargain, a meal or a road: ordinary speech, not argument. The sixth is the one that matters most after Lesson 12's: writing the longest text in the language, I typed *lakin* twice without deciding to, and the checker refused it both times. This paragraph said five of seven, and the seventh, until September 5, 2026 — the day one row was withdrawn from the table above it and the sentence reading that table was not. The contrast is not a literary device here. It is what people say about weather and prices. ## Candidates, run through the rules Every shape below was put through [`check.py`](../check.py)'s own tests — alphabet, four-letter minimum, consonant runs, attested vowel pairs, minimal pairs against all 300 roots, and the l/r ban. | Shape | Source | Verdict | |---|---|---| | **lakin** | Arabic *lākin*, Persian, Turkish *lakin*, Urdu/Hindi *lekin*, and Swahili *lakini* | passes everything | | lakini | the Swahili form of the same word | passes; one syllable and two letters longer | | tapi | Indonesian, Malay | passes; reaches one family | | tetapi | the formal Indonesian form | passes; six letters for one family | | ancak | Turkish | passes; reaches one family | | pero | Spanish, Portuguese | passes the rules, but stands one sound from *por*, which is a preposition in constant use | | demo | Japanese | passes the rules; rejected as a false friend — *demo* is an English word worldwide | | amma | Persian, Arabic | passes the rules; rejected — *ama* is the verb "to love", and a listener would have to hear a doubled consonant to tell a conjunction from a verb | | bali | invented shape, for comparison | **fails**: minimal pair with *nali*, "where" | **lakin reaches five families with one word** — Semitic, Iranian, Turkic, Indo-Aryan and, through *lakini*, Bantu. Nothing else on the list reaches more than one. That is unusual enough to be the whole argument. Two costs come with it, and both should be said plainly: - **The balance rule would count it as Semitic.** Origin is assigned by the first language named in the entry, so a word that actually reaches five families would be scored for one. That is a flaw in the measure, not in the word, and [balance.md](../dictionary/balance.md) would misreport it. - **It would be the longest word in its class.** *aur* is three letters and *o* is one; the four-letter minimum for new roots makes any new conjunction the odd one out. The rule exists because the short space is full, and this is the first case where it visibly costs something. ## What the rule would be Nothing new. *aur* stands between what it joins, and a contrast word would stand in the same place: | | | |---|---| | Tarik duan **aur** safari asan. | The road is short and the journey easy. | | Tarik duan **lakin** safari muskil. | The road is short but the journey hard. | No position rule, no new class, no change to any sentence already written. The whole cost is one root. ## The case for doing nothing [conjunction.md](conjunction.md) raises it and it deserves a fair hearing. Plenty of speech gets by with juxtaposition — *the road was short; the journey was hard* — and the six sentences above are all readable as they stand. A language that adds a word for every shade of English risks becoming a relexed English, which is the thing this project has refused from the start. Against that: **the invented *ma* in Lesson 12**. The other five pages rephrased; that one reached for a word that did not exist and put it in a lesson. That is what a real gap does. ## What is measured and what is judgement Measured: the six pages and their sentences, which `check.py` recounts; the candidates' pass or failure against the phonotactic rules; the family reach. Judgement, and the founder's: whether the language takes a contrast word at all, whether *lakin* is the shape, and whether it belongs in A1 as root 301 or waits for A2. --- # Proposal: where a frequency adverb stands **Nothing here is adopted.** This page measures what the material actually does, sets out the three answers and what each would cost, and leaves the decision to the founder. It matters more than its size suggests: *kadang* is one of only two roots in the dictionary that [cannot be written at all](../dictionary/proposal-a2.md), because any sentence using it would settle this question by accident. ## The record said one thing and the corpus says another [adverbs.md](adverbs.md) records that *daima* and *kadang* "sit in the adverb slot with everything else". Counted on September 4, 2026 across the lessons, the texts, the grammar files and the phrasebook, **two of thirteen do**. | Position | Uses | Examples | |---|---|---| | **before the verb** | 5 | *Mualim daima kara libro* · *Ta kadang lai* · *Mi kadang go in tren* | | **the adverb slot** | 2 | *Mi kula daima pan* · *Anak proba daima* | | **after the object** | 6 | *Mi nomu halib daima* · *Mi ingat yu daima* · *Anak-anak kula telur kadang* | Thirteen sentences, three positions, and one lesson — Lesson 19 — uses all three by itself. Nobody chose this; the hand followed whichever source language it was thinking in. One of the six depends on a question that is also open: *Kita rabota daima in skola* counts as "after the object" only because [*rabota* is a noun in the dictionary](verb-chains.md). Read it as a verb and that sentence moves into the adverb slot. Two open questions touching one sentence is worth knowing before either is answered. ## The three answers, costed | | Rule | Sentences to rewrite | Argument for | Argument against | |---|---|---|---|---| | **A** | the adverb slot: *Mi kula **daima** pan* | 11 of 13 | one rule for every adverb; [adverbs.md](adverbs.md) already says so | the most rewriting, and it puts the commonest use — the bare *Ta kadang lai* — in the position it is least often written | | **B** | before the verb, where the tense markers stand: *Ta **kadang** lai* | 8 of 13 | frequency is about time, and *daima kara* would sit where *suda kara* sits | a second word before the verb, so the order of *no*, *suda* and *daima* has to be settled too | | **C** | last in the sentence, like *tena*: *Mi nomu halib **daima*** | 7 of 13 | cheapest, and it matches the one neighbour that has already settled itself | the tail of a sentence is where [place and time](place.md) already go, and a fourth thing there is a crowd | ## *tena* settled itself and nobody noticed *tena*, "again", is not a frequency word — it counts repetitions rather than naming a rate — but it is the nearest neighbour in the dictionary and it has been consistent from the start: **32 of its 32 uses are last in the sentence**, without one exception. *Mi lai tena*, *Anak katab tena*, *Yamur no lai tena*, *Din lai tena*. No rule ever granted that. The writing simply never wavered, which is the opposite of what happened with *daima*, and it is the strongest evidence available for answer **C** — the language has already been doing it with the word next door. ## What is measured and what is judgement Measured: the thirteen sentences and their three positions, the cost of each answer, and *tena*'s 32 of 32 — [`check.py`](../check.py) recounts all of them, so this page cannot drift the way [adverbs.md](adverbs.md) did. That sentence was not true when it was written. The three positions and *tena* were recounted; the total and the three costs were prose, and a cost is the whole point of the page — each answer's price is every sentence that is not already standing where it would put them. All five figures are computed from the corpus as of September 5, 2026. They were right, which is luck: the last three days moved *insan genc* through Lesson 25 and rewrote five contrast sentences, and any of that could have touched one. Judgement, and the founder's: which position, whether *daima* and *kadang* behave alike, and whether *tena* is bound by the same answer or stays a separate thing that happens to sit in the same place. --- # Proposal: "want to be" plus an adjective **This is not a rule.** Nothing here has been adopted. It is the fourth of the briefings, after [stress](proposal-stress.md), [the imperative and the fragment](proposal-sentence-types.md) and [marking a name](proposal-names.md). With it, all four of the questions the language was already relying on are prepared. It was called the smallest of the four, on the strength of two pages having tripped over it. Measured, it is not small. Counted across the material — the twenty-six lessons, the thirty texts and the phrasebook — which is the scope every other count in this repository uses, and not this page, whose illustrations are the only *modal + adjective* strings in existence. | | | |---|---| | Amadunia sentences in the material | 1895 | | of those, verbless adjective predicates | **405 — one in five** | | modal followed by a verb, working normally | 59 | | modal followed by an adjective | **0, and none is possible** | | combinations with no legal form | **111** — three modals against 37 adjectives | The construction is not rare and the modals are not rare. They simply cannot meet. [`check.py`](../check.py) recounts every row from the material and the dictionary, so the figures follow the language rather than this page. The first three read 1481, 291 and 61 until September 5, 2026, and the fourth said "in any file" — which, read literally, is four, because this page prints *Ca mau garam* and *Mi mau senang* to argue that they cannot exist. Only the 111 was ever recounted. ## The gap *es* is a verb and chains like any other, so a noun predicate works: ``` Mi mau es doktor. I want to be a doctor. Ta saufa es rafiki yu. She will be your friend. ``` An adjective is its own predicate and takes no verb at all, so a modal has nothing to attach to: ``` Ca garam. The tea is hot. Ca mau garam. ✗ reads as "the tea wants heat" ``` "Wants to be hot" cannot be built. ## The asymmetry that explains it **Tense works with adjectives. Modals do not.** | | | |---|---| | Ca **suda** garam. | The tea was hot. ✓ | | Ca **saufa** garam. | The tea will be hot. ✓ | | Ca **mau** garam. | ✗ | *suda* and *saufa* are particles: they attach to whatever the predicate is, verb or adjective or place word. *mau*, *bisa* and *lasim* are verbs, and a verb needs a verb to chain to. That is the whole of the problem, and it is worth noticing that the language already contains the shape that solves it. ## Where it bit Twice, and both were written around rather than fixed: - [Text 3](../texts/text-3-lingua-ini.md) wanted *Amadunia mau es paling asan*, "wants to be the easiest". It was rewritten as *mau es lingua paling asan* — a noun was supplied so *es* had something legal to take. - Lesson 21 wanted *Legis lasim es benar*, "the law must be true". The modal was dropped: *Legis benar*. Both original forms put *es* in front of an adjective, which [copula.md](copula.md) forbids, and both were caught when every sentence in the repository was checked against that rule. ## The options **1. Allow *es* before an adjective inside a verb chain.** *Mi mau es senang.* Cost: an exception. "es never precedes an adjective" would become "es never precedes an adjective except after a modal", and design rule 1 is zero exceptions. It is the cheapest fix and the most expensive principle. **2. Supply a noun.** What text 3 does. Cost: nothing, but it only works when a suitable noun exists, and "wants to be happy" has none. **3. Let a modal take an adjective directly.** *Mi mau senang.* Cost: nothing structurally — the modal's complement becomes "a verb or an adjective". The risk is that *mau senang* also reads as "wants happiness", since an adjective can head a phrase. **4. Make *bisa* and *lasim* particles rather than verbs.** They would then sit where *suda* and *saufa* sit and attach to anything: *Ca lasim garam*, "the tea must be hot". **Neither ever takes a noun object** — they are pure modals — so nothing is lost. *mau* cannot join them, because *Mi mau pan* is real and must keep working. Cost: the language gains a second particle slot but no new word, and the verb-chain rule shrinks to cover *mau* alone. **5. A separate verb meaning "to become".** Turkish *olmak*, German *werden*, Spanish *ponerse* all do this and all take adjectives. Cost: a word, and a second thing that means nearly what *es* means. ## What a decision needs to say Which of the five; whether the answer also covers *no mau senang* (not wanting to be happy) and questions (*Yu mau senang?*); whether text 3 and Lesson 21 should be restored to what they were trying to say; and whether *punya*, which is a verb taking a noun, is affected at all. --- **All the briefings are now written** — four then, eight now. None decides anything. Together with [the ordering](README.md) they are what the founder needs to settle the four places where Amadunia has been using something it never granted itself. --- # Proposal: marking a name **This is not a rule.** Nothing here has been adopted. It is the briefing for the third of the four questions the language is already relying on without permission, after [stress](proposal-stress.md) and [the imperative and the fragment](proposal-sentence-types.md). ## The problem, measured Two dictionary roots are used as people's names throughout the material: **Sol** and **Luma**, which are also the words for *sun* and *light*. This page first said Sol was used as a name 33 times and Luma 18. Those figures cannot be reproduced, and the reason is the problem itself: separating a name-use from a sun-use is a judgement, not a measurement, and it is exactly the judgement no rule currently supports. Counting every occurrence instead gives Sol 51 and Luma 19. **Forty-eight sentences are formally ambiguous.** That number *is* reproducible, and [`check.py`](../check.py) now recounts it: a sentence whose first word is *Sol* or *Luma*, where every word in it is a dictionary word and nothing says which reading is meant. Ten are in each of the two stories, four in Lesson 12, three in Lesson 11, and the rest scattered across ten more files. It was thirty-nine until September 4, 2026, when [text 15](../texts/text-15-sol-lai.md) was written for Lesson 10 and opened with *Sol lai* — the sun comes, or Sol comes. The text was not written about this question and reached the ambiguity twice on its own, which is the strongest argument on this page: the collision is not in the old material, it is in the words. The forty-second arrived the same way. [Text 20](../texts/text-20-in-skola.md) had written *Hora du*, hour two, for noon; the language has no clock, so the line became *Sol upar* — the sun is up. Avoiding one open question walked straight into another, and no one chose either. | | Reading A | Reading B | |---|---|---| | Sol hao. | Sol is good. | The sun is good. | | Sol kabir. | Sol is big. | The sun is big. | | Sol asfar. | Sol is yellow. | The sun is yellow. | | Sol es doktor. | Sol is a doctor. | The sun is a doctor. | And the same string means different things in different texts. [Story 1](../texts/story-1-anak-espera-sol.md) writes *Sol beri luma* four times meaning **the sun gives light**; [Story 2](../texts/story-2-safari-por-pahar.md) writes *Sol sema* meaning **a child named Sol says**. Nothing in the language distinguishes them. ## Why capital letters are not the answer They are what the material uses today, and they work on a page. But design rule 2 is **written as spoken**, and a capital letter cannot be spoken. A listener hearing *sol hao* has nothing at all. This is the only open question where the status quo contradicts one of the five founding rules directly. ## Why it cannot be avoided by choosing better names The obvious dodge — say a name may not be an existing root — does not survive contact with the world. A world auxiliary language has to carry the name of anyone who has one, and people are called Sol, Luna, Kanta, Mira and Nadia. The dictionary has 300 roots today and will have a thousand; every name will eventually collide with something. **The problem gets worse as the language succeeds.** ## The options **A personal article — a small word before the name.** This is a real and widespread device, not an invention: Tagalog *si*, Portuguese *o/a*, Catalan *en/na*, Māori *a*, Fijian *o*, Ancient Greek, southern German. Roughly it would be *X Sol hao* — "Sol is good" — with *X* saying that what follows is a name. Cost: one word, and it must survive the minimal-pair rule, which has killed every two-letter candidate the language has tried. It would probably have to be three or four letters. **Capitals only — keep the status quo and say so.** Cost: nothing, but it leaves the language unspeakable at exactly the point where people introduce themselves, and it breaks rule 2. **Nothing — let context do it.** Chinese and English manage. But they manage partly because their name stock is recognisable as names, and Amadunia's is not: *Sol* is a perfectly ordinary word of the language. **A shape rule for names.** Names must end in a sound no root ends in, or take a fixed ending. Cost: it mangles the name, which is the one thing a person will not accept about their own name. ## What a decision needs to say Whether the marker appears with every name or only where ambiguity is possible; whether it also marks place names (*Amadunia* itself is a name and appears unmarked five times in [text 3](../texts/text-3-lingua-ini.md)); whether it is stressed; and what happens to the forty-eight sentences already written — most of which, on inspection, would simply gain one word. That last number was **thirty-six** on this page until September 5, 2026, while the paragraph six screens above it said thirty-nine and then forty-two. The page that exists to give the founder one reproducible figure was carrying two, and [`check.py`](../check.py) was holding only the first of them. It now holds every spelled count on the page to the same recount, which is what should have happened the day the headline was fixed. ## The other question still waiting *"Want to be" plus an adjective* has not been briefed yet. It is the smallest of the four: *es* is a verb and chains freely, but an adjective is its own predicate and takes no verb, so *mau* has nothing to attach to and "wants to be the easiest" cannot be built. Both [text 3](../texts/text-3-lingua-ini.md) and Lesson 21 hit it and were rewritten around it. Recorded in [copula.md](copula.md). --- # Proposal: first, second, eighth **Nothing here is adopted.** This page gathers what has been measured about ordinal numbers so the decision can be made on evidence. It is on this list because of a measurement, not a wish: [answering it would remove the need for nineteen roots](../dictionary/proposal-a2.md), which no other open question can say. ## The evidence | Where | What stopped | |---|---| | [Text 6](../texts/text-6-seti-din.md) | a story told in days — ill seven days, medicine eight days — wanted *on the eighth day* and could only say *for eight days* | | [Story 2](../texts/story-2-safari-por-pahar.md) | records it plainly: *"the first morning" and "the second morning" cannot be told apart from "one morning" and "two mornings"* | | the days of the week | **no name for any of them**, and none needed if a day can be numbered | | the months | **no name for any of them**, likewise | Seven weekday names and twelve month names is nineteen roots. A language built on economy would number them, as Chinese does — 星期一, week-one — and as Japanese does with its months. It cannot, because numbering the second day is naming an ordinal. ## The one candidate that is disqualified before it starts **An ending.** *-inci*, *-th*, *-ème*. Every natural language the dictionary draws on has one, and Amadunia cannot: it has **no affix at all** — no conjugation, no plural ending, no case — and design rule 2 is that no word changes shape. An ordinal ending would be the first affix in the language, and [the imperative was decided on exactly this ground](sentence-types.md): a distinct verb form was rejected outright for being an inflection. ## What is left | | How it would work | Cost | |---|---|---| | **Position** | a number **before** its noun counts, a number **after** it orders: *du din* is two days, *din du* is the second day | no new word, and no new kind of rule — the adjective, the owner and the demonstrative all follow their noun already, and the number is the only modifier that precedes | | **A particle** | one new root before the number, like *suda* before a verb | a root, plus a position rule, plus a word in every date anybody says | | **A word for each** | *first*, *second*, *third* as roots | unbounded; disqualified by the same argument that killed a plural ending | | **Nothing** | the ambiguity stays and context carries it | free, and it leaves nineteen roots to find, two closed sets empty, and story 2's complaint standing | **Position is the cheapest and it has a warning attached.** Two sentences in the corpus had a number after its noun — *Anak-anak du sini* and *Es hotel du sini* — and both meant a count, not an order. They were corrected on September 4, 2026 and [a check now refuses that shape](../check.py). If position is adopted, the check is what would have to change, and those two sentences are the evidence that a writer reaches for the slot by accident. The other warning is that position would carry meaning in a third way — after word order and after the copula's noun-or-not — and each of those is a thing a learner has to hold. ## What is measured and what is judgement Measured: the two texts that stopped, the two empty sets, the nineteen roots, the absence of any affix in the language, and the two accidental uses of the slot that position would need. Judgement, and the founder's: whether the language orders numbers at all, and if so whether position is the way. Nothing here decides it. --- # Proposal: the imperative and the fragment **Decided on September 3, 2026. Both rules are in [sentence-types.md](sentence-types.md):** a sentence that begins with a verb and has no subject is a command, and any constituent may stand alone as an utterance. This page is kept as the record of what the decision was made on. Both were already in use when it was written. Neither had ever been granted, and not one of the sentences counted below needed rewriting once they were. ## What is actually in use **The imperative.** This read "17 sentences" until September 5, 2026, and the figure added two unlike things. Seven were commands across Lessons 13, 16 and 19 and [the phrasebook](../phrasebook.md); the other ten were in [text 8](../texts/text-8-kaifa-suru-ca.md), a recipe written deliberately *without* an imperative to find out what that costs — every step of it says *yu* where the rest of the world says nothing. Those ten are the sharpest evidence on the page and they are not the imperative in use. They are the places it was needed and unavailable, which is the opposite claim, printed under a heading that says **what is actually in use**. Recounted, the two halves are these. Text 8 has **12** sentences that open with *yu* and a verb, not ten. And over the four files the seven was drawn from, as they stood on the day this page was written, a subject-less verb-initial sentence occurs **10 times** and says **8 different things** — *Anda yamin, Anda tarik ini, Buka ain yu, Funga libro aur otur, Go kiri, Go yamin, Lai kan ta, Sema tena*. Seven is neither. **And the block below prints five, under a sentence that called them seven.** It has printed five since the day the page was written. The two that are missing cannot be recovered: three of the eight above — *Anda yamin*, *Funga libro aur otur*, *Go kiri* — are absent from it, and nothing says which two of those three were meant. Recorded rather than repaired, and the sentence says five now, because five is what is there: ``` Anda tarik ini. Walk this road. Go yamin. Go right. Buka ain yu! Open your eyes! Sema tena, plis. Say it again, please. Lai kan ta. Come and see her. ``` The form in use is a **bare verb with no subject**. Lessons have written it that way since Lesson 10 and no rule anywhere permits it. **The fragment — 44 utterances** in dialogue and the phrasebook: *Ya, ok.* *No, bas.* *Corba, plis.* *Sasa.* *Besok.* *Madad!* *Ke?* *Harga?* ## A distinction the count had to be corrected for Counting fragments naively gives fifty-seven, and most are not fragments at all. A grammar lesson that prints *dom mi*, *tri anak* or *nama yu* in a table is **citing a phrase**, not claiming it is a sentence — the same way an English grammar prints "the big house". Only utterances in conversation are the real phenomenon, and there are 44 of them. That distinction matters for whatever rule is written: it should govern what may be *said alone*, not what may be *printed alone*. ## What the corpus scan counts that nobody said This page drew a distinction by hand — a grammar lesson printing *dom mi* in a table is citing a phrase, not claiming it is a sentence — and the scanner that produces every corpus figure in the repository has never known it. Measured on September 6, 2026: **21 runs come from explanatory prose** in the lessons and the phrasebook rather than from a table, a quoted line, a code block or a numbered practice item. They divide three ways. **Six are real** — the closing *Mi ama dunia. Yu ama dunia. Kita ama dunia.* of Lesson 23 and of the phrasebook, three sentences each. **Eleven are citations**: *Mi kula*, *yu kula*, *ta kula* in Lesson 02, *Pagi hao* and *Rat mal* in Lesson 03, *Tri hora* in Lesson 16, *Buka mulut* and *No go skola besok* in Lesson 26, and three more. **Five are fragments** cut by markdown emphasis and not phrases at all: *no es* comes from a line of English prose about the copula, and *kita kula sasa* is what is left when a quoted conditional wraps across two lines. **Three rules were tried and all three misclassify.** *Presented at the head of its line* drops the second clause of a legitimate practice item — *Din kabir, rat keci* loses *rat keci* — and the second and third sentences of the closing line above. *Wrapped in emphasis* drops that closing line entirely, since it is italic. *Unbalanced asterisks* finds 33 runs across the repository and among them the front page's fourteen-word full-order example, a real sentence whose emphasis merely wraps. A rule that costs real sentences to remove citations is worse than the residue, so none was adopted. Sixteen things nobody said, in 1799 — under one per cent. Recorded rather than corrected, and [`check.py`](../check.py) holds the count so it cannot grow quietly. This record lives on a grammar page because [texts/README.md](../texts/README.md), where it belongs by subject, is inside the material the scanner counts: writing these examples out there added ten more of the thing being measured, and the checker caught it. ## The imperative: the slot is free This is the same shape as two decisions already made. Existential *es* works because **Amadunia never drops a subject** — so a sentence beginning with *es* can only be existential. Verb chains work because verb-next-to-verb meant nothing before. In both cases the language took an empty slot for nothing. A sentence beginning with a **verb** is the same empty slot. Amadunia has no subject-dropping, so a bare initial verb cannot be anything else. Ratifying what is already written would cost no word and create no ambiguity. The alternatives all cost something: | Option | Who does it | Cost | |---|---|---| | Bare verb — what is in use | Chinese, Indonesian, English, Swahili | nothing | | A particle after the verb | Indonesian *-lah*, Japanese | a word, and a new position rule | | Reuse *plis* as the marker | — | makes every command polite, which is not the same thing | | A distinct verb form | Spanish, Arabic, Hindi | breaks "no word ever changes shape" | The last is disqualified outright: a separate imperative form would be the first inflection in the language. ## The fragment: what a rule would have to allow Every language permits fragments; none of the source languages requires a full sentence to answer a question. The options are about **how much** to permit: | Option | What it says | |---|---| | Any constituent may stand alone as an utterance | Describes what already happens. Costs nothing. | | Only answers and greetings may | Would make *Corba, plis* and *Harga?* illegal, and those are what a phrasebook is | | Nothing may — full sentences only | Would make the phrasebook impossible and no language does it | The first is almost certainly right, but it should be **written down** rather than left to work by accident, because a learner reading the grammar today finds nothing that lets them say *Sasa.* ## What a decision needs to say For the imperative: the rule in one sentence; that no subject appears; whether a subject *may* appear for emphasis (*Yu go!*); how negation works with it (*No go* — presumably, since *no* precedes the predicate); and whether *plis* has any grammatical role or is only polite. For fragments: which constituents may stand alone; whether a fragment answer inherits the question's tense (*— Yu suda kula? — Ya.*); and whether the existing conversations in the lessons need any rewriting, which on this evidence they do not. --- # Proposal: where the stress falls **Decided on September 3, 2026. The rule is in [stress.md](stress.md): the second-to-last syllable of every word, and a vowel pair is one syllable.** This page is kept as the record of what the decision was made on. One figure in it was understated. It says penultimate and initial agree on 209 of the 300 roots; they agree on **266 — 89%** — because the 57 one-syllable roots agree trivially and were left out of the count. ## Why this is first [The poem](../texts/text-5-uan.md) showed what the absence costs. Read its opening line stressing *DIN lai*, then *din LAI*: both are correct Amadunia and they are different poems. Two people cannot agree on the metre of a single line, so nothing can be sung, scanned, or taught by ear. Every other open question limits what can be *said*; this one limits what can be *spoken*. ## A prior question that has to go first You cannot say "stress the second-to-last syllable" until you have said what a syllable is. In 28 of the 300 roots — 9% — the count depends on whether an attested vowel pair is one nucleus or two: | | one nucleus | two | |---|---|---| | *dunia* | du-nia (2) | du-ni-a (3) | | *akua* | a-kua (2) | a-ku-a (3) | | *mausim* | mau-sim (2) | ma-u-sim (3) | | *familia* | fa-mi-lia (3) | fa-mi-li-a (4) | The affected roots are the ones containing *ai, ao, au, ia, ua*: *ain, akua, auto, bai, dunia, familia, fai, kimbia, lai, mau, mausim, mualim, naiti, nau, sungai, tair* and twelve others. **This is now recorded as an open question in [phonology.md](phonology.md).** It has to be settled first, or settled at the same time. ## What the dictionary looks like | Syllables | Roots | |---|---| | 1 | 57 | | 2 | **209** | | 3 | 34 | Counting vowel pairs as one nucleus. Seven in ten roots are two syllables. ## The options **Penultimate — the second-to-last syllable.** Esperanto, Swahili, Indonesian and Malay, Polish, Quechua. Two of Amadunia's three largest source families already do this, and so does the one constructed language that has actually been spoken for a century. **Initial — always the first.** Czech, Hungarian, Finnish, Icelandic, Latvian. **Final — always the last.** Turkish, Persian, French, Armenian, Hebrew. **None — leave it free.** What the language does today. It is why the poem has no metre. ## The finding that makes this easier than it looks **For 209 of the 300 roots, penultimate and initial are the same syllable.** A two-syllable word has its first syllable as its penultimate one. The two options can only disagree on the 34 three-syllable roots — *bahari, habari, kalima, kamisa, kemarin, madina, negara, numero, polisi, problema, restoran, safari, samaki, sapatu, telinga, uhuru* and the other eighteen. So the practical choice is between penultimate-or-initial on one side and final on the other, and that separates 243 of the 300 roots — every word of more than one syllable, 81%. Two corrections, September 5, 2026. The list above named *kucing* and *muskil* among the three-syllable roots and both have two syllables — *ku-cing*, *mus-kil* — by the definition this same page spends a section establishing. They are replaced by *kemarin* and *problema*, which are three. And the last sentence read *even that separates only a third of the vocabulary*: the choice between penultimate and final separates every root of two syllables or more, which is 81% of the dictionary and not a third. A third is nobody's count of anything here — the number beside it in the paragraph is 34, and 34 of 300 is a ninth. The sentence was the one that made the decision look cheap, and it was the one nobody checked. ## And it is free **No two words in Amadunia are told apart by stress, and none ever can be** — that would be a minimal pair, which the language forbids. Whatever is chosen, no existing word changes meaning and no text has to be rewritten. Nothing is at risk. That is exactly why it has been easy to defer for a day, and exactly why it should be settled now: the cost of choosing is nil and the cost of not choosing is that the language cannot be spoken in metre. ## What a decision needs to say Following the shape of the other files in this directory: the rule in one sentence; whether a vowel pair is one syllable or two; which languages already do it and how many people that is; the options rejected and why; and what it leaves open — clitics, whether *mi-mi* is stressed once or twice, and whether the grammar particles *es, no, aur, o, in, dari, por* carry stress at all. --- # Proposal: may a root hold two jobs? **Nothing here is adopted.** The design rule says one root, one job. Two roots break it, both inherited from before the rule was written down, and [verb-chains.md](verb-chains.md) records them as one open question. This page separates them, because **they are not the same decision** — one is free and one is expensive. ## *madad* — the decision costs nothing *madad* is glossed "help; to help". It has **never been used in a sentence**: not in a lesson, not in a text, not in the phrasebook. Counted across the whole material on September 4, 2026, the only line a scanner finds is the phrasebook's list of bare words separated by dots, which is not an utterance. [`check.py`](../check.py) recounts that, so the claim cannot go stale. That is unusual and it matters. Every other open question has sentences at stake. This one has none: **whichever way it goes, nothing has to be rewritten.** [Lesson 08](../lessons/lesson-08-getting-by.md) teaches the word and then deliberately shows no sentence containing it, which is why. ### The imperative rule, settled after this question was opened, tips it The cry *Madad!* is the word's commonest use in every source language. Since September 3, 2026, [a verb with no subject is a command](sentence-types.md) — so if *madad* is a verb, *Madad!* is already a grammatical sentence and needs no rule at all. As a noun it is a one-word fragment, which the same page also grants, so both readings survive the cry. What separates them is the other sentence. [Text 18](../texts/text-18-gusa-in-tarik.md) is a scene about a police officer helping a driver and cannot say so: *Polisi madad sofer* is the sentence the story is about. As a verb that line becomes writable. As a noun, "to help" needs a new root and the language has none. | If *madad* is | *Madad!* | *Polisi madad sofer* | What A2 must add | |---|---|---|---| | a verb | a command | writable | a noun for "help", if one is ever wanted | | a noun | a fragment | impossible | a verb for "to help" | **Two is now a measured figure and not an assumption.** Swept September 6, 2026: every root standing straight after a pronoun subject, which is the verb slot and nothing else — a noun beside a noun would be ambiguous with possession, and after a pronoun it cannot be. Only *rabota* comes back. The count had never been checked, and it was wrong three days earlier: *kanta* was doing both jobs in [text 21](../texts/text-21-uan-umur.md), where a doubled verb read as *sing a song*, and adjacency rather than the verb slot is what found it. [`check.py`](../check.py) now holds both shapes, so a third root cannot join *madad* and *rabota* without saying so. ## *rabota* — the decision costs twenty sentences *rabota* is glossed "work", one noun, and is used both ways. Recounted inside the material — the lessons, the texts and the phrasebook — **eleven places put it straight after a subject** (*Mi rabota sini*, *Ta-ta rabota*, *Kita rabota cok*) and **nine put a verb or a quantity in front of it** (*Mi suru rabota*, *Kita finis rabota*, *Mi suda naiti rabota baru*). Twenty in all, and the convention is places rather than sentences, because a cost is what would have to be rewritten and the same sentence printed twice is two rewrites. Counting each distinct sentence once instead gives ten and seven, seventeen in all; [`check.py`](../check.py) holds both figures. [verb-chains.md](verb-chains.md) reports thirteen and six over a wider scope that includes the grammar files; the figures differ by where the count is drawn, not by what the sentences say. **This heading said eighteen until September 5, 2026, and eighteen is not any reading of the material.** Every occurrence gives 11 and 9; every distinct sentence gives 10 and 7. The page had taken its eleven from the first count and its seven from the second and added them — which is how a number no method produces gets written down, and why the two conventions are now named rather than assumed. The nine are exactly the ambiguous ones. *Ta suda lasim rabota* is "she had to work" if *rabota* is a verb and "she had to have work" if it is a noun, and nothing on the page decides which. The costs are already measured in [verb-chains.md](verb-chains.md): read as a verb, one site breaks; read as the noun the dictionary calls it, ten do. That asymmetry is not an argument by itself — the cheaper repair is not the same as the right answer — but it is the number the decision has to face. ## The policy behind both Every root added since the rule was written has one job. These two are the inheritance, not the practice. Three ways out: | | What it means | |---|---| | **Split them** | each root takes one job, and the other job waits for A2 or goes without | | **Allow two jobs where position disambiguates** | *Mi rabota* is a verb because a subject precedes it; *suru rabota* is a noun because a verb does. This is what English does and it works — but it makes position carry meaning in a second, new way, and the language has spent 300 roots keeping one job per shape | | **Leave it open** | the current state: the words are taught and the sentences avoided. It has cost one word its entire use | ## What is measured and what is judgement Measured: *madad*'s zero sentences, *rabota*'s split above, the one-site against ten-site repair cost, and the fact that every root added since has one job. `check.py` recounts the first two. Judgement, and the founder's: whether a root may ever hold two jobs, and if not, which job each of these two keeps. --- # Questions *Taught in [Lesson 07](../lessons/lesson-07-questions.md).* *Status: settled — decided September 2, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## Yes / no questions: rising tone, written with ? A yes/no question is a statement said with a rising tone and written with a question mark. Nothing is added, nothing moves. | Statement | Question | |---|---| | Yu mau ca. | Yu mau ca? | | Dom yu kabir. | Dom yu kabir? | | Ta suda lai. | Ta suda lai? | This is what the language has done since its first sentence (*Hi! Ok?*). It is confirmed, not changed. ## Question words: stay where the answer goes To ask *what, who, where, when, why, how, how many*, put the question word **in the exact slot the answer would occupy**. Word order does not change. | Question | Answer | | |---|---|---| | Yu kan **ke**? | Yu kan anak. | What do you see? — You see a child. | | **Kim** lai? | Rafiki lai. | Who is coming? — A friend is coming. | | Yu go **nali**? | Mi go dom. | Where are you going? — I'm going home. | | Yu saufa lai **kab**? | Mi saufa lai din. | When will you come? — I'll come by day. | | Yu mau ca **porke**? | — | Why do you want tea? | | Yu go **kaifa**? | Mi go taksi. | How do you go? — I go by taxi. | | Yu kan **berapa** anak? | Mi kan tri anak. | How many children do you see? — I see three. | Read the question and the answer side by side: they are the same sentence with one word swapped. That is the whole rule. ## The seven words | Word | Meaning | Source | |---|---|---| | ke | what | Spanish *qué*, Portuguese *que*, Italian *che*, French *que* | | kim | who | Turkish, Azeri, Uzbek, Kazakh *kim* | | nali | where | Chinese *nǎlǐ* 哪里 | | kab | when | Hindi/Urdu *kab*, Bengali *kobe* | | porke | why | Spanish *por qué*, Portuguese *por que*, Italian *perché*, French *pourquoi* — all literally "for what" | | kaifa | how | Arabic *kayfa* كيف | | berapa | how many, how much | Indonesian/Malay *berapa* | Six families in seven words. Each was chosen to be recognised on sight by the largest possible number of people while no one family supplies more than two. ## Two details **berapa behaves like a number.** It sits where a number sits — before the noun — and, like a number, leaves the noun single: *berapa anak*, never *berapa anak-anak*. This follows from the plural rule; nothing new. **A question word needs no rising tone.** *Kim lai?* is already a question because *kim* is in it. The question mark is written; the voice may rise or not. ## Sources for the two rules *Rising tone for yes/no* is the primary or a fully standard method in Spanish, Portuguese, Italian, Russian, Hindi, Indonesian, Persian, Hebrew, Greek and every Arabic dialect — most of the world outside East Asia. *Question word in place* (no fronting) is how Chinese, Japanese, Korean, Turkish, Hindi, Indonesian, Tamil, Persian and Swahili ask: *ni kan shénme?* (you see what?), *sen ne gördün?* (you what saw?), *kamu lihat apa?* (you see what?). By speaker count it is the majority pattern on Earth. English, Romance and Arabic move the question word to the front instead; that movement is a rule with its own exceptions in every language that has it, and it is the one thing this design leaves out. ## Why this and not the alternatives **Why no yes/no particle.** Every particle with real reach collides with a word already settled — and not just one or two of them: | Candidate | Source | Collision | |---|---|---| | ma | Chinese *ma* 吗 | minimal pair with *ba*, *ca*, *mi*, *ta*, *ya* | | ka | Japanese *ka* か, Korean *-kka* | minimal pair with *ba*, *ca*, *ke*, *ta*, *ya* — five at once | | mi | Turkish *mi* | is already "I" | | ne | Latin *-ne*, Japanese *ne* | minimal pair with *ke*, *no* | | mai | Thai *mái* | minimal pair with *bai*, *fai*, *lai*, *mal*, *mau* | | ba | Chinese *ba* 吧, Tagalog *ba* | is already 8 | | ni | Korean *-ni*, Yoruba *ni* | minimal pair with *hi*, *mi*, *no* | | kya | Hindi *kya* | doubles as "what"; two jobs for one word | | hal | Arabic *hal* | minimal pair with *hao*, *mal* — good and bad at once | | apa | Indonesian *apa(kah)* | minimal pair with *ama* | Ten candidates from ten families, ten collisions. As with possession, the language is saying it does not want a word here — and it does not need one, because rising tone is already attested from the first lesson and is standard across most of the world. A tonal-language speaker who finds rising intonation awkward can rely on the question mark in writing and on context in speech, exactly as they do in colloquial Mandarin. **Why in-place, not fronting.** Fronting changes word order, and Amadunia has exactly one word order. In-place keeps it. It is also the majority pattern by speakers and the easier one to learn: the learner builds the statement they already know and swaps one word. **Candidates for the seven words that were rejected:** | Candidate | Meaning | Reason rejected | |---|---|---| | ne (Turkish) | what | minimal pair with *ke*, *no* — including the *ke* that was chosen for the same job | | che (Persian) | what | not spellable: *c* is already the *ch* sound and the alphabet has no digraph. Written legally as *ce* it is a minimal pair with *ca* and *ke* | | mana (Indonesian) | where | minimal pair with *mama*, *mano* | | do (Japanese) | how | minimal pair with *du*, *go*, *no* | | kak (Russian) | how | minimal pair with *kab*, *kan* | | kam (Arabic) | how many | minimal pair with *kab*, *kan*, *kim* | | kitne (Hindi) | how many | **reason withdrawn — see below** | | kuando / kuanto (Spanish) | when / how many | contain *uan* (1); and would be a minimal pair with each other | | kapan (Indonesian) | when | ends in *pan* (bread) | | itsu (Japanese) | when | sound, not sense: recognised by ~125M against *kab*'s ~700M | *kab* and *berapa* each share a first syllable with a settled word (*kabir*, *beri*). Neither is a minimal pair — the words differ in length and stress — and both were kept for their reach. ### The collisions were recounted on September 6, 2026 Both tables above were recomputed against the dictionary, one candidate at a time. **Two rows were exact** — *ka* against five, and *ce* against *ca* and *ke*. **Eight named one collision where there were two, three or five**: *ne* collides with *ke* as well as *no*, *mai* with *bai*, *fai*, *lai* and *mal* as well as *mau*, *kam* with *kab* and *kim* as well as *kan*. Every row names all of them now. Two of the eight were understated in a way that mattered more than arithmetic. *hal* was rejected for being "close to *hao*" and *apa* for sitting "inside *papa*" — a resemblance and a substring, neither of which is a rule here. Both are outright minimal pairs, *hal* with *hao* and *mal* both, and *apa* with *ama*. The decisive reason was available and a softer one was written down. And one row's reason does not hold at all. **"*kitne* begins with *kita*"** — it does not: *kitne* is k-i-t-n-e and *kita* is k-i-t-a. *kitne* has no minimal pair in the dictionary and contains no root, so the page now offers no reason for rejecting it. That is left as a gap in the record rather than filled in: *berapa* is settled and nothing here reopens it, but the reason a candidate was turned away is the part of a decision worth keeping, and this one was never true. The most pointed of the corrections is the smallest. The Turkish *ne* was rejected for colliding with *no*; it also collides with *ke*, the word this page chose for the same job. Both are "what", one letter apart. **It is not only this page.** The check written for these two tables was pointed at every rejection table in `grammar/`, and it found the same habit in six more: *ga* in [tense.md](tense.md) named *go* and collides with five; *ki* in [subordination.md](subordination.md) named *ke* and collides with three; [place.md](place.md) under-named nine rows of fifteen. Of the **51 such rows in `grammar/`, 23 across seven pages named one collision where the dictionary held more**, and none of them named a collision that was not real — the reasons were true and short, never wrong. Every row is complete now, and the checker requires the whole set rather than a member of it. ## Open questions - **May a question word stand in a modifier's slot?** The swap rule produces it without anyone deciding: the answer is *Libro ini*, this book, the unknown is the *ini*, and swapping it gives *Libro ke?* — which [Lesson 16](../lessons/lesson-16-school-and-time.md) writes and glosses "Which book?". English does the same with *what book*, so nothing looks wrong, and the language has no word for *which*. But this page describes the question word as standing where an **answer** stands, and *ini* is a modifier rather than an answer. One sentence in the whole course does it. Found September 5, 2026 while checking that a question word used as a noun predicate takes *es*; that check exempts the two-word case for exactly this reason, and if the answer is no, *Libro ke?* has to be rewritten. - ~~**The imperative was never granted.**~~ — granted September 3, 2026 in [sentence-types.md](sentence-types.md): a sentence that begins with a verb and has no subject is a command, and that is also what tells it from a statement. *Anda tarik ini*, *Buka ain yu*, *Otur sini* stand unchanged. Found by [the fourth text](../texts/text-4-in-madina-baru.md). - ~~**Fragments.**~~ — settled the same day in [sentence-types.md](sentence-types.md): any constituent may stand alone as an utterance. *Ke?* *Harga?* *Ponte?* are legal by rule now rather than by accident, and a fragment answer inherits the tense and subject of the question. - ~~Adverb position~~ — settled in [place.md](place.md): place first, then time, both at the end. *kab, porke, kaifa* stay where they are. - **Lesson 6 asked *Nama yu?*** with no question word. That is a yes/no-shaped question that context turns into "what is your name". The explicit form is *Nama yu es ke?* Both are allowed; whether the short form should be discouraged in teaching is open. --- # Grammar Twenty-one rules and eight briefings, one decision each. **Six briefings are open**: [a mark for a name](proposal-names.md), ["want to be" plus an adjective](proposal-modal-adjective.md), [a word for "but"](proposal-but.md), [where a frequency adverb stands](proposal-frequency.md), [whether a root may hold two jobs](proposal-two-jobs.md) and [first, second, eighth](proposal-ordinals.md). Two are closed: [stress](proposal-stress.md) and [the imperative and the fragment](proposal-sentence-types.md), both decided on September 3, 2026. Every page records what was settled, why, which candidates were rejected and on what ground, and what it left open. ## The rules | | | |---|---| | [phonology](phonology.md) | Twenty letters, one sound each. No digraph, no diacritic. At most two consonants in a row, never three vowels, and no two roots a single sound apart. | | [stress](stress.md) | The second-to-last syllable of every word. A vowel pair is one syllable. | | [sentence-types](sentence-types.md) | A verb with no subject is a command. Any constituent may stand alone. | | [pronunciation](pronunciation.md) | What each of the twenty letters sounds like. *g* never softens; *s* never voices. | | [numbers](numbers.md) | Eleven words count to a hundred. No irregular number anywhere. | | [tense](tense.md) | *suda* past, *saufa* future, present unmarked — before the verb. The verb never changes. | | [plural](plural.md) | Say the noun twice. Single after a number. The unmarked noun is neutral. | | [possession](possession.md) | The owner follows the thing owned, with nothing added. Owner before adjective. | | [definiteness](definiteness.md) | No articles. A bare noun is neither *a* nor *the*; context decides. | | [word-formation](word-formation.md) | Two roots join for a number or a plural, and for nothing else. No compound word exists. | | [questions](questions.md) | Rising tone for yes/no. The question word stays where the answer would go. | | [pronouns](pronouns.md) | Six forms. No case, no gender, no politeness. *mi-mi* excludes you, *kita* includes you. | | [copula](copula.md) | *es* before a noun, nothing before an adjective or place. At the front with no subject it means "there is". | | [negation](negation.md) | *no* goes in front of what it denies, and before the tense particle. | | [conjunction](conjunction.md) | *aur* joins nouns, adjectives and sentences alike. *o* is or. | | [demonstratives](demonstratives.md) | *ini* this, *itu* that — after the noun, last in the phrase. | | [place](place.md) | *in* at, *dari* from, *por* to. Motion verbs take their destination bare. Place first, then time, both at the end. | | [verb-chains](verb-chains.md) | Verbs simply follow one another. Tense and negation attach to the first. | | [comparison](comparison.md) | *lebi / kurang / paling* before the adjective, *dari* for than, *kadar* for as-as. | | [subordination](subordination.md) | A clause needs no marker. *porke* because, *kab* when, *agar* if. | | [adverbs](adverbs.md) | An adjective straight after the verb describes the verb, before the object. | ### Not a rule | | | |---|---| | [proposal-ordinals.md](proposal-ordinals.md) | First, second, eighth. No affix is possible, so the choice is position or a particle; answering it would remove the need for nineteen roots. **Open — the founder's call.** | | [proposal-two-jobs.md](proposal-two-jobs.md) | May a root hold two jobs? *madad* has never been used in a sentence, so its decision is free; *rabota* stands in twenty places. **Open — the founder's call.** | | [proposal-frequency.md](proposal-frequency.md) | Where *daima* and *kadang* stand. Thirteen uses in three positions; *tena* settled itself 32 of 32. **Open — the founder's call.** | | [proposal-but.md](proposal-but.md) | A word for "but". Six pages wanted one and Lesson 12 invented one; *lakin* reaches five families. **Open — the founder's call.** | | [proposal-stress.md](proposal-stress.md) | The briefing the open questions put first. **Decided September 3, 2026** — the rule is [stress.md](stress.md); this is the record of what it was decided on. | | [proposal-sentence-types.md](proposal-sentence-types.md) | The imperative and the fragment. **Decided September 3, 2026** — the rules are [sentence-types.md](sentence-types.md). | | [proposal-names.md](proposal-names.md) | Marking a name. Forty-eight sentences are ambiguous today, and capital letters cannot be spoken. **Open — the founder's call.** | | [proposal-modal-adjective.md](proposal-modal-adjective.md) | "Want to be" plus an adjective. Called the smallest of the four until it was measured: 405 verbless adjective predicates and 59 working modals in the material, and not one place the two may meet. **Open — the founder's call.** | *Three claims in the tables above were the last copies of things corrected elsewhere,* and the third is the reason this paragraph is not about numbers alone. The row for ["want to be" plus an adjective](proposal-modal-adjective.md) called it *the smallest of the four* — a judgement its own briefing opens by withdrawing, in the sentence **"Measured, it is not small."** The same row was also the only open briefing in the table with no **Open — the founder's call.** on it, so the page said six were open and marked five. Both corrected September 6, 2026. *And four figures were stale.* The index said *rabota* has **eighteen** at stake — a figure [its own briefing](proposal-two-jobs.md) had already withdrawn as "not any reading of the material" — and it gave *tena*'s figure as it stood before the recount, one short of the thirty it stood at that day. Both corrected September 6, 2026, and *tena*'s stood twice on the briefing itself, twelve lines apart, one corrected and one not. The other two were the modal row's, **377** verbless adjective predicates and **60** working modals against a briefing that had been recounted to 400 and 59; they were corrected later the same day, by which point this paragraph had been complaining about exactly that for a day while carrying two instances of it four lines above. An index is where a withdrawn number goes to survive: it repeats a page's claim in one clause, links to it, and is read by nobody who is checking that page. Every page stating any of the four is checked now. **The full order of a sentence:** > subject → *no* → *suda/saufa* → verb → adverb → object → place → time → clause Every rule above appears in this one sentence, and not a word in it changes shape: > *Mi no suda kara hao libro ini in dom rat ini porke mi sedih.* > I did not read this book well at home last night because I was sad. ## Where to start The list below is grouped by file. Ordered instead by what each one unblocks, and by the evidence for it, the top of the list is: **~~1. Where the stress falls.~~ Settled September 3, 2026** — the second-to-last syllable of every word, and a vowel pair is one syllable. [The rule](stress.md), and [what it was decided on](proposal-stress.md). It blocked more than anything else here: until it was made, *DIN lai* and *din LAI* were both correct Amadunia and two people could not agree on the metre of a line. **2, 3, 4, 5. The four the language is already using without permission.** These were not left open on purpose. Each is a form the material relies on that no rule ever granted, and each was found by writing something that was not a lesson: - ~~**The imperative.**~~ **Settled September 3, 2026** — a verb with no subject is a command. [The rule](sentence-types.md). Fourteen sentences had been written that way since Lesson 10 and none needed changing. - ~~**Fragments.**~~ **Settled the same day** — any constituent may stand alone as an utterance. Forty-four were already in use. - **A mark for a name.** *Sol es genc* is "Sol is young" and "the sun is young" alike; found by [the second story](../texts/story-2-safari-por-pahar.md). - **"Want to be" plus an adjective.** [Text 3](../texts/text-3-lingua-ini.md) and Lesson 21 both reached for it and both were rewritten around it. **6. How `r` is made.** Less urgent than stress only because no two words in the language are told apart by it — any *r* is understood today. **7 onward, and the demand is thinner than it looks.** These four sit together on the list, but the material has reached for them exactly three times between them: | | Asked for by | Where | |---|---|---| | *very* | once | [the phrasebook](../phrasebook.md), which lost it from "thank you very much" and "this is very good" | | *never* | once | [story 2](../texts/story-2-safari-por-pahar.md), which wanted "I will never forget this morning" | | ordinal numbers | once | [story 2](../texts/story-2-safari-por-pahar.md) | | *all, some, none* | **never** | nothing has yet tried to say it | They appear in the *Still open* line at the foot of six lessons, which is one list repeated, not six demands. Counting mentions instead of findings overstates all four by roughly five times, and invents a case for *all, some, none* that does not exist. The rule that came out of checking this: **evidence is a page that tried to say something and could not, not a line on a list.** Two items are not grammar at all and sit outside this order: *madad* doing two jobs, and the thinness of Sino-Tibetan and Japonic recorded in [balance.md](../dictionary/balance.md). Both are vocabulary policy. ## Open questions — 40 of them Gathered from the files above so they can be read in one place. Each is recorded where the decision would live. That sentence was not true until September 3, 2026. *All, some, none* and the ordinal numbers appeared in the demand table below and in no rule file at all, so [`check.py`](../check.py) — which counts what the files record — could not see them and the count was right about the wrong set. Both are now in [plural.md](plural.md) and [numbers.md](numbers.md). The table and the list are two views of one thing and nothing enforces that they agree; matching a table row like *a word for then* to a bullet in a file is not something a script can be trusted to do, so it is kept by hand and said so here. ### What the writing has actually asked for The list below is grouped by file, which says nothing about which questions matter. This does. The measure is narrow and deliberate: **a page that tried to say something and could not**, recorded at the time in a text's gaps, the phrasebook's list, or a lesson's note. First counted September 3, 2026 and recounted September 5. | Question | Pages that reached for it | |---|---| | ~~the imperative~~ | settled September 3, 2026 — 3 pages had reached for it | | a word for *then* | **6** — [text 8](../texts/text-8-kaifa-suru-ca.md), [text 12](../texts/text-12-tarik-por-skola.md), [text 13](../texts/text-13-kula-una.md), [text 14](../texts/text-14-yamur-aur-ca.md), [text 21](../texts/text-21-uan-umur.md), [text 22](../texts/text-22-agar-yamur-lai.md) | | *although, before, after, until* | 3 — [text 7](../texts/text-7-surat-por-mama.md), [text 8](../texts/text-8-kaifa-suru-ca.md), [text 26](../texts/text-26-dom-eski.md) | | ordinal numbers | 3 — [text 6](../texts/text-6-seti-din.md), [story 2](../texts/story-2-safari-por-pahar.md), [text 27](../texts/text-27-kalima-por-anak-anak.md) | | *very* | 2 — [the phrasebook](../phrasebook.md), and a sentence the founder asked for, [recorded with the gaps](../dictionary/README.md#a-third-way-of-finding-one) | | *all, some, none, every* | **2** — [text 19](../texts/text-19-kamra-mi.md) wanted *every* morning, [text 27](../texts/text-27-kalima-por-anak-anak.md) wanted *everyone* | | whether *daima* covers *forever* | 1 — [text 10](../texts/text-10-mila-tahun.md) | | whether a time noun takes *in* | 1 — [text 7](../texts/text-7-surat-por-mama.md), which could not say "in the month I arrive" | | *never* | 1 — [story 2](../texts/story-2-safari-por-pahar.md) | | reflexives | 1 — [story 2](../texts/story-2-safari-por-pahar.md) | | superlative within a named group | 1 — [text 23](../texts/text-23-kim-paling-hayai.md) | | indirect questions | 0 | | where frequency adverbs sit | 0 | **A word for *then* leads, and it has led since the second recount.** The September 3 count had it at one page; five more have recorded it since. **Recounted in full on September 6, 2026, and seven of the thirteen rows moved.** *never*, reflexives and the superlative in a named group all stood at zero and all three had been recorded in a text's gaps — two of them in [story 2](../texts/story-2-safari-por-pahar.md), which is where the mark for a name was found as well. *then* gained [text 22](../texts/text-22-agar-yamur-lai.md), *every* gained [text 27](../texts/text-27-kalima-por-anak-anak.md), the subordinators gained [text 26](../texts/text-26-dom-eski.md) and the ordinals gained text 27. The row for *whether a time noun takes* in named the phrasebook and the page that could not say it is [text 7](../texts/text-7-surat-por-mama.md). **Two rows are machine-counted and the rest are not**, and the two are the ones whose claim is written the same way everywhere: *no word for then* and *no word for every*. Both counters were a page short until September 6, 2026, because they required the italics the claim is usually written in — text 22 wrote *then* in quotation marks and text 27 wrote *everyone*. A counter that sees one spelling of a claim flatters the table it is checking. Two questions have never been reached for by anything anyone wrote. That is not an argument for dropping them — a language needs *never* eventually — but it is an argument about order, and it is a much shorter list than the five this paragraph claimed for three days. **Two caveats, and they matter.** Stress and marking a name score zero here and are the first and third briefings, because neither was found by writing: stress was found by asking how a word is said aloud, and the name problem by scanning every sentence for ambiguity. Absence from this table means nothing was blocked while writing, not that the question is small. And eleven of the thirteen rows are **not machine-checked**. They rest on matching topic words against gap sections, and the topic list would be hand-kept — the arrangement that has already gone wrong three times in this repository. Those eleven are a dated measurement with its method stated, to be re-run rather than trusted, and re-running them on September 6, 2026 moved five of them. This paragraph said the whole table was unchecked while the paragraph above it said `check.py` counts the pages; both were half right and the page said two things. **[adverbs](adverbs.md)** - **Adverbs of degree on adverbs** — "very quickly" — needs a word for "very". *cok* is "much"; whether it covers "very" is untested. Wanted twice now: by [the phrasebook](../phrasebook.md), which lost it from "thank you very much", and on September 4, 2026 by *Mi parolas Esperanton tre flue*, which came out as *Mi sema hao Amadunia* — I speak Amadunia well, with both the degree and the manner gone. - **Whether *daima* covers "forever".** Glossed "always", from Arabic *dāʾiman*, which carries both. *Natura hidup daima* would say nature lives forever. One word for habit and endlessness, where most languages use two. - **Frequency adverbs.** *daima* (always) and *kadang* (sometimes) are adverbs by meaning but not adjectives. This page used to say they sit in the adverb slot with everything else; **counted on September 4, 2026, two of their thirteen uses do.** Five stand before the verb and six after the object, and Lesson 19 alone uses all three positions. [A briefing measures it and costs the three answers](proposal-frequency.md). Until it is decided the word *kadang* cannot be written at all, because any sentence using it would settle the question by accident. **[comparison](comparison.md)** - **Superlative among a named group** — "the biggest of the three" — has no form yet. *paling kabir dari tri* is untested. - **Whether two degree words may stack.** Lesson 18's conversation read *Cok lebi hao* for "much better" until September 4, 2026. *cok*, *lebi*, *kurang* and *paling* all stand in the same slot in front of what they scale, and nothing grants putting two there at once. The line now reads *Lebi hao*, which says less. Found by reading the lesson rather than by a check, and left open rather than granted. **[conjunction](conjunction.md)** - **Joining more than two clauses** — whether long chains want a different rhythm — is a question for when there is more prose. - **There is no word for "but".** *aur* joins and *o* chooses; nothing contrasts. Found on September 4, 2026 in [Lesson 12](../lessons/lesson-12-colors-directions.md), which had been saying *Ma hao!* — an invented word — for "but good!". The lesson now says *Hao!* and says less. **[copula](copula.md)** - **"want to be" plus an adjective has no form.** *es* is a verb and can be chained — *mau es doktor*, "want to be a doctor". An adjective is its own predicate and takes no verb, so *mau* has nothing to attach to: "wants to be the easiest" cannot be built. Both [text 3](../texts/text-3-lingua-ini.md) and Lesson 21 hit this and both were rewritten around it — text 3 by supplying a noun (*mau es lingua paling asan*), Lesson 21 by dropping the modal. Whether *es* is allowed before an adjective inside a verb chain, or something else is needed, is undecided. Found September 3, 2026 by checking every sentence in the repository against the copula rule. - **A noun with an adjective is now both a sentence and a phrase.** *Dom merah* is "the house is red" by the copula rule, and since [the fragment rule](sentence-types.md) of September 3, 2026 it is also "the red house" standing alone. 154 two-word utterances in the repository have this shape. Found September 5, 2026 by reading Lesson 12 end to end: its table glosses *Dom merah.* one way and its conversation glossed the same string the other way, eighty lines apart. Recorded, not briefed — context has carried it everywhere else. **[demonstratives](demonstratives.md)** - A third distance (*that over there*, Indonesian *sana*, Japanese *are*) is not provided. *situ* covers it for now. **[negation](negation.md)** - **Nothing, nobody, never.** Not yet designed. Whether they are built from *no* plus an existing word, or given their own roots, is open. - **Contrastive ellipsis — "X, not Y" — has no form.** *no* goes before a verb, an adjective, a place word or *es*, never before a noun, and five sentences were writing *X, no Y* anyway — a lesson practice item, two lesson tables, the phrasebook and this index's own demonstratives page. All five are full clauses now and all five are longer. Deciding it would give them back at no cost in words. Found September 5, 2026, the day after one of them was promoted onto a rule page for a different reason and nobody looked at its *no*. **[numbers](numbers.md)** - **Ordinal numbers have no form.** *First, second, eighth* cannot be built, and nothing in the cardinal system suggests a derivation. [Text 6](../texts/text-6-seti-din.md) is the clearest case: a story that runs on days wanted *on the eighth day* and could only say *for eight days*. **It also blocks two whole vocabulary sets**: the language names no weekday and no month, and numbering them — the economical answer, and Chinese's — needs an ordinal. Answering this question would remove the need for nineteen roots — seven weekday names and twelve month names. - **A number cannot count a pronoun.** *The three of us* has no form: numbers go before what they count — *tri anak* — and a pronoun takes neither that shape nor the reverse. Found September 5, 2026 in [text 19](../texts/text-19-kamra-mi.md), which had written *Kita tri sini* and glossed it "the three of us are here". The invention was a shape, not a word, and every check the repository had was looking for words. - Whether the separator between groups is a hyphen or a space is used inconsistently above (*du-des-uan* vs *pat-des fai*, both inherited from the README) and needs a single rule. **[phonology](phonology.md)** - ~~Whether a vowel pair is one syllable or two~~ — settled in [stress.md](stress.md): a vowel pair is one nucleus, so *dunia* is two syllables. Counting a syllable means counting vowel groups. - ~~Whether stress is fixed (and if so, on which syllable)~~ — settled in [stress.md](stress.md) on September 3, 2026: **the second-to-last syllable of every word**. This bullet outlived the decision by a day, and the day is the point: [the briefing it points at](proposal-stress.md) had said *Decided* at the top of it the whole time. - Whether the two-consonant sequences above are the complete permitted set, or merely the ones used so far, is not yet decided. **[place](place.md)** - **Whether a time word may open a sentence.** The order puts time at the end and grants fronting to a subordinate clause alone, and eight sentences front a time word instead — *sasa ta hao*, *Besok mi saufa go dom mi* — two of them in lessons. Every one is a contrast, which is what fronting is for in the languages this one borrows from. **Open — the founder's call.** - **Does a time expression take *in*?** Six sentences say *in pagi*; twenty-odd say *rat ini* with no preposition. Nobody decided; the two forms grew side by side. - **With an object, the two jobs of *una* collide.** Alone after a verb it is "together" — *Kita kula una* — and before a noun it is "with". Put an object in and the reading flips: *Kita kula una pan* is "we eat with bread", never "we eat the bread together". Found by [text 13](../texts/text-13-kula-una.md), September 4, 2026. - **A place word cannot attach to a noun, and cannot take one.** *sini, situ, upar, sub, kiri, yamin* are whole place expressions and go last, so **the people here**, **the soup there** and **below the house** have no form. Four sentences were writing them anyway, in a text, a lesson's source and this rule's own illustration; all four are rewritten and the gap is recorded. *Ponte in sungai* had been glossed "the bridge is **over** the river" in two places — *in* is at or in. Found September 5, 2026 by counting every adjacent pair of word classes in the repository and reading the rare ones. - *Through, without, about, before, after, **near*** — the second tier of prepositions — are for later. **near** has been wanted by three pages and again by [text 12](../texts/text-12-tarik-por-skola.md). **[plural](plural.md)** - **All, some, none.** *cok* covers "many" and stands before the noun; there is nothing for *all*, *some* or *none*, and nothing has yet tried to say them. - Whether doubling may also carry the distributive sense some source languages give it (*din-din* = "every day"), or stays strictly plural. **[possession](possession.md)** - **Alienable vs inalienable.** Some languages distinguish "my hand" from "my house". Amadunia currently does not; whether it ever should is open. **[pronouns](pronouns.md)** - **Reflexives** (*myself*) and **reciprocals** (*each other*) are not yet designed. - Whether *mi-mi* and *ta-ta* may ever be shortened in fast speech is a question for later, once there is speech. **[pronunciation](pronunciation.md)** - **How `r` is made.** A tap (Spanish *pero*, Japanese, Turkish, Indonesian), a - ~~Where the stress falls~~ — settled in [stress.md](stress.md): the second-to-last syllable of every word, and it changes no sound on the pronunciation page. **[questions](questions.md)** - **May a question word stand in a modifier's slot?** The swap rule produces *Libro ke?* — the answer is *Libro ini*, the unknown is the *ini* — and [Lesson 16](../lessons/lesson-16-school-and-time.md) glosses it "Which book?". The language has no word for *which*, and this page describes the question word as standing where an **answer** stands, not a modifier. One sentence in the course does it. Found September 5, 2026. - ~~**The imperative was never granted.**~~ — granted September 3, 2026 in [sentence-types.md](sentence-types.md): a sentence that begins with a verb and has no subject is a command, and that is also what tells it from a statement. *Anda tarik ini*, *Buka ain yu*, *Otur sini* stand unchanged. Found by [the fourth text](../texts/text-4-in-madina-baru.md). - ~~**Fragments.**~~ — settled the same day in [sentence-types.md](sentence-types.md): any constituent may stand alone as an utterance. *Ke?* *Harga?* *Ponte?* are legal by rule now rather than by accident, and a fragment answer inherits the tense and subject of the question. - **Lesson 6 asked *Nama yu?*** with no question word. That is a yes/no-shaped question that context turns into "what is your name". The explicit form is *Nama yu es ke?* Both are allowed; whether the short form should be discouraged in teaching is open. **[sentence-types](sentence-types.md)** - **Whether a command can be softened or strengthened by grammar.** English has *do sit down* and *just go*; Amadunia has the voice, the exclamation mark and *plis*, and nothing else. Untested. - **There is no way to command someone to *be* something.** *Es doktor* is taken by the existential, so *be a doctor*, *be quick*, *be careful* have no form. Found the day the command rule was made, by asking what it does to *es*. **[stress](stress.md)** - **Whether the short grammar words reduce in speech.** The rule stresses every word, so *Mi ama dunia* has three stresses in three words. Most languages let *es, no, aur, in, dari, por* lean on the word beside them. A fact about how it is spoken, not how it is written. **[subordination](subordination.md)** - **Reported questions** — "I asked where you were" — puts a question word inside a clause. Untested. - **Although, before, after, until** — the second tier of joiners. Not yet decided. **[tense](tense.md)** - Whether *suda* also covers the perfect ("I have eaten") or whether that needs its own marker. - Whether a habitual ("I eat every day") ever needs marking, or whether the unmarked present carries it. - **What tenses an adjective predicate.** The marker goes before the verb and an adjective predicate has none, so *the house was big* has nowhere to put *suda* — and the material has been putting it straight in front of the adjective since Lesson 11, in five sentences no rule grants. **Open — the founder's call.** **[definiteness](definiteness.md)** - **Is anything needed when context does not settle it?** *Mi kan dom* is both "I see a house" and "I see the house". *ini*, an owner and a number each force it, and each says more than "definite". **[word-formation](word-formation.md)** - **Does the language form compound words at all?** It never has — all 37 hyphenated forms in the repository are numbers or doubled plurals — and the front page had been leaning on the idea that it does. Juxtaposition is already taken by possession, and the hyphen already has two jobs. **[verb-chains](verb-chains.md)** - **Words that are both noun and verb.** *madad* is glossed "help; to help", so *Mi mau madad* is "I want help" **or** "I want to help". *rabota* is glossed "work" and is used both ways, in [twenty places](proposal-two-jobs.md) and as a noun in 6. Two roots, not one, and the second is a core A1 word: read as a verb one site breaks, read as a noun ten do. The chain rule made this visible for the first time. Nothing on this list has been invented to make a lesson or a text work. When something was needed and undecided, the lesson said so. Four of them were different in kind — the imperative, the mark for a name, how *r* is made, and "want to be" plus an adjective; the imperative was settled on September 3, 2026 and three remain. They were not left open on purpose. Each is something the language had been using or relying on that no rule ever granted, and all four were found by writing something that was not a lesson. --- # Commands and fragments *Fragments are taught in [Lesson 01](../lessons/lesson-01-greetings.md), where the first conversation is made of them; commands in [Lesson 10](../lessons/lesson-10-a-day.md), where the first one appears.* *Status: settled — decided September 3, 2026.* Two rules, decided together because they are the same question asked twice: what counts as a sentence when it is not a statement. ## A command is a verb with no subject **A sentence that begins with a verb and has no subject is a command.** | | | |---|---| | Kula pan. | Eat bread. | | Anda tarik ini. | Walk this road. | | Go yamin. | Go right. | | Buka ain yu! | Open your eyes! | | Otur sini. | Sit here. | **No word was added and no word changed shape.** The slot was empty and the language took it, exactly as it did with the existential *es* and with verb chains. Amadunia never drops a subject, so a sentence starting with a verb could not have been anything else. Fourteen sentences in the lessons, the phrasebook and text 4 were already written this way, since Lesson 10, and not one of them needs rewriting. ### No subject, ever *Yu go* is "you go" — a statement, however forcefully it is said. A command has no subject and that is what makes it one. Emphasis is carried by the voice and the exclamation mark, not by putting the pronoun back: | | | |---|---| | Go! | Go! | | Yu go. | You go. — a statement about you | This is the line that keeps the rule free of exceptions. Letting a subject appear "for emphasis" would make *Yu go* two sentences at once, and the language has spent three hundred roots avoiding exactly that. ### The one verb it does not apply to *es* is a verb — [copula.md](copula.md) says so, and it takes the tense particles like any other. So *Es doktor* is a sentence beginning with a verb and no subject, and by the rule above it should be a command: *be a doctor*. It is not. **A sentence-initial *es* is the existential**, settled first and standing in **44** sentences: *Es hotel in madina*, there is a hotel in the city. That read thirty until September 5, 2026, counted the day the rule was made and not since; [`check.py`](../check.py) counts it now, over the lessons, the texts and the phrasebook. The narrower rule wins over the broader one, which is how any two rules that overlap have to be ordered, and it is not an exception carved out for a word — *es* is simply already spoken for in that position. The cost is named in the open questions below: there is no way to tell someone to **be** something. ### Negation needs no new rule *no* goes in front of what it denies, which in a command is the verb. | | | |---|---| | No kula pan. | Do not eat bread. | | No go! | Do not go! | ### *plis* is politeness, not grammar *Kula, plis* is a polite command; *Kula* is a command. *plis* adds nothing grammatical, marks nothing, and may be left out of any command without changing what the sentence is. ## Any constituent may stand alone **A phrase may be a whole utterance when what it answers is clear.** | | | |---|---| | — Yu mau ke? — Ca. | — What do you want? — Tea. | | — Kab? — Besok. | — When? — Tomorrow. | | Corba, plis. | Soup, please. | | Madad! | Help! | | Harga? | Price? | Forty-four of these are already in the lessons and the phrasebook. The rule describes them; it does not change one. **Said alone, not printed alone.** A grammar table that prints *dom mi* or *tri anak* is citing a phrase, the way an English grammar prints "the big house". It is not claiming that a phrase is a sentence, and this rule has nothing to say about it. Counting naively gives fifty-seven fragments; forty-four is the count of things anybody actually said. **Those two numbers cannot be reproduced, and the reason is the distinction itself.** Separating a phrase somebody said from a phrase a table cites is a judgement about the page it sits on, and no rule stated here makes it mechanical — which is why the sentence needs two numbers at all. A scan of utterances with no predicate, taking a quoted turn or a phrasebook row as "said", gives 74 and 30 rather than 57 and 44. That is a different method, not a correction, and it is recorded rather than substituted: replacing a figure with one from a definition invented to check it is how a page ends up with a number no method produces. The existential count above **is** mechanical — a sentence whose first word is *es* — and is checked. **A fragment answer inherits the question.** *— Yu suda kula? — Ya.* means yes, I ate. The tense, the subject and the verb all come from the question, and nothing needs to be repeated or marked. ## Where the two rules touch A bare verb is a command. A fragment may be any constituent, and a verb is one. So *Kula* alone is a command at the start of an exchange and an answer in the middle of one: > — Yu mau suru ke? > — Kula. That is context doing the work, and it is what English does with *Eat.* too. It is the one place these two rules overlap, and it is written here rather than left to be discovered. ## Sources Bare-verb commands: Chinese, Indonesian, Malay, English, Swahili, Turkish (*git*), Persian. It is the majority pattern on Earth and the one that costs nothing. Fragments: every language. None of the source languages requires a full sentence to answer a question. ## Rejected **A particle after the verb** — Indonesian *-lah*, Japanese. Costs a word and a new position rule to say what word order already says. **Reusing *plis* as the marker.** It would make every command polite, and a warning is not a request. **A distinct verb form** — Spanish *come*, Arabic, Hindi. Disqualified outright: it would be the first inflection in a language whose second design rule is that no word ever changes shape. **Requiring a full sentence** — no fragments. No language does this, and it would make the phrasebook impossible to write. ## Open questions - **Whether a command can be softened or strengthened by grammar.** English has *do sit down* and *just go*; Amadunia has the voice, the exclamation mark and *plis*, and nothing else. Whether that is enough is untested. - **There is no way to command someone to *be* something.** *Es doktor* is taken by the existential — there is a doctor — so *be a doctor*, *be quick*, *be careful* have no form. Every language has these and they are among the commonest things anyone says. Found on September 3, 2026 by asking what the command rule does to *es*, the day the command rule was made. **Evidenced by writing on September 4**: [text 12](../texts/text-12-tarik-por-skola.md), a mother walking a child to a road, wanted it three times on one page — *be careful*, *be quick*, *do not be afraid* — and wrote two commands, a time word, and nothing at all in their places. *Do not be afraid* fails twice over, because *korku* is a noun as well. --- # Stress *Taught on [the front page](../README.md) and in [Lesson 01](../lessons/lesson-01-greetings.md).* *Status: settled — decided September 3, 2026.* ## The rule **Stress falls on the second-to-last syllable of every word.** A one-syllable word carries it on the syllable it has. There is no exception, no marked word, and nothing to memorise per word. | | | |---|---| | **DU**-nia | world | | a-ma-**DU**-nia | the language's own name | | **MER**-si | thank you | | ka-**LI**-ma | word | | sa-**FA**-ri | journey | | **MI** | I | | **ES** | is | ## What counts as a syllable **One syllable per vowel group. A vowel pair is one nucleus, not two.** This is the question that had to be settled first, and it is settled by the phonology rather than by choice: only five vowel pairs exist — *ai, ao, au, ia, ua* — three vowels in a row are forbidden, and those five are exactly the sequences the source languages pronounce as a single glide. Counting them as one nucleus makes syllable-counting mechanical: **count the vowel groups.** | | Syllables | | |---|---|---| | *dunia* | 2 | du-nia, not du-ni-a | | *akua* | 2 | a-kua | | *mausim* | 2 | mau-sim | | *familia* | 3 | fa-mi-lia | | *uan* | 1 | one vowel group | Under this definition the 300 roots are 57 of one syllable, 209 of two and 34 of three. ## What stress does to the sound **It makes the syllable louder and a little higher. It does not change it.** No vowel reduces, none lengthens, and none shifts quality. The *a*'s in a-ma-**DU**-nia are the same *a* whether stressed or not, and *dunia* said with the beat on *DU* has exactly the sounds spelled in it. This is how Indonesian, Swahili and Japanese work, and it is the opposite of English, Russian and Portuguese, where unstressed vowels collapse toward a schwa. That is not a stylistic preference. Design rule 2 says the language is written as it is spoken; a reduced vowel would break it the first time anyone said a word out loud, because the letter on the page would no longer be the sound in the air. Stress had to be the kind that leaves the spelling true, and this kind is. See [pronunciation.md](pronunciation.md), where the vowels are. ## The hyphen does not join words for stress A hyphen joins two words for meaning — a plural, or a number. Each half keeps its own stress. | Written | Said | | |---|---|---| | anak-anak | **A**-nak **A**-nak | children | | mi-mi | both halves stressed, each on its only syllable | we, not including you | | des-uan | **DES** **U**-an | eleven | ## Why the second-to-last **It is what the largest source families already do.** Indonesian and Malay, Swahili, Polish and Quechua stress the penultimate as a rule, and it is the default in Spanish and Italian. Austronesian is the family with the widest reach in this dictionary, and Niger-Congo is not far behind. **It is what the one constructed language that has actually been spoken for a century does.** Esperanto has stressed the penultimate since 1887. **It is the most forgiving choice available.** In **266 of the 300 roots — 89% — the second-to-last syllable is also the first**, because 57 roots have one syllable and 209 have two. A learner whose own language stresses the beginning of a word is right about nine words in ten without being taught anything, and can only go wrong on the 34 three-syllable roots. ## Rejected **Initial — always the first syllable.** Czech, Hungarian, Finnish, Icelandic, Latvian. It agrees with the penultimate on 89% of the dictionary anyway, so it buys almost nothing, and it disagrees with Austronesian and Swahili — the two families this language leans on hardest — on the third where it differs. **Final — always the last.** Turkish, Persian, French, Armenian, Hebrew. A real tradition and a large one, but it is the option that disagrees with the penultimate everywhere except one-syllable words, and it disagrees with Esperanto, Indonesian, Malay and Swahili at once. Choosing it would mean the widest-reach family in the dictionary is also the one whose speakers say every word wrong. **Free — no rule.** What the language did until today, and [the poem](../texts/text-5-uan.md) is why it could not stay that way: *DIN lai* and *din LAI* are different lines, so nothing could be sung, scanned or taught by ear. ## What it unblocked, the same day [The poem](../texts/text-5-uan.md) was written while this was undecided and said so — being said aloud was "the thing the language is least ready for". It was scanned within the hour, and two things came out of it that nobody could have seen before. Its paired lines match to the syllable, 6 and 6, 4 and 4, 6 and 6, 5 and 5, though the poem was built on parallel words and nothing counted beats. And its closing line, *Kita ama dunia una*, is the only regular line in it: **X . X . X . X .**, four trochees with nothing left over. That was luck rather than craft, because when the line was written no rule could say where its beats were. ## Nothing was at risk — and what it cost anyway **No two words in Amadunia are told apart by stress, and none ever can be** — that would be a minimal pair, which the language forbids. No word changed meaning, and no text needed rewriting. But a rule that fits one tradition cannot fit them all, and it is worth saying plainly who pays. Taking the first source each entry names, and counting only roots of two syllables or more: | | Roots | |---|---| | First source stresses the penultimate — Indonesian, Malay, Swahili, Spanish, Italian, Polish | **87**, and they keep their beat | | First source stresses the last — Turkish, Persian, French | **34**, and their beat moves | *anahtar* is a-nah-**TAR** in Turkish and a-**NAH**-tar here. *pencere* is pen-ce-**RE** and becomes pen-**CE**-re. *ates*, *yatak*, *korku*, *burun*, *bage*, *mersi* all shift the same way. A Turkish or Persian speaker meets eleven per cent of the dictionary saying their own word with the beat in the wrong place. That is the real price of the decision, it is not zero, and nothing makes it smaller: any single rule moves somebody's beat. It is paid to the smaller side — 87 against 34, two and a half to one — and what is lost is recognition of rhythm, never meaning, because meaning never rode on stress here in the first place. The ratio read *nearly three to one* until September 5, 2026. Eighty-seven to thirty-four is 2.56, which rounds to three only if you round it once; and the rounding went the way that made the price look smaller, on the one page whose stated job is to say plainly who pays. **Two thirty-fours, and they are not the same thirty-four.** This page says a learner "can only go wrong on the 34 three-syllable roots", and then that 34 roots have their beat moved. The two sets share four words — *anahtar*, *pencere*, *restoran*, *telefon* — and nothing else. **Thirty of the roots whose beat moves have two syllables**, which is to say they are among the roots the paragraph above has just promised an initial-stress learner cannot get wrong: that learner is right about where the beat falls and a Turkish speaker is wrong about the same word, because the two are being told different things. Both counts are correct and the collision is chance — 34 out of 300 twice over — but one page printing one number for two sets reads as one set. Found on September 5, 2026 by listing the members instead of comparing the totals, which is the only way it could have been found. (French is counted as final-stress by simplification: it stresses the phrase rather than the word, so *mersi* has no fixed word beat to move.) ## Open questions - **Whether the short grammar words reduce in speech.** The rule stresses every word, so *Mi ama dunia* is **MI A**-ma **DU**-nia, with three stresses in three words. Most languages let particles like *es, no, aur, in, dari, por* lean on the word beside them instead. Whether Amadunia does is a fact about how it is spoken, not about how it is written, and nothing here decides it. --- # Subordination — that, because, if, when *Taught in [Lesson 18](../lessons/lesson-18-comparing-and-joining.md).* *Status: settled — decided September 3, 2026.* ## A clause can be an object. Nothing marks it. | | | |---|---| | Mi fikir yu hao. | I think you are good. | | Mi bil ta suda lai. | I know he came. | | Mi mau yu kula. | I want you to eat. | | Ta sema dom mi kabir. | She says my house is big. | There is no word for "that". A whole sentence simply stands where an object would stand. **And that is why the rule was mismeasured, twice.** Every count of subordination in this repository looked for *porke*, *kab* and *agar*, because those are the words there are to look for. The unmarked clause has none, so half the rule was invisible to the instrument, and the eleven sentences that carry it had never been counted anywhere. The other half of the mistake was counting too much: *kab* and *porke* are also the question words *when* and *why*, and in a question they stand last, where the answer will stand, marking nothing. Ten sentences in the material are that shape and every one was being read as a subordinate clause. The material holds **50 sentences** with a marker and **11 without one**. [`check.py`](../check.py) finds them now, by the one shape that cannot be read two ways — a verb, then a pronoun, then a predicate, as in *Mi fikir yu hao*. A pronoun takes no adjective and heads no phrase, so nothing else can follow it there. **It is a floor and not a count:** *Ta sema dom mi kabir* is a clause too, and it is indistinguishable by machine from a noun with an adjective, which is [the same ambiguity](README.md) recorded for *Dom merah*. This is the same reasoning that settled [verb chains](verb-chains.md): the slot was empty, so the language took it for nothing. And it answers the question those left open — *Mi mau yu kula*, "I want **you** to eat", is a clause object, not a verb chain. English drops "that" freely (*I think you're good*), and Chinese, Indonesian, Thai and Vietnamese never had one. ## Because is *porke* | | | |---|---| | Mi mau ca **porke** ca hao. | I want tea because tea is good. | | Ta no lai **porke** yamur. | She is not coming because of rain. | **No new word.** *porke* has been "why" since Lesson 7, and Spanish separates the two only by a space — *por qué* asks, *porque* answers. One word does both jobs here, as it does in most of the world. ## When is *kab* | | | |---|---| | Mi lai **kab** yu sema. | I come when you speak. | | **Kab** yamur lai, mi in dom. | When rain comes, I am at home. | **No new word.** *kab* has been the question "when" since Lesson 7. English, Spanish, French and Turkish all use one form for both. ## If is *agar* | | | |---|---| | Mi go market **agar** yamur no lai. | I'll go to the market if it doesn't rain. | | **Agar** yu mau, kita kula sasa. | If you want, we eat now. | The one new word. Persian and Urdu *agar*, Hindi *agar*, Turkish *eğer* — some five hundred million speakers. ## Where the clause goes **After the main clause**, in the same place as time phrases. It may come first for emphasis, and then a comma marks the break: | | | |---|---| | Mi in dom kab yamur lai. | I am at home when it rains. | | Kab yamur lai, mi in dom. | When it rains, I am at home. | Both are correct. The following form is the one to teach. ## Why no word for "that" Every candidate collided, and none was needed: | Candidate | Source | Reason rejected | |---|---|---| | ki | Persian, Turkish, Hindi | minimal pair with *hi*, *ke*, *mi* — including *ke*, the question word it would stand beside | | ke | Spanish *que*, Persian | already "what" | | bahwa | Indonesian | one family, and long for the commonest joiner in a language | | anna | Arabic | minimal pair risk against *nama*, *nyama* families and no reach outside Semitic | ## Open questions - **Reported questions** — "I asked where you were" — puts a question word inside a clause. Tested on September 3, 2026 and **no rule forbids it**: *Mi tanya yu nali*, *Mi bil ta go nali*, *Ta sema kim lai* pass every check in [`check.py`](../check.py). The first of those three read *Mi tanya yu **es** nali* until September 5, 2026, and it did pass every check — because no check covered it. [copula.md](copula.md) says *es* stands before a noun and before nothing else, and its own illustration is *Hotel nali?*, "where is the hotel", marked *place word, no es*. The check that holds this had a hand-written list of what may not follow *es* and that list was missing *nali*, *kiri* and *yamin*. It reads the Place group now. The sentence offered as evidence that the composition is unforbidden was the one thing on this page the rules forbid. The two settled rules already compose — a clause stands where an object stands, and a question word stands where the answer will stand — which is the same argument this page used above to resolve *Mi mau yu kula*. What is left to decide is whether that composition is granted or merely unforbidden, and whether a clause with a question word in it needs anything to keep it from being heard as a question, given that questions are marked by tone alone. - **Although, before, after, until** — the second tier of joiners. Not yet decided. --- # Tense *Taught in [Lesson 04](../lessons/lesson-04-tense.md).* *Status: settled — decided September 2, 2026. Reversible if the founder disagrees; see "Why this and not the alternatives" below.* ## The rule **The verb never changes.** Time is shown by one invariant particle placed directly before the verb. The present is unmarked. | Time | Particle | Example | English | |---|---|---|---| | present | — | Mi kula pan. | I eat bread. | | past | **suda** | Mi **suda** kula pan. | I ate bread. | | future | **saufa** | Mi **saufa** kula pan. | I will eat bread. | That is the whole system. There is nothing to conjugate, no irregular verbs, and no agreement with the subject: *mi suda kula*, *yu suda kula*, *ta suda kula*, *kita suda kula* — the verb *kula* is the same word every time. ## Where the particle goes Always immediately before the verb, after the subject. - Ta **suda** lai dom. — She came home. - Rafiki **saufa** nomu ca. — The friend will drink tea. - Anak **suda** kan sol. — The child saw the sun. Never after the verb, never at the end of the sentence. ## When the particle may be left out Once the time is clear, it does not need to be repeated. If a sentence has already been placed in the past, the sentences that follow may drop *suda* until the time changes. > Mi **suda** go dom. Mi kula pan. Mi nomu ca. > — I went home. I ate bread. I drank tea. Including the particle in every sentence is never wrong. Leaving it out is allowed only when the time has already been set. ## Sources | Particle | Source | Meaning there | |---|---|---| | suda | Indonesian/Malay *sudah* | already, done — used exactly this way to mark completed action | | saufa | Arabic *sawfa* (سوف) | the standard future particle in Modern Standard Arabic | Two different families, neither already dominant in the vocabulary. Both are semantically transparent in their source: *suda* literally says "already", *saufa* literally says "will". ## Why this and not the alternatives **Why particles, not endings.** Design rule 3 forbids conjugation. A particle is a separate word that never changes; an ending would attach to the verb and, in every natural language that has them, eventually breeds irregulars. **Why an unmarked present.** It is the most common tense in speech and the one a beginner needs first. Making the default the shortest form is what Chinese, Indonesian, Malay, Thai and Vietnamese all do, and it is the single feature that makes those languages easy to start. **Why two particles, not one.** Some languages mark only "not-present" and let context sort past from future. That saves a word but costs clarity, and clarity is cheaper. **Candidates rejected, and why:** | Candidate | Source | Reason rejected | |---|---|---| | akan (future) | Indonesian *akan* | metathesis pair with *anak* (child) — a-k-a-n vs a-n-a-k | | ta (future) | Swahili *-ta-* | *ta* is already he/she | | ga (future) | Hindi *-gā* | minimal pair with *ba*, *ca*, *go*, *ta*, *ya* — and *go* is the verb it would most often stand next to | | hui (future) | Chinese *huì* | introduces the vowel sequence *ui*, not yet attested in the language | | le (past) | Chinese *le* | one light syllable, easy to miss in speech; and Chinese *le* is post-verbal, so the borrowing would misrepresent the source | | mau (future) | Indonesian colloquial | *mau* is already "to want"; using it for both breeds ambiguity | | go (future) | English "going to" | same problem: *go* is already "to go" | ## Open questions - Whether *suda* also covers the perfect ("I have eaten") or whether that needs its own marker. - Whether a habitual ("I eat every day") ever needs marking, or whether the unmarked present carries it. - **What tenses an adjective predicate.** This rule puts the marker in front of the verb; [the copula rule](copula.md) puts nothing in front of an adjective, so *Dom kabir* — the house is big — has no verb to mark and no way to say *was*. **The material has been answering it since Lesson 11 without a rule:** the marker goes straight in front of the adjective, where the verb would stand. | Where | Sentence | | |---|---|---| | [Lesson 11](../lessons/lesson-11-being.md) | *Dom suda kabir* | the house was big | | [Lesson 11](../lessons/lesson-11-being.md) | *Ca saufa garam* | the tea will be hot | | [Lesson 11](../lessons/lesson-11-being.md) | *Din saufa garam* | the day will be hot | | [Lesson 20](../lessons/lesson-20-colours-and-health.md) | *Ta suda bimar* | she was sick | | [text 16](../texts/text-16-berapa-harga.md) | *Mi fikir samaki saufa hao* | I think the fish will be good | Five sentences, and the lesson that teaches the copula teaches three of them. Nothing grants the shape: this page says *before the verb* and there is no verb, and copula.md says *nothing before an adjective* and this is something. It is the same kind of thing as the imperative before September 3 — a rule the writing settled by using it, waiting for somebody to say so — and it is the founder's to grant or refuse. Found September 6, 2026 by [text 26](../texts/text-26-dom-eski.md), which is a memory, needed it in its second sentence, and worked round it because an open shape is not for new writing. [`check.py`](../check.py) holds the count at five so the shape cannot spread while the question is open. --- # Verb chains — want to, can, must *Taught in [Lesson 17](../lessons/lesson-17-wanting-and-able.md), after [Lesson 16](../lessons/lesson-16-school-and-time.md) found it missing.* *Status: settled — decided September 3, 2026.* ## The rule **Put the verbs one after another. Nothing goes between them.** | | | |---|---| | Mi **mau kula** pan. | I want to eat bread. | | Yu **bisa otur** sini. | You can sit here. | | Ta **lasim go** skola. | He must go to school. | | Mi **mau studi kara**. | I want to learn to read. | There is no word for "to". There is nothing to add and nothing to change — the verbs simply stand in a row. ## Why nothing goes between them Two nouns side by side already mean something in Amadunia: the second owns the first. That is why *es* and *aur* had to exist. Two **verbs** side by side meant nothing at all — the slot was empty. So the language could take it for free. This is the opposite of the copula and the conjunction, and it is worth noticing: those two decisions added words because the slot was taken. This one adds nothing because the slot was open. ## Tense and negation go at the front Both attach to the **first** verb and cover the whole chain: | | | |---|---| | Mi **suda** mau kula. | I wanted to eat. | | Yu **saufa** bisa kara. | You will be able to read. | | Mi **no** mau kula. | I do not want to eat. | | Mi **no suda** mau kula. | I did not want to eat. | The full order is unchanged from every other rule in the language: **no → suda/saufa → verb → verb → object.** ## Where *no* sits changes the meaning Because *no* goes in front of what it denies, moving it moves the meaning. This falls out of the existing rule; nothing new is needed. The last row read "She can stay away" until September 5, 2026 — a paraphrase with no negative in it at all, in the one table whose whole job is to show where the negative went. *lai* is *to come*, and *no lai* is *not come*, not *stay away*, which says something about intention the sentence does not. | | | |---|---| | Mi **no** mau kula. | I do not want to eat. | | Mi mau **no** kula. | I want *not* to eat. | | Ta **no** bisa lai. | She cannot come. | | Ta bisa **no** lai. | She is able *not* to come. | ## The object comes last After the final verb, as always: | | | |---|---| | Mi mau kupi libro. | I want to buy a book. | | Kita lasim funga dom. | We must close the house. | | Yu bisa kara libro ini. | You can read this book. | ## Sources Bare verb juxtaposition is what most of the world's speakers do: | Language | | |---|---| | Chinese | *wǒ xiǎng chī* — I want eat | | Indonesian, Malay | *saya mau makan* — I want eat | | Thai, Vietnamese, Lao | the same, with no marker | | Hausa, Yoruba, Akan | serial verbs, the construction's home ground | | Haitian Creole | *mwen vle manje* | English *to*, Spanish *a* and *de*, French *à* and *de* are the minority pattern — and an unpredictable one. Which preposition follows which verb is pure memorisation in all of them, exactly the kind of exception the language exists to avoid. ## Why no marker for "to" Amadunia's verb has one form. There is no infinitive to distinguish from anything else, so a marker would mark a difference the language does not have. And every candidate collided anyway: | Candidate | Source | Reason rejected | |---|---|---| | to | English | minimal pair with *go*, *no*, *ta* | | ku | Swahili *ku-* | minimal pair with *du*, *ke*, *yu* | | de | Spanish, French | minimal pair with *du*, *ke* | | a | Spanish | 69 roots end in *-a*; the join would blur | | na | Irish, Hindi | minimal pair with *ba*, *ca*, *no*, *ta*, *ya* | | za | Polish | the alphabet has no *z* | ## The two words this unlocks | Word | Meaning | Source | |---|---|---| | bisa | can, to be able | Indonesian and Malay *bisa* | | lasim | must, to have to | Arabic *lāzim*, spoken natively in Turkish *lazım*, Persian *lāzem*, Urdu and Hindi *lāzim*, Swahili *lazima*, Indonesian *lazim* | *lasim* is written with *s* because the alphabet has no *z*. It is the one adaptation in the word. ## Open questions - ~~A different subject~~ — settled in [subordination.md](subordination.md): *Mi mau yu kula* is a clause standing where an object stands, and needs no marker. - **Words that are both noun and verb.** *madad* is glossed "help; to help", so *Mi mau madad* is "I want help" **or** "I want to help". The chain rule made this visible for the first time. [A briefing now separates the two roots](proposal-two-jobs.md): *madad* has never been used in a sentence at all, so its decision costs nothing, while *rabota* stands in twenty places. The dictionary was audited on September 3, 2026 by reading glosses, and that audit said madad was the only case in all 300 roots. **It was wrong,** because it read the dictionary and not the sentences. *rabota* is glossed "work", one noun, and is used both ways — *Mi rabota sini* is I work here, *Kita finis rabota* is we finish the work. [The briefing counts the places](proposal-two-jobs.md) and `check.py` recomputes them; this page said "13 places and 6" over its own wider scope until September 5, 2026, and two pages counting one word over two scopes with nothing recomputing either is how the briefing came to print a total no method produces. It reproduces the madad problem exactly: *Ta suda lasim rabota* is both "she had to work" and "she had to have work." Two others looked like cases and are not: *bisa* ("can, to be able") and *lasim* ("must, to have to") are one modal verb under two English glosses, not a noun beside a verb. So the open question covers two roots, not one, and the second is a core A1 word standing in [twenty places](proposal-two-jobs.md). Nothing here decides it. *madad* is treated as a verb by check.py, which keeps it out of the early lessons rather than letting one answer by accident. *rabota* cannot be treated as either, because the course already needs both readings at once: Lesson 08 needs the verb (*Mi rabota sini*, I work here) and Lesson 10 needs the noun (*Mi suru rabota*, literally I do work). The two readings conflict inside the course, so the check names it as pending instead of picking a side. What each answer costs, measured rather than guessed. Read as a verb, **one** site breaks: Lesson 10's *Mi suru rabota*, where the lesson's own gloss says "literally: I do work" and the word is plainly an object. Read as the noun the dictionary already calls it, **ten** sites break — every *Mi rabota sini* in the lessons, the grammar files and the phrasebook, each needing an *es* that would make it "I am work". A third thing only a reader can judge: *naiti rabota baru*, to find new work, puts an adjective on it, and only a noun takes one. The cheaper repair is not the same as the right answer, and nothing here makes the choice. So the open question is not a review of many entries but a decision about one word — and behind it, whether a root may ever hold both jobs. Every root added since has been given one job only. --- # Word formation *The two joinings are taught in [Lesson 05](../lessons/lesson-05-plural.md) and on [the front page](../README.md); nothing else joins.* **Settled: two roots may be joined in exactly two ways, and both were settled elsewhere. There is no third way, and no compound word exists.** | Joining | Means | Where it was settled | |---|---|---| | *des-uan*, *du-des-uan* | eleven, twenty-one | [numbers.md](numbers.md) | | *anak-anak*, *dom-dom* | children, houses | [plural.md](plural.md) | ## What the corpus actually contains Every hyphenated form in the repository is collected and counted by [`check.py`](../check.py) on every run, so this page cannot fall behind the corpus — it did once, by one form, when three texts were written after it. There are **37**, and every single one is a number or a reduplicated plural: - 17 number shapes: *des-du*, *du-des*, *ba-des*, *du-sen*, *du-mila* … - 20 reduplications: *anak-anak*, *ain-ain*, *doktor-doktor*, *hora-hora* … - nothing else. File names are not counted. *surat-por-mama* and *anak-katab* are hyphenated because file names are, and they are not forms of the language. **Not one compound word has ever been formed.** The language has no *airport*, no *bookshop*, no *railway* — no word built out of two words. ## What has been called compounding is not compounding The front page explained the 300-root target by saying compounding did the rest, and [text 6](../texts/text-6-seti-din.md) called *mesin ambil foto korpo anak* "the compounding the design was betting on". Both were corrected; this page was the reason, and it is the last page still describing the old state as the present one, until September 5, 2026. That sentence means *a machine takes a picture of the child's body*, and it does say X-ray without a new root — but it is a **sentence**, not a word. Nothing was formed. What carried it was the possession rule and a verb. That distinction matters for the vocabulary argument. Three hundred roots go further than three hundred words because things can be **described** with what is already there, not because new words can be **built**. Description is a weaker claim than word-formation, and it is the true one. ## Open questions - **Does the language form compound words at all?** It never has. The front page leaned on the idea that it does until this page was written, and says the opposite now. The decision has two hard parts. First, **juxtaposition is already taken**: *foto korpo* cannot mean "body-photo" as one word, because the possession rule already makes it "the body's picture". Second, the hyphen is already doing two jobs, and a third would make *des-uan*, *anak-anak* and a compound three unrelated things wearing one mark. Whatever the answer, it decides how far 300 roots really reach, so it belongs with [the A2 briefing](../dictionary/proposal-a2.md). Found by scanning every hyphen in the repository after noticing that "compounding" appeared four times and had no page. --- # Lesson 1 — Greetings *Everything in this lesson uses only settled vocabulary.* ## New words | Word | Meaning | |---|---| | hi | hello | | salam | peaceful greeting | | ok | ok | | ya | yes | | no | no | | bai | goodbye | ## Where the stress goes **The second-to-last syllable, in every word, always.** A word with one syllable takes it there — *hi*, *ok*, *ya*, *no*, *bai* are all one syllable, so there is nothing to decide. *salam* has two: **SA**-lam. A vowel pair counts as one syllable, so *bai* is one and not two. There is no exception anywhere in the language and no word has to be learned with its stress. [The rule](../grammar/stress.md). ## Your first conversation > — Hi! Ok? > — Ya, ok! Bai! Literally: *Hello! Ok?* — *Yes, ok! Goodbye!* That is a complete, correct Amadunia exchange. There is nothing hidden in it: no conjugation to learn, no gendered form to choose, no article to place. Not one of those four lines is a full sentence, and none needs to be: [any phrase may stand alone](../grammar/sentence-types.md) when what it answers is clear. That is the rule from the first line of the first lesson, and the same page grants the commands you will meet in [Lesson 10](lesson-10-a-day.md). ## Two ways to greet *hi* is the everyday greeting — quick, neutral, usable with anyone. *salam* carries the older sense of wishing peace. Use it when you want the greeting to mean something. > — Salam! > — Salam! Ok? > — Ya, ok. ## Saying no *no* answers exactly as *ya* does, in the same slot: > — Ok? > — No. ## Practice Say these out loud. Every letter has one sound, so read them exactly as written. 1. Hi! 2. Salam! Ok? 3. Ya, ok! Bai! 4. Hi! ... No. ... Bai! ## What you can already say With six words you can open a conversation, ask how someone is, answer either way, and close. That is the whole point of the design. **Next lesson:** [water, food and home](lesson-02-food-and-home.md) — the first nouns, the first verbs, and the subject-verb-object order they stand in. That line read "pronouns and the verb *ama* — waiting on the pronoun and word-order decisions" until September 5, 2026. It was written on the first day, before either decision existed; both were settled on September 2 and [pronouns](../grammar/pronouns.md) got a lesson of their own at [09](lesson-09-pronouns.md). The first sentence of the course pointed at a lesson that was never written and at questions that had been answered for three days. --- # Lesson 2 — Water, food, home *Prerequisite: [Lesson 1](lesson-01-greetings.md).* ## New words | Word | Meaning | |---|---| | akua | water | | ca | tea | | pan | bread, food | | dom | house, home | | sol | sun | | nama | name | | kula | to eat | | lai | to come | | go | to go | | mau | to want | | beri | to give | | hao | good | | bas | enough, stop | ## Building a sentence The README already gives the pattern: > *Mi ama dunia.* — I love the world. Subject, then verb, then object. Every sentence below uses that same shape — nothing new is being introduced. | Amadunia | English | |---|---| | Mi kula pan. | I eat bread. | | Yu mau akua. | You want water. | | Ta go dom. | She goes home. | | Kita lai dom. | We come home. | | Mi beri pan. | I give bread. | | Ta mau ca. | He wants tea. | Notice what did *not* happen: the verb never changed. *Mi kula*, *yu kula*, *ta kula*, *kita kula* — one form, always. Notice something else missing. There is no word for *a* and none for *the*. *pan* is *bread*, *a bread* or *the bread*, and which one it is comes from the situation, the way it does in Russian, Turkish, Hindi, Indonesian, Chinese and Swahili. You will see the same noun translated both ways throughout this course. That is not sloppiness — it is [the rule](../grammar/definiteness.md). ## Asking Lesson 1 asked a question by tone alone: *Ok?* The same works with a full sentence. > — Yu mau akua? > — Ya, mi mau akua. > — Yu mau ca? > — Ya! Mi mau ca. > — Yu mau pan? > — No, bas. *bas* closes things politely: enough, that's all, stop. ## Describing > Sol hao. > Akua hao. > Dom hao. ## Practice Say each aloud, then cover the English. 1. Mi mau pan. — *I want bread.* 2. Ta lai dom. — *She comes home.* 3. Kita kula una. — *We eat together.* 4. Sol hao! — *The sun is good!* 5. Yu beri akua? — *Will you give water?* ## What you can already say Two lessons, **40 roots** — nineteen taught in these two lessons and the other twenty-one the numbers and first words on [the front page](../README.md). You can greet someone, ask their state, say what you want, eat, come, go, give, and call something good. No tense, no agreement, no gender has been needed yet. That count read "about twenty words" until September 5, 2026, counting only the two word tables while every other lesson counts what the learner knows. It was the one lesson using its own convention, and the one whose figure nothing checked, because the check reads a number that follows the word *lessons*. Past and future are covered in [Lesson 4](lesson-04-tense.md), plurals in [Lesson 5](lesson-05-plural.md), questions in [Lesson 7](lesson-07-questions.md). --- # Lesson 3 — People, day and night *Prerequisite: [Lesson 2](lesson-02-food-and-home.md).* ## New words | Word | Meaning | |---|---| | anak | child | | rafiki | friend | | din | day | | rat | night | | pagi | morning | | aksam | evening | | kan | to see | | sema | to say, to speak | | bil | to know | | nomu | to drink | | kabir | big | | keci | small | | mal | bad | | doktor | doctor | ## Talking about people The sentence shape has not changed since Lesson 2 — subject, verb, object. | Amadunia | English | |---|---| | Mi kan rafiki. | I see a friend. | | Anak nomu akua. | The child drinks water. | | Ta bil nama. | She knows the name. | | Rafiki sema. | The friend speaks. | ## Describing A quality word follows what it describes: | Amadunia | English | |---|---| | Dom kabir. | The house is big. | | Anak keci. | The child is small. | | Ca hao. | The tea is good. | | Rat mal. | The night is bad. | There is no separate word for *is*. *Dom kabir* is complete as it stands. ## The parts of the day > — Din hao? > — Ya, din hao. Sol kabir! > — Rat hao? > — No. Rat mal. | Amadunia | English | |---|---| | Pagi hao. | The morning is good. | | Aksam hao. Rat mal. | The evening is good. The night is bad. | **Four words for the day, and until September 4, 2026 this lesson had two.** *pagi* used to arrive in Lesson 22 and *aksam* in Lesson 24 — nineteen and twenty-one lessons after *din* and *rat*, for words of exactly the same kind. [Text 15](../texts/text-15-sol-lai.md), written for Lesson 10, opens *Sol lai* — the sun comes — because it could not say *in pagi*. **Saying *in the morning* cannot be done yet:** *Sol kabir in pagi* needs *in*, which arrives in [Lesson 15](lesson-15-pointing-placing.md), and [`check.py`](../check.py) refused this lesson when the sentence was first put in it. Until Lesson 15 these four words stand as subjects and as things described: *Pagi hao*, *Rat mal*. ## Asking for help > — Salam! Mi mau doktor. > — Ok. Doktor lai. ## Practice 1. Mi kan anak. — *I see a child.* 2. Rafiki nomu ca. — *The friend drinks tea.* 3. Doktor bil. — *The doctor knows.* 4. Dom keci. — *The house is small.* 5. Din kabir, rat keci. — *The day is big, the night is small.* — *cang* and *duan*, long and short, arrive in Lessons 13 and 16. ## What is still open Tense, plurals, possession and questions have all since been settled — see Lessons [4](lesson-04-tense.md), [5](lesson-05-plural.md), [6](lesson-06-possession.md) and [7](lesson-07-questions.md). Sentence 4 above, first offered as a guess, turned out to follow the rule. The rising-tone question used since Lesson 1 was confirmed as the rule. --- # Lesson 4 — Yesterday, today, tomorrow *Prerequisite: [Lesson 3](lesson-03-people.md).* ## New words | Word | Meaning | |---|---| | suda | past marker — put it before the verb | | saufa | future marker — put it before the verb | | kemarin | yesterday | | besok | tomorrow | **Two of them are the entire tense system of Amadunia.** The other two are the words that say when, and this lesson kept its name for twenty lessons without them: *kemarin* used to arrive in Lesson 24 and *besok* in Lesson 18, so a learner met the past marker twenty lessons before the word *yesterday*. Corrected on September 4, 2026. There is no word for **today**. It is *din ini*, this day, and *ini* arrives in [Lesson 15](lesson-15-pointing-placing.md) — so of the three words in this lesson's title, two are here and the third is a phrase you can build later. ## The verb never changes Take any sentence you already know and drop one word in front of the verb. | Amadunia | English | |---|---| | Mi kula pan. | I eat bread. | | Mi **suda** kula pan. | I ate bread. | | Mi **saufa** kula pan. | I will eat bread. | *kula* stayed *kula*. It will always stay *kula*. ## Everyone, every time | | past | present | future | |---|---|---|---| | I | mi **suda** go | mi go | mi **saufa** go | | you | yu **suda** go | yu go | yu **saufa** go | | she / he | ta **suda** go | ta go | ta **saufa** go | | we | kita **suda** go | kita go | kita **saufa** go | Read down any column: the verb is identical in every row. Read across any row: only the particle moves. ## Where the particle goes Directly before the verb, after the subject. Nowhere else. - Rafiki **suda** lai. — The friend came. - Anak **saufa** nomu akua. — The child will drink water. - Doktor **suda** kan ta. — The doctor saw her. ## Saying when The particle says *whether* it has happened; the time word says *when*. They sit at opposite ends of the sentence — the particle before the verb, the time word at the end. | Amadunia | English | |---|---| | Mi suda kula pan kemarin. | I ate bread yesterday. | | Mi saufa kula pan besok. | I will eat bread tomorrow. | | Rafiki suda lai kemarin. | The friend came yesterday. | | Kita saufa go besok. | We will go tomorrow. | Both words are optional. *Mi suda kula pan* is already past; *kemarin* only says which past. ## A conversation > — Yu suda kula? > — Ya, mi suda kula pan. Yu? > — No. Mi saufa kula. > — Ok. Mi saufa beri ca. *Did you eat? — Yes, I ate bread. You? — No. I will eat. — Ok. I'll give (you) tea.* ## Not repeating yourself Once the time is set, you may leave the particle off until the time changes: > Mi **suda** go dom. Mi kula pan. Mi nomu ca. Mi kan rafiki. > — I went home. I ate bread. I drank tea. I saw a friend. Only the first sentence carries *suda*; the rest inherit it. Putting *suda* in every sentence is also correct — just longer. ## Practice 1. Ta suda sema. — *She spoke.* 2. Kita saufa kan sol. — *We will see the sun.* 3. Anak suda nomu akua. — *The child drank water.* 4. Yu saufa lai dom? — *Will you come home?* 5. Mi suda bil nama. — *I knew the name.* ## What you can already say Four lessons in: greet, ask, want, eat, drink, come, go, give, see, speak, know — in past, present and future, for any subject, without a single verb ever changing shape. Plurals are covered in [Lesson 5](lesson-05-plural.md), possession in [Lesson 6](lesson-06-possession.md), questions in [Lesson 7](lesson-07-questions.md). --- # Lesson 5 — More than one *Prerequisite: [Lesson 4](lesson-04-tense.md).* ## New words | Word | Meaning | |---|---| | cok | many, much | One word. The rest of this lesson needs no new vocabulary at all. ## Say it twice To say "more than one", say the noun twice with a hyphen between. | One | More than one | |---|---| | anak | anak-anak | | dom | dom-dom | | rafiki | rafiki-rafiki | | doktor | doktor-doktor | Nothing is added. Nothing inside the word changes. There is no word in Amadunia this does not work on. ## In a sentence | Amadunia | English | |---|---| | Anak-anak kula pan. | The children eat bread. | | Mi kan dom-dom. | I see houses. | | Rafiki-rafiki suda lai. | The friends came. | | Doktor-doktor saufa sema. | The doctors will speak. | ## When you already said how many The numbers are on [the front page](../README.md), and no lesson teaches them, because there is nothing to teach: twelve words with not one exception between them. This is where they are first used. If there is a number, do not double the noun. The number already told everyone. | Amadunia | English | |---|---| | tri anak | three children | | du dom | two houses | | cok rafiki | many friends | | des doktor | ten doctors | So: *anak-anak* **or** *tri anak* — never *tri anak-anak*. ## When you do not need it at all A plain noun is neither singular nor plural — it is just the thing. *Mi kan anak* is "I see a child" or "I see children", whichever is obvious from where you are. Double the noun only when the "more than one" actually matters and nothing else in the sentence says it. ## A conversation > — Yu suda kan anak-anak? > — Ya. Tri anak. Anak-anak keci! > — Rafiki-rafiki lai? > — Cok rafiki saufa lai. *Did you see the children? — Yes. Three children. The children are small! — Are the friends coming? — Many friends will come.* ## Practice 1. Anak-anak nomu akua. — *The children drink water.* 2. Mi kan du sol? — *Do I see two suns?* 3. Rafiki-rafiki suda kula. — *The friends ate.* 4. Cok din, cok rat. — *Many days, many nights.* 5. Doktor bil dom-dom. — *The doctor knows the houses.* ## What you can already say Five lessons, **59 roots**. Past, present, future; one thing or many; and a grammar in which no word has ever changed its shape. Possession is covered in [Lesson 6](lesson-06-possession.md), questions in [Lesson 7](lesson-07-questions.md). --- # Lesson 6 — Mine, yours, theirs *Prerequisite: [Lesson 5](lesson-05-plural.md).* ## New words | Word | Meaning | |---|---| | es | is, am, are — before a noun | One word, and it earns its place at the end of this lesson. *es* joins a subject to a **noun**: *Nama mi es Luma*, my name is Luma. It never goes before an adjective — *dom mi keci*, my house is small, has no *es* in it and never will. That is the whole rule, and [grammar/copula.md](../grammar/copula.md) gives the reasoning. ## The owner comes second To say whose something is, put the owner right after it. | Amadunia | English | |---|---| | dom mi | my house | | nama yu | your name | | anak ta | her child, his child | | rafiki kita | our friend | You already know *mi*, *yu*, *ta*, *kita*. There is no separate word for *my* — the ordinary pronoun does it. ## Works with any noun The owner does not have to be a pronoun. | Amadunia | English | |---|---| | dom rafiki | the friend's house | | nama anak | the child's name | | ca doktor | the doctor's tea | ## Owner, then description If you also want to describe the thing, the owner comes first, then the adjective. | Amadunia | English | |---|---| | dom mi kabir | my big house | | anak ta keci | his small child | | rafiki yu hao | your good friend | ## Chains Each word belongs to the one before it. Keep going as long as you need. | Amadunia | English | |---|---| | dom rafiki mi | my friend's house | | anak rafiki yu | your friend's child | | nama anak rafiki mi | my friend's child's name | ## In a sentence | Amadunia | English | |---|---| | Mi kan dom yu. | I see your house. | | Rafiki mi saufa lai. | My friend will come. | | Anak ta suda nomu ca. | His child drank tea. | | Kita bil nama doktor. | We know the doctor's name. | ## With plurals Double the noun, then add the owner: | Amadunia | English | |---|---| | anak-anak mi | my children | | dom-dom rafiki | the friend's houses | | tri anak ta | her three children | ## A conversation > — Nama yu? > — Nama mi es Luma. Nama yu? > — Nama mi es Sol. Dom yu kabir? > — No, dom mi keci. Rafiki mi lai! > — Rafiki yu hao? > — Ya, rafiki mi hao. *Your name? — My name is Luma. Yours? — My name is Sol. Is your house big? — No, my house is small. My friend is coming! — Is your friend good? — Yes, my friend is good.* ## Practice 1. Dom mi keci. — *My house is small.* (You met this sentence in Lesson 3 as a guess. It was right.) 2. Mi mau ca yu. — *I want your tea.* 3. Anak-anak rafiki mi kula pan. — *My friend's children eat bread.* 4. Doktor suda kan anak ta. — *The doctor saw her child.* 5. Nama rafiki yu? — *Your friend's name?* ## What you can already say Six lessons, **60 roots**, and still not one word has changed its shape. You can now say whose anything is, to any depth, using only words you already had. Questions are covered in [Lesson 7](lesson-07-questions.md). --- # Lesson 7 — Asking *Prerequisite: [Lesson 6](lesson-06-possession.md).* ## New words | Word | Meaning | |---|---| | ke | what | | kim | who | | nali | where | | kab | when | | porke | why | | kaifa | how | | berapa | how many, how much | | taksi | taxi | | foto | photo | The last two are not question words. They are the first of the already-global words — the ones design rule 5 says stay as they are, because most of the world can already say them. *taksi* turns up in the examples below; you learned both the moment you read them. ## You already know how to ask yes or no Since Lesson 1 you have been raising your voice at the end of a sentence to make it a question. That is the rule, and it is the whole rule. | | | |---|---| | Yu mau ca. | You want tea. | | Yu mau ca? | Do you want tea? | Nothing is added. Nothing moves. ## For everything else: swap one word Take the sentence that would be the answer. Replace the thing you do not know with the question word. Do not move anything. | Answer | Question | | |---|---|---| | Yu kan **anak**. | Yu kan **ke**? | What do you see? | | **Rafiki** lai. | **Kim** lai? | Who is coming? | | Yu go **dom**. | Yu go **nali**? | Where are you going? | | Yu lai **din**. | Yu lai **kab**? | When do you come? | | — | Yu go **kaifa**? | How do you go? | | Yu kan **tri** anak. | Yu kan **berapa** anak? | How many children do you see? | Look down the two columns. Every question is its own answer with one word changed. There is no other pattern to learn. **One row has no answer yet, and that is honest.** *Yu go kaifa?* asks how, and the language cannot yet say *by taxi* or *quickly*: the adverb arrives in [Lesson 12](lesson-12-colors-directions.md) and *una*, with, in [Lesson 15](lesson-15-pointing-placing.md). The question word is taught here because it belongs with the others; its answers come later. ## Why *porke* goes at the end: > — Yu mau ca porke? > — Mi mau ca. Ca hao! *Why do you want tea? — I want tea. Tea is good!* ## How many *berapa* works exactly like a number: it goes before the noun, and the noun stays single. | | | |---|---| | tri anak | three children | | berapa anak? | how many children? | | cok rafiki | many friends | | berapa rafiki? | how many friends? | ## Asking a name Lesson 6 asked *Nama yu?* — your name? — and that works. The full form puts the question word where the name would go: > — Nama yu es ke? > — Nama mi es Sol. Both are correct. The short one leans on tone; the full one does not need it. ## A conversation > — Salam! Nama yu es ke? > — Nama mi es Luma. Yu es kim? > — Mi es rafiki Sol. Sol nali? > — Sol suda go dom. > — Ta saufa lai kab? > — Rat. Yu mau ca? > — Ya! Berapa rafiki suda nomu ca yu? > — Cok! Cok rafiki suda nomu ca mi. *Hello! What's your name? — My name is Luma. Who are you? — I'm Sol's friend. Where is Sol? — Sol went home. — When will she come? — Tonight. Do you want tea? — Yes! How many friends drank your tea? — Many! Many friends drank my tea.* ## Practice 1. Yu kula ke? — *What do you eat?* 2. Kim suda kan anak? — *Who saw the child?* 3. Doktor nali? — *Where is the doctor?* 4. Rafiki yu saufa lai kab? — *When will your friend come?* 5. Yu bil nama ta kaifa? — *How do you know her name?* 6. Berapa dom? — *How many houses?* 7. Ta go porke? — *Why does she go?* 8. Foto nali? — *Where is the photo?* ## What you can already say Seven lessons, **69 roots**. Statements, past and future, one or many, whose, and now every kind of question — and still not one word has ever changed its shape or its place. Pronouns are covered in [Lesson 9](lesson-09-pronouns.md), the copula in [Lesson 11](lesson-11-being.md), negation in [Lesson 14](lesson-14-not-and.md), answering a negative question in [Lesson 15](lesson-15-pointing-placing.md). --- # Lesson 8 — Getting by *Prerequisite: [Lesson 7](lesson-07-questions.md).* ## New words | Word | Meaning | |---|---| | mersi | thank you | | plis | please | | pardon | sorry, excuse me | | madad | help | | pesa | money | | market | market, shop | | rabota | work | | insan | person | | sini | here | | sasa | now | | garam | hot | | barid | cold | Most of these you already know. *mersi*, *pardon*, *plis*, *market* travel the world as they are; *insan*, *madad*, *pesa* are shared across a dozen languages each. Amadunia only wrote them down. **Why *madad* is not in a sentence here.** It is glossed "help; to help", a noun and a verb at once. Until that is settled, *Mi mau madad* is two sentences at the same time — *I want help* and *I want to help* — so this lesson teaches the word and waits. See [the briefing](../grammar/proposal-two-jobs.md). This paragraph called *madad* **the only such root** until September 5, 2026. There are two, and the second is *rabota*, taught in the table above and used three times on this page — *Mi rabota sini*, *Yu rabota nali*, *Insan-insan rabota sasa*. The audit that produced "the only" read the dictionary's glosses and not the sentences, and [verb-chains.md](../grammar/verb-chains.md) recorded the correction on September 4 without this lesson hearing about it. *rabota* cannot be held back the way *madad* is: the course needs both of its readings already, which is exactly what makes its decision the expensive one. ## The three words that open every door | | | |---|---| | Plis. | Please. | | Mersi. | Thank you. | | Pardon. | Sorry. / Excuse me. | They work alone, and they work at either end of a sentence: > — Pardon! Market nali? > — Market sini. > — Mersi! *Excuse me! Where is the market? — The market is here. — Thank you!* ## Asking for help > — Madad! Plis! > — Ke? > — Mi mau doktor. > — Ok. Doktor saufa lai. *Help! Please! — What? — I want a doctor. — Ok. A doctor will come.* ## At the market > — Salam! Yu mau ke? > — Mi mau ca garam. Berapa pesa? > — Du pesa. > — Ok. Mersi! *Hello! What do you want? — I want hot tea. How much money? — Two pesa. — Ok. Thanks!* *berapa pesa* — literally "how many money" — is the whole of asking a price. ## Here and now | | | |---|---| | Mi sini. | I am here. | | Rafiki mi sini? | Is my friend here? | | Mi mau ca sasa. | I want tea now. | | Doktor saufa lai sasa. | The doctor will come now. | ## Hot and cold | | | |---|---| | Ca garam. | The tea is hot. | | Akua barid. | The water is cold. | | Din garam, rat barid. | The day is hot, the night is cold. | | Mi mau akua barid, plis. | I want cold water, please. | ## Work and people | | | |---|---| | Mi rabota sini. | I work here. | | Yu rabota nali? | Where do you work? | | Insan hao. | A good person. / The person is good. | | Cok insan sini. | Many people here. | ## Practice 1. Pardon, yu bil market nali? — *Excuse me, do you know where the market is?* 2. Mersi, rafiki! — *Thanks, friend!* 3. Mi mau ca, plis. — *I want tea, please.* 4. Berapa pesa? — Tri pesa. — *How much? — Three.* 5. Insan-insan rabota sasa. — *The people are working now.* 6. Ca yu garam? — No, barid. — *Is your tea hot? — No, cold.* ## What you can already say Eight lessons, **81 roots**. You can now be polite, ask for help, find a place, ask a price, say where you are and when, and describe heat and cold — enough to get through a day somewhere you have never been. **A note on word order.** *sini* and *sasa* go at the end of the sentence, place before time. This became the rule in [Lesson 15](lesson-15-pointing-placing.md). --- # Lesson 9 — We, you all, they *Prerequisite: [Lesson 8](lesson-08-getting-by.md).* ## New words None. Every form in this lesson is built from words you have. ## Say it twice You know *mi*, *yu*, *ta*. You know that saying a word twice makes it "more than one". Put those together: | one | more than one | |---|---| | mi — I | mi-mi — we | | yu — you | yu-yu — you all | | ta — he, she, it | ta-ta — they | | Amadunia | English | |---|---| | Ta-ta suda lai. | They came. | | Yu-yu mau ca? | Do you all want tea? | | Mi-mi rabota sini. | We work here. | | Dom ta-ta nali? | Where is their house? | ## Two kinds of "we" You also know *kita*. It means "we" too — but a particular we. | | | | |---|---|---| | mi-mi | we, **not** you | me and my people | | kita | we, **with** you | you and me, together | > — Mi-mi go market sasa. Yu-yu? > — No, mi-mi sini. > — Ok. Kita saufa kula una! *We're going to the market now. You all? — No, we're staying here. — Ok. We'll all eat together!* The last line said *Kita saufa kula una rat* — "tonight" — until September 6, 2026. [Lesson 15](lesson-15-pointing-placing.md) puts a noun after *una* to mean *with*, so *una rat* is "with the night", and the time word had to come out: at Lesson 09 there is no *in* to hold it apart. The last line uses *kita* because the speaker means everyone — both groups. ## One word, three jobs *ta* is he, she, **and it**. | | | |---|---| | Doktor lai. Ta hao. | The doctor is coming. She is good. | | Ca garam. Mi mau ta. | The tea is hot. I want it. | | Anak keci. Ta suda kula? | The child is small. Has he eaten? | ## Nothing changes The same pronoun whether it does the action or receives it: | | | |---|---| | Mi kan ta. | I see her. | | Ta kan mi. | She sees me. | | Yu-yu bil mi-mi? | Do you all know us? | | Mi-mi bil yu-yu. | We know you all. | No *I / me*, no *she / her*, no *we / us*. One form each. ## A conversation > — Salam! Yu-yu es kim? > — Mi-mi es rafiki Sol. Sol sini? > — Ya, ta sini. Ta-ta rabota. > — Ta-ta es kim? > — Sol. Rafiki ta. Anak-anak ta. Kita kan ta-ta? > — Ya! Mersi! *Hello! Who are you all? — We're Sol's friends. Is Sol here? — Yes, she's here. They're working. — Who are they? — Sol. Her friend. Her children. Do we see them? — Yes! Thanks!* ## Practice 1. Ta-ta mau akua. — *They want water.* 2. Yu-yu suda kan doktor? — *Did you all see the doctor?* 3. Mi-mi saufa lai kab? — *When will we come?* (we, not you) 4. Kita saufa lai sasa. — *We'll come now.* (all of us) 5. Nama ta-ta es ke? — *What are their names?* 6. Pan garam. Mi mau ta! — *The bread is hot. I want it!* ## What you can already say Nine lessons, **81 roots**. You can now talk about anyone — one person or many, with you or without you, a person or a thing — and the words for them have never once changed their shape. **Two corrections, September 5, 2026.** Practice 5 read *Nama ta-ta ke?* with without *es*. A question word stands where the answer stands, and the answer to that one is a name — a noun, which takes *es*. Lessons 07 and 11 and the phrasebook all write *Nama yu es ke?*; this was the only one of the seven without it. And the conversation answered *Ta-ta es kim?* with a comma list — *Sol, rafiki ta, anak-anak ta*. [conjunction.md](../grammar/conjunction.md) lets commas stand in for all but the last, so a list still needs *aur* at the end, and *aur* does not arrive until [Lesson 14](lesson-14-not-and.md). The lesson had written something it had no way to write. It is three separate answers now, which the fragment rule grants and which this lesson can say. *there* is covered in [Lesson 10](lesson-10-a-day.md), the copula in [Lesson 11](lesson-11-being.md), *not* in [Lesson 14](lesson-14-not-and.md), *this* and *that* in [Lesson 15](lesson-15-pointing-placing.md). --- # Lesson 10 — A day *Prerequisite: [Lesson 9](lesson-09-pronouns.md).* ## New words | Word | Meaning | |---|---| | lala | to sleep | | otur | to sit | | suru | to do, to make | | ambil | to take | | kupi | to buy | | situ | there | | sar | head | | ain | eye | | mano | hand | | familia | family | | hotel | hotel | | telefon | telephone | ## Here and there *sini* has its pair now. | | | |---|---| | Mi otur sini. | I sit here. | | Yu otur situ. | You sit there. | | Hotel nali? — Situ. | Where is the hotel? — There. | | Familia mi situ. | My family is there. | ## Doing things | | | |---|---| | Yu suru ke? | What are you doing? | | Mi suru rabota. | I'm working. — *literally: I do work* | | Mi ambil telefon. | I take the telephone. | | Ta suda kupi pan. | She bought bread. | | Kita saufa kupi ca. | We'll buy tea. | ## Telling someone to do it Leave the subject out and the sentence becomes a command. Nothing is added. | | | |---|---| | Otur sini. | Sit here. | | Kula pan. | Eat bread. | | Lai sini. | Come here. | *Yu otur sini* is different — it is a statement about you. A command has no subject at all, and that is what makes it one. [The rule](../grammar/sentence-types.md). The third row read *No go!* until September 6, 2026, which put *no* in front of a predicate four lessons before [Lesson 14](lesson-14-not-and.md) teaches it to go there, in a single line with nothing said about it. [Lesson 26](lesson-26-telling-and-answering.md) is where a command is denied, and it explains the rule it is using. ## The body | | | |---|---| | Sar mi garam. | My head is hot. | | Ain ta kabir. | Her eyes are big. — *or:* her eye is big | | Mano-mano anak keci. | The child's hands are small. | | Mi kan yu. Ain mi hao. | I see you. My eyes are good. | *ain* on its own is neutral — one eye or two, as the situation makes obvious. *ain-ain* when the "two" matters. ## A day, start to finish > Din. Mi ambil telefon. Mi suru rabota. > Mi go market. Mi kupi pan. Mi kupi ca. Mi kupi akua. > Mi lai dom. Familia mi situ. Kita kula una. > Rat. Mi otur. Mi lala. *Day. I take the phone. I work. I go to the market. I buy bread. I buy tea. I buy water. I come home. My family is there. We eat together. Night. I sit. I sleep.* Every verb in that story is in its one and only form. Not one of them changed for *I*, for *we*, for past or present. ## A conversation > — Pardon, hotel nali? > — Situ. Yu-yu es familia? > — Ya, mi-mi es familia. Anak-anak mi-mi saufa lala. > — Ok. Hotel hao. Yu-yu saufa lala. > — Mersi! *Excuse me, where is the hotel? — There. Are you a family? — Yes, we're a family. Our children will sleep. — Ok. The hotel is good. You'll sleep. — Thanks!* ## Practice 1. Yu lala kab? — *When do you sleep?* 2. Mi suda kupi telefon. — *I bought a telephone.* 3. Otur sini, plis. — *Sit here, please.* 4. Sar ta mal? — *Is his head bad?* (does his head hurt?) 5. Familia yu situ? — *Is your family there?* 6. Ta-ta suru ke sasa? — *What are they doing now?* ## What you can already say Ten lessons, **93 roots** — past the first milestone of 80. You can now tell the story of a whole day, name the body, point here and there, buy and take and make and sleep. **A note on word order.** *situ*, like *sini* and *sasa*, sits at the end of its sentence — place before time. This became the rule in [Lesson 15](lesson-15-pointing-placing.md). --- # Lesson 11 — Being *Prerequisite: [Lesson 10](lesson-10-a-day.md).* ## No new word | Word | Meaning | Taught in | |---|---|---| | es | is, am, are — before a noun | [Lesson 6](lesson-06-possession.md) | Nothing new here. *es* arrived five lessons ago, in one line at the foot of the possession lesson, because Lesson 7 could not ask *Yu es kim?* without it. It is the most widely shared word on Earth — Spanish *es*, English *is*, German *ist*, Russian *est'*, Hindi *hai*, Persian *ast*, Latin *est* — and you have been using it since. This lesson is where it gets its rule stated in full. This page called it a new word until September 5, 2026, which meant a learner was told twice that the same word was new. ## Before a noun: es | | | |---|---| | Mi es doktor. | I am a doctor. | | Yu es rafiki mi. | You are my friend. | | Sol es anak. | Sol is a child. | | Kita es familia. | We are a family. | | Nama mi es Luma. | My name is Luma. | ## Before anything else: nothing You have been doing this since Lesson 3. Nothing changes. | | | |---|---| | Dom kabir. | The house is big. | | Ca garam. | The tea is hot. | | Mi sini. | I am here. | | Familia mi situ. | My family is there. | ## Why the difference Put two nouns together and Amadunia already reads them: the second one owns the first. | | | |---|---| | doktor rafiki mi | my friend's doctor | | Rafiki mi **es** doktor. | My friend is a doctor. | Without *es* you have said whose doctor. With *es* you have said what your friend is. An adjective cannot own anything, so it never needs the help. ## Was, will be *es* is a verb. Put the tense word in front of it: | | | |---|---| | Mi suda es anak. | I was a child. | | Yu saufa es doktor. | You will be a doctor. | Adjectives and place words take the tense word themselves — no *es*: | | | |---|---| | Dom suda kabir. | The house was big. | | Ca saufa garam. | The tea will be hot. | | Ta-ta suda sini. | They were here. | ## Asking | | | |---|---| | Nama yu es ke? | What is your name? | | Yu es kim? | Who are you? | | Ta es doktor? | Is she a doctor? | | Dom kabir? | Is the house big? | ## A conversation > — Salam! Nama yu es ke? > — Nama mi es Sol. Mi es doktor. Yu? > — Mi es rafiki Luma. Luma sini? > — No, Luma situ. Ta suda es anak keci — sasa ta es doktor! > — Ta hao? > — Ya, ta hao. *Hello! What's your name? — My name is Sol. I'm a doctor. You? — I'm Luma's friend. Is Luma here? — No, Luma is there. She was a small child — now she's a doctor! — Is she good? — Yes, she is good.* ## Practice 1. Mi es insan. — *I am a person.* 2. Ain ta kabir. — *Her eyes are big.* 3. Yu-yu es familia? — *Are you a family?* 4. Ta suda es rafiki mi. — *He was my friend.* 5. Hotel es dom? — *Is a hotel a house?* 6. Din saufa garam. — *The day will be hot.* ## What you can already say Eleven lessons, **93 roots** — not one of them new on this page. You can say what anything is, was or will be, and what it is like, and the whole of it rests on one word you already had. Negation is covered in [Lesson 14](lesson-14-not-and.md) — *Mi no es doktor*. See the open questions in [grammar/](../grammar/). --- # Lesson 12 — Colours and directions *Prerequisite: [Lesson 11](lesson-11-being.md).* ## New words | Word | Meaning | |---|---| | merah | red | | asul | blue | | upar | up, above | | sub | down, below | | baru | new | | eski | old | | hayai | fast | | dugu | sibling | | kara | to read | | katab | to write | | anda | to walk | | musik | music | ## Colours describe, so they take no *es* A colour is an adjective. It follows the noun, and nothing stands between them. | | | |---|---| | Dom merah. | The house is red. | | Akua asul. | The water is blue. | | Ca mi merah. | My tea is red. | | Mi kan telefon asul. | I see a blue telephone. | Compare with a noun, which does take *es*: | | | |---|---| | Sol es doktor. | Sol is a doctor. | | Sol merah. | Sol is red. | ## Up and down | | | |---|---| | Sol upar. | The sun is above. | | Anak sub. | The child is below. | | Familia mi upar. | My family is upstairs. | | Ain mi kan upar. | My eyes look up. | ## New and old | | | |---|---| | Telefon mi baru. | My telephone is new. | | Hotel eski. | The hotel is old. | | Mi suda kupi dom baru. | I bought a new house. | | Rafiki eski hao. | An old friend is good. | Two adjectives in a row? Put the owner first, then describe: *dom mi baru* — my new house. ## Reading and writing *kara* and *katab* come from the same tradition: *kara* gave the world *Quran* — "that which is read" — and *katab* gave *kitab*, the word for "book" in Arabic, Turkish, Hindi, Urdu, Swahili and Indonesian alike. | | | |---|---| | Mi kara. | I read. | | Yu katab ke? | What are you writing? | | Anak suda kara. | The child read. | | Doktor saufa katab nama mi. | The doctor will write my name. | ## Walking, fast An adjective placed straight after the verb describes the verb, not the noun. That is the whole rule — Amadunia has no separate class of adverbs and no ending to add, the way English adds *-ly*. | | | |---|---| | Mi anda market. | I walk to the market. | | Ta anda hayai. | She walks fast. | | Kita suda anda cok. | We walked a lot. | | Taksi hayai. | The taxi is fast. | ## Sibling — one word for both *dugu* is a brother **or** a sister, exactly as *ta* is he or she. Amadunia does not split people by gender anywhere. | | | |---|---| | Dugu mi es doktor. | My sibling is a doctor. | | Dugu-dugu mi situ. | My siblings are there. | | Mi es dugu yu. | I am your sibling. | ## A conversation > — Salam! Dom yu nali? > — Dom mi situ, upar. Dom merah. > — Dom baru? > — No, eski. Hao! > — Dugu yu situ? > — Ya. Ta kara, mi katab. Kita ama musik. *Hello! Where is your house? — My house is there, above. The house is red. — Is it new? — No, old. Good! — Is your sibling there? — Yes. She reads, I write. We love music.* ## Practice 1. Telefon mi es baru. — *careful:* **wrong.** *baru* is an adjective — say *Telefon mi baru.* 2. Sol upar, anak sub. — *The sun is above, the child below.* 3. Yu kara ke? — *What are you reading?* 4. Dugu mi anda hayai. — *My sibling walks fast.* 5. Mi suda katab nama yu. — *I wrote your name.* 6. Musik yu merah? — *Is your music red?* (nonsense, but grammatical) ## What you can already say Twelve lessons, **105 roots** — past one hundred. You can colour things, place them above and below, call them new or old or fast, read, write, walk, and talk about your sibling without ever naming a gender. *not* is covered in [Lesson 14](lesson-14-not-and.md), *this* and *that* in [Lesson 15](lesson-15-pointing-placing.md). --- # Lesson 13 — Weather and directions *Prerequisite: [Lesson 12](lesson-12-colors-directions.md).* ## New words | Word | Meaning | |---|---| | angin | wind | | yuki | snow | | yamin | right | | kiri | left | | cang | long | | sundar | beautiful | | asan | easy | | buka | to open | | faham | to understand | | studi | to learn | | espera | to wait | | kimbia | to run | ## Weather Weather words are nouns. Describe them with an adjective and no *es*: | | | |---|---| | Angin barid. | The wind is cold. | | Yuki asul. | The snow is blue. | | Din garam, rat barid. | The day is hot, the night is cold. | | Yuki sundar! | The snow is beautiful! | But say **what** something is, and *es* comes back: | | | |---|---| | Yuki es akua. | Snow is water. | ## Right and left | | | |---|---| | Dom mi yamin. | My house is on the right. | | Market kiri. | The market is on the left. | | Mano yamin, mano kiri. | Right hand, left hand. | | Anda yamin, plis. | Walk right, please. | *yamin* comes from Arabic *yamīn*, "the right hand" — the same word that named **Yemen**, the land on Arabia's right. ## Long and beautiful and easy | | | |---|---| | Din cang. | The day is long. | | Rat cang. Mi espera. | The night is long. I wait. | | Dugu mi sundar. | My sibling is beautiful. | | Amadunia asan! | Amadunia is easy! | ## Five new verbs This heading said **four** until September 5, 2026, and the closing line named the same four — *kimbia*, to run, was taught in the table above, shown in the last row below, and left out of both counts. | | | |---|---| | Mi buka telefon. | I open the telephone. | | Yu faham? | Do you understand? | | Mi studi Amadunia. | I am learning Amadunia. | | Kita espera dugu yu. | We are waiting for your sibling. | | Anak kimbia hayai. | The child runs fast. | *espera* is Spanish and Portuguese *esperar*, to wait and to hope — the same root that gave **Esperanto** its name, "one who hopes". Amadunia takes the word and keeps the hope. ## With tense Nothing new. The particle goes before the verb, and the verb never changes: | | | |---|---| | Mi suda studi. | I learned. | | Yu saufa faham. | You will understand. | | Ta-ta suda espera cok. | They waited a lot. | | Anak suda kimbia sub. | The child ran down. | ## A conversation > — Salam! Yu faham Amadunia? > — Ya! Mi studi sasa. Amadunia asan. > — Hao! Dom yu nali? > — Dom mi yamin, upar. Dom merah. > — Angin barid sasa. Yuki? > — Ya, yuki. Yuki sundar! > — Kita anda? Mi espera yu. > — Ok! Mi kimbia! *Hello! Do you understand Amadunia? — Yes! I'm learning now. Amadunia is easy. — Good! Where is your house? — My house is on the right, above. The red house. — The wind is cold now. Snow? — Yes, snow. The snow is beautiful! — Shall we walk? I'll wait for you. — Ok! I'm running!* ## Practice 1. Angin hayai. — *The wind is fast.* 2. Yu faham mi? — *Do you understand me?* 3. Market yamin, hotel kiri. — *The market is on the right, the hotel on the left.* 4. Mi suda studi din cang. — *I studied a long day.* 5. Buka ain yu! — *Open your eyes!* 6. Dugu mi kimbia hayai. — *My sibling runs fast.* ## What you can already say Thirteen lessons, **117 roots**. Weather, directions, and five verbs that carry a lot: open, understand, learn, wait, run. You can now say you are learning the language, in the language. *not* and *and* are covered in [Lesson 14](lesson-14-not-and.md); *this*, *that* and place marking in [Lesson 15](lesson-15-pointing-placing.md). --- # Lesson 14 — Not, and *Prerequisite: [Lesson 13](lesson-13-weather-directions.md).* ## New words | Word | Meaning | |---|---| | aur | and | One new word. The other half of this lesson is a word you have had since Lesson 1. ## *no* was waiting all along You have used *no* to answer since your first conversation. It does a second job with no change at all: put it in front of what you are denying. | | | |---|---| | Mi kula pan. | I eat bread. | | Mi **no** kula pan. | I do not eat bread. | | Mi es doktor. | I am a doctor. | | Mi **no** es doktor. | I am not a doctor. | | Dom kabir. | The house is big. | | Dom **no** kabir. | The house is not big. | | Mi sini. | I am here. | | Mi **no** sini. | I am not here. | There is no separate negative form of any verb. *kula* stays *kula*. ## With past and future *no* comes first, then the time word: | | | |---|---| | Mi **no suda** kula. | I did not eat. | | Yu **no saufa** lai. | You will not come. | | Ta **no suda es** anak. | She was not a child. | The order never changes: **no → suda/saufa → verb**. ## Answering, as before > — Yu mau ca? > — No. > — Yu mau ca? > — No, mi no mau ca. Mi mau akua. ## *aur* joins | | | |---|---| | Mi kupi pan **aur** ca. | I buy bread and tea. | | Angin **aur** yuki lai. | Wind and snow come. | | Dom mi keci **aur** sundar. | My house is small and beautiful. | | Mi lai **aur** mi kan yu. | I come and I see you. | Nouns, adjectives, whole sentences — one word for all three. For a list, put *aur* between each pair: > Mi kupi pan **aur** ca **aur** akua. In writing you may use commas for all but the last: *pan, ca aur akua*. The full form is never wrong. ## Why you could not just say *pan ca* Because you already learned what that means. Two nouns together is possession — *pan ca* is "the tea's bread". The slot was taken, so "and" needed a word of its own. ## A conversation > — Salam! Yu mau ca aur pan? > — Mi mau ca. Mi no mau pan, mersi. > — Yu no kula sasa? > — No. Mi suda kula. Mi mau ca aur akua. > — Ok. Ca garam aur akua barid. > — Mersi! Yu es rafiki hao. *Hello! Do you want tea and bread? — I want tea. I don't want bread, thanks. — You're not eating now? — No. I ate. I want tea and water. — Ok. Hot tea and cold water. — Thanks! You're a good friend.* ## Practice 1. Mi no faham. — *I do not understand.* 2. Anak no suda lala. — *The child did not sleep.* 3. Sol aur luma. — *The sun and the light.* 4. Dom mi no kabir aur no keci. — *My house is not big and not small.* 5. Yu no es doktor? — *Are you not a doctor?* 6. Mi studi aur mi faham. — *I study and I understand.* ## What you can already say Fourteen lessons, **118 roots**. With one new word and one old word doing new work, you can now deny anything and join anything — the two things a language cannot do without. *This*, *that*, place marking and answering a negative question are all covered in [Lesson 15](lesson-15-pointing-placing.md). --- # Lesson 15 — Pointing and placing *Prerequisite: [Lesson 14](lesson-14-not-and.md).* ## New words | Word | Meaning | |---|---| | ini | this | | itu | that | | in | at, in, on | | dari | from | | por | to, for | | o | or | Six little words. After this lesson the grammar you need to get through a day is complete. ## This and that They go after the noun, and after everything else in the phrase: | | | |---|---| | dom ini | this house | | anak itu | that child | | Ca ini garam. | This tea is hot. | | Insan itu es doktor. | That person is a doctor. | | dom mi kabir ini | this big house of mine | On their own they are a subject, and before a noun they take *es*: | | | |---|---| | Ini es dom mi. | This is my house. | | Itu es ke? | What is that? | | Ini hao, itu mal. | This is good, that is bad. | You have met their shape before: *s-ini* is here, *s-itu* is there. *din ini* is today. *rat ini* is tonight. ## At, from, for | | | |---|---| | Mi in dom. | I am at home. | | Anak lala in dom. | The child sleeps in the house. | | Mi lai dari market. | I come from the market. | | Yu dari nali? | Where are you from? | | Mi beri pan por yu. | I give bread to you. | | Ca ini por kim? — Por mi! | Who is this tea for? — For me! | *por* is the *por* in *porke* — "for what", why. Going somewhere needs no word at all. After *go, lai, anda, kimbia*, the place is the destination: | | | |---|---| | Mi go market. | I go to the market. | | Ta lai dom. | She comes home. | ## Or | | | |---|---| | Ca o akua? | Tea or water? | | Mi mau pan o ca. | I want bread or tea. | | Yu lai din ini o din itu? | Are you coming today or that day? | ## With *una* was "together". Put a noun after it and it is "with": | | | |---|---| | Kita kula una. | We eat together. | | Kita kula una familia. | We eat with the family. | | Mi go market una rafiki mi. | I go to the market with my friend. | ## Answering *no*-questions | | | |---|---| | — Yu no mau ca? | Don't you want tea? | | — No. | I don't. | | — Ya. | I do. | *ya* and *no* answer the **fact**, never the shape of the question. *Ya* means it is so; *no* means it is not. When it matters, answer with the verb: *Mi mau.* / *Mi no mau.* ## Where the little phrases go Place first, then time, both at the end: | | | |---|---| | Mi kula pan in dom din ini. | I eat bread at home today. | | Kita saufa lai dari market rat ini. | We'll come from the market tonight. | ## A conversation > — Salam! Yu dari nali? > — Mi dari dom. Yu go nali? > — Mi go market. Yu lai una mi? > — Ya! Mi mau pan o ca. > — Pan ini hao. Pan por familia yu? > — Ya, por familia mi. Aur ca por mi! > — Kita go sasa. Market situ, kiri. *Hello! Where are you from? — From home. Where are you going? — To the market. Coming with me? — Yes! I want bread or tea. — This bread is good. Bread for your family? — Yes, for my family. And tea for me! — Let's go now. The market is there, on the left.* ## Practice 1. Dom ini kabir aur dom itu keci. — *This house is big and that house is small.* 2. Mi lai dari market una dugu mi. — *I come from the market with my sibling.* 3. Ca ini por yu o por mi? — *Is this tea for you or for me?* 4. Anak-anak lala in dom rat ini. — *The children sleep at home tonight.* 5. Yu no faham? — Ya, mi faham. — *Don't you understand? — Yes, I do.* 6. Ini es ke? — Itu es telefon. — *What is this? — That is a telephone.* ## What you can already say Fifteen lessons, **124 roots**. You can point at things, place them, say where they came from and who they are for, offer a choice, go somewhere with someone, and answer any question without ambiguity. That is the grammar of A1, and it is done. Existence — how to say "there is" — and comparison were both settled the next day; see [Lesson 18](lesson-18-comparing-and-joining.md). --- # Lesson 16 — School and time *Prerequisite: [Lesson 15](lesson-15-pointing-placing.md).* ## New words | Word | Meaning | |---|---| | skola | school | | libro | book | | kursi | chair | | tarik | road, way | | hora | hour | | tahun | year | | yamur | rain | | fikir | to think | | funga | to close | | tena | again | | duan | short | | lebi | more | ## At school | | | |---|---| | Mi studi in skola. | I study at school. | | Skola ini kabir. | This school is big. | | Anak-anak in skola sasa. | The children are at school now. | | Mi kara libro in skola. | I read a book at school. | ## Books and chairs | | | |---|---| | Libro ini sundar. | This book is beautiful. | | Mi otur in kursi. | I sit on a chair. | | Libro itu por dugu mi. | That book is for my sibling. | | Mi suda kupi libro dari market. | I bought a book from the market. | ## Thinking | | | |---|---| | Mi fikir. | I think. | | Mi fikir cok. | I think a lot. | | Yu fikir ke? | What do you think? | | Ta suda fikir aur ta suda faham. | She thought and she understood. | ## Open and close *buka* you have had since Lesson 13. Its pair arrives now: | | | |---|---| | Mi buka libro. | I open the book. | | Mi funga libro. | I close the book. | | Buka ain yu! | Open your eyes! | | Skola funga rat ini. | The school closes tonight. | ## Again | | | |---|---| | Mi lai tena. | I come again. | | Sema tena, plis. | Say it again, please. | | Rafiki mi saufa lai tena. | My friend will come again. | | Mi no studi tena. | I do not study again. | *tena* goes at the end, where time words go. ## Long and short, more | | | |---|---| | Tarik ini cang. | This road is long. | | Tarik itu duan. | That road is short. | | Din cang, rat duan. | The day is long, the night is short. | | Mi mau lebi ca. | I want more tea. | | Lebi libro, plis! | More books, please! | *lebi* goes before the noun, exactly where *cok* goes. ## Hours and years | | | |---|---| | Uan hora. | One hour. | | Tri hora. | Three hours. | | Mi suda studi du tahun. | I studied for two years. | | Anak itu **es** fai tahun. | That child is five years old. | Remember: after a number the noun stays single. *Tri hora*, never *tri hora-hora*. The age line carried no *es* until September 5, 2026. *fai tahun* is a noun phrase, and a noun predicate takes *es* — *itu* closes the phrase before it, so *Anak itu fai tahun* was not a sentence at all. ## Rain | | | |---|---| | Yamur lai. | Rain comes. | | Yamur barid. | The rain is cold. | | Mi no go market — yamur! | I'm not going to the market — rain! | | Yamur aur angin din ini. | Rain and wind today. | ## A conversation > — Salam! Yu go nali? > — Mi go skola. Yu? > — Mi in dom. Yamur situ! > — Ya, yamur aur angin. Tarik dari dom por skola cang. > — Berapa hora? > — Du hora. Mi no anda in yamur. > — Yu otur in dom aur yu kara libro. > — Libro ke? > — Libro ini! Mi suda kara tena aur tena. Libro hao. > — Mersi! Mi saufa kara sasa. *Hello! Where are you going? — To school. You? — I'm at home. Rain out there! — Yes, rain and wind. The road from home to school is long. — How many hours? — Two hours. I don't walk in the rain. — You sit at home and you read a book. — Which book? — This book! I've read it again and again. It's a good book. — Thanks! I'll read it now.* ## Practice 1. Mi kara libro in skola din ini. — *I read a book at school today.* 2. Tarik dari market por dom duan. — *The road from the market to home is short.* 3. Yu fikir ke? — *What do you think?* 4. Funga libro aur otur. — *Close the book and sit.* 5. Mi no suda studi tahun itu. — *I did not study that year.* 6. Lebi ca, plis. Aur pan! — *More tea, please. And bread!* ## What you can already say Sixteen lessons, **136 roots**. School, books, roads, hours, years, rain — and *tena*, which lets anything happen twice. **A word this lesson could not add.** *bisa* — "can, be able" — was chosen for this batch and dropped, because *Yu bisa otur* would have put one verb in front of another and that had never been decided. It was settled the next day: see [Lesson 17](lesson-17-wanting-and-able.md). Every lesson up to this one keeps to one verb at a time. --- # Lesson 17 — Wanting, being able, having to *Prerequisite: [Lesson 16](lesson-16-school-and-time.md).* ## New words | Word | Meaning | |---|---| | bisa | can, to be able | | lasim | must, to have to | ## Just put the verbs together | | | |---|---| | Mi mau kula. | I want to eat. | | Mi mau kula pan. | I want to eat bread. | | Yu bisa otur. | You can sit. | | Yu bisa otur sini. | You can sit here. | | Ta lasim go skola. | He must go to school. | | Kita lasim funga libro. | We must close the book. | There is no word for the "to" of *to eat*. Nothing goes between the verbs, and neither verb changes. (*por* is the preposition "to, for", as in *por yu* — a different word doing a different job.) Until now you could only say *Mi mau pan* — "I want bread". Now you can want to **do** things. ## Past and future The time word goes in front of the first verb: | | | |---|---| | Mi suda mau kula. | I wanted to eat. | | Yu saufa bisa kara. | You will be able to read. | | Ta suda lasim rabota. | He had to work. | ## Saying no *no* also goes in front: | | | |---|---| | Mi no mau kula. | I do not want to eat. | | Ta no bisa lai. | She cannot come. | | Yu no lasim go. | You do not have to go. | | Mi no suda mau studi. | I did not want to study. | ## Moving *no* moves the meaning You already know *no* goes in front of what it denies. Put it in a different place and it denies a different thing: | | | |---|---| | Mi **no** mau kula. | I do not want to eat. | | Mi mau **no** kula. | I want *not* to eat. | Nothing new to learn — the old rule, applied where there are now two verbs. ## Three verbs | | | |---|---| | Mi mau studi kara. | I want to learn to read. | | Yu bisa studi katab. | You can learn to write. | | Kita lasim studi faham. | We must learn to understand. | There is no limit. The verbs line up, the object goes last. ## A conversation > — Salam! Yu mau go market una mi? > — Mi no bisa. Mi lasim studi. > — Yu studi ke? > — Mi studi Amadunia. Mi mau kara libro ini. > — Libro cang? > — Ya, cang. Mi lasim kara du hora. > — Yu bisa kara in dom mi. Ca aur pan situ! > — Mersi! Mi saufa lai tena. *Hello! Do you want to go to the market with me? — I can't. I have to study. — What are you studying? — Amadunia. I want to read this book. — Is the book long? — Yes, long. I have to read for two hours. — You can read at my house. Tea and bread there! — Thanks! I'll come again.* ## Practice 1. Mi mau nomu ca. — *I want to drink tea.* 2. Anak no bisa lala. — *The child cannot sleep.* 3. Kita lasim anda tena. — *We must walk again.* 4. Yu suda mau kupi libro itu? — *Did you want to buy that book?* 5. Ta bisa sema aur ta bisa kara. — *She can speak and she can read.* 6. Mi no lasim rabota din ini. — *I do not have to work today.* ## What you can already say Seventeen lessons, **138 roots**. Want, can, must — and any verb after any of them. This was the last thing the grammar was missing. "I want *you* to eat" — a second subject inside the sentence — turns out to be a clause standing where an object stands, and needs no marker; existence and comparison were settled at the same time. All three are in [Lesson 18](lesson-18-comparing-and-joining.md). --- # Lesson 18 — Comparing and joining *Prerequisite: [Lesson 17](lesson-17-wanting-and-able.md).* ## New words | Word | Meaning | |---|---| | kurang | less | | paling | most | | kadar | as ... as | | agar | if | Four words. Everything else in this lesson is made from words you already have. **Also introduced here** — these appear in the sentences below, and nothing before this lesson has taught them: | Word | Meaning | |---|---| | hafta | week | | kafe | coffee | | kanta | to sing | | kotor | dirty | | madina | city | | negara | country | | problema | problem | | sedih | sad | ## Bigger, smaller, biggest Put the degree word in front of the adjective, exactly where *cok* goes: | | | |---|---| | Dom ini kabir. | This house is big. | | Dom ini **lebi** kabir. | This house is bigger. | | Dom ini **kurang** kabir. | This house is less big. | | Dom ini **paling** kabir. | This house is the biggest. | ## Than is *dari* *dari* has meant "from" since Lesson 15. It is also "than": | | | |---|---| | Dom mi lebi kabir **dari** dom yu. | My house is bigger than yours. | | Mi lebi hayai **dari** dugu mi. | I am faster than my sibling. | | Ca ini kurang garam **dari** ca itu. | This tea is less hot than that one. | Half the world says "bigger *from* you" — Arabic, Persian, Turkish, Hindi, Indonesian all do. ## As ... as | | | |---|---| | Dom mi kabir **kadar** dom yu. | My house is as big as yours. | | Mi hayai **kadar** yu. | I am as fast as you. | No degree word here — equality is not a degree. ## There is Start a sentence with *es* and no subject: | | | |---|---| | **Es** hotel in madina. | There is a hotel in the city. | | **No es** ca. | There is no tea. | | **Es** dom kabir sini? | Is there a big house here? | | **Suda es** skola in negara ini. | There was a school in this country. | No new word. Amadunia never leaves out a subject, so an *es* at the front can only mean "there is". ## Putting a whole sentence inside another Anywhere an object goes, a sentence can go. Nothing joins them: | | | |---|---| | Mi fikir **yu hao**. | I think you are good. | | Mi bil **ta suda lai**. | I know he came. | | Mi mau **yu kula**. | I want you to eat. | | Ta sema **dom mi kabir**. | She says my house is big. | There is no word for the "that" of *I think that…*, and none is needed. (*itu* is the demonstrative "that", as in *dom itu* — a different word.) ## Because, when, if Two of these you already know: | | | |---|---| | Mi mau ca **porke** ca hao. | I want tea because tea is good. | | Mi lai **kab** yu sema. | I come when you speak. | | Mi go market **agar** yamur no lai. | I'll go to the market if it doesn't rain. | *porke* was "why" and is now also "because" — Spanish separates them by a space alone. *kab* was "when" as a question and is "when" as a joiner. Only *agar* is new. The joined part normally comes second. Put it first and a comma marks the break: *Agar yu mau, kita kula sasa.* ## How something is done You have been doing this since [Lesson 12](lesson-12-colors-directions.md): an adjective straight after the verb describes the verb. What is new here is where it goes when the verb has an object — the adjective still comes first, before the object. | | | |---|---| | Ta anda **hayai**. | He walks fast. | | Mi kara **hao**. | I read well. | | Kita rabota **cok**. | We work a lot. | Keep it against the verb, before the object — that is what keeps it clear: | | | |---|---| | Mi kara **hao** libro. | I read a book well. | | Mi kara libro **hao**. | I read a good book. | ## A conversation > — Salam! Es kafe in madina ini? > — Ya, es. Kafe itu paling hao. > — Lebi hao dari kafe in tarik mi? > — Lebi hao! Aur kurang kotor. > — Mi fikir yu es rafiki hao. > — Mi lai una yu agar yu mau. > — Mi mau! Kita go kab yu bisa. > — Sasa. Mi no rabota din ini porke hafta baru lai besok. *Hello! Is there a café in this city? — Yes, there is. That café is the best. — Better than the café on my street? — Better! And less dirty. — I think you're a good friend. — I'll come with you if you want. — I want to! Let's go when you can. — Now. I'm not working today because a new week comes tomorrow.* ## Practice 1. Libro ini lebi cang dari libro itu. — *This book is longer than that one.* 2. Es problema? — No, no es problema. — *Is there a problem? — No, there's no problem.* 3. Mi bil yu suda kara libro ini. — *I know you have read this book.* 4. Kita lasim go agar yu sedih. — *We must go if you are sad.* 5. Ta kanta paling hao in skola. — *She sings the best in the school.* 6. Mi no mau nomu kafe porke kafe kotor. — *I don't want to drink the coffee because it's dirty.* ## What you can already say Eighteen lessons, **150 roots**. You can compare anything, say what exists, put a whole thought inside a sentence, and give a reason, a time or a condition. That is the grammar of A2. What is still open is kept in one place, with what each one blocks and how often the material has actually reached for it: [the open questions](../grammar/README.md). --- # Lesson 19 — Home and nature *Prerequisite: [Lesson 18](lesson-18-comparing-and-joining.md).* ## New words | Word | Meaning | |---|---| | pintu | door | | pencere | window | | tabla | table | | yatak | bed | | pohon | tree | | flor | flower | | bahari | sea | | pahar | mountain | | sungai | river | | animal | animal | | kucing | cat | | samaki | fish | | tair | bird | | auto | car | | tren | train | | mualim | teacher | | polisi | police | | kamisa | shirt | | sapatu | shoe | | korku | fear | | gusa | anger | | sabar | patience | | daima | always | | kadang | sometimes | | mila | 1000 | **Also introduced here** — these appear in the sentences below, and nothing before this lesson has taught them: | Word | Meaning | |---|---| | baid | far | | hidup | to live | | punya | to have | ## In the house | | | |---|---| | Dom mi punya du pintu. | My house has two doors. | | Pencere ini kabir. | This window is big. | | Libro in tabla. | The book is on the table. | | Mi lala in yatak mi. | I sleep in my bed. | | Es kursi kadar pintu? | Are there as many chairs as doors? | ## Outside | | | |---|---| | Pohon itu paling cang. | That tree is the tallest. | | Flor ini merah. | This flower is red. | | Bahari lebi kabir dari sungai. | The sea is bigger than the river. | | Pahar itu baid. | That mountain is far. | | Es cok pohon in pahar. | There are many trees on the mountain. | ## Animals | | | |---|---| | Kucing mi keci. | My cat is small. | | Samaki hidup in bahari. | Fish live in the sea. | | Tair kimbia upar. | The bird runs upward. | | Animal-animal in pahar. | The animals are on the mountain. | ## Getting there | | | |---|---| | Mi go madina in auto. | I go to the city by car. | | Tren lebi hayai dari auto. | The train is faster than the car. | | Mi kadang go in tren. | I sometimes go by train. | ## People who work | | | |---|---| | Mualim mi hao. | My teacher is good. | | Polisi in tarik. | The police are on the road. | | Mualim daima kara libro. | The teacher always reads books. | ## Feelings *korku*, *gusa* and *sabar* are things you have, not things you are: | | | |---|---| | Mi punya korku. | I have fear. — *I am afraid.* | | Ta punya gusa. | She has anger. — *She is angry.* | | Mi punya sabar. | I have patience. | | Mi no punya korku porke yu sini. | I am not afraid because you are here. | ## Always and sometimes The adverb rule would put them straight after the verb. The course does not, consistently — see below: | | | |---|---| | Mi kula daima pan. | I always eat bread. | | Ta kadang lai. | She sometimes comes. | | Kita rabota daima in skola. | We always work at school. | **Where these two stand is not settled, and this lesson shows three positions.** One sentence above puts the word before its verb, one in the adverb slot after it, and [Lesson 20](lesson-20-colours-and-health.md) puts one after the object. Thirteen sentences across the course, three positions, because nobody has decided which is right. [The briefing measures it](../grammar/proposal-frequency.md) and costs the three answers. Until it is decided, read all three as correct and do not copy one over the others. *The examples are named by position rather than quoted here on purpose: the briefing counts every sentence that uses these words, and a note quoting one would count itself as evidence.* ## A thousand | | | |---|---| | mila | 1000 | | du-mila | 2000 | | mila sen | 1100 | ## A conversation > — Dom yu nali? > — In pahar. Es pohon-pohon aur uan sungai keci. > — Sundar! Yu punya animal? > — Ya, uan kucing. Ta kadang kimbia in pohon. > — Mi punya korku dari animal kabir. Mi no punya korku dari kucing. > — Kucing mi paling keci in madina! Lai kan ta. > — Mi lai agar mi bisa go in tren. *Where is your house? — On the mountain. There are trees and one small river. — Beautiful! Do you have animals? — Yes, one cat. She sometimes runs in the trees. — I'm afraid of big animals, not of cats. — My cat is the smallest in the city! Come and see her. — I'll come if I can go by train.* ## Practice 1. Pencere mi lebi keci dari pintu. — *My window is smaller than my door.* 2. Es samaki in sungai ini? — *Are there fish in this river?* 3. Mualim daima punya sabar. — *A teacher always has patience.* 4. Mi kupi kamisa aur sapatu in market. — *I buy a shirt and shoes at the market.* 5. Tren lai kab mi in madina. — *The train comes when I am in the city.* 6. Pahar itu lebi cang dari mila metro. — *That mountain is more than a thousand metres tall.* — **careful:** *metro* is not a word yet. ## What you can already say Nineteen lessons, **178 roots**. A house with doors and windows, a mountain with trees, animals, ways of travelling, and the first three feelings. What is still open is kept in one place: [the open questions](../grammar/README.md). --- # Lesson 20 — Colours and health *Prerequisite: [Lesson 19](lesson-19-home-and-nature.md).* ## New words | Word | Meaning | | Word | Meaning | |---|---|---|---|---| | siya | black | | kisan | farmer | | putih | white | | tacir | merchant | | asfar | yellow | | sofer | driver | | yesil | green | | bimar | sick | | asman | sky | | ilac | medicine | | badal | cloud | | halib | milk | | rambut | hair | | telur | egg | | telinga | ear | | legum | vegetable | | dente | tooth | | bola | ball | | benar | true | | oyun | game | | kalima | word | | surat | letter | | numero | number | | uhuru | freedom | | umur | life, age | | | | **Also introduced here** — these appear in the sentences below, and nothing before this lesson has taught them: | Word | Meaning | |---|---| | beca | to sell | | fruta | fruit | | ide | idea | | mulut | mouth | ## Four more colours Colours follow the noun, like every adjective, and take no *es*: | | | |---|---| | Rambut mi siya. | My hair is black. | | Yuki putih. | Snow is white. | | Sol asfar. | The sun is yellow. | | Pohon yesil. | The tree is green. | | Mi mau kamisa asul. Mi no mau kamisa merah. | I want a blue shirt. I do not want a red shirt. | ## Sky and cloud | | | |---|---| | Asman asul din ini. | The sky is blue today. | | Es cok badal. | There are many clouds. | | Badal siya — yamur lai! | Black clouds — rain is coming! | | Asman lebi kabir dari bahari. | The sky is bigger than the sea. | ## More body | | | |---|---| | Rambut ta cang aur siya. | Her hair is long and black. | | Mi punya du telinga aur uan mulut. | I have two ears and one mouth. | | Dente anak keci putih. | The small child's teeth are white. | ## Being ill | | | |---|---| | Mi bimar. | I am sick. | | Ta suda bimar, sasa ta hao. | She was sick, now she is well. | | Doktor beri ilac por mi. | The doctor gives medicine to me. | | Mi no bisa go skola porke mi bimar. | I can't go to school because I'm sick. | ## Work | | | |---|---| | Papa mi es kisan. | My father is a farmer. | | Tacir itu beca fruta in market. | That merchant sells fruit at the market. | | Sofer punya auto merah. | The driver has a red car. | | Mualim mi paling hao. | My teacher is the best. | ## Food | | | |---|---| | Mi nomu halib daima. | I always drink milk. | | Es telur aur legum in dom? | Are there eggs and vegetables at home? | | Anak-anak kula telur kadang. | The children sometimes eat eggs. | *Where *daima* and *kadang* stand is [not settled](../grammar/proposal-frequency.md); this course uses three positions and none of them is the rule yet.* ## Playing | | | |---|---| | Bola mi merah. | My ball is red. | | Kita suru oyun in skola. | We play a game at school. | | Oyun ini lebi asan dari oyun itu. | This game is easier than that one. | ## Words and numbers | | | |---|---| | Amadunia punya tri-sen kalima. | Amadunia has 300 words. | | Numero ini kabir. | This number is big. | | Mi katab surat por rafiki mi. | I write a letter to my friend. | | Kalima ini benar. | This word is true. | That first line said *du-sen tri-des pat* — 234 — until September 5, 2026, which was the size of the dictionary on the day it was written. The repository has twenty-odd checks that recount a number and not one of them could read this one, because it is written in Amadunia. ## Life and freedom | | | |---|---| | Umur mi sundar. | My life is beautiful. | | Uhuru es ide paling kabir. | Freedom is the biggest idea. | | Mi mau uhuru por insan-insan. | I want freedom for people. | ## A conversation > — Salam! Yu hao? > — No, mi bimar. Sar mi garam. > — Yu suda go doktor? > — Ya. Ta suda beri ilac por mi. Mi lasim nomu ilac tri din. > — Mi saufa kupi halib aur telur por yu. > — Mersi! Yu es rafiki paling hao. > — Mi lai besok. Agar yu lebi hao, kita suru oyun. *Hello! Are you well? — No, I'm sick. My head is hot. — Did you go to the doctor? — Yes. She gave me medicine. I have to drink the medicine for three days. — I'll buy milk and eggs for you. — Thanks! You're the best friend. — I'll come tomorrow. If you're better, we'll play a game.* ## Practice 1. Badal siya in asman. — *Black clouds in the sky.* 2. Mi no bimar sasa. — *I am not sick now.* 3. Kisan itu punya cok animal. — *That farmer has many animals.* 4. Telinga mi keci, mulut mi kabir. — *My ears are small, my mouth is big.* 5. Mi katab kalima ini in surat. — *I write this word in the letter.* 6. Uhuru lebi hao dari pesa. — *Freedom is better than money.* ## What you can already say Twenty lessons, **207 roots**. Six colours, the sky and its clouds, being ill and getting better, four kinds of work, and the first large ideas the language can hold — *umur*, *uhuru*, *benar*. What is still open is kept in one place: [the open questions](../grammar/README.md). --- # Lesson 21 — Things and ideas *Prerequisite: [Lesson 20](lesson-20-colours-and-health.md).* ## New words | Word | Meaning | | Word | Meaning | |---|---|---|---|---| | kamra | room | | harga | price | | anahtar | key | | posta | mail | | lampu | lamp | | kertas | paper | | sabun | soap | | mesin | machine | | bage | garden | | regalo | gift | | hutan | forest | | lingua | language | | petra | stone | | habari | news | | corba | soup | | safari | journey | | minyak | oil | | umid | hope | | burun | nose | | sansi | chance | | kulit | skin | | amani | peace | | dulce | sweet | | arte | art | | legis | law | | | | **Also introduced here** — these appear in the sentences below, and nothing before this lesson has taught them: | Word | Meaning | |---|---| | kuat | strong | ## In the room | | | |---|---| | Kamra mi keci aur putih. | My room is small and white. | | Es lampu in kamra. | There is a lamp in the room. | | Anahtar in tabla. | The key is on the table. | | Mi punya sabun aur akua. | I have soap and water. | ## Outside the house | | | |---|---| | Bage mi punya cok flor. | My garden has many flowers. | | Hutan itu lebi kabir dari bage mi. | That forest is bigger than my garden. | | Petra ini siya aur kuat. | This stone is black and strong. | ## Eating | | | |---|---| | Corba garam. | The soup is hot. | | Mi mau corba una pan. | I want soup with bread. | | Minyak no es akua. | Oil is not water. | | Ca ini dulce. | This tea is sweet. | ## Buying | | | |---|---| | Harga ini kabir. | This price is big. | | Berapa harga? | How much is the price? | | Mi kupi kertas in market. | I buy paper at the market. | | Mesin ini lebi baru dari mesin itu. | This machine is newer than that one. | ## Sending and giving | | | |---|---| | Mi katab surat aur mi beri ta por posta. | I write a letter and give it to the mail. | | Ta beri regalo por mi. | She gives a gift to me. | | Habari ini benar? | Is this news true? | ## Big words | | | |---|---| | Amadunia es lingua. | Amadunia is a language. | | Lingua ini asan. | This language is easy. | | Mi punya umid. | I have hope. | | Amani lebi hao dari gusa. | Peace is better than anger. | | Legis benar por insan-insan. | The law is true for people. | | Arte aur musik sundar. | Art and music are beautiful. | ## A journey | | | |---|---| | Safari mi cang. | My journey is long. | | Mi suda suru safari dari madina por pahar. | I made a journey from the city to the mountain. | | Agar mi punya sansi, mi go tena. | If I have the chance, I'll go again. | ## A conversation > — Yu suda kan habari? > — No. Suda es ke? > — Mualim mi saufa go safari. Ta go hutan in pahar. > — Berapa din? > — Des din. Ta punya umid ta kan animal-animal situ. > — Mi mau go una ta! > — Yu bisa. Harga no kabir. Mi saufa katab surat por ta. > — Mersi! Ini es regalo paling hao. *Did you see the news? — No. What happened? — My teacher will go on a journey. She's going to the forest in the mountains. — How many days? — Ten days. She hopes she'll see animals there. — I want to go with her! — You can. The price isn't big. I'll write a letter to her. — Thanks! This is the best gift.* ## Practice 1. Anahtar mi in kamra. Anahtar mi no in bage. — *My key is in the room. My key is not in the garden.* 2. Corba ini lebi dulce dari ca. — *This soup is sweeter than the tea.* 3. Es kertas aur lampu in kamra? — *Is there paper and a lamp in the room?* 4. Mi punya umid porke yu sini. — *I have hope because you are here.* 5. Lingua ini paling asan in dunia. — *This language is the easiest in the world.* 6. Safari lebi hao dari pesa. — *A journey is better than money.* 7. Kulit ta barid. Burun ta keci. — *His skin is cold. His nose is small.* 8. Mi punya sis libro. — *I have six books.* ## What you can already say Twenty-one lessons, **233 roots**. A room with its things, a garden and a forest, buying and sending, and the words a language needs before it can talk about itself — *lingua*, *legis*, *arte*, *amani*, *umid*. What is still open is kept in one place: [the open questions](../grammar/README.md). --- # Lesson 22 — Doing and feeling *Prerequisite: [Lesson 21](lesson-21-things-and-ideas.md).* ## New words | Word | Meaning | | Word | Meaning | |---|---|---|---|---| | kirim | to send | | ates | fire | | naiti | to find | | tanah | earth, ground | | ingat | to remember | | udara | air | | unut | to forget | | mensis | month | | finis | to finish | | | | | proba | to try | | cehra | face | | gasil | to wash | | sakil | heavy | | tabak | to cook | | hafif | light in weight | | kanis | dog | | genc | young | | farasi | horse | | amik | deep | | masal | story | | golos | voice | | mimpi | dream | | grupo | group | | yalan | lie | | | | ## Eight new verbs | | | |---|---| | Mi kirim surat por rafiki mi. | I send a letter to my friend. | | Mi suda naiti anahtar mi! | I found my key! | | Yu ingat nama ta? | Do you remember her name? | | Mi suda unut. | I forgot. | | Kita finis rabota sasa. | We finish the work now. | | Mi mau proba corba ini. | I want to try this soup. | | Ta gasil mano ta. | He washes his hands. | | Mama mi tabak hao. | My mother cooks well. | ## Remember and forget They are opposites and they behave the same: | | | |---|---| | Mi ingat yu daima. | I always remember you. | | Mi no unut yu. | I do not forget you. | | Ta suda unut kalima ini. | He forgot this word. | *Where* daima *stands is [not settled](../grammar/proposal-frequency.md): this course uses three positions and none of them is the rule yet.* ## Two animals | | | |---|---| | Kanis mi siya aur keci. | My dog is black and small. | | Farasi lebi kabir dari kanis. | A horse is bigger than a dog. | | Kanis kimbia in hutan. | The dog runs in the forest. | ## Fire, earth, air | | | |---|---| | Ates garam. | Fire is hot. | | Akua, ates, udara aur tanah. | Water, fire, air and earth. | | Es ates in dom! | There is a fire in the house! | | Udara barid in rat. | The air is cold at night. | | Tanah in bage mi hao. | The ground in my garden is good. | ## Morning and month | | | |---|---| | Pagi ini sundar. | This morning is beautiful. | | Mi kula in pagi. | I eat in the morning. | | Uan tahun punya des-du mensis. | One year has twelve months. | | Mensis ini cang. | This month is long. | ## Heavy and light, young and deep | | | |---|---| | Libro ini sakil, libro itu hafif. | This book is heavy, that book is light. | | Anak itu genc. | That child is young. | | Bahari amik. | The sea is deep. | | Kanis mi lebi hafif dari farasi yu. | My dog is lighter than your horse. | ## Face, voice, story, dream | | | |---|---| | Cehra ta sundar. | Her face is beautiful. | | Golos yu hao. | Your voice is good. | | Mama mi sema masal por mi in rat. | My mother tells me a story at night. | | Mi suda kan mimpi. | I saw a dream. — *I dreamed.* | | Grupo ini punya des insan. | This group has ten people. | ## True and false | | | |---|---| | Habari ini benar. | This news is true. | | Habari itu es yalan. | That news is a lie. | | Mi no sema yalan. | I do not tell lies. | ## A conversation > — Pagi hao! Yu ingat kita go hutan din ini? > — Mi suda unut! Mi lasim gasil aur kula. > — Mi tabak corba. Yu kula, sasa kita go. > — Mersi! Kanis mi lai una kita? > — Ya. Ta genc aur ta mau kimbia. > — Hutan itu amik. Mi punya korku keci. > — Mi una yu. No es korku — mi ingat tarik. *Good morning! Do you remember we're going to the forest today? — I forgot! I have to wash and eat. — I'll cook soup. You eat, now we go. — Thanks! Is my dog coming with us? — Yes. He's young and he wants to run. — That forest is deep. I have a little fear. — I'm with you. There is no fear — I remember the way.* ## Practice 1. Mi suda naiti libro mi in kamra. — *I found my book in the room.* 2. Kanis no bisa kara masal. — *A dog cannot read a story.* 3. Petra sakil, kertas hafif. — *Stone is heavy, paper is light.* 4. Mi ingat cehra ta aur golos ta. — *I remember her face and her voice.* 5. Kita finis oyun kab pagi lai. — *We finish the game when morning comes.* 6. Mimpi mi es yalan o benar? — *Is my dream a lie or true?* ## What you can already say Twenty-two lessons, **257 roots**. Eight more verbs, the four elements, a dog and a horse, and the words for what a face and a voice are — enough to tell someone a story. What is still open is kept in one place: [the open questions](../grammar/README.md). The word for *danger* that this lesson's conversation had to borrow English for arrived at the 300-root milestone: it is *bahaya*, taught in [Lesson 25](lesson-25-doing-and-being.md). The gap was found here and is closed. --- # Lesson 23 — Everything so far *Prerequisite: [Lesson 22](lesson-22-doing-and-feeling.md).* ## No new words This lesson adds nothing. It is the whole of Amadunia on one page, at the point where the dictionary reached three hundred roots and A1 was complete. ## The one sentence This sentence fills all nine slots of the order at once: > **Mi no suda kara hao libro ini in dom rat ini porke mi sedih.** > *I did not read this book well at home last night because I was sad.* Read it piece by piece: | Piece | What it does | Rule | |---|---|---| | **Mi** | subject | comes first | | **no** | denies what follows | before the predicate | | **suda** | past | after *no*, before the verb | | **kara** | verb | never changes shape | | **hao** | adverb | adjective straight after the verb | | **libro** | object | after the adverb | | **ini** | this | last in the noun phrase | | **in dom** | place | preposition before the noun | | **rat ini** | time | place first, then time | | **porke mi sedih** | reason | the joined clause comes last | **subject → no → suda/saufa → verb → adverb → object → place → time → clause** Nothing in that sentence changed its form. Not one word. ## The whole grammar, in fifteen lines | | | |---|---| | **Word order** | subject, verb, object | | **Adjectives** | after the noun — *dom kabir* | | **Possession** | owner after the thing — *dom mi*; owner before adjective — *dom mi kabir* | | **This, that** | last in the phrase — *dom mi kabir ini* | | **Plural** | say the noun twice — *anak-anak*; single after a number — *tri anak* | | **Tense** | *suda* past, *saufa* future, present unmarked — before the verb | | **Copula** | *es* before a noun; nothing before an adjective or place | | **There is** | *es* at the front, no subject — *Es hotel in madina* | | **Negation** | *no* before the predicate — *Mi no suda kula* | | **Questions** | rising tone; the question word stays where the answer goes | | **Verb chains** | verbs simply follow — *Mi mau kula pan* | | **Comparison** | *lebi/kurang/paling* before the adjective, *dari* for than | | **Pronouns** | *mi*, *yu*, *ta* — *ta* is he, she or it; the plural rule gives the rest | | **Commands** | a verb with no subject — *Anda!*; and any piece may stand alone — *Sasa.* | | **Stress** | the beat is on the second-to-last syllable, always — *a-ma-DU-nia* | Add *aur* and *o* for joining, *in/dari/por* for place, *una* for with, *porke/kab/agar* for because, when and if — and that is the entire language. The last three lines were added on September 5, 2026. The command and the fragment were granted on September 3 and this page — the one that claims to be the whole of Amadunia — went two days without them; the stress rule was settled the same day and fared the same. [Lesson 26](lesson-26-telling-and-answering.md) teaches the two of them that need teaching. A summary page is the easiest page in a repository to leave behind, because nothing in it ever looks wrong. ## What has never happened In three hundred roots and twenty-three lessons: - **No word has ever changed shape.** No conjugation, no case, no agreement, no plural ending, no gender. - **No rule has an exception.** Every one of the fifteen above holds everywhere, without a list of irregulars. - **No letter is silent** and no letter has two readings. ## Read a whole story You now know enough to read [Anak espera sol](../texts/story-1-anak-espera-sol.md) without help — [the reading ladder](reading-ladder.md) opens it at Lesson 13, eleven lessons before anything else. Try the Amadunia first and check the English only when you are stuck. This page sent you to [Safari por pahar](../texts/story-2-safari-por-pahar.md) until September 5, 2026. The ladder opens that one at Lesson 25, two lessons after this one — the page was recommending a story it had not taught you to read, and the ladder had been saying so all along. ## What the language still cannot do Being honest about this is part of the design: | Missing | What it blocks | |---|---| | A mark for a name | *Sol genc* is "Sol is young" **and** "the sun is young" | | *never* | only *no daima* and *no saufa*, which are weaker | | *all, some, none* | quantity beyond *cok* and *lebi* | | ordinal numbers | *first*, *second* | | *very* | *cok* is "much"; whether it covers "very" is untested | | reflexive | *Mi kan mi* reads as "I see me", not clearly "myself" | | *although, before, after, until* | the second tier of joiners | Each is recorded as an open question in [grammar/](../grammar/). None was invented to make a lesson easier. ## Where the language stands Twenty-three lessons, and **257 roots** of the three hundred — Lessons 24 and 25 have the rest. Every root in the dictionary carries a sourced etymology, drawn from Austronesian, Semitic, Latin and Romance, Indo-Aryan, Iranian, Turkic, Niger-Congo, Slavic, Sino-Tibetan, Japonic, Greek and Germanic. No two words differ by a single sound unless that contrast is safe in nearly every language on Earth. No *l* against *r* anywhere. Twenty letters. One sound each. *Mi ama dunia. Yu ama dunia. Kita ama dunia.* --- # Lesson 24 — The table and the city *Prerequisite: [Lesson 23](lesson-23-everything-so-far.md).* These words entered the dictionary when it passed 180 roots and never got a lesson of their own. This one and [the next](lesson-25-doing-and-being.md) close that gap. No new grammar — everything here uses rules you already have. Twenty-three words are new here. Seven more are in the sentences below because this is where they group, and they are listed apart, under the lesson that actually taught them. Until September 5, 2026 all thirty sat in one table headed **New words**, which told a learner that a word from Lesson 18 was new in Lesson 24. ## New words | Word | Meaning | | Word | Meaning | |---|---|---|---|---| | nyama | meat | | bulan | moon | | risi | rice | | stela | star | | namak | salt | | tempat | place | | sukar | sugar | | ponte | bridge | | gelas | glass | | banka | bank | | piring | plate | | hospital | hospital | | sikin | knife | | restoran | restaurant | | kaki | foot, leg | | mausim | season | | kalb | heart | | momen | moment | | korpo | body | | | | | nasab | relative | | | | | tempo | time | | | | **Also introduced here** — these appear in the sentences below, and nothing before this lesson has taught them: | Word | Meaning | |---|---| | dekat | near | | lihat | to look, to watch | ## Words you already have Seven of the words in this lesson are not new. They belong to the table and the city as much as the rest, so they appear in the sentences below — but you met each of them earlier, and this is where each was taught. | Word | Meaning | Taught in | |---|---|---| | besok | tomorrow | [Lesson 4](lesson-04-tense.md) | | hafta | week | [Lesson 18](lesson-18-comparing-and-joining.md) | | kafe | coffee | [Lesson 18](lesson-18-comparing-and-joining.md) | | madina | city | [Lesson 18](lesson-18-comparing-and-joining.md) | | negara | country | [Lesson 18](lesson-18-comparing-and-joining.md) | | fruta | fruit | [Lesson 20](lesson-20-colours-and-health.md) | | mulut | mouth | [Lesson 20](lesson-20-colours-and-health.md) | *besok* is the plainest of them: it was moved to Lesson 4 on September 4, 2026, because that lesson is called *Yesterday, today, tomorrow*. The move was made and this table was not trimmed, so for a day the word was new in two places. ## On the table | | | |---|---| | Es gelas aur piring in tabla. | There is a glass and a plate on the table. | | Sikin in tabla. | The knife is on the table. | | Mi kula nyama aur risi. | I eat meat and rice. | | Namak aur sukar putih. | Salt and sugar are white. | | Ca lebi dulce dari kafe. | Tea is sweeter than coffee. | | Mi nomu kafe in pagi. | I drink coffee in the morning. | | Mi kupi fruta in market. | I buy fruit at the market. | ## The body | | | |---|---| | Kaki mi cang. | My legs are long. | | Korpo mi bimar. | My body is sick. | | Kalb mi hao. | My heart is good. | | Mulut aur burun in cehra. | The mouth and nose are in the face. | ## The city | | | |---|---| | Madina ini sundar. | This city is beautiful. | | Es banka in madina ini? | Is there a bank in this city? | | Hospital in tarik ini. | The hospital is on this road. | | Mi kula in restoran. | I eat at a restaurant. | | Ponte in sungai. | The bridge is at the river. | | Tempat ini hao. | This place is good. | | Negara mi kabir. | My country is big. | | Nasab mi in madina itu. | My relatives are in that city. | ## The sky at night | | | |---|---| | Bulan aur stela in asman. | The moon and the stars are in the sky. | | Bulan putih. Stela keci. | The moon is white. The stars are small. | | Mi lihat bulan kab rat lai. | I look at the moon when night comes. | ## Time | | | |---|---| | Uan hafta punya seti din. | One week has seven days. | | Mi suda espera ba hora. | I waited eight hours. | | Skola ini punya nau kamra. | This school has nine rooms. | | Uan tahun punya pat mausim. | One year has four seasons. | | Aksam lai. Sol go. | Evening comes. The sun goes. | | Mi go skola besok. | I go to school tomorrow. | | Mi suda rabota kemarin. | I worked yesterday. | | Tempo go hayai. | Time goes fast. | | Momen ini sundar. | This moment is beautiful. | ## A conversation > — Yu mau kula in restoran besok? > — Ya! Restoran nali? > — In tarik kabir. > — Ok. Mi lai in aksam. > — Hao. Ta punya nyama, risi, fruta — aur kafe paling hao in madina. > — Mi no nomu kafe. Mi nomu ca. > — Ta punya ca tena! *Do you want to eat at a restaurant tomorrow? — Yes! Where is the restaurant? — Near the bridge, on the big road. — Ok. I'll come in the evening. — Good. They have meat, rice, fruit — and the best coffee in the city. — I don't drink coffee. I drink tea. — They have tea too!* ## Practice 1. Es sikin aur piring in tabla. — *There is a knife and a plate on the table.* 2. Madina ini punya du ponte. — *This city has two bridges.* 3. Mi lihat bulan aur stela. — *I look at the moon and the stars.* 4. Uan tahun punya des-du mensis aur pat mausim. — *A year has twelve months and four seasons.* 5. Nasab mi rabota in hospital. — *My relative works at a hospital.* 6. Kemarin aksam, besok pagi. — *Yesterday evening, tomorrow morning.* ## What you can already say Twenty-four lessons. Everything on a table, most of a city, the night sky, and the whole calendar from a moment to a season. **Next:** [Lesson 25](lesson-25-doing-and-being.md) has the verbs and the qualities that came with the same batch. --- # Lesson 25 — Doing and being *Prerequisite: [Lesson 24](lesson-24-the-table-and-the-city.md).* The other half of the words that entered at 180 roots without a lesson of their own. Again, no new grammar. Twenty words are new here: nine verbs and eleven qualities. Twelve more are gathered with them because this is where they group — you met each of those earlier, and they are listed apart, under the lesson that taught them. Until September 5, 2026 all thirty-two sat in one table headed **New words**. ## New words | Word | Meaning | | Word | Meaning | |---|---|---|---|---| | dengar | to hear | | plen | full | | senti | to feel | | kosong | empty | | tanya | to ask | | safi | clean | | respon | to answer | | muskil | difficult | | leta | to bring | | senang | happy | | pone | to put | | yanlis | wrong | | mati | to die | | bahaya | danger | | start | to begin | | sukut | silence | | stop | to stop | | historia | history | | | | | kultura | culture | | | | | natura | nature | ## Words you already have | Word | Meaning | Taught in | |---|---|---| | kanta | to sing | [Lesson 18](lesson-18-comparing-and-joining.md) | | kotor | dirty | [Lesson 18](lesson-18-comparing-and-joining.md) | | problema | problem | [Lesson 18](lesson-18-comparing-and-joining.md) | | sedih | sad | [Lesson 18](lesson-18-comparing-and-joining.md) | | baid | far | [Lesson 19](lesson-19-home-and-nature.md) | | hidup | to live | [Lesson 19](lesson-19-home-and-nature.md) | | punya | to have | [Lesson 19](lesson-19-home-and-nature.md) | | beca | to sell | [Lesson 20](lesson-20-colours-and-health.md) | | ide | idea | [Lesson 20](lesson-20-colours-and-health.md) | | kuat | strong | [Lesson 21](lesson-21-things-and-ideas.md) | | dekat | near | [Lesson 24](lesson-24-the-table-and-the-city.md) | | lihat | to look, to watch | [Lesson 24](lesson-24-the-table-and-the-city.md) | The two from Lesson 24 are the sharpest case: that lesson says in as many words that **nothing before it** has taught *dekat* and *lihat*, and then this page, the very next one, called them new again. ## Fourteen verbs | | | |---|---| | Mi lihat bulan. | I look at the moon. | | Yu dengar musik? | Do you hear the music? | | Mi senti korku. | I feel fear. | | Mi punya du kanis. | I have two dogs. | | Ta tanya nama mi. | He asks my name. | | Mi respon: "Nama mi es Sol." | I answer: "My name is Sol." | | Tacir beca fruta in market. | The merchant sells fruit at the market. | | Mama mi leta ca. | My mother brings tea. | | Mi pone libro in tabla. | I put the book on the table. | | Anak kanta hao. | The child sings well. | | Insan hidup, insan mati. | People live, people die. | | Kita start rabota in pagi. | We start work in the morning. | | Mi stop kab mi bimar. | I stop when I am sick. | ## Full and empty, clean and dirty | | | |---|---| | Gelas plen. Gelas itu kosong. | The glass is full. That glass is empty. | | Akua ini safi. | This water is clean. | | Kamisa mi kotor. Mi gasil ta. | My shirt is dirty. I wash it. | | Piring safi in tabla. | The clean plates are on the table. | ## Near and far, strong and hard | | | |---|---| | Ponte dekat. Pahar baid. | The bridge is near. The mountain is far. | | Farasi kuat. | The horse is strong. | | Tarik ini muskil. | This road is difficult. | | Lingua ini no muskil. Lingua ini asan. | This language is not difficult. It is easy. | ## Happy and sad | | | |---|---| | Mi senang porke yu sini. | I am happy because you are here. | | Ta sedih porke rafiki ta suda go. | She is sad because her friend went. | | Sukut in hutan. | There is silence in the forest. | ## Big words | | | |---|---| | Ide ini baru. | This idea is new. | | Es problema? No, no es problema. | Is there a problem? No, there's no problem. | | Habari ini yanlis. | This news is wrong. | | Es bahaya in tarik itu. | There is danger on that road. | | Mi studi historia aur kultura. | I study history and culture. | | Natura sundar. | Nature is beautiful. | ## A conversation > — Yu senang o sedih? > — Mi senang! Mi suda naiti rabota baru. > — Nali? > — In skola baru. Mi saufa es mualim anak-anak. > — Muskil? > — Ya, muskil. Aur hao. Mi ama anak-anak. > — Mi senti yu saufa es mualim paling hao. *Are you happy or sad? — I'm happy! I've found a new job. — Where? — At a new school. I'll be the children's teacher. — Difficult? — Yes, difficult. And good. I love children. — I feel you will be the best teacher.* ## Practice 1. Mi dengar golos yu. — *I hear your voice.* 2. Gelas ini plen, gelas itu kosong. — *This glass is full, that one is empty.* 3. Mi tanya, ta respon. — *I ask, she answers.* 4. Es bahaya porke tarik kotor. — *There is danger because the road is dirty.* 5. Mi senti mi bisa kanta! — *I feel I can sing!* 6. Natura, historia, kultura — tri ide kabir. — *Nature, history, culture — three big ideas.* ## What you can already say Twenty-five lessons and **300 roots** — every one now taught somewhere. [Lesson 26](lesson-26-telling-and-answering.md) adds no words and teaches the two rules that were granted after this page was written. [`check.py`](../check.py) enforces that from here on: a word added to the dictionary without a lesson will fail the build. --- # Lesson 26 — Telling and answering *Prerequisite: [Lesson 25](lesson-25-doing-and-being.md).* No new words. Two rules that this course has been using since Lesson 1 and Lesson 10 without ever teaching them, because until September 3, 2026 neither had been granted: **the command**, and **the one-word answer**. Both are in [sentence-types.md](../grammar/sentence-types.md), and both cost the language nothing — no word was added and no word changed shape. ## A command is a verb with no subject | | | |---|---| | Otur sini. | Sit here. | | Buka pintu. | Open the door. | | Espera momen. | Wait a moment. | | Lai dom mi. | Come to my house. | | Dengar mualim. | Listen to the teacher. | | Kirim surat por mama yu. | Send a letter to your mother. | That is the whole rule. Amadunia never leaves a subject out of a statement, so a sentence that begins with a verb can only be one thing. ## The subject is what makes it a statement | | | |---|---| | Go! | Go! | | Yu go. | You go. | | Kula! | Eat! | | Yu kula pan. | You eat bread. | *Yu go* is a statement about you, however loudly it is said. Putting the pronoun back "for emphasis" would make one sentence mean two things, and this language does not do that. Emphasis is the voice and the mark: **Go!** ## Saying no to a command *no* goes in front of what it denies, exactly as it does in a statement. | | | |---|---| | No go. | Do not go. | | No kula nyama itu. | Do not eat that meat. | | No lala sasa. | Do not sleep now. | | No kimbia hayai. | Do not run fast. | The order is *no* → verb → adverb, which is the order you already know. ## *plis* is politeness, not grammar | | | |---|---| | Otur, plis. | Sit, please. | | Beri akua, plis. | Give water, please. | | Madad, plis. | Help, please. | Take *plis* away and the sentence is still a command. Add it and nothing changes grammatically — which is why a warning does not take it: *No go!* ## One word is an answer A phrase may be a whole turn when what it answers is clear. > — Yu mau ke? > — Ca. > — Kab? > — Besok. > — Nali? > — In dom. > — Yu suda kula? > — Ya. The last one means *yes, I ate*. The tense, the subject and the verb all come from the question; nothing is repeated and nothing is marked. ## A command and an answer can be the same word > — Yu mau suru ke? > — Kula. *Kula* at the start of an exchange is **eat!**; *Kula* answering a question is **eating**. Context does the work, and English does the same with *Eat.* ## A conversation > — Mama, mi bimar. > — Otur sini. Buka mulut. > — Ya. > — Yu bimar. No go skola besok. > — Mi mau go! > — No. Lala sasa. Mi saufa leta ca aur ilac. > — Mersi, mama. *Mother, I'm ill. — Sit here. Open your mouth. — Yes. — You are ill. Don't go to school tomorrow. — I want to go! — No. Sleep now. I'll bring tea and medicine. — Thank you, mother.* Seven of those eleven sentences are commands or short turns — four commands (*Otur sini*, *Buka mulut*, *No go skola besok*, *Lala sasa*) and three turns that are a single word of answer: *Ya*, *No*, and the thanks that closes the exchange, whose second word is the person addressed rather than part of what is said. The three were unnamed until September 6, 2026, and naming them is what makes the seven checkable. Counted strictly — a sentence of exactly one word — there are two, *Ya* and *No*, and the total is six. Both readings are defensible and only one of them was written down, which is the difference between a figure and a claim. ## Practice 1. Buka pencere, plis. — *Open the window, please.* 2. No nomu akua itu. — *Do not drink that water.* 3. Lai sini aur otur. — *Come here and sit.* 4. — Yu mau ca o kafe? — Ca. — *Do you want tea or coffee? — Tea.* 5. Kara hao libro ini. — *Read this book well.* — the adverb stays with its verb; *Kara libro ini hao* would be "read this good book". 6. Yu kara libro ini. — *You read this book.* — a statement, because it has a subject. ## What this lesson could not do **There is no way to tell someone to *be* something.** *Be quick*, *be careful*, *be a doctor* have no form: a sentence-initial *es* is already the existential, so *Es doktor* is "there is a doctor". The mother above says *Otur sini* and *Lala sasa* because she cannot say *be still*. [That is the open question the command rule left](../grammar/sentence-types.md), and [text 12](../texts/text-12-tarik-por-skola.md) wanted it three times on one page. --- # The reading ladder *How much of the writing you can read after each lesson.* Derived from the material itself and regenerated by [`check.py`](../check.py), so it follows the course rather than a claim about it. The measure is plain: every word inside an Amadunia sentence in [the thirty texts](../texts/), counted with repeats, against the words the lessons have taught by that point. ## After each lesson | After lesson | Share of the texts' words you know | |---|---| | 01 | 23% | | 02 | 34% | | 03 | 47% | | 04 | 48% | | 05 | 49% | | 06 | 50% | | 07 | 52% | | 08 | 56% | | 09 | 56% | | 10 | 59% | | 11 | 59% | | 12 | 63% | | 13 | 66% | | 14 | 67% | | 15 | 73% | | 16 | 77% | | 17 | 78% | | 18 | 80% | | 19 | 85% | | 20 | 88% | | 21 | 91% | | 22 | 95% | | 23 | 95% | | 24 | 97% | | 25 | 100% | | 26 | 100% | **Nearly a quarter of everything after one lesson — 23%.** That figure has been 23, then 22, then 23 again in three days, printed four lines above a table that is regenerated: it moves whenever a text moves the denominator, and it moved up when [text 28](../texts/text-28-din-mi.md) and [text 29](../texts/text-29-yu-mau-ke.md) were written inside the first eight lessons. [The front page](../README.md) had the same number as *a quarter*, with no figure at all — rounding 22 up by three on the page a reader meets first. Both are computed from this table now. That is the frequency curve doing the work — [ten roots carry a third of the corpus](../dictionary/frequency.md), and the first lesson teaches the commonest of them. Half by Lesson 06, three quarters by Lesson 16. **Four lessons leave the figure unchanged: 09, 11, 23 and 26.** None of them teaches a new root at all — pronouns, the copula, a review and the two sentence rules are grammar, not vocabulary. Nothing is wrong with any of the four; the measure cannot see what they add. **This list has now changed six times in six days, and no lesson has changed once.** Lessons 05 and 17 have been on it and off it repeatedly as the texts moved the denominator underneath them: 05 left when [text 28](../texts/text-28-din-mi.md) was written inside the first six lessons, 17 came back the same day with [text 29](../texts/text-29-yu-mau-ke.md), and 17 left again with [text 30](../texts/text-30-katab-hao.md), which is written inside the first twelve and so lifts Lesson 17's row along with the rest. A lesson's figure on this page is a fact about the corpus as much as about the lesson — the same reason **Half by Lesson 06** replaced *Lesson 07* above, which had replaced *Lesson 08*, which had replaced *Lesson 07*, without any of those lessons being touched. The list is derived and will keep moving; what is worth reading is the four that teach no root, and those four do not move. **The last stretch is steep because it was built that way.** Lessons 24 and 25 were written to sweep up the roots that had reached the dictionary without a lesson, so they carry 95% to 100% between them. ## When each text opens A text is listed against the first lesson after which every word in it has been taught. | Text | Readable after | |---|---| | [Din mi](../texts/text-28-din-mi.md) | 06 | | [Yu mau ke?](../texts/text-29-yu-mau-ke.md) | 08 | | [Sol lai](../texts/text-15-sol-lai.md) | 10 | | [Katab hao](../texts/text-30-katab-hao.md) | 12 | | [Anak espera sol](../texts/story-1-anak-espera-sol.md) | 13 | | [In skola](../texts/text-20-in-skola.md) | 16 | | [Yamur aur ca](../texts/text-14-yamur-aur-ca.md) | 18 | | [Kamra mi](../texts/text-19-kamra-mi.md) | 21 | | [In madina baru](../texts/text-4-in-madina-baru.md) | 24 | | [Lingua ini](../texts/text-3-lingua-ini.md) | 24 | | [Uan](../texts/text-5-uan.md) | 24 | | [Anak katab](../texts/text-11-anak-katab.md) | 25 | | [Berapa harga](../texts/text-16-berapa-harga.md) | 25 | | [Gusa in tarik](../texts/text-18-gusa-in-tarik.md) | 25 | | [Kaifa suru ca](../texts/text-8-kaifa-suru-ca.md) | 25 | | [Kula una](../texts/text-13-kula-una.md) | 25 | | [Mila tahun](../texts/text-10-mila-tahun.md) | 25 | | [Pagi in madina](../texts/text-9-pagi-in-madina.md) | 25 | | [Safari por pahar](../texts/story-2-safari-por-pahar.md) | 25 | | [Seti din](../texts/text-6-seti-din.md) | 25 | | [Surat por mama](../texts/text-7-surat-por-mama.md) | 25 | | [Tarik por skola](../texts/text-12-tarik-por-skola.md) | 25 | | [Tren aur farasi](../texts/text-17-tren-aur-farasi.md) | 25 | | [Uan umur](../texts/text-21-uan-umur.md) | 25 | | [Agar yamur lai](../texts/text-22-agar-yamur-lai.md) | 25 | | [Kim paling hayai](../texts/text-23-kim-paling-hayai.md) | 25 | | [Mama mama mi](../texts/text-24-mama-mama-mi.md) | 25 | | [Tarik por hospital](../texts/text-25-tarik-por-hospital.md) | 24 | | [Dom eski](../texts/text-26-dom-eski.md) | 25 | | [Kalima por anak-anak](../texts/text-27-kalima-por-anak-anak.md) | 25 | **[Din mi](../texts/text-28-din-mi.md) is the first thing a learner can read, at Lesson 06** — 30 roots and five adjectives, with no conjunction, no negation, no question word, no copula and no preposition at all. It was written after this table showed that the course taught for nine lessons before it offered anything to read. **[Yu mau ke?](../texts/text-29-yu-mau-ke.md) is the second rung, at Lesson 08** — 37 roots, and a conversation rather than a description: Lesson 07 hands over the question words and almost every line of it is a question or an answer to one. **[Sol lai](../texts/text-15-sol-lai.md) is the third rung, at Lesson 10** — 89 roots, no prepositions, no *ini* or *itu*, no verb chain and no *porke*. It held the bottom of this table until text 28 was written, and it was itself written when the bottom was Lesson 13. **[Katab hao](../texts/text-30-katab-hao.md) is the fourth rung, at Lesson 12** — 21 roots, written for the gap between 10 and 13 and for the rule Lesson 12 grants: an adjective straight after the verb describes the verb. **[Story 1](../texts/story-1-anak-espera-sol.md) comes next, at Lesson 13** — written at 113 roots against Lessons 1–13, which is why it opens early. **[In skola](../texts/text-20-in-skola.md) is the sixth rung, at Lesson 16**, written for the five lessons between 13 and 18 that had nothing new to read. **[Yamur aur ca](../texts/text-14-yamur-aur-ca.md) is the seventh rung, at Lesson 18,** and it exists because this page was measured. The first version of this table had story 1 at 13 and then nothing until 24: eleven lessons with no new reading. It was written inside the 147 roots the course has taught by Lesson 18, to put a rung in the middle of that gap, and *Sol lai* was written for the gap below it. **[Kamra mi](../texts/text-19-kamra-mi.md) is the eighth rung, at Lesson 21**, written for what was then the largest gap left in this table — six lessons between 18 and 24 with nothing new to read. Everything else waits for Lesson 24 or 25, because every other text was written at 300 roots. That is a fact about how the texts were made, not about how hard they are: [*Uan*](../texts/text-5-uan.md) uses 21 roots and [*Kaifa suru ca*](../texts/text-8-kaifa-suru-ca.md) uses 19, but each reaches for one word that arrives late. **Knowing every word is not the same as understanding the sentence.** A text needs the rules as well, and the two do not have to arrive together — the verb chain does not come until Lesson 17, existence and comparison until 18. **Each row is the later of the two**, so it is what a reader actually needs. For twenty-nine of the thirty texts the two halves make no difference: the last word a text needs arrives at or after its last rule, and the widest margin is [text 17](../texts/text-17-tren-aur-farasi.md), which needs nothing after Lesson 15 and waits until 25 for a word. **[Text 28](../texts/text-28-din-mi.md) is the one exception, and it is why this page changed.** Every word in it is taught by Lesson 04 and it uses possession, which is Lesson 06. Until September 6, 2026 this table held the vocabulary figure alone, with a separate check that no text's grammar had overrun it — a check that passed because no text ever had. It would have offered text 28 three lessons early. One number, computed from both halves, and [`check.py`](../check.py) requires this page to name any text where they differ. --- # Lessons Twenty-six lessons, in order. Each assumes the one before it and nothing else. A lesson never uses a rule that has not been decided. Where one was needed and missing, the lesson says so at the foot of the page — that is how [verb chaining](../grammar/verb-chains.md), [place marking](../grammar/place.md) and the word for *danger* were all found. Between them the lessons teach every one of the 300 roots. [`check.py`](../check.py) enforces that: a word added to the dictionary without a lesson fails the build. | | Lesson | What it introduces | |---|---|---| | 1 | [Greetings](lesson-01-greetings.md) | The first six words, and a whole conversation made of them. | | 2 | [Water, food, home](lesson-02-food-and-home.md) | Subject, verb, object; the adjective after its noun. | | 3 | [People, day and night](lesson-03-people.md) | People, day and night. The first sentence that needed a rule the language did not have. | | 4 | [Yesterday, today, tomorrow](lesson-04-tense.md) | Tense: *suda* and *saufa*, and a verb that never changes. | | 5 | [More than one](lesson-05-plural.md) | Plural by saying the noun twice; single after a number. | | 6 | [Mine, yours, theirs](lesson-06-possession.md) | Possession with nothing added — *dom mi*. | | 7 | [Asking](lesson-07-questions.md) | The question words, each standing where its answer would. The yes/no question by tone has been in use since Lesson 1. | | 8 | [Getting by](lesson-08-getting-by.md) | Getting through a day somewhere new — please, thanks, sorry, help, money. | | 9 | [We, you all, they](lesson-09-pronouns.md) | The full pronoun set, built from the plural rule at no cost. | | 10 | [A day](lesson-10-a-day.md) | A whole day told start to finish, in one tense each way. | | 11 | [Being](lesson-11-being.md) | *es*, and why an adjective never takes it. | | 12 | [Colours and directions](lesson-12-colors-directions.md) | Colours and directions; the first deliberately wrong sentence, to drill the copula. | | 13 | [Weather and directions](lesson-13-weather-directions.md) | Weather and left and right; the lesson that found place marking missing. | | 14 | [Not, and](lesson-14-not-and.md) | *no* for not — a word from Lesson 1 doing a second job — and *aur* for and. | | 15 | [Pointing and placing](lesson-15-pointing-placing.md) | *ini*, *itu*, and the three prepositions. Six words that finished A1 grammar. | | 16 | [School and time](lesson-16-school-and-time.md) | School and time; the lesson that found verb chaining missing and held a word back. | | 17 | [Wanting, being able, having to](lesson-17-wanting-and-able.md) | Wanting, being able, having to — verbs simply following one another. | | 18 | [Comparing and joining](lesson-18-comparing-and-joining.md) | Comparing, existence, whole clauses inside sentences, and how something is done. | | 19 | [Home and nature](lesson-19-home-and-nature.md) | Home and nature; feelings as things you have, not things you are. | | 20 | [Colours and health](lesson-20-colours-and-health.md) | Colours completed, illness, and four kinds of work. | | 21 | [Things and ideas](lesson-21-things-and-ideas.md) | A room and its things, and the first words big enough to describe the language itself. | | 22 | [Doing and feeling](lesson-22-doing-and-feeling.md) | Eight verbs, the four elements, and the words a story needs. | | 23 | [Everything so far](lesson-23-everything-so-far.md) | The whole language on one page, and an honest list of what it still cannot do. | | 24 | [The table and the city](lesson-24-the-table-and-the-city.md) | The table, the city and the calendar — twenty-three words that reached the dictionary without a lesson, and seven more revisited. | | 25 | [Doing and being](lesson-25-doing-and-being.md) | Nine verbs and eleven qualities, the other half of the same gap, with twelve earlier words gathered beside them. | | 26 | [Telling and answering](lesson-26-telling-and-answering.md) | No new words: the command and the one-word answer, both used since Lesson 1 and neither taught until the rules were granted. | ## Four words that were in the wrong lessons On September 4, 2026 four time words were moved to the lessons that are about them. Each move is the same argument: a lesson that names a thing should teach the words for it. | Word | Was | Now | Why | |---|---|---|---| | kemarin | 24 | [04](lesson-04-tense.md) | the lesson is called *Yesterday, today, tomorrow* and taught neither | | besok | 18 | [04](lesson-04-tense.md) | the same | | pagi | 22 | [03](lesson-03-people.md) | the lesson is called *People, day and night* and taught two of the day's four parts | | aksam | 24 | [03](lesson-03-people.md) | the same | A learner used to meet the past marker twenty lessons before the word *yesterday*, and *rat* nineteen lessons before *pagi*. It showed in the material: [text 15](../texts/text-15-sol-lai.md), written for Lesson 10, opens *Sol lai* — the sun comes — because it could not say *in pagi*. ## When you can start reading [The reading ladder](reading-ladder.md) measures it: after one lesson you know nearly a quarter of every word in the texts, half by Lesson 06, and [story 1](../texts/story-1-anak-espera-sol.md) becomes readable at Lesson 13 — eleven lessons before anything else. The page is derived from the course and regenerated by [`check.py`](../check.py). ## After the lessons Read the [texts](../texts/) — thirty pieces written in Amadunia with nothing invented to make them work: ten stories, two arguments, two dialogues, a poem, a letter, a recipe, a walk to school, a shared meal, a room, a life, a journey that may not happen, a notice on a school wall and a day made of six lessons. Each ends with what the language could not say. The [phrasebook](../phrasebook.md) is the practical page. The [grammar files](../grammar/) give the reasoning behind every rule, including the candidates that were rejected and why. ## Where each rule starts Every rule a sentence can be scanned for, and the lesson that grants it. **This table is the single place these numbers live**: [`check.py`](../check.py) reads it, holds every earlier lesson to it, and dates the [reading ladder](reading-ladder.md) by it, so the syllabus and the checker cannot disagree. | Rule | Introduced in | |---|---| | number | 01 | | tone question | 01 | | tense | 04 | | plural | 05 | | copula | 06 | | possession | 06 | | question | 07 | | command | 10 | | adverb | 12 | | conjunction | 14 | | negation | 14 | | demonstrative | 15 | | place | 15 | | una | 15 | | comparison | 16 | | verb chain | 17 | | existence | 18 | | subordination | 18 | **Five of these are the ones that drifted, and they are the five that arrive with no word at all** — possession, the command, the adverb, the verb chain and the existential. Nothing to look up and nothing to spot: verb chains ran in three lessons before Lesson 17 granted them, and Lesson 18 introduced the adverb rule with a sentence Lesson 12 had already used. The other thirteen arrive with a word, and a lesson using a word no lesson has taught is stopped by a different rule. **Thirteen of these numbers lived in the checker until September 6, 2026**, and three of the thirteen were wrong: the copula was dated to Lesson 11, where the rule is stated in full, when *es* arrives at the foot of Lesson 06; comparison to Lesson 18 when *lebi* arrives in 16; and every question mark to Lesson 07 when asking by tone has been in the lessons since Lesson 01. A number the founder cannot read is a number nobody checks. --- # Texts Original writing in Amadunia. Every text here is written using **only** the roots in the [dictionary](../dictionary/dictionary.md) and **only** the grammar that has been settled in [grammar/](../grammar/). Nothing is invented to make a sentence work. When a text cannot say something, that is recorded at the end of it as a gap — which is the point. A lesson shows what the language can do; a text finds out what it cannot. **Eighteen rules can be spotted in a sentence**, and they each stand in **at least eight** texts. That is not the same as every settled rule, and the two sentences used to be one: the claim read *every settled rule is exercised here* with the narrower figure beneath it, which is an overclaim followed by its own qualification. There are twenty-one rule pages. Fourteen of them are measured here, four more rows are sub-rules of those pages — existence, *una*, the command and the tone question — and **seven rules cannot be seen by a sentence scanner at all**: definiteness because it is marked by nothing, word-formation because it is an absence, phonology, pronunciation and stress because they are properties of every word rather than of any sentence, and word order because it is true of all of them. Pronouns are the exception and are simply not tracked: the scanner could count them and does not. **A question mark is two rules, and the scan counted it as one.** Asking with a question word — *ke*, *kim*, *nali*, *kab*, *kaifa*, *berapa* — is Lesson 07. Asking by tone alone has been in the lessons since Lesson 01, and the syllabus is checked for exactly that. Of the material's 259 questions, 164 use no question word, so dating every one of them to Lesson 07 charged six lessons to any text that asked one — both figures recounted here. They are two rows now, and the tone question stands in more texts than the question word does. **Possession was on the invisible list until September 6, 2026**, on the grounds that it is marked by nothing — and being marked by nothing is not the same as being invisible. A noun with a pronoun straight after it is possession and cannot be anything else, which is a shape a scanner can see; it now stands in 23 texts, second only to place. It was the rule the reading ladder needed most, because it arrives in Lesson 06 and [a text was written to that lesson](text-28-din-mi.md). **The question rule was missing from this table until September 6, 2026**, and not because it is hard to see. Its detector looked for a question mark in the sentence, and the scanner splits sentences *on* the question mark, so the mark was never there to find. Counted properly, the question rule stands in fourteen texts, seventh of sixteen. **And the material itself was prepared nine different ways.** Nine scans read the lessons, the texts and the phrasebook, and each prepared the text itself: five removed the whole *New words* section of a lesson, four kept it. The difference is one sentence and four words — a line of prose under that heading in Lesson 06 — so the repository's two headline corpus sizes were being measured over two different corpora, the sentence count over one and the word count over the other. There is one definition now, and it removes the table rows rather than the section: a gloss is not a use, and a sentence is a sentence wherever it is printed. Merged September 6, 2026, and no check enforces it — a structural rule stated precisely enough to test turned out to need four exceptions, which is a rule not worth writing. **It was missing because the test existed twice.** The reading ladder needs the same judgement — which rules a sentence uses — and had its own copy, which read the raw line and had always worked. Compared line by line on September 6, 2026, the two copies had drifted **three** ways: the table could not see a question, the ladder could not see a *negative* command, and the ladder had no number detector at all. Only the first had reached a published figure; the ladder's two gaps were latent, and the rows did not move when they were closed. There is one copy now, and [`check.py`](../check.py) refuses a second. Two copies of one judgement is not a risk that a rule will be measured wrongly — it is a guarantee that two pages will eventually disagree, and the only question is which figure is read first. This table is regenerated by [`check.py`](../check.py) from the same scan that holds the floor, and each rule's detector is written down in the script beside it — a count is only as good as what it counted. | Rule | Texts | |---|---:| | place | 25 | | possession | 25 | | tense | 20 | | conjunction | 19 | | negation | 19 | | number | 18 | | copula | 17 | | tone question | 16 | | demonstrative | 15 | | plural | 13 | | question | 12 | | subordination | 12 | | verb chain | 12 | | command | 11 | | existence | 11 | | adverb | 10 | | una | 9 | | comparison | 8 | That is a different sentence from the one this page carried for a day. The first count asked whether each rule appeared *at all*, and the answer was yes while two rules stood in one text and two, which is not the same as being exercised. **And the sentence that replaced it went stale too.** It read "Place leads at eighteen; subordination is the thinnest at five, then comparison and commands at six and the adverb rule at seven" — and against the scan that was already in `check.py`, place stands in nineteen texts, commands in seven and the adverb rule in eight. Only the floor was ever checked; the four named figures were prose, and three of them drifted as the texts were edited this week. The table above replaced them on September 5, 2026 and is regenerated, so it cannot. **Two rules were thin and each was given a text of its own**, which is how the floor got to five. The adverb rule stood in one text until [text 11](text-11-anak-katab.md) was written for it and used it eleven times in eighteen sentences. Commands stood in two — once in [text 4](text-4-in-madina-baru.md), and once only as an illustration of what [text 8](text-8-kaifa-suru-ca.md) *would* say — until [text 12](text-12-tarik-por-skola.md) was written for them and used eleven in twenty-seven sentences. Both rules were introduced by the lessons well after the material had started using them, and writing a page for each is what turned that round. *una* was the third, at 21 sentences, and [text 13](text-13-kula-una.md) answered it. **Subordination was the fourth** — and writing its text is what showed the table had been measuring the rule by half of itself. It stood at five, the thinnest of the fifteen, and [text 22](text-22-agar-yamur-lai.md) was written for it on September 6, 2026. That page then pointed out that a clause can be the object of a verb with *nothing* marking it, so the detector — which looks for *porke*, *kab* and *agar* — could not see that shape at all. Taught to see it, the scan puts subordination in **seven** texts, and the thinnest rule is **comparison at six**. So the premise was wrong and the text was still worth writing: subordination now stands in seven texts instead of five, and eleven sentences that exercise it were being counted nowhere. But the honest form of the sentence this paragraph used to carry is that **a rule with no word in it cannot be found by a scanner that reads words** — and comparison, not subordination, is what the next text should be written for. ## How many shapes the language actually uses Vocabulary has been counted since the beginning; sentence *shape* never had been. Every sentence in the lessons, the texts and the phrasebook was reduced to its parts of speech — R for a pronoun, V a verb, N a noun, A an adjective, P a preposition, and so on — and the shapes counted. **1895 sentences, 449 distinct shapes**, and the fifteen commonest cover 42% of them. The leaders are *RVN* at 8.3% (*Ta-ta lihat stela-stela*), *NA* at 7.5% (*Sol kabir*) and *NV* at 4.8% (*Anak sema*). [`check.py`](../check.py) recounts the two headline numbers. **1582 of the 1895 are distinct — 84%.** The corpus is not padded with repeats: the most repeated sentence appears nine times across six files and is *Mi sema*, "I say", which is a quotation frame rather than a sentence anybody is making. It read *Mi sema* at seven until September 6, 2026, and the line went stale without anything noticing, because the two headline numbers above were checked and the sentence under them was prose; it is derived now. That matters for [the frequency list](../dictionary/frequency.md), which counts every occurrence: if a sixth of the material were copies of itself the curve would be measuring the copying. The measurement was made to test a claim and refuted it. [Text 21](text-21-uan-umur.md), the long one, said at first that the shapes drone at that length. They do not: it uses 52 distinct shapes, more than any other text, and only 8% of its adjacent sentence pairs share a shape, which puts it fifteenth of thirty. The most repetitive pages are [the recipe](text-8-kaifa-suru-ca.md) at 28% and [the poem](text-5-uan.md) at 25%, both of which repeat on purpose. **Four texts headed a summary with a sentence count nobody could reproduce.** *Twenty-five sentences, 45 roots* opens text 14; the roots were right and the sentences were not, and the same shape stood on texts 15, 19 and 20. Measured by the convention [`check.py`](../check.py) writes down — a full stop, question mark or exclamation mark inside the text's own code block — three of the four were wrong by three or four, and none of them matched the line count either. The root half was right every time, because the foot of each page states it and that figure is checked; the sentence half sat in a phrasing the sentence check did not match. **A number is checked when its wording is checked, not when its subject is.** Corrected September 6, 2026, and the phrasing is covered now. ### What the corpus scan counts that nobody said Every figure on this page comes from one scanner, and on September 6, 2026 it was found reading a row of vocabulary as two sentences. Excluding that left a smaller residue, measured rather than guessed: **21 runs, 55 words**, come from explanatory prose in the lessons and the phrasebook rather than from a table, a quoted line, a code block or a numbered practice item. Six of the 22 are real sentences; the other 16 are a page naming a form to talk about it, or a piece of one left behind when a quoted phrase wraps across two lines. **The full list, and the three rules for telling a citation from an utterance that were tried and rejected, are [on the briefing that owns the distinction](../grammar/proposal-sentence-types.md#what-the-corpus-scan-counts-that-nobody-said)** — and they are there rather than here for the reason the record itself makes: this page is inside the material the scanner counts, so writing the examples out would have added ten more of exactly the thing being measured. It did, in the first draft, and the checker said so. **A gap has to be a sentence that stopped.** Before a text records one, the rule page that would grant the thing has to be read. [Text 10](text-10-mila-tahun.md) claimed the language could not compare one duration with another; [comparison.md](../grammar/comparison.md) has a section headed *With verbs* granting exactly that, and the claim survived a day before being withdrawn. Nothing in [`check.py`](../check.py) can catch this — it verifies that a "no word for X" note is still true, because words are in a list, but a rule is not — so it is a rule for writing rather than a check. These are original compositions, not translations. Translating an existing book would mean either reproducing a copyrighted work or waiting for a vocabulary the language does not yet have — usually both. | Text | Roots used | Written with | |---|---|---| | [Anak espera sol](story-1-anak-espera-sol.md) | 34 | 113 roots, Lessons 1-13 — a story | | [Safari por pahar](story-2-safari-por-pahar.md) | 76 | 300 roots, Lessons 1-22 — a story | | [Lingua ini](text-3-lingua-ini.md) | 43 | 300 roots, Lessons 1-23 — an argument, not a story | | [In madina baru](text-4-in-madina-baru.md) | 62 | 300 roots, Lessons 1-23 — a dialogue | | [Uan](text-5-uan.md) | 21 | 300 roots, Lessons 1-23 — a poem | | [Seti din](text-6-seti-din.md) | 58 | 300 roots, Lessons 1-23 — a story, written to reach the unreached roots | | [Surat por mama](text-7-surat-por-mama.md) | 54 | 300 roots, Lessons 1-23 — a letter | | [Kaifa suru ca](text-8-kaifa-suru-ca.md) | 19 | 300 roots, Lessons 1-23 — instructions, written without the imperative the language lacks | | [Pagi in madina](text-9-pagi-in-madina.md) | 56 | 300 roots, Lessons 1-23 — an errand, written from the frequency list to reach the tail | | [Mila tahun](text-10-mila-tahun.md) | 37 | 300 roots, Lessons 1-23 — an argument, for the abstract words a scene cannot reach | | [Anak katab](text-11-anak-katab.md) | 31 | 300 roots, Lessons 1-23 — a portrait, written for the least-exercised rule | | [Tarik por skola](text-12-tarik-por-skola.md) | 36 | 300 roots, Lessons 1-25 — a walk to school, written for the imperative granted the day before | | [Kula una](text-13-kula-una.md) | 43 | 300 roots, Lessons 1-26 — a shared meal, written for *una*, the rule that had become the thinnest | | [Yamur aur ca](text-14-yamur-aur-ca.md) | 45 | **147 roots, Lessons 1-18** — a rainy night, written for a learner in the middle of the course | | [Sol lai](text-15-sol-lai.md) | 42 | **89 roots, Lessons 1-10** — a day at home and at the market, the lowest rung on the reading ladder | | [Berapa harga](text-16-berapa-harga.md) | 62 | 300 roots, Lessons 1-26 — bargaining at the market, written for the words no text had ever used | | [Tren aur farasi](text-17-tren-aur-farasi.md) | 73 | 300 roots, Lessons 1-26 — a journey to a farm, written for the same reason and reaching 27 more | | [Gusa in tarik](text-18-gusa-in-tarik.md) | 43 | 300 roots, Lessons 1-26 — a blocked street, the last five roots that could be written | | [Kamra mi](text-19-kamra-mi.md) | 48 | **232 roots, Lessons 1-21** — a room, written for the largest gap left on the reading ladder | | [In skola](text-20-in-skola.md) | 42 | **136 roots, Lessons 1-16** — a school day, written for the gap between Lessons 13 and 18 | | [Uan umur](text-21-uan-umur.md) | 122 | 300 roots, Lessons 1-26 — a whole life, and **the first long text**: 88 sentences against a previous longest of 34 | | [Agar yamur lai](text-22-agar-yamur-lai.md) | 62 | 300 roots, Lessons 1-26 — a journey that may not happen, **written for subordination**, which turned out not to be the thinnest rule | | [Kim paling hayai](text-23-kim-paling-hayai.md) | 50 | 300 roots, Lessons 1-26 — children ranking animals, **written for comparison**, the thinnest rule once subordination stopped being it | | [Mama mama mi](text-24-mama-mama-mi.md) | 59 | 300 roots, Lessons 1-26 — a portrait of one person, and the first text to find a gap in the prepositions rather than the vocabulary | | [Tarik por hospital](text-25-tarik-por-hospital.md) | 45 | 300 roots, Lessons 1-24 — directions to a stranger, **written for commands** and as a second witness to the missing preposition | | [Dom eski](text-26-dom-eski.md) | 45 | 300 roots, Lessons 1-26 — a memory, and the first page to find that an adjective predicate cannot be put in the past | | [Kalima por anak-anak](text-27-kalima-por-anak-anak.md) | 39 | 300 roots, Lessons 1-26 — a notice on a school wall, **written for the imperative and against a missing *everyone***, and one of the four texts that use no *mi* | | [Din mi](text-28-din-mi.md) | 30 | **30 roots, Lessons 1-06** — one day from sunrise to night, written for the nine lessons that had nothing to read; the lowest rung on the ladder | | [Yu mau ke?](text-29-yu-mau-ke.md) | 37 | **37 roots, Lessons 1-08** — a market conversation, the second rung on the ladder and the first text that is mostly questions | | [Katab hao](text-30-katab-hao.md) | 21 | **21 roots, Lessons 1-12** — a parent teaching a child to write, the fourth rung and the text for the adverb rule | --- # Anak espera sol *The child waits for the sun.* An original story, written with 34 roots — of the 113 the language had on September 2, 2026 — and no grammar beyond Lesson 13. --- ``` Din. Anak keci otur. Sol upar. Sol kabir. Sol garam. Sol beri luma. Anak kan upar. Anak sema: "Sol, yu es kim?" Sol beri luma. Anak sema: "Yu es rafiki mi?" Sol beri luma. Din cang. Sol beri luma. Rat lai. Sol go. Dunia barid. Anak espera. Anak espera cok. Angin lai. Yuki lai. Anak keci sini. Din baru lai. Sol lai upar. Anak kimbia. Anak sema: "Salam! Mi suda espera yu!" Sol beri luma. Anak faham: rafiki go, rafiki lai. Anak otur. Anak kan sol. Anak hao. ``` --- ## Line by line | Amadunia | English | |---|---| | Din. Anak keci otur. Sol upar. | Day. A small child sits. The sun is above. | | Sol kabir. Sol garam. Sol beri luma. | The sun is big. The sun is hot. The sun gives light. | | Anak kan upar. Anak sema: "Sol, yu es kim?" | The child looks up. The child says: "Sun, who are you?" | | Sol beri luma. | The sun gives light. | | Anak sema: "Yu es rafiki mi?" | The child says: "Are you my friend?" | | Sol beri luma. Din cang. Sol beri luma. | The sun gives light. The day is long. The sun gives light. | | Rat lai. Sol go. Dunia barid. | Night comes. The sun goes. The world is cold. | | Anak espera. Anak espera cok. | The child waits. The child waits a lot. | | Angin lai. Yuki lai. Anak keci sini. | Wind comes. Snow comes. The small child is here. | | Din baru lai. Sol lai upar. | A new day comes. The sun comes above. | | Anak kimbia. Anak sema: "Salam! Mi suda espera yu!" | The child runs. The child says: "Hello! I waited for you!" | | Sol beri luma. | The sun gives light. | | Anak faham: rafiki go, rafiki lai. | The child understands: friends go, friends come. | | Anak otur. Anak kan sol. Anak hao. | The child sits. The child looks at the sun. The child is well. | ## What the language did well **The sun never answers in words, and it did not need to.** *Sol beri luma* repeated four times carries the whole weight of the story. A language with no way to say "the sun replied warmly" has to find another way, and repetition — which every oral tradition on Earth uses — was there waiting. **Nothing had to be invented.** Every sentence is subject-verb-object, adjectives follow their noun, *es* appears only before a noun, and *suda* appears exactly once, where the child looks back at the waiting. The rules held for a whole story without a single exception. ## What the language could not say Writing this found four gaps that thirteen lessons had not: 1. ~~**No "and".**~~ *Angin lai. Yuki lai.* should have been one sentence — "wind and snow came". There was no conjunction in the language at all. **Closed September 3, 2026:** *aur*, see [conjunction.md](../grammar/conjunction.md). The story is left as written, from before the word existed. 2. ~~**No "not".**~~ The story contains no negation because none was possible; the sun's silence had to be shown by what it did instead. **Closed September 3, 2026:** *no*, see [negation.md](../grammar/negation.md). The silence is better as it stands, so the story keeps it. 3. ~~**No "again".**~~ *Rafiki lai* is "friends come"; "friends come back" needed a word the language did not have. **Closed September 3, 2026:** *tena*, Swahili. *Rafiki lai tena*. 4. ~~**No place marking.**~~ *Sol lai upar* leaned on word order alone. **Closed September 3, 2026:** *in*, *dari* and *por*, see [place.md](../grammar/place.md). All four gaps this story found are now closed; it stands as written. Gaps 1 and 2 are grammar decisions for the founder. Gaps 3 and 4 are partly vocabulary. ## Roots used 34 of 300 in the dictionary today — and 34 of the 113 that existed when this was written. They were: angin, anak, baru, barid, beri, cang, cok, din, dunia, es, espera, faham, garam, go, hao, kabir, kan, keci, kim, kimbia, lai, luma, mi, otur, rafiki, rat, salam, sema, sini, sol, suda, upar, yu, yuki. --- # Safari por pahar *Journey to the mountain.* The second story written in Amadunia, and the first with the full A1 vocabulary behind it. Written the day the dictionary reached 300 roots. --- ``` Pagi. Du anak sini: Sol aur Luma. Sol genc. Luma lebi genc dari Sol. Ta-ta mau go pahar. Mama ta-ta sema: "Pahar amik. Es bahaya." Sol sema: "Mi-mi punya sabar aur umid." Mama leta pan, akua, aur uan lampu por ta-ta. Ta-ta anda. Tarik cang. Din garam, udara hafif. Ta-ta lihat pohon-pohon yesil aur flor merah. Kanis keci lai una ta-ta. Din finis. Aksam lai. Asman siya. Ta-ta lihat stela-stela. Ta-ta lihat bulan putih. Luma sema: "Mi punya korku." Sol sema: "Mi una yu." Aur sukut lai. Ates keci garam. Ta-ta lala in tanah. Pagi baru lai. Sol asfar in asman. Ta-ta upar in pahar. Ta-ta lihat dunia: bahari baid, madina keci, hutan yesil, sungai hafif. Luma sema: "Mi no punya korku sasa." Ta-ta lai dom. Mama espera in pintu. Mama tanya: "Yu-yu suda lihat ke?" Sol sema: "Mi-mi suda lihat dunia." Luma sema: "Mi-mi suda lihat mi-mi." ``` --- ## Line by line | Amadunia | English | |---|---| | Pagi. Du anak sini: Sol aur Luma. | Morning. Two children here: Sol and Luma. | | Sol genc. Luma lebi genc dari Sol. | Sol is young. Luma is younger than Sol. | | Ta-ta mau go pahar. | They want to go to the mountain. | | Mama ta-ta sema: "Pahar amik. Es bahaya." | Their mother says: "The mountain is deep. There is danger." | | Sol sema: "Mi-mi punya sabar aur umid." | Sol says: "We have patience and hope." | | Mama leta pan, akua, aur uan lampu por ta-ta. | Mother brings bread, water and one lamp for them. | | Ta-ta anda. Tarik cang. Din garam, udara hafif. | They walk. The road is long. The day is hot, the air is light. | | Ta-ta lihat pohon-pohon yesil aur flor merah. | They look at green trees and red flowers. | | Kanis keci lai una ta-ta. | A small dog comes with them. | | Din finis. Aksam lai. Asman siya. | The day finishes. Evening comes. The sky is black. | | Ta-ta lihat stela-stela. Ta-ta lihat bulan putih. | They look at the stars. They look at the white moon. | | Luma sema: "Mi punya korku." | Luma says: "I have fear." | | Sol sema: "Mi una yu." | Sol says: "I am with you." | | Aur sukut lai. | And silence comes. | | Ates keci garam. Ta-ta lala in tanah. | A small fire is hot. They sleep on the ground. | | Pagi baru lai. Sol asfar in asman. | A new morning comes. The sun is yellow in the sky. | | Ta-ta upar in pahar. Ta-ta lihat dunia: | They are up on the mountain. They look at the world: | | bahari baid, madina keci, hutan yesil, sungai hafif. | the sea far, the city small, the forest green, the river light. | | Luma sema: "Mi no punya korku sasa." | Luma says: "I have no fear now." | | Ta-ta lai dom. Mama espera in pintu. | They come home. Mother waits at the door. | | Mama tanya: "Yu-yu suda lihat ke?" | Mother asks: "What did you see?" | | Sol sema: "Mi-mi suda lihat dunia." | Sol says: "We saw the world." | | Luma sema: "Mi-mi suda lihat mi-mi." | Luma says: "We saw ourselves." | ## What changed since the first story The first story had 113 roots and could not say "and", "not", "again" or where anything was. This one has all four, and the difference shows most in the last line. *Mi-mi suda lihat mi-mi* — "we saw us" — was impossible before; the pronoun plural did not exist. *Aur sukut lai* is the sentence the first story could not have written at all: it needs *aur* for the join and *sukut* for a thing that is an absence. ## What the language still could not say 1. **No mark for a name.** *Sol genc* is "Sol is young" **and** "the sun is young", because *sol* is also the word for sun. *Luma* is likewise the word for light. The story leans on capital letters, which help a reader and do nothing for a listener. Whether Amadunia wants a particle before a name, or a rule that names are simply learned, is undecided. 2. **No "never".** Luma should be able to say "I will never forget this morning". *no saufa unut* is "will not forget", which is weaker and not the same. 3. **No ordinal.** "The first morning" and "the second morning" cannot be told apart from "one morning" and "two mornings". 4. **No reflexive.** The last line, *Mi-mi suda lihat mi-mi*, works only because the plural pronoun repeats. "I saw myself" would be *Mi suda kan mi*, which reads as "I saw me" and is not clearly reflexive. Gaps 2, 3 and 4 are on the open list already. Gap 1 is new, and it is the largest thing the language has been missing without noticing. ## Roots used 76 of 300. --- # Lingua ini *This language.* The third text, and the first that is not a story. Stories exercise narrative — one thing after another. This one argues, compares and supposes, which the grammar had never been tested on. Written at 300 roots, with no grammar beyond Lesson 23. It uses 43 of them. --- ``` Amadunia es lingua baru. Ta punya tri-sen kalima. Kalima-kalima lai dari dunia. Ta-ta no lai dari uan tempat. Cok insan lihat kalima aur ta-ta ingat: "Mi bil kalima ini." Kalima no sakil in lingua ini. Yu kula. Mi kula. Ta kula. Kita kula. Mi suda kula. Mi saufa kula. Mi no kula. Ini es lingua asan. Es lingua-lingua lebi kabir dari Amadunia. Es lingua-lingua lebi eski. Amadunia no mau es lingua paling kabir. Amadunia mau es lingua paling asan. Agar yu bil des kalima, yu bisa sema. Agar yu bil sen kalima, yu bisa sema masal. Agar yu bil tri-sen kalima, yu bisa sema umur yu. Lingua no por uan insan. Lingua por dunia. Mi ama dunia. Yu ama dunia. Kita ama dunia. ``` --- ## Line by line | Amadunia | English | |---|---| | Amadunia es lingua baru. | Amadunia is a new language. | | Ta punya tri-sen kalima. | It has three hundred words. | | Kalima-kalima lai dari dunia. | The words come from the world. | | Ta-ta no lai dari uan tempat. | They do not come from one place. | | Cok insan lihat kalima aur ta-ta ingat: "Mi bil kalima ini." | Many people look at a word and they remember: "I know this word." | | Kalima no sakil in lingua ini. | A word is not heavy in this language. | | Yu kula. Mi kula. Ta kula. Kita kula. | You eat. I eat. She eats. We eat. | | Mi suda kula. Mi saufa kula. Mi no kula. | I ate. I will eat. I do not eat. | | Ini es lingua asan. | This is an easy language. | | Es lingua-lingua lebi kabir dari Amadunia. | There are languages bigger than Amadunia. | | Es lingua-lingua lebi eski. | There are languages older. | | Amadunia no mau es lingua paling kabir. | Amadunia does not want to be the biggest language. | | Amadunia mau es lingua paling asan. | Amadunia wants to be the easiest language. | | Agar yu bil des kalima, yu bisa sema. | If you know ten words, you can speak. | | Agar yu bil sen kalima, yu bisa sema masal. | If you know a hundred words, you can tell a story. | | Agar yu bil tri-sen kalima, yu bisa sema umur yu. | If you know three hundred words, you can tell your life. | | Lingua no por uan insan. Lingua por dunia. | A language is not for one person. A language is for the world. | | Mi ama dunia. Yu ama dunia. Kita ama dunia. | I love the world. You love the world. We love the world. | ## What this text tested that the stories did not **Existence.** *Es lingua-lingua lebi kabir* — "there are languages bigger" — puts the existential *es* in front of a plural in front of a comparison. Three rules stacked, and the sentence stays flat. **Supposition.** The three *agar* lines are the first conditionals written outside a lesson. Putting the clause first and the main clause second, with a comma, reads naturally. **A verb chain across the copula.** *Amadunia no mau es paling asan* runs negation, then a verb, then *es*, then a superlative. Nothing needed adding. **A paradigm.** *Yu kula. Mi kula. Ta kula. Kita kula.* The point of those four lines is that they are boring, and the language has to be able to be boring on purpose. ## What the language still could not say Four new gaps, none of them small: 1. **No word for "change".** The central claim of the whole language — *the word never changes* — cannot be said in it. The text had to show the paradigm and let the reader conclude it. This is the most striking absence found so far. 2. **No name for any country or people.** The text wanted to say the words come from Chinese, Hindi, Arabic, Swahili. It could only say *dari dunia*, "from the world". Amadunia has no way to name a place on Earth. 3. **No word for a letter of the alphabet.** *Twenty letters, one sound each* is on the front page of the project and is unsayable in the language it describes. *surat* is a letter you post, which is a different thing. 4. **No word for a part of speech** — no "verb", no "noun", no "word class". A language cannot yet teach itself. Gap 1 also touches the still-open question of proper nouns: *Amadunia* appears five times here as a name, with nothing marking it as one. A reader manages; a listener is guessing. ## Roots used 43 of 300. --- # In madina baru *In a new city.* The fourth text, and the first dialogue. The two stories tell; the third argues; this one is two people talking, which turns out to be the hardest of the four for the grammar. Written at 300 roots, with no grammar beyond Lesson 23. --- ``` — Pardon! Yu punya tempo? — Ya. Ke? — Mi baru in madina ini. Mi naiti hotel. — Es du hotel sini. Uan lebi kabir, uan lebi keci. — Harga? — Hotel kabir: sis-des pesa por uan rat. Hotel keci: tri-des. — Mi mau hotel keci. Ta nali? — Anda tarik ini. Kab yu lihat ponte, go yamin. — Ponte? — Ya. Ponte in sungai. Yu no bisa unut ta — ta merah. — Mersi! Aur es restoran in tempat ini? — Es. Restoran keci. Corba paling hao in madina. — Yu sema benar? — Mi sema benar daima. Mi es tacir. Mi no bisa sema yalan. — Ok! Mi go sasa. Aksam hao! — Aksam hao. Agar yu naiti problema, lai tena. Mi in market daima. ``` --- ## Line by line | Amadunia | English | |---|---| | — Pardon! Yu punya tempo? | Excuse me! Do you have time? | | — Ya. Ke? | Yes. What? | | — Mi baru in madina ini. Mi naiti hotel. | I am new in this city. I am looking for a hotel. | | — Es du hotel sini. Uan lebi kabir, uan lebi keci. | There are two hotels here. One bigger, one smaller. | | — Harga? | The price? | | — Hotel kabir: sis-des pesa por uan rat. Hotel keci: tri-des. | The big hotel: sixty for one night. The small one: thirty. | | — Mi mau hotel keci. Ta nali? | I want the small hotel. Where is it? | | — Anda tarik ini. Kab yu lihat ponte, go yamin. | Walk this road. When you see a bridge, go right. | | — Ponte? | A bridge? | | — Ya. Ponte in sungai. Yu no bisa unut ta — ta merah. | Yes. The bridge on the river. You cannot forget it — it is red. | | — Mersi! Aur es restoran in tempat ini? | Thanks! And is there a restaurant in this place? | | — Es. Restoran keci. Corba paling hao in madina. | There is. A small restaurant. The soup is the best in the city. | | — Yu sema benar? | Are you telling the truth? | | — Mi sema benar daima. Mi es tacir. Mi no bisa sema yalan. | I always tell the truth. I am a merchant. I cannot tell lies. | | — Ok! Mi go sasa. Aksam hao! | Ok! I'm going now. Good evening! | | — Aksam hao. Agar yu naiti problema, lai tena. Mi in market daima. | Good evening. If you find a problem, come again. I am at the market always. | ## What a dialogue tested that the others did not **Turns are short.** A story controls its own length; a conversation does not. Eight of the sixteen turns here are under six words, and three are one word. **The joke lands.** *Mi es tacir. Mi no bisa sema yalan* — a merchant claiming he cannot lie — works because the reader supplies the irony, not the grammar. A language that could not carry that would be a poor one. **Numbers in the wild.** *Sis-des pesa por uan rat* is the first time a compound number, a currency and a preposition have stood together outside a lesson. Nothing about it needed explaining. ## What the language still could not say Two of these are new, and one of them is uncomfortable. 1. **There is no rule for a fragment.** *Ke?* *Harga?* *Ponte?* — three turns here are a single word, which is what conversation is made of. Every one of them is understood, and not one is covered by anything in [grammar/](../grammar/). They work by accident. 2. **The imperative was never decided.** *Anda tarik ini*, *go yamin*, *lai tena* are commands. Bare verbs have been used as commands since Lesson 10 — *Otur sini, plis*, *Buka ain yu!* — and **no rule anywhere says they are**. The lessons taught a form the grammar never granted. This is the same kind of oversight as the missing mark for names, and it was found the same way: by writing something that was not a lesson. 3. **No word for "per".** *Sixty per night* had to become *por uan rat*, "for one night". Understandable, and not the same thing. 4. **No interjections.** A laugh, an *oh*, a hesitation. The dialogue reads more clipped than any real one because the language has no way to make a noise that is not a word. ## Roots used 62 of 300. --- # Uan *One.* The fifth text, and the first that is not prose. The stories tell, the third text argues, the fourth is two people talking. This one is meant to be said aloud, which turns out to be the thing the language is least ready for. Written at 300 roots, with no grammar beyond Lesson 23. --- ``` Din lai. Sol upar. Rat lai. Bulan upar. Din lai tena. Mi anda in tarik. Yu anda in tarik. Kita anda in uan tarik. Mi lihat sol. Yu lihat sol. Kita lihat uan sol. Mi punya kalima. Yu punya kalima. Kita punya uan lingua. Mi ama dunia. Yu ama dunia. Kita ama dunia una. ``` --- ## Line by line | Amadunia | English | |---|---| | Din lai. Sol upar. | Day comes. The sun is above. | | Rat lai. Bulan upar. | Night comes. The moon is above. | | Din lai tena. | Day comes again. | | Mi anda in tarik. | I walk on the road. | | Yu anda in tarik. | You walk on the road. | | Kita anda in uan tarik. | We walk on one road. | | Mi lihat sol. | I see the sun. | | Yu lihat sol. | You see the sun. | | Kita lihat uan sol. | We see one sun. | | Mi punya kalima. | I have words. | | Yu punya kalima. | You have words. | | Kita punya uan lingua. | We have one language. | | Mi ama dunia. | I love the world. | | Yu ama dunia. | You love the world. | | Kita ama dunia una. | We love the world together. | ## Metre — readable only since September 3, 2026 This poem was written while stress was still undecided, and it said so: being said aloud was "the thing the language is least ready for". [Stress was settled today](../grammar/stress.md) — the second-to-last syllable of every word, a vowel pair counting as one — so the poem can now be scanned. It was, and it holds two things nobody could have seen before. **The paired lines match to the syllable.** Nothing enforced it; the poem was built on parallel words, not counted beats. | | Syllables | |---|---| | Mi anda in tarik. / Yu anda in tarik. | 6 and 6 | | Mi lihat sol. / Yu lihat sol. | 4 and 4 | | Mi punya kalima. / Yu punya kalima. | 6 and 6 | | Mi ama dunia. / Yu ama dunia. | 5 and 5 | Each third line is longer than the pair it closes — 8, 6, 7, 8 — so every verse widens at the end. **The last line is the only regular one in the poem.** ``` Ki-ta a-ma du-nia u-na X . X . X . X . ``` Four beats, every other syllable, with no syllable left over: a line of four trochees. Every other line in the poem is irregular — *Din lai. Sol upar.* is four beats in five syllables, `XXXX.` — and the closing line, the one that says *we love the world together*, is the one that falls into step. That was luck, not craft. It could not have been craft, because when the line was written there was no rule that could tell anyone where its beats were. ## What the design gives a poem for free **Parallelism is exact.** *Mi ama · yu ama · kita ama* — three subjects, one verb, not a letter different. English cannot do this: *I love, you love, she loves*. Nor can Spanish, Arabic, Russian or Hindi, all of which bend the verb to its subject. Amadunia's promise that no word ever changes shape turns out to be a poetic resource and not only a learner's convenience, and this is the first text to use it on purpose. **Rhyme is cheap.** Half the three hundred roots end in a vowel and 23% end in *-a* alone; *-ar* and *-an* close another 8% between them. A rhyme is always available. That is worth saying plainly rather than boasting about: **a rhyme that costs nothing is worth little.** In a language where nearly a quarter of all words end in the same sound, rhyming is not a craft. The line read *a quarter* until September 6, 2026 — five lines under the 23% it was rounding, and the second page found this week where a fraction with no figure beside it overstated the figure four lines above it. If Amadunia ever grows a poetry, its music will have to come from rhythm and repetition, not from endings. ## What the language could not do 1. **There is no rhythm, because stress has never been decided.** [phonology.md](../grammar/phonology.md) records that as open, and until now it was an abstraction. Read the first line aloud twice, stressing *DIN lai* and then *din LAI*: both are correct Amadunia, and they are different poems. Two readers cannot agree on the metre of a single line. **This is the gap that matters most for anything meant to be spoken**, and a poem is where it stops being theoretical. 2. **No word for "change".** The first stanza is about a day turning into a night and back, and the language cannot name what is happening. Already [recorded](text-3-lingua-ini.md), and felt harder here. 3. **Fragments once more.** *Din lai. Sol upar.* are full sentences, but a poem wants to break a line where it likes, and nothing says whether it may. ## Roots used 21 of 300 — the fewest of any text here, which is the point. A poem should be sayable by someone three lessons in. --- # Seti din *Seven days.* The sixth text. A child is ill, and is taken to the hospital. It was written to reach the part of the dictionary nothing had reached. Five roots — *kulit*, *yanlis*, *foto*, *ba*, *nau* — appeared nowhere in a sentence in the whole repository: they existed as a dictionary row and nothing else. Forty-five more appeared exactly once. Most of them were body parts, illness words, and the high numbers. A doctor's visit is where those three meet without being forced together. Written at 300 roots, with no grammar beyond Lesson 23. --- ``` Anak bimar. Ta bimar seti din. Sar ta garam. Kulit ta garam tena. Ta no mau kula. Ta no mau lala. Mama ta espera. Anak no lebi hao. Mama aur anak go hospital pagi ini. Es cok insan in hospital. Nau insan espera in kamra baru. Ta-ta espera du hora. Doktor buka mulut anak. Ta lihat dente. Ta lihat ain aur telinga. Ta senti kulit anak. Kulit garam. Mesin ambil foto korpo anak. Doktor lihat foto. Mama tanya: "Kalb anak mal?" Doktor respon: "No. Kalb anak hao. Korpo ta kuat. Ide yu yanlis." Doktor beri ilac por mama. "Anak nomu ilac ba din." Anak nomu ilac ba din. Ta lala cok. Ta kula tena. Kulit ta no garam. Ta no bimar. Anak go skola tena. Ta senang. ``` --- ## Line by line | Amadunia | English | |---|---| | Anak bimar. | The child is ill. | | Ta bimar seti din. | He is ill seven days. | | Sar ta garam. | His head is hot. | | Kulit ta garam tena. | His skin is hot too. | | Ta no mau kula. | He does not want to eat. | | Ta no mau lala. | He does not want to sleep. | | Mama ta espera. | His mother waits. | | Anak no lebi hao. | The child is no better. | | Mama aur anak go hospital pagi ini. | Mother and child go to the hospital this morning. | | Es cok insan in hospital. | There are many people in the hospital. | | Nau insan espera in kamra baru. | Nine people wait in the new room. | | Ta-ta espera du hora. | They wait two hours. | | Doktor buka mulut anak. | The doctor opens the child's mouth. | | Ta lihat dente. | He looks at the teeth. | | Ta lihat ain aur telinga. | He looks at the eyes and the ears. | | Ta senti kulit anak. | He feels the child's skin. | | Kulit garam. | The skin is hot. | | Mesin ambil foto korpo anak. | A machine takes a picture of the child's body. | | Doktor lihat foto. | The doctor looks at the picture. | | Mama tanya: "Kalb anak mal?" | The mother asks: "Is the child's heart bad?" | | Doktor respon: "No." | The doctor answers: "No." | | Kalb anak hao. | The child's heart is good. | | Korpo ta kuat. | His body is strong. | | Ide yu yanlis. | Your idea is wrong. | | Doktor beri ilac por mama. | The doctor gives the mother medicine. | | "Anak nomu ilac ba din." | "The child drinks the medicine for eight days." | | Ta lala cok. | He sleeps a lot. | | Ta kula tena. | He eats again. | | Kulit ta no garam. | His skin is not hot. | | Ta no bimar. | He is not ill. | | Anak go skola tena. | The child goes to school again. | | Ta senang. | He is happy. | ## What the text found **A machine takes a photo of a body.** *Mesin ambil foto korpo anak.* The language has no word for an X-ray and did not need one: a machine, a picture, a body. Nothing was lost here — though calling it compounding, as this note first did, was wrong: it is a sentence, and [the language forms no compound words](../grammar/word-formation.md). **No copula anywhere in twelve predicates.** *Sar ta garam*, *Korpo ta kuat*, *Kulit garam* — every one of these is a noun and an adjective with nothing between them. English needs *is* twelve times here; Amadunia needs it zero times, because *es* is only for noun predicates. It appears once in the whole text, and there it means *there are*. **The high numbers work like the low ones.** *seti din*, *ba din*, *nau insan* — seven, eight and nine behave exactly as *du* does, which is the whole claim of the number system and had never actually been shown. ## Gaps Three things this text wanted to say and could not. None of them was invented to make the text work. **The perfect is not in the language, and this line was borrowing it.** *Ta bimar seti din* was glossed "he **has been** ill seven days" until September 6, 2026. The Amadunia has no tense marker at all and no word for *for*: it is an unmarked predicate with a bare duration after it, and the English perfect was doing work the sentence does not do. Whether *suda* covers the perfect is [an open question](../grammar/tense.md), and a gloss is not the place to answer it. It reads "he is ill seven days" now, which is awkward English and exact Amadunia. **There is no word for pain.** (First on [the list](../dictionary/README.md#words-the-writing-has-asked-for).) The child is ill for seven days and never says where it hurts, because there is no way to. The nearest root is *mal*, bad, which is not the same thing — a bad leg is not a leg that hurts. This is a dictionary gap, not a grammar gap, and it is the first one found by writing rather than by counting. Recorded for the vocabulary after A1. **Ordinals are still open, and this text is the clearest case yet.** The story runs on days: ill seven days, medicine eight days, back to school after that. It wants *on the eighth day* and *on the ninth day*, and can only say *for eight days*. The line "Anak go skola tena" — the child goes to school again — is a retreat from a date to an adverb. See the open question in [grammar/README.md](../grammar/README.md). **~~A doctor cannot tell a patient to open their mouth.~~ It can now.** The examination is written from outside — *Doktor buka mulut anak*, the doctor opens the child's mouth — because on the day it was written there was no settled imperative to write it from inside. This paragraph was one of the sites that argued for one, and unlike the sites counted in the briefing it was found by writing rather than by grep. The rule arrived on September 3, 2026: [a verb with no subject is a command](../grammar/sentence-types.md), so *Buka mulut* is now sayable. The text is left as written, because what it records is what the absence cost. ## Roots used 58 of 300. --- # Surat por mama *A letter to mother.* The seventh text, and the first that is written to someone. The stories tell, the third argues, the fourth is two people talking, the fifth is meant to be said aloud. A letter is none of those: one person, absent, addressing another person, absent, about a third place. It is the form that leans hardest on tense and on the second person, and it turned out to be the form the language is furthest from being able to finish. Written at 300 roots, with no grammar beyond Lesson 23. --- ``` Salam mama, Mi in madina baru sasa. Dom mi keci aur safi. Es pencere kabir in kamra mi. Mi lihat asman dari pencere. Insan-insan hao sini. Mi punya du rafiki baru. Ta-ta studi in skola una mi. Mi ingat dom kita. Mi ingat ca yu. Mi ingat golos papa. Mi sedih porke yu-yu baid. Mi suda anda por madina kemarin. Mi suda lihat ponte cang aur pahar kabir. Madina ini sundar. Aur ta no es dom mi. Mi saufa lai dom. Mi espera din itu. Mi hao sini. No es korku. Anak yu ``` --- ## Line by line | Amadunia | English | |---|---| | Salam mama, | Dear mother, | | Mi in madina baru sasa. | I am in the new city now. | | Dom mi keci aur safi. | My house is small and clean. | | Es pencere kabir in kamra mi. | There is a big window in my room. | | Mi lihat asman dari pencere. | I look at the sky from the window. | | Insan-insan hao sini. | The people are good here. | | Mi punya du rafiki baru. | I have two new friends. | | Ta-ta studi in skola una mi. | They study at school with me. | | Mi ingat dom kita. | I remember our house. | | Mi ingat ca yu. | I remember your tea. | | Mi ingat golos papa. | I remember father's voice. | | Mi sedih porke yu-yu baid. | I am sad because you are far. | | Mi suda anda por madina kemarin. | I walked through the city yesterday. | | Mi suda lihat ponte cang aur pahar kabir. | I saw a long bridge and a big mountain. | | Madina ini sundar. | This city is beautiful. | | Aur ta no es dom mi. | And it is not my home. | | Mi saufa lai dom. | I will come home. | | Mi espera din itu. | I wait for that day. | | Mi hao sini. | I am well here. | | No es korku. | There is no fear. | | Anak yu | Your child | ## What the text found **The three tenses carry the whole letter.** *Mi suda anda* yesterday, *Mi lihat* now, *Mi saufa lai* to come — three particles, no verb changing shape once. A letter in English needs walked, walk, will walk, saw, see, will see; this one needs *suda*, nothing, and *saufa*, and they never touch the verb. **Negation does the work of a whole sentence twice.** *Aur ta no es dom mi* — and it is not my home — is the turn the letter is built around, and *No es korku*, there is no fear, is the reassurance it ends on. The existential *es* and the negator *no* were settled separately and combine without a rule being added for the combination. ## Gaps Four things this letter wanted and could not have. None was invented. **A letter cannot be signed.** (*Love as a noun* is on [the list](../dictionary/README.md#words-the-writing-has-asked-for).) *With love* has no form: *ama* is the verb to love, and there is no noun for love. One root does one job, so the verb cannot be borrowed as a noun the way English borrows it — that is the rule working as designed, and the cost is that the most common closing in the world's correspondence cannot be written. The letter signs *Anak yu*, your child, which is a fact rather than a feeling. **It cannot say "I miss you".** The nearest the language reaches is *Mi sedih porke yu-yu baid*, I am sad because you are far, which is a paraphrase of the cause, not the feeling. This is a vocabulary gap and joins the missing word for pain that [text 6](text-6-seti-din.md) found. Both were found by writing, which is now the second time that has happened and the first time it has happened twice to the same kind of word — the ones about how a person feels. **It cannot say "before I came" or "after I arrive".** ([subordination.md](../grammar/subordination.md) keeps *although, before, after, until* open.) A letter is made of those. *although*, *before*, *after* and *until* are all on the open list, so every sentence here had to be arranged so that no clause needed one. That is why the middle of the letter is four short sentences instead of two long ones. **It could not say "in the month I arrive" without answering a live question.** Whether a plain time noun takes *in* was [only just recorded as open](../grammar/place.md), so the letter says *Mi espera din itu*, I wait for that day, and stays out of it. ## Roots used 54 of 300. --- # Kaifa suru ca *How to make tea.* The eighth text, and the first that tries to tell someone how to do something. Three stories, an argument, a dialogue, a poem and a letter came before it. This is the genre the language cannot yet write, and it was written anyway, to find out exactly what that costs. Written at 300 roots, with no grammar beyond Lesson 23, and **with no imperative**, because on the day it was written there was none. There is one now — [a verb with no subject is a command](../grammar/sentence-types.md) — and this text is what decided it. It is left in its original form on purpose; see [Gaps](#gaps) for what its opening would be today. --- ``` Yu ambil akua. Yu pone akua in gelas. Akua barid sasa. Yu pone gelas in ates. Yu espera. Akua garam sasa. Yu ambil ca. Yu pone ca in akua garam. Yu espera tri momen. Ca garam. Ca hao. Yu mau sukar? Yu pone sukar. Yu mau halib? Yu pone halib. Yu nomu ca sasa. ``` --- ## Line by line | Amadunia | English | |---|---| | Yu ambil akua. | You take water. | | Yu pone akua in gelas. | You put the water in a glass. | | Akua barid sasa. | The water is cold now. | | Yu pone gelas in ates. | You put the glass on the fire. | | Yu espera. | You wait. | | Akua garam sasa. | The water is hot now. | | Yu ambil ca. | You take tea. | | Yu pone ca in akua garam. | You put the tea in the hot water. | | Yu espera tri momen. | You wait three moments. | | Ca garam. Ca hao. | The tea is hot. The tea is good. | | Yu mau sukar? Yu pone sukar. | Do you want sugar? You put sugar. | | Yu mau halib? Yu pone halib. | Do you want milk? You put milk. | | Yu nomu ca sasa. | You drink the tea now. | ## What it cost **Sixteen sentences, twelve of them starting with *yu*.** Ten are instructions that would drop the subject in any language that has an imperative — English says *take water, put it in a glass, wait* and needs the word once; Turkish, Spanish, Indonesian, Swahili and Chinese each do the same in their own way. Amadunia says *yu* ten times, because a sentence without a subject would be an imperative, and there is no rule granting one. It is not unreadable. Recipes written as *you take, you put, you wait* exist in English cookbooks and are ordinary in spoken instruction. But it is the language's first genre that comes out visibly worse than the rest, and every line of this page is evidence for a decision that is already waiting. **A statement and a command looked identical, and that is what the rule fixed.** When this was written, *Yu pone sukar* was "you put sugar" — a description of what you do, or an instruction to do it, with nothing to tell them apart, and in a recipe that ambiguity is the whole text. The rule settled it by making the subject the difference: *Yu pone sukar* is the statement, *Pone sukar* is the command. This paragraph is the argument that decided it and is left standing as one. ## Gaps **~~There is no imperative.~~ Granted on September 3, 2026,** and this page is what decided it: it more than doubled the count of places one was wanted, from seven to seventeen. [The rule](../grammar/sentence-types.md) — a verb with no subject is a command, no word added. The text is left exactly as written, because it is the record of what the absence cost. What its opening would now be: ``` Ambil akua. Pone akua in gelas. Pone gelas in ates. Espera. ``` Four sentences, four subjects gone. **There is no word for *then*.** The steps of a recipe are a sequence, and the language can only set them side by side. *tena* is "again", not "next". Nothing here says which step follows which — the order on the page is doing work that the grammar is not. Recorded as a vocabulary gap alongside [the others found by writing](../dictionary/README.md#words-the-writing-has-asked-for). **It cannot say "wait until it is hot".** ([subordination.md](../grammar/subordination.md).) *until* is on the open list, so the text says *Yu espera* and then, separately, *Akua garam sasa* — the water is hot now. Two sentences where every language on Earth uses one. ## Roots used 19 of 300. --- # Pagi in madina *A morning in the city.* The ninth text, and the first written from a list rather than from an idea. [The frequency page](../dictionary/frequency.md) ended by noting that thirty-seven roots appear in exactly one sentence in the whole corpus, and that this is where the next text should look. Eighteen of them are here. An errand is the right shape for it: a bed, soap, shoes, a driver, the post office, the bank, rice and salt do not belong to one theme, but they belong to one morning. Written at 300 roots, with no grammar beyond Lesson 23. --- ``` Pagi lai. Mi start din. Mi go dari yatak. Mi gasil mano mi una sabun. Mi pone sapatu in kaki mi. Mi anda por tarik. Tair kanta in asman. Auto lai. Sofer sema: "Yu go nali?" Mi respon: "Mi go posta." Mi kirim surat in posta. Mi go banka. Mi ambil pesa. Mi kupi risi, legum, minyak aur namak in market. Legum ini baru. Mi beri fai pesa. Grupo anak-anak suru oyun una bola in tarik. Mi lihat ta-ta. Mi senang. Mi lai dom. Mi stop anda. Din finis. ``` --- ## Line by line | Amadunia | English | |---|---| | Pagi lai. | Morning comes. | | Mi start din. | I begin the day. | | Mi go dari yatak. | I get up from the bed. | | Mi gasil mano mi una sabun. | I wash my hands with soap. | | Mi pone sapatu in kaki mi. | I put shoes on my feet. | | Mi anda por tarik. | I walk along the road. | | Tair kanta in asman. | A bird sings in the sky. | | Auto lai. | A car comes. | | Sofer sema: "Yu go nali?" | The driver says: "Where are you going?" | | Mi respon: "Mi go posta." | I answer: "I am going to the post office." | | Mi kirim surat in posta. | I send a letter at the post office. | | Mi go banka. | I go to the bank. | | Mi ambil pesa. | I take money. | | Mi kupi risi, legum, minyak aur namak in market. | I buy rice, vegetables, oil and salt at the market. | | Legum ini baru. | These vegetables are fresh. | | Mi beri fai pesa. | I give five coins. | | Grupo anak-anak suru oyun una bola in tarik. | A group of children play a game with a ball in the road. | | Mi lihat ta-ta. | I watch them. | | Mi senang. | I am happy. | | Mi lai dom. | I come home. | | Mi stop anda. | I stop walking. | | Din finis. | The day is done. | ## What the text found **Writing from a list is not the same as being counted.** Eighteen of the thirty-seven appear in this text, but the first draft moved only seventeen of them. *legum* sat alone between two commas — *risi, legum, minyak aur namak* — and the frequency measure counts a word only inside a run of two or more, so a word in the middle of a list is named rather than used. It has its own sentence now. The measure is right and the draft was wrong: a list is not a use. **The commonest words did not get commoner.** But *mi* still opens fifteen of the twenty-two sentences, because a first-person errand cannot avoid it. The distribution is a fact about the language, not about the material. **A verb pair carried the frame.** *Mi start din* and *Mi stop anda* open and close the day. *start* and *stop* are already-global loans that had been sitting unused since Lesson 25 taught them, and they turn out to do exactly what they do everywhere else. **One morning holds words from six themes.** A bed, soap, a shoe, a foot, a driver, the post, a bank, rice, oil, salt, a bird, a ball. A thematic wordlist separates those; a life does not. That is an argument for writing texts rather than filling categories, and it is the same argument the vocabulary gaps made from the other side — [the words the writing asked for](../dictionary/README.md#words-the-writing-has-asked-for) were never the words a category would have suggested. ## Gaps **It cannot say "I got up".** (*To get up, to stand* is on [the list](../dictionary/README.md#words-the-writing-has-asked-for).) *Mi go dari yatak*, I go from the bed, is a paraphrase of a position change the language has no verb for. Rising, standing, sitting down — *otur* is to sit, and there is nothing for its opposite. **It cannot count money.** *Mi beri fai pesa* is "I give five money"; whether *pesa* pluralises, and what a coin is as against an amount, is untouched. [The clock could not be asked either](../phrasebook.md), and this is the same kind of hole: the language names things well and measures them poorly. **"Along the road" is *por tarik*, and that is a guess.** *por* is the preposition for *to, for*; using it for a path is not something any rule grants. [Place](../grammar/place.md) settles direction and location and says nothing about a route, and the second tier of prepositions — through, without, about — is recorded there as being for later. This sentence is the first to want one. ## Roots used 56 of 300. --- # Mila tahun *A thousand years.* The tenth text. [Text 9](text-9-pagi-in-madina.md) was written from the frequency list and emptied most of its tail; what stayed behind was all of one kind — *amani*, *arte*, *historia*, *legis*, *natura*, *numero*, *sansi*, *mati*, *proba*, *mila*. Peace, art, history, law, nature, number, chance, to die, to try, a thousand. Nine texts about tea, hospitals, letters and errands could not reach a single one of them, and that is not a failure of those texts. Abstract words need an argument, not a scene. This is the second text with an argument in it, after [text 3](text-3-lingua-ini.md), and the argument is what a language shares with a mountain. Written at 300 roots, with no grammar beyond Lesson 23. --- ``` Insan hidup sen tahun. Insan mati. Pahar hidup lebi dari insan. Pahar hidup mila tahun. Natura no mati. Insan suda katab historia. Historia no mati. Arte in dom eski. Kalima in libro eski. Legis lai dari insan. Legis no lai dari natura. Amani lai dari insan. Numero no mati. Uan es uan in dunia. Lingua mati kab insan stop sema. Mi proba sema. Yu proba sema. Kita punya sansi. ``` --- ## Line by line | Amadunia | English | |---|---| | Insan hidup sen tahun. | A person lives a hundred years. | | Insan mati. | A person dies. | | Pahar hidup lebi dari insan. | A mountain lives longer than a person. | | Pahar hidup mila tahun. | A mountain lives a thousand years. | | Natura no mati. | Nature does not die. | | Insan suda katab historia. | People wrote history. | | Historia no mati. | History does not die. | | Arte in dom eski. | The art is in the old house. | | Kalima in libro eski. | The words are in the old book. | | Legis lai dari insan. | Law comes from people. | | Legis no lai dari natura. | Law does not come from nature. | | Amani lai dari insan. | Peace comes from people. | | Numero no mati. | Number does not die. | | Uan es uan in dunia. | One is one, in the world. | | Lingua mati kab insan stop sema. | A language dies when people stop speaking. | | Mi proba sema. | I try to speak. | | Yu proba sema. | You try to speak. | | Kita punya sansi. | We have a chance. | ## What the text found **Abstraction costs nothing extra.** *Legis lai dari insan* is four words and so is *Mi kupi pan in market*. There is no register here, no formal vocabulary and no longer forms for weightier things — the same seventeen rules carry a shopping list and a claim about where law comes from. That is not an achievement of this text; it is what having no conjugation and no articles buys, and this is the first page where it can be seen plainly. **Negation does the arguing.** Four of the eighteen sentences are *no* sentences, and they are the spine of it: nature does not die, history does not die, law does not come from nature, number does not die. The language has one negator, it goes in one place, and it turns out to be enough to build a case with. **The last three sentences change person.** *Mi proba*, *Yu proba*, *Kita punya* — I, you, we. The pronouns never change shape, so the shift is carried by nothing but the words themselves, and the inclusive *kita* does the work that in English needs a choice about whether "we" includes the reader. ## Gaps **A gap claimed here on the first draft was not one.** The draft said the language could not compare durations — that *lebi* scales adjectives and nouns and nothing scales a verb — and [comparison.md](../grammar/comparison.md) has a section headed *With verbs* saying exactly the opposite: the same words work on a verb, in the adverb slot, *Mi kara lebi dari yu*. The text now says *Pahar hidup lebi dari insan*, a mountain lives longer than a person, which is what it wanted in the first place. A gap asserted without reading the rule page is not a gap. **It cannot say "for a thousand years".** *mila tahun* is a bare time expression, and whether that means a point or a span is [the question recorded in place.md](../grammar/place.md) about *in*. Here the span reading is the only sensible one, which is evidence but not a decision. **It said "does not die" four times where one word might have done.** The first draft recorded this as a missing word for *forever*. That was the second wrong claim on this page: *daima*, always, is in the dictionary, and *Natura hidup daima* would say nature lives forever. Whether one word should carry both habit and endlessness is a real question and it is now [on the list](../grammar/adverbs.md), next to the identical one about *cok* and "very". What is not true is that the language had nothing to reach for. ## Roots used 37 of 300. --- # Anak katab *A child writes.* The eleventh text, written for one rule. [The rules are all exercised somewhere](README.md), but the adverb — an adjective straight after the verb, describing the verb — appeared in one text out of ten. It is also the rule [the lessons used from Lesson 10 and did not introduce until Lesson 12](../grammar/adverbs.md), so it has been the least examined thing in the language twice over. Manner is what a lesson about learning to write is made of. This is that lesson, told rather than taught. Written at 300 roots, with no grammar beyond Lesson 23. --- ``` Mualim katab hao. Anak lihat. Anak proba katab. Anak katab mal. Mualim sema asan: "Yu proba tena." Anak katab tena. Mano ta no kuat. Anak studi kuat. Anak proba daima. Anak katab hao sasa. Anak kara hao libro ta. Mualim lihat kertas. Mualim senang. Mualim sema: "Yu suda studi hao." Anak kanta hayai. Anak kimbia dari skola. Anak katab nama ta in kertas baru. ``` --- ## Line by line | Amadunia | English | |---|---| | Mualim katab hao. | The teacher writes well. | | Anak lihat. | The child watches. | | Anak proba katab. | The child tries to write. | | Anak katab mal. | The child writes badly. | | Mualim sema asan: | The teacher says simply: | | "Yu proba tena." | "You try again." | | Anak katab tena. | The child writes again. | | Mano ta no kuat. | His hand is not strong. | | Anak studi kuat. | The child studies hard. | | Anak proba daima. | The child keeps trying. | | Anak katab hao sasa. | The child writes well now. | | Anak kara hao libro ta. | He reads his book well. | | Mualim lihat kertas. | The teacher looks at the paper. | | Mualim senang. | The teacher is happy. | | Mualim sema: "Yu suda studi hao." | The teacher says: "You have learned well." | | Anak kanta hayai. | The child sings fast. | | Anak kimbia dari skola. | The child runs from the school. | | Anak katab nama ta in kertas baru. | He writes his name on new paper. | ## What the text found **One rule carries the whole story — eleven times in seventeen sentences.** *katab hao*, *katab mal*, *katab tena*, *studi kuat*, *kanta hayai*, *kara hao libro* — the same adjective-after-verb in every one, and the story is the difference between *mal* and *hao*. Nothing is added to make an adjective describe a verb: the position does it, and the word does not change. **The extension earns its keep in one line.** *Anak kara hao libro ta* — the adverb comes before the object, which is the part [Lesson 18](../lessons/lesson-18-comparing-and-joining.md) adds to the rule. Without it, *kara libro hao* would read as a good book rather than reading well, and the sentence would be about the book. ## Gaps **There is no word for *slowly*.** ([the list](../dictionary/README.md#words-the-writing-has-asked-for).) The child writing badly is the point of the first half, and the natural sentence is that the child writes *slowly* rather than badly, which is not the same thing and not unkind in the same way. *hayai* is fast and has no opposite. The text used *mal*, bad, which says more than it means to. **It cannot say "better than before".** ([comparison.md](../grammar/comparison.md) grants comparing with *dari*; what is missing is a word for the earlier moment.) *lebi hao* is better, and comparing it to an earlier state needs *dari* with something to point at — a moment rather than a thing. The text says *Anak katab hao sasa*, the child writes well now, and lets *sasa* carry the change. ## Roots used 31 of 300. --- # Tarik por skola *The road to school.* The twelfth text, and the first written **for the imperative**. The rule was granted on September 3, 2026 — [a sentence that begins with a verb and has no subject is a command](../grammar/sentence-types.md) — and until this page nothing had been written to use it. [Text 8](text-8-kaifa-suru-ca.md) was written deliberately without one, to measure what the absence cost; this is the other half of that experiment. A mother walks a child to school. It is the situation a command belongs to: somebody who knows the road telling somebody who does not. Written at 300 roots, with no grammar beyond Lesson 25. --- ``` Pagi. Mama aur anak go skola. Ambil mano mi. Anda una mi. No kimbia. Stop! Lihat tarik. Es auto. Auto go hayai. Espera sini. Auto suda go. Tarik kosong sasa. Go sasa. Anak: Mi kan skola! Mama: Ya. Skola in tarik ini. Anak: Mi mau kimbia. Mama: Kimbia. No kimbia hayai. Anak kimbia skola. Mama lihat anak. Mama: Studi hao! Dengar mualim! Anak: Ya, mama! Bai! Mama: Bai. Mi espera yu in dom. ``` --- ## Line by line | Amadunia | English | |---|---| | Pagi. | Morning. | | Mama aur anak go skola. | A mother and a child go to school. | | Ambil mano mi. | Take my hand. | | Anda una mi. | Walk with me. | | No kimbia. | Do not run. | | Stop! | Stop! | | Lihat tarik. | Look at the road. | | Es auto. | There is a car. | | Auto go hayai. | The car goes fast. | | Espera sini. | Wait here. | | Auto suda go. | The car has gone. | | Tarik kosong sasa. | The road is empty now. | | Go sasa. | Go now. | | Mi kan skola! | I see the school! | | Skola in tarik ini. | The school is on this road. | | Mi mau kimbia. | I want to run. | | Kimbia. No kimbia hayai. | Run. Do not run fast. | | Anak kimbia skola. | The child runs to school. | | Mama lihat anak. | The mother watches the child. | | Studi hao! | Study well! | | Dengar mualim! | Listen to the teacher! | | Ya, mama! Bai! | Yes, mother! Goodbye! | | Bai. Mi espera yu in dom. | Goodbye. I will wait for you at home. | ## What the imperative did **Eleven of the twenty-seven sentences are commands**, and not one of them needed a word that did not already exist. *Ambil mano mi*, *Anda una mi*, *No kimbia*, *Stop*, *Lihat tarik*, *Espera sini*, *Go sasa*, *Kimbia*, *No kimbia hayai*, *Studi hao*, *Dengar mualim*. Put beside [text 8](text-8-kaifa-suru-ca.md), where twelve of sixteen sentences open with *yu* because nothing else could carry an instruction, the difference is the whole argument for the rule. Here *yu* appears **once**, in *Mi espera yu in dom*, where it is a real object and not a stand-in for a missing form. **The other half of the same rule turned up without being asked for.** Five of the twenty-seven are fragments — *Pagi.*, *Ya.*, *Ya, mama!*, *Bai!*, *Bai.* — which is the [constituent-alone rule](../grammar/sentence-types.md) settled on the same day. Neither was planned into this page; a road and a goodbye produced them. **Negation needed nothing new either.** *No kimbia* and *No kimbia hayai* put *no* in front of the verb exactly as a statement does, and *No kimbia hayai* also carries an adverb — the order is *no* → verb → adverb, which is the same order the full sentence rule gives. **A command and a statement never collide here.** *Mi mau kimbia* is a statement because it has a subject; *Kimbia* is a command because it has none. The two stand in consecutive lines of the dialogue and no reader can mix them. ## Gaps **There is still no way to tell someone to *be* something** ([sentence-types.md](../grammar/sentence-types.md) records it), and this page wanted it three times: - *Be careful* at the road — written as *Stop! Lihat tarik*, two commands for one. - *Be quick* when the road cleared — written as *Go sasa*, which says when and not how. - *Do not be afraid* of the car — impossible twice over: *korku* is a noun, and a sentence-initial *es* is the existential. That is [the open question the command rule left](../grammar/sentence-types.md), and this is the first page to hit it by writing rather than by reasoning. Three wants on one page is the strongest evidence it has. **No preposition for *near*.** *The school is near the tree* could not be written, so the text says *Skola in tarik ini* — the school is on this road — which is true but is not what a person pointing at a tree would say. *dekat* is an adjective, and using it in front of a noun would be inventing a preposition. [Already on the open list](../grammar/place.md), found again here. **No word for *then*.** The car passes, the road empties, the child runs; the order on the page is doing the work. Recorded with [the other gaps the writing has asked for](../dictionary/README.md#words-the-writing-has-asked-for), where [text 8](text-8-kaifa-suru-ca.md) put it first. ## Roots used 36 of 300. --- # Kula una *Eating together.* The thirteenth text, and the first written **for *una***. Counted across the lessons, the texts and the phrasebook on September 4, 2026, *una* stood in 21 sentences — fewer than any other settled rule, and fewer than half of what the adverb rule had before [text 11](text-11-anak-katab.md) was written for it. *una* does two jobs. Alone after a verb it is "together"; with a noun after it, it is "with". Of the 21, seventeen were "with" and **four** were "together", so the rarer of the two jobs is what this page mostly does. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Aksam. Familia mi in dom. Mama tabak corba una legum. Papa gasil piring una sabun. Mi leta pan dari market una dugu mi. Rafiki mi lai. Kanis ta lai una ta. Kita otur una. Kita kula una. Corba garam. Pan baru. Mama sema: "Kula, plis." Kita kula aur sema aur kanta. Anak keci lala in kursi. Papa leta ta por yatak. Rat. Es sukut in dom. Mi senang porke kita una. ``` --- ## Line by line | Amadunia | English | |---|---| | Aksam. | Evening. | | Familia mi in dom. | My family is at home. | | Mama tabak corba una legum. | Mother cooks soup with vegetables. | | Papa gasil piring una sabun. | Father washes the plates with soap. | | Mi leta pan dari market una dugu mi. | I bring bread from the market with my sibling. | | Rafiki mi lai. | My friend comes. | | Kanis ta lai una ta. | His dog comes with him. | | Kita otur una. | We sit together. | | Kita kula una. | We eat together. | | Corba garam. Pan baru. | The soup is hot. The bread is fresh. | | Mama sema: "Kula, plis." | Mother says: "Eat, please." | | Kita kula aur sema aur kanta. | We eat and talk and sing. | | Anak keci lala in kursi. | The small child sleeps in the chair. | | Papa leta ta por yatak. | Father carries him to bed. | | Rat. Es sukut in dom. | Night. There is silence in the house. | | Mi senang porke kita una. | I am happy because we are together. | ## What *una* did **Seven of the eighteen sentences use it**, and between them they cover both jobs and three kinds of "with" that English keeps apart: | | | |---|---| | una legum | an ingredient | | una sabun | an instrument | | una dugu mi, una ta | a companion | | otur una, kula una, kita una | together | The last row is the one that needed writing. Four sentences in the whole corpus had used the bare adverb; this page adds three more and one of them, *Mi senang porke kita una*, uses it with no verb at all — the shape the copula rule licenses, and the shape that ["Mi una yu"](../lessons/lesson-22-doing-and-feeling.md) was corrected into on the same day. ## Gaps **With an object, the two jobs of *una* collide.** *Kita kula una* is "we eat together", and *Kita kula una pan* should be "we eat together, bread" — but it reads as "we eat with bread", because *una* followed by a noun is the preposition. There is no way to say *we eat the bread together* in one sentence. This page avoids the collision by keeping every *together* sentence free of an object, which is a way of writing around a hole, not a solution. [Recorded as an open question in place.md](../grammar/place.md). **This page's claim about *but* did not survive inspection, and is withdrawn.** It said *Corba garam. Pan baru.* would rather be "the soup was hot **but** the bread was old". The text is in the present, and *baru* is **new** — the table above glosses the line "The soup is hot. The bread is fresh." Two pieces of praise, which *aur* joins perfectly well. The note described a sentence this page never wrote, and [the briefing](../grammar/proposal-but.md) counted it as one of seven pages of evidence until September 5, 2026. It is six now, and the six are unaffected. A gap note is evidence the founder decides on; it has to quote the page it is on. **No word for *then*** ([the list](../dictionary/README.md#words-the-writing-has-asked-for))**.** The child falls asleep and is carried to bed; the order on the page is again doing what the grammar cannot. ## Roots used 43 of 300. --- # Yamur aur ca *Rain and tea.* The fourteenth text, and the first written **for a learner in the middle of the course**. [The reading ladder](../lessons/reading-ladder.md) showed the problem: [story 1](story-1-anak-espera-sol.md) opens at Lesson 13 and the next text does not open until Lesson 24. Eleven lessons with nothing new to read. This one uses **only the 147 roots the lessons have taught by Lesson 18**, and no rule beyond that lesson. It is not a simplified text — it is a text written inside a smaller language, which is a different thing and a harder one. --- ``` Rat. Yamur in madina. Mi anda in tarik. Angin barid. Mi mau ca garam. Es kafe keci sini. Mi go kafe. Mi otur in kursi. Rafiki mi lai. Ta kotor aur barid. Mi sema: "Yu barid?" Ta sema: "Mi no bil tarik por dom mi." Mi kan telefon mi. Telefon mi no hao. Kita no bisa sema una telefon. Mi sema: "Bas. Kita anda una." Kita nomu ca. Ca garam. Kita espera. Yamur no lai tena. Kita anda in tarik una. Mi bil tarik ini. Dom ta situ, upar. Ta sema: "Mersi, rafiki." Ta go dom ta. Mi go dom mi. Rat. Mi no sedih. ``` --- ## Line by line | Amadunia | English | |---|---| | Rat. Yamur in madina. | Night. Rain in the city. | | Mi anda in tarik. | I walk on the road. | | Angin barid. | The wind is cold. | | Mi mau ca garam. | I want hot tea. | | Es kafe keci sini. | There is a small cafe here. | | Mi go kafe. Mi otur in kursi. | I go to the cafe. I sit in a chair. | | Rafiki mi lai. | My friend comes. | | Ta kotor aur barid. | She is dirty and cold. | | Mi sema: "Yu barid?" | I say: "Are you all right?" | | Ta sema: "Mi no bil tarik por dom mi." | She says: "I do not know the road to my house." | | Mi kan telefon mi. Telefon mi no hao. | I look at my phone. My phone is not good. | | Kita no bisa sema una telefon. | We cannot speak with a telephone. | | Mi sema: "Bas. Kita anda una." | I say: "Enough. We walk together." | | Kita nomu ca. Ca garam. | We drink tea. The tea is hot. | | Kita espera. Yamur no lai tena. | We wait. The rain does not come again. | | Kita anda in tarik una. | We walk on the road together. | | Mi bil tarik ini. | I know this road. | | Dom ta situ, upar. | Her house is there, above. | | Ta sema: "Mersi, rafiki." | She says: "Thank you, friend." | | Ta go dom ta. Mi go dom mi. | She goes to her house. I go to mine. | | Rat. Mi no sedih. | Night. I am not sad. | ## What a smaller language cost Twenty-nine sentences, 45 roots, and **every one of them taught by Lesson 18**. The grammar is all there — existence, the verb chain, *una*, a command, a question, a subordinate clause — because Lesson 18 is where the last of those arrives. What is missing is words, and the shape of what is missing is worth recording, because none of it is a hole in the language: | Wanted | Arrives in | Written instead | |---|---|---| | *senang*, happy | Lesson 25 | *Mi no sedih* — I am not sad | | *tanya*, to ask | Lesson 25 | *Mi sema: "Yu barid?"* | | *punya*, to have | Lesson 25 | left out; the phone is described, not owned | | *pencere*, a window | Lesson 21 | the cafe has no window | **The ending is the clearest case.** The story wants *I am happy*, and at Lesson 18 the only feeling word is *sedih*. So it ends on *Mi no sedih*, which is a colder sentence than the one intended and is the honest one to write. A learner who reaches Lesson 25 can come back and see what the language gained. ## Gaps These are gaps in the **course order**, not in the language: every word above exists and is taught later. That distinction matters, and this is the first text able to make it — every other text was written at 300 roots, where the question cannot come up. The two standing gaps are here as well. **No word for *then***: the friend arrives, the tea is drunk, the rain stops, and the page's order does the work. **No word for *but***: *Ca garam* would rather be "the tea is hot but the night is cold". Both are [recorded with the others](../dictionary/README.md#words-the-writing-has-asked-for). ## Roots used 45 of 300. --- # Sol lai *The sun comes.* The fifteenth text, and the lowest rung on [the reading ladder](../lessons/reading-ladder.md): it uses **only the 89 roots the lessons have taught by Lesson 10**, and no rule beyond that lesson. Ten lessons in, a learner has no prepositions, no *ini* or *itu*, no verb chains, no comparison and no subordinate clauses — those arrive between Lessons 12 and 18. What is there is a subject, a verb, an object, an adjective after its noun, the two tense markers, the plural, possession, a question and a command. This page is what that is enough for. --- ``` Sol lai. Din hao. Mama mi suda kupi pan. Papa mi nomu ca. Mama sema: "Kula pan!" Anak keci lala. Mi kan ta. Mi sema: "Sasa, anak! Sol lai." Anak ambil mano mi. Kita go market. Market kabir. Cok insan situ. Mi kupi ca. Anak kan foto. Pesa mi keci sasa. Rafiki mi lai. Ta sema: "Hi! Yu hao?" Mi sema: "Ya. Familia mi hao." Ta sema: "Yu suru ke?" Mi sema: "Mi suru rabota." Kita go dom. Rat lai. Familia mi sini. Anak lala. ``` --- ## Line by line | Amadunia | English | |---|---| | Sol lai. Din hao. | The sun comes. The day is good. | | Mama mi suda kupi pan. | My mother bought bread. | | Papa mi nomu ca. | My father drinks tea. | | Mama sema: "Kula pan!" | Mother says: "Eat bread!" | | Anak keci lala. | The small child sleeps. | | Mi kan ta. | I see him. | | Mi sema: "Sasa, anak! Sol lai." | I say: "Now, child! The sun is up." | | Anak ambil mano mi. | The child takes my hand. | | Kita go market. | We go to the market. | | Market kabir. | The market is big. | | Cok insan situ. | There are many people there. | | Mi kupi ca. | I buy tea. | | Anak kan foto. | The child sees a photograph. | | Pesa mi keci sasa. | My money is small now. | | Rafiki mi lai. | My friend comes. | | Ta sema: "Hi! Yu hao?" | She says: "Hello! Are you well?" | | Mi sema: "Ya. Familia mi hao." | I say: "Yes. My family is well." | | Ta sema: "Yu suru ke?" | She says: "What are you doing?" | | Mi sema: "Mi suru rabota." | I say: "I'm working." | | Kita go dom. | We go home. | | Rat lai. Familia mi sini. | Night comes. My family is here. | | Anak lala. | The child sleeps. | ## What ten lessons are enough for Twenty-seven sentences and 42 roots. A morning, an errand, a meeting, an evening — the shape of a day, told with no rule the course has not reached. **Four things the page had to do without**, and none of them is missing from the language: | Not yet taught | Arrives in | Written instead | |---|---|---| | *in*, *dari*, *por* | Lesson 15 | *Kita go market* — a movement verb takes its target bare, so going somewhere works; **being** somewhere does not | | *ini*, *itu* | Lesson 15 | nothing is pointed at; the photograph is *foto*, never *this photograph* | | verb chains | Lesson 17 | *Mi suru rabota*, not *Mi mau suru rabota* | | *porke*, *kab* as joiners | Lesson 18 | every sentence stands alone; nothing is because of anything | **The last line of the page is where the absence shows.** *My family is at home* cannot be written at Lesson 10: *dom* is a noun, a noun predicate needs *es*, and *Familia mi es dom* would say my family is a house. The page says *Familia mi sini* — my family is here — which is true, and is not the same sentence. **Nothing is because of anything.** Without *porke* the child wakes, the market is crowded and the money runs short, and the page cannot say that one follows from another. That is the single largest difference between this text and [text 14](text-14-yamur-aur-ca.md), eight lessons later. ## Roots used 42 of 300. --- # Berapa harga *How much does it cost.* The sixteenth text, written for **the words the writing had never reached for**. Counted on September 4, 2026: **55 of the 300 roots appeared in no text at all** — they were taught in a lesson, drilled once, and never used again. The surprise was what they are. Not obscure nouns: *o* (or), *kadar* (as … as), *kurang* (less), *lasim* (must), *berapa* (how many), *kaifa* (how), *fikir* (to think). Function words — the joints of the language — exercised in lessons and never once in a sentence anybody wrote for its own sake. A market is where a person says *how much*, *less*, *or*, and *as … as*, so a market is what this is. --- ``` Mi lasim go market hafta ini. Tacir beca fruta, telur aur nyama. Mi tanya: "Berapa harga fruta?" Ta respon: "Des pesa." Mi fikir. Harga ini kabir kadar harga in madina baru. Mi sema: "Kurang, plis." Tacir sema: "Ba pesa. Bas." Ta tanya: "Yu mau nyama o samaki?" Mi respon: "Samaki, plis. Nyama lebi eski." Mi kupi fruta, telur aur samaki. Mano mi plen. Fruta ini dulce. Tacir funga market in aksam. Insan-insan go dom. Kaifa mi tabak samaki? Mi no bil. Mi fikir mi lasim tanya mama mi. Mi pone samaki in tabla. Mi ambil sikin. Mausim ini barid, aur cok badal in asman. Besok mi saufa tabak samaki una mama mi. Mi fikir samaki saufa hao. ``` --- ## Line by line | Amadunia | English | |---|---| | Mi lasim go market hafta ini. | This week I have to go to the market. | | Tacir beca fruta, telur aur nyama. | The merchant sells fruit, eggs and meat. | | Mi tanya: "Berapa harga fruta?" | I ask: "How much is the fruit?" | | Ta respon: "Des pesa." | He answers: "Ten money." | | Mi fikir. | I think. | | Harga ini kabir kadar harga in madina baru. | This price is as big as the price in the new city. | | Mi sema: "Kurang, plis." | I say: "Less, please." | | Tacir sema: "Ba pesa. Bas." | The merchant says: "Eight money. Enough." | | Ta tanya: "Yu mau nyama o samaki?" | He asks: "Do you want meat or fish?" | | Mi respon: "Samaki, plis. Nyama lebi eski." | I answer: "Fish, please. The meat is older." | | Mi kupi fruta, telur aur samaki. | I buy fruit, eggs and fish. | | Mano mi plen. | My hands are full. | | Fruta ini dulce. | This fruit is sweet. | | Tacir funga market in aksam. | The merchant closes the market in the evening. | | Insan-insan go dom. | The people go home. | | Kaifa mi tabak samaki? | How do I cook fish? | | Mi no bil. | I do not know. | | Mi fikir mi lasim tanya mama mi. | I think I have to ask my mother. | | Mi pone samaki in tabla. Mi ambil sikin. | I put the fish on the table. I take a knife. | | Mausim ini barid, aur cok badal in asman. | This season is cold, and there are many clouds in the sky. | | Besok mi saufa tabak samaki una mama mi. | Tomorrow I will cook the fish with my mother. | | Mi fikir samaki saufa hao. | I think the fish will be good. | ## What it reached **Twenty-one of the fifty-five**, counted rather than guessed: *badal*, *beca*, *berapa*, *besok*, *dulce*, *fikir*, *fruta*, *funga*, *hafta*, *kadar*, *kaifa*, *kurang*, *lasim*, *mausim*, *nyama*, *o*, *plen*, *samaki*, *sikin*, *tabla*, *telur*. **Thirty-four are still unused** — one scene about buying food reached two fifths of them, which says something about the rest. The function words behaved. *o* joined two nouns inside a question and needed nothing; *kadar* compared two prices with no degree word, as [comparison.md](../grammar/comparison.md) says it must; *kurang* stood alone as a whole turn, which the fragment rule allows and no lesson had shown; *lasim* chained twice, once with *go* and once with *tanya*. ## Gaps **No word for *but*, again.** *Nyama lebi eski* wants to be "the meat is cheaper **but** older", and the bargaining scene is where a contrast belongs most. Ninth on [the list](../dictionary/README.md#words-the-writing-has-asked-for) and now asked for by three separate pages. **A price cannot be named.** *Des pesa* is "ten money", because *pesa* is the substance and there is no coin. [Recorded from text 9](../dictionary/README.md#words-the-writing-has-asked-for); a market text makes it worse, because bargaining is entirely about the difference between an amount and a price. **No word for *cheap* or *expensive*.** ([the list](../dictionary/README.md#words-the-writing-has-asked-for).) The whole exchange runs on *kabir* and *kurang* — big and less — and says *this price is as big as the price in the new city* where a person would say *this is dear*. Found here, and the first time a text has wanted it. ## Roots used 62 of 300. --- # Tren aur farasi *The train and the horse.* The seventeenth text, and the second written for **the roots no text had ever used**. [Text 16](text-16-berapa-harga.md) took the market words and the function words and brought the count from 55 down to 34. This one goes to the country after the rest. A journey was the obvious shape, because most of what was left is out there: a train, a taxi, a farmer, a horse, a cat, a garden, stones, a key, a shirt. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Uan mensis suda go. Mi suda punya mimpi: safari kabir. Mi ambil taksi. Sofer leta mi por tren. Tren anda hayai. Tren anda pat hora. Mi kan negara mi. Petra kabir in pahar. Asman asul. Nasab mi espera sini. Ta es kisan. Cehra ta hao. Rambut ta siya. Ta buka pintu una anahtar. Ta punya farasi aur kucing keci. Animal-animal in bage. Bage sub, dom upar. Mi beri regalo por ta: kamisa asul. Ta senang. Mi senang. Kita otur. Ta sema habari negara. Kita dengar musik. Musik ini es kultura ta. Rat lai. Mi lala. Mi punya mimpi tena. Besok mi saufa go dom mi. Tarik duan aur safari no muskil. Mi fikir uhuru es safari. ``` --- ## Line by line | Amadunia | English | |---|---| | Uan mensis suda go. | One month has gone. | | Mi suda punya mimpi: safari kabir. | I had a dream: a big journey. | | Mi ambil taksi. | I take a taxi. | | Sofer leta mi por tren. | The driver brings me to the train. | | Tren anda hayai. | The train goes fast. | | Tren anda pat hora. | The train travels four hours. | | Mi kan negara mi. | I see my country. | | Petra kabir in pahar. | Big stones on the mountain. | | Asman asul. | The sky is blue. | | Nasab mi espera sini. | My relative waits here. | | Ta es kisan. | He is a farmer. | | Cehra ta hao. Rambut ta siya. | His face is good. His hair is black. | | Ta buka pintu una anahtar. | He opens the door with a key. | | Ta punya farasi aur kucing keci. | He has a horse and a small cat. | | Animal-animal in bage. | The animals are in the garden. | | Bage sub, dom upar. | The garden is below, the house above. | | Mi beri regalo por ta: kamisa asul. | I give a gift to him: a blue shirt. | | Ta senang. Mi senang. | He is happy. I am happy. | | Kita otur. | We sit. | | Ta sema habari negara. | He tells the news of the country. | | Kita dengar musik. | We listen to music. | | Musik ini es kultura ta. | This music is his culture. | | Rat lai. Mi lala. | Night comes. I sleep. | | Mi punya mimpi tena. | I dream again. | | Besok mi saufa go dom mi. | Tomorrow I will go to my house. | | Tarik duan aur safari no muskil. | The road is short and the journey is not difficult. | | Mi fikir uhuru es safari. | I think freedom is a journey. | ## What it reached **Twenty-seven roots used here for the first time in any text**: *anahtar*, *animal*, *asul*, *bage*, *cehra*, *duan*, *farasi*, *habari*, *kamisa*, *kisan*, *kucing*, *kultura*, *mensis*, *mimpi*, *musik*, *muskil*, *nasab*, *negara*, *pat*, *petra*, *rambut*, *regalo*, *safari*, *sub*, *taksi*, *tren*, *uhuru*. The draft called it sixteen; the list under it had twenty-seven in it, and the count is the thing that was wrong. **Seven roots are now left**, and two of them cannot be used at all yet: *madad*, whose class is undecided, and *kadang*, whose position as a frequency adverb is an open question. Writing either would answer a question this text has no standing to answer. The other five are *burun*, *dekat*, *gusa*, *kiri* and *polisi*. *sub* was the one that needed care. It is a place word, like *sini* and *situ*, not a preposition: *Bage sub* is "the garden is below", and *sub dom* would be inventing a fourth preposition beside *in*, *dari* and *por*. The line was written the way the language allows and reads better for it. *pat*, four, was the last digit no text had used. It is in *Tren anda pat hora* — four hours — which also leaves duration unmarked, the way [Lesson 13](../lessons/lesson-13-weather-directions.md) already does with *din cang*. ## Gaps **No word for a station.** (Not on [the list](../dictionary/README.md#words-the-writing-has-asked-for) and not in [the A1 checklist](../dictionary/a1-checklist.md) either — a station is a thing neither method had reached.) The taxi brings me *por tren*, to the train, because there is nowhere for a train to be. Every journey text will hit this. **Duration has no marker.** *Tren anda pat hora* is "the train goes four hours", and nothing distinguishes that from a time when. The language has been doing this since Lesson 13 and has never decided it; it belongs beside [the question about *in* and time](../grammar/place.md). **No word for *but*, a fourth time.** *Tarik duan aur safari no muskil* wants "the road is short **but** the journey is not easy". Every text that has tried to contrast two clauses has wanted it — [the ninth gap](../dictionary/README.md#words-the-writing-has-asked-for), now asked for by four pages. ## Roots used 73 of 300. --- # Gusa in tarik *Anger on the road.* The eighteenth text, and the last of the sweep. Two days ago 55 roots had never appeared in any text; [text 16](text-16-berapa-harga.md) took twenty-one and [text 17](text-17-tren-aur-farasi.md) took twenty-seven. Five were left that could be written: *burun*, *dekat*, *gusa*, *kiri*, *polisi*. They are all here, in the one scene that holds them — a blocked street. Two more remain and are not here on purpose. See the gaps below. --- ``` Madina in aksam. Cok auto in tarik. Auto no anda. Sofer punya gusa. Ta sema kuat: "Kaifa mi go dom?" Polisi lai. Ta sema: "Sabar, plis." Polisi suda funga tarik ini. Ta sema: "Go kiri." Sofer tanya: "Market dekat?" Polisi respon: "Ya, dekat. Pat momen." Anak keci in auto. Burun ta barid. Mama ta funga pencere. Auto anda kiri. Tarik baru duan. Sofer no punya gusa sasa. Polisi anda in tarik tena. ``` --- ## Line by line | Amadunia | English | |---|---| | Madina in aksam. | The city in the evening. | | Cok auto in tarik. | Many cars on the road. | | Auto no anda. | The cars do not move. | | Sofer punya gusa. | The driver has anger. | | Ta sema kuat: "Kaifa mi go dom?" | He says strongly: "How do I get home?" | | Polisi lai. | A police officer comes. | | Ta sema: "Sabar, plis." | She says: "Patience, please." | | Polisi suda funga tarik ini. | The police have closed this road. | | Ta sema: "Go kiri." | She says: "Go left." | | Sofer tanya: "Market dekat?" | The driver asks: "Is the market near?" | | Polisi respon: "Ya, dekat. Pat momen." | She answers: "Yes, near. Four moments." | | Anak keci in auto. | A small child is in the car. | | Burun ta barid. | His nose is cold. | | Mama ta funga pencere. | His mother closes the window. | | Auto anda kiri. | The car goes left. | | Tarik baru duan. | The new road is short. | | Sofer no punya gusa sasa. | The driver has no anger now. | | Polisi anda in tarik tena. | The officer walks the road again. | ## What it reached The five: *burun*, *dekat*, *gusa*, *kiri*, *polisi*. **Every root in the dictionary that can be written has now been written in a text**, and [`check.py`](../check.py) holds that. *gusa* is the interesting one. It is a noun — anger is something a person *has*, the way *korku* is — so *Sofer punya gusa* and, at the end, *Sofer no punya gusa sasa*. English would say he *was* angry and then *calmed down*; Amadunia says he had a thing and then did not. The frame is the language's, not a translation of ours, and it came out of the grammar rather than being designed. *dekat* stayed an adjective. *Market dekat?* is "is the market near?", a predicate, and nowhere does this page write *dekat market* — that would be [the preposition the language does not have](../grammar/place.md). ## Gaps **Two roots are left and neither can be written.** *madad* — its class is undecided, so a sentence using it would answer [the open question](../grammar/verb-chains.md) by accident. *kadang* — where a frequency adverb stands is [open](../grammar/adverbs.md), so it cannot appear in a sentence without settling that. This scene wanted both. *Polisi madad sofer*, the officer helps the driver, is the sentence the story is actually about, and it cannot be written. *Kadang tarik funga*, sometimes the road is closed, is what the officer would say. Two open questions, two sentences, and the words sitting in the dictionary unable to work. **No word for *but*, a fifth time** ([the briefing](../grammar/proposal-but.md))**.** *Tarik baru duan* wants "the new road is short **but** slow". ## Roots used 43 of 300. --- # Kamra mi *My room.* The nineteenth text, written for the largest gap left on [the reading ladder](../lessons/reading-ladder.md): between Lesson 18 and Lesson 24 there was nothing to read. It uses **only the roots taught by Lesson 21**, which is where a room and its things arrive, so a learner who has just finished that lesson reads a page made of what they have just learned. It opened at 22 for one day: the word was *pagi*, morning, which the course did not teach until Lesson 22 — until [Lesson 03 took it back](../lessons/lesson-03-people.md), beside *din* and *rat*. --- ``` Kamra mi keci. Es uan pencere. Pencere kabir. Mi kan pohon aur asman. In pagi sol lai in kamra. Es yatak, tabla aur kursi. Libro, kertas aur lampu in tabla. Lampu mi eski. Ta beri luma hao. Mi kara libro in yatak. Mi katab surat in tabla. Kalima mi keci, ide mi kabir. Flor in pencere. Flor merah aur putih. Mi beri akua por flor in pagi. Rafiki mi lai. Kita nomu ca aur sema. Kamra keci. Kita sini. In rat mi funga pencere. Kamra barid. Mi lala hao. Kamra ini keci. Mi ama kamra mi. ``` --- ## Line by line | Amadunia | English | |---|---| | Kamra mi keci. | My room is small. | | Es uan pencere. | There is one window. | | Pencere kabir. | The window is big. | | Mi kan pohon aur asman. | I see a tree and the sky. | | In pagi sol lai in kamra. | In the morning the sun comes into the room. | | Es yatak, tabla aur kursi. | There is a bed, a table and a chair. | | Libro, kertas aur lampu in tabla. | A book, paper and a lamp are on the table. | | Lampu mi eski. Ta beri luma hao. | My lamp is old. It gives good light. | | Mi kara libro in yatak. | I read a book in bed. | | Mi katab surat in tabla. | I write a letter at the table. | | Kalima mi keci, ide mi kabir. | My words are small, my ideas are big. | | Flor in pencere. Flor merah aur putih. | Flowers are in the window. The flowers are red and white. | | Mi beri akua por flor in pagi. | I give water to the flowers in the morning. | | Rafiki mi lai. Kita nomu ca aur sema. | My friend comes. We drink tea and talk. | | Kamra keci. Kita sini. | The room is small. We are here. | | In rat mi funga pencere. | At night I close the window. | | Kamra barid. Mi lala hao. | The room is cold. I sleep well. | | Kamra ini keci. Mi ama kamra mi. | This room is small. I love my room. | ## What Lesson 21 is enough for Twenty-four sentences, 48 roots, and no rule the course has not reached: the existential, possession, a subordinate-free description, an adverb, a conjunction. The vocabulary is the lesson's own — *kamra*, *pencere*, *lampu*, *yatak*, *kertas*, *kalima*, *ide*, *flor* — which is the point of writing at this rung rather than at 300. ## Gaps **A room with no walls and no floor.** The dictionary has a house, a room, a door, a window, a bed, a table, a chair and a lamp, and not one surface to put them against. *Kamra mi keci* is as close as the language gets to describing the shape of a room. Three words are missing and nothing here can stand in for them, so the description goes around the room's contents and never touches the room. Recorded as [a gap the writing asked for](../dictionary/README.md#words-the-writing-has-asked-for). **No word for *every*.** *Mi beri akua por flor in pagi* says "in the morning", where a person waters plants **every** morning. *all, some, none* is [on the open list](../grammar/plural.md) and this is the first page to want the fourth member of that family. **No word for *but*, a sixth time.** *Kamra keci. Kita sini* is two sentences where one was meant: the room is small **but** we fit. The draft of this page wrote it with *lakin*, the candidate in [the briefing](../grammar/proposal-but.md), to see how the sentence would look if the briefing were accepted — and [`check.py`](../check.py) rejected it, correctly. **A text invents nothing.** A briefing is where a word that does not exist may be written; a text is where the absence has to be felt. The line stands as two sentences, and the gap is recorded rather than papered over. **A number cannot count a pronoun, and this page had assumed it could.** The line was *Kita tri sini*, "the three of us are here", until September 5, 2026. Numbers go before what they count — *tri anak* — and nothing in the language attaches one to *kita*: *tri kita* is not a shape either, and *kita tri* would read as possession if it read as anything. The line is *Kita sini* now, and "the three of us" is [recorded as an open question in numbers.md](../grammar/numbers.md). It was written by reflex from English and stood for a day, which is what the rule about texts inventing nothing exists to prevent — the invention here was not a word but a shape, and no check was looking for shapes. ## Roots used 48 of 300. --- # In skola *At school.* The twentieth text, written for the gap between Lesson 13 and Lesson 18 on [the reading ladder](../lessons/reading-ladder.md). It uses **only the 136 roots the lessons have taught by Lesson 16**, which is the lesson called *School and time* — so a learner who has just finished it reads a page about what it just taught. No verb chains, no existential *es*, no *porke* joining two clauses, no comparison of adjectives: all four arrive in Lessons 17 and 18. --- ``` Pagi. Anak-anak go skola. Skola in tarik ini. Skola kabir. Anak-anak otur in kursi. Mi buka libro mi. Mi kara. Rafiki mi katab in libro ta. Mi no faham. Mi sema: "Kaifa?" Rafiki mi sema: "Kan ini." Mi kan. Sasa mi faham. Sol upar. Kita kula pan in skola. Anak-anak kimbia in tarik. Aksam. Mi go dom. Mi kara libro tena. Mi studi. Papa mi sema: "Yu studi hao." Mi mau libro baru. Pesa mi keci. Besok mi saufa go skola tena. ``` --- ## Line by line | Amadunia | English | |---|---| | Pagi. Anak-anak go skola. | Morning. The children go to school. | | Skola in tarik ini. Skola kabir. | The school is on this road. The school is big. | | Anak-anak otur in kursi. | The children sit in chairs. | | Mi buka libro mi. Mi kara. | I open my book. I read. | | Rafiki mi katab in libro ta. | My friend writes in his book. | | Mi no faham. | I do not understand. | | Mi sema: "Kaifa?" | I say: "How?" | | Rafiki mi sema: "Kan ini." | My friend says: "Look at this." | | Mi kan. Sasa mi faham. | I look. Now I understand. | | Sol upar. Kita kula pan in skola. | The sun is up. We eat bread at school. | | Anak-anak kimbia in tarik. | The children run in the road. | | Aksam. Mi go dom. | Evening. I go home. | | Mi kara libro tena. Mi studi. | I read the book again. I study. | | Papa mi sema: "Yu studi hao." | My father says: "You study well." | | Mi mau libro baru. Pesa mi keci. | I want a new book. My money is small. | | Besok mi saufa go skola tena. | Tomorrow I will go to school again. | ## What sixteen lessons are enough for Twenty-four sentences, 42 roots. A day at school told with the plain sentence: subject, verb, object, an adjective after its noun, a preposition, a demonstrative, the plural, possession, a command, a question and the two tense markers. Nothing here is a workaround; it is simply a smaller language, and the smaller language is the one the learner has at that point. **One sentence carries the whole of the missing grammar.** *Mi no faham. Mi sema: "Kaifa?"* is two sentences because at Lesson 16 there is no *porke* to join them — *I did not understand, so I asked how* has to be said twice. Every other absence in this page is the same shape: the rule arrives one or two lessons later. ## Gaps **Noon cannot be named.** The draft wrote *Hora du* for "two o'clock", which is "hour two" and claims a clock the language does not have — [the gap is on the list](../dictionary/README.md#words-the-writing-has-asked-for). The line says *Sol upar*, the sun is up, which is what a school day at 136 roots can say about the middle of it. **A school with no teacher.** ([The syllabus](../lessons/README.md) records which lesson teaches what.) *mualim* is not taught until Lesson 19, so the only people in this school are children and a friend. Lesson 16 is called *School and time* and gives the building, the book and the chair; the person who teaches is filed in Lesson 19 among doors, windows, trees and animals. Whether that word should move is a syllabus question and this page does not answer it — it records that a school text written at Lesson 16 cannot put a teacher in the room. Nothing else was wanted that the language lacks: the vocabulary of a school day at 136 roots is enough for a school day. ## Roots used 42 of 300. --- # Uan umur *One life.* The twenty-first text, and **the first long one**. Counted before it was written: twenty texts, 1875 words of running Amadunia, and the longest was thirty-four sentences. Everything the language had was a page. Length is its own test. A short text can hide what a long one cannot — who is who after forty sentences with no gender and no case, whether a tense holds across a whole life, whether six sentence shapes become a drone. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Anak lai in dom keci. Mama senang. Papa senang. Dom in tarik keci. Es bage sub. Anak kula, lala, kanta. Mama sema kalima. Anak dengar. Papa katab nama anak in libro. Din-din go. Anak anda. Anak kimbia in bage. Anak kan flor, pohon aur tair in asman. Anak tanya: "Ini es ke?" Mama respon. Anak tanya tena: "Porke?" Papa kanta. Anak go skola. Skola in madina keci. Mualim beri libro. Anak buka libro. Kalima in libro muskil. Anak no faham. Mualim sema: "Kara tena." Anak kara tena. Sasa anak faham. Anak katab nama ta. Anak katab kalima baru. Anak studi numero aur historia. Rafiki otur dekat. Rafiki genc aur hayai. Kita kula pan una. Kita kimbia una. Tahun-tahun go. Anak genc sasa. Insan genc mau kan dunia. Insan genc sema: "Mi mau go madina kabir." Mama sedih. Papa sema: "Go." Insan genc ambil tren. Tren anda cok hora. Insan genc kan pahar, sungai, hutan aur bahari. Madina kabir. Cok insan in tarik. Insan genc naiti kamra keci. Kamra kotor. Insan genc gasil kamra una sabun aur akua. Insan genc naiti rabota in market. Tacir beca fruta aur nyama. Insan genc ambil fruta. Insan genc beca fruta. Insan-insan lai. Insan-insan kupi. Aksam lai. Insan genc funga market. Pesa keci. Insan genc kula pan aur corba. Insan genc lala. Insan genc punya mimpi. Insan genc kan insan baru in market uan din. Insan baru kupi legum aur telur. Insan genc sema: "Salam." Insan sema: "Salam." Ta-ta sema. Ta-ta kanta. Ta-ta anda in tarik. Insan genc sema: "Mi ama yu." Insan sema: "Mi ama yu." Ta-ta naiti dom keci in madina. Es pencere kabir in kamra. Ta-ta pone tabla aur du kursi in kamra. Anak lai. Anak keci kula, lala, kanta. Historia go tena. Tahun-tahun go. Rambut siya sasa putih. Anak-anak kabir sasa. Anak-anak go madina baid. Es sukut in dom. Ta-ta otur in bage. Ta-ta kan sol. Ta-ta ingat din-din eski. Insan sema: "Mi ingat yu in skola." Insan genc — no genc sasa — respon: "Mi ingat yu in market." Ta-ta kanta tena. Rat lai. Stela in asman. Insan mati. Es sukut in dom tena. Anak-anak lai. Anak-anak kan bage. Anak keci tanya: "Kim es insan ini?" Mama respon: "Ta es nasab yu. Ta suda hidup sini." Anak keci kan foto. Anak keci ingat nama. Insan mati. Kalima no mati. Nama in libro. Historia in kalima. Anak keci sema nama tena. ``` --- ## Line by line | Amadunia | English | |---|---| | Anak lai in dom keci. | A child comes into a small house. | | Mama senang. Papa senang. | The mother is happy. The father is happy. | | Dom in tarik keci. | The house is on a small road. | | Es bage sub. | There is a garden below. | | Anak kula, lala, kanta. | The child eats, sleeps, sings. | | Mama sema kalima. Anak dengar. | The mother says words. The child listens. | | Papa katab nama anak in libro. | The father writes the child's name in a book. | | Din-din go. | The days go. | | Anak anda. Anak kimbia in bage. | The child walks. The child runs in the garden. | | Anak kan flor, pohon aur tair in asman. | The child sees flowers, trees and a bird in the sky. | | Anak tanya: "Ini es ke?" Mama respon. | The child asks: "What is this?" The mother answers. | | Anak tanya tena: "Porke?" Papa kanta. | The child asks again: "Why?" The father sings. | | Anak go skola. Skola in madina keci. | The child goes to school. The school is in a small town. | | Mualim beri libro. Anak buka libro. | The teacher gives a book. The child opens the book. | | Kalima in libro muskil. Anak no faham. | The words in the book are difficult. The child does not understand. | | Mualim sema: "Kara tena." | The teacher says: "Read again." | | Anak kara tena. Sasa anak faham. | The child reads again. Now the child understands. | | Anak katab nama ta. | The child writes his name. | | Anak katab kalima baru. | The child writes new words. | | Anak studi numero aur historia. | The child studies numbers and history. | | Rafiki otur dekat. | A friend sits near. | | Rafiki genc aur hayai. | The friend is young and quick. | | Kita kula pan una. Kita kimbia una. | We eat bread together. We run together. | | Tahun-tahun go. Anak genc sasa. | The years go. The child is young now. | | Insan genc mau kan dunia. | The young one wants to see the world. | | Insan genc sema: "Mi mau go madina kabir." | He says: "I want to go to a big city." | | Mama sedih. Papa sema: "Go." | The mother is sad. The father says: "Go." | | Insan genc ambil tren. Tren anda cok hora. | He takes a train. The train travels many hours. | | Insan genc kan pahar, sungai, hutan aur bahari. | He sees mountains, rivers, forest and sea. | | Madina kabir. Cok insan in tarik. | The city is big. Many people are in the road. | | Insan genc naiti kamra keci. Kamra kotor. | He finds a small room. The room is dirty. | | Insan genc gasil kamra una sabun aur akua. | He washes the room with soap and water. | | Insan genc naiti rabota in market. | He finds work at the market. | | Tacir beca fruta aur nyama. | The merchant sells fruit and meat. | | Insan genc ambil fruta. Insan genc beca fruta. | He takes fruit. He sells fruit. | | Insan-insan lai. Insan-insan kupi. | People come. People buy. | | Aksam lai. Insan genc funga market. | Evening comes. He closes the market. | | Pesa keci. Insan genc kula pan aur corba. | The money is little. He eats bread and soup. | | Insan genc lala. Insan genc punya mimpi. | He sleeps. He has a dream. | | Insan genc kan insan baru in market uan din. | One day he sees a new person at the market. | | Insan baru kupi legum aur telur. | The new person buys vegetables and eggs. | | Insan genc sema: "Salam." Insan sema: "Salam." | He says: "Peace." The person says: "Peace." | | Ta-ta sema. Ta-ta kanta. | They talk. They sing. | | Ta-ta anda in tarik. | They walk in the road. | | Insan genc sema: "Mi ama yu." | He says: "I love you." | | Insan sema: "Mi ama yu." | The person says: "I love you." | | Ta-ta naiti dom keci in madina. | They find a small house in the city. | | Es pencere kabir in kamra. | There is a big window in the room. | | Ta-ta pone tabla aur du kursi in kamra. | They put a table and two chairs in the room. | | Anak lai. | A child comes. | | Anak keci kula, lala, kanta. | The small child eats, sleeps, sings. | | Historia go tena. | The story goes again. | | Tahun-tahun go. Rambut siya sasa putih. | The years go. The black hair is white now. | | Anak-anak kabir sasa. | The children are big now. | | Anak-anak go madina baid. | The children go to a far city. | | Es sukut in dom. Ta-ta otur in bage. | There is silence in the house. They sit in the garden. | | Ta-ta kan sol. Ta-ta ingat din-din eski. | They see the sun. They remember the old days. | | Insan sema: "Mi ingat yu in skola." | The person says: "I remember that morning at school." | | Insan genc — no genc sasa — respon: "Mi ingat yu in market." | The young one — not young now — answers: "I remember you at the market." | | Ta-ta kanta tena. | They sing again. | | Rat lai. Stela in asman. | Night comes. Stars are in the sky. | | Insan mati. Es sukut in dom tena. | The person dies. There is silence in the house again. | | Anak-anak lai. Anak-anak kan bage. | The children come. The children see the garden. | | Anak keci tanya: "Kim es insan ini?" | A small child asks: "Who is this person?" | | Mama respon: "Ta es nasab yu. Ta suda hidup sini." | The mother answers: "She is your relative. She lived here." | | Anak keci kan foto. Anak keci ingat nama. | The small child sees a photograph. The small child remembers the name. | | Insan mati. Kalima no mati. | A person dies. Words do not die. | | Nama in libro. Historia in kalima. | The name is in a book. The story is in words. | | Anak keci sema nama tena. | The small child says the name again. | ## What length found **101 sentences, 417 words, 122 roots** — three times the longest text before it, and a fifth of everything the language has ever written in one page. **Who is who survived, and it should not have been obvious.** The language has no gender and no case, so *ta* is he, she and it alike, and forty sentences separate the child from the old person it becomes. What holds it together is that the subject is named again whenever it changes — *anak*, *insan genc*, *insan*, *ta-ta* — which the language forces by never dropping a subject. The one rule that looked like a cost, no pronoun dropping, is what makes a long story readable. **The middle name was wrong for a day.** The second stage was *Genc* — the young one — in seventeen sentences, and *genc* is an adjective. Nothing in the language lets an adjective stand as a noun, and letting it would cost a settled rule: an adjective predicate takes no *es* and a noun predicate must have one, so *Ta genc* would stop having one reading. It is *insan genc* now, a noun with an adjective after it, which is the shape the language actually has and says exactly what was meant. Found September 5, 2026 by counting adjacent word classes across the repository; every word in all seventeen sentences existed, which is why nothing had seen them. [Recorded in copula.md](../grammar/copula.md). **The tense held by not being marked.** The whole life is told in the unmarked present, and *suda* appears once — *Ta suda hidup sini*, she lived here — spoken by someone standing in the past's garden. Marking every sentence would have been correct and unreadable. That the unmarked present carries a whole life is [an open question answered by use](../grammar/tense.md) rather than by decision, and it is worth saying out loud. **The drone is not real, and I claimed it was.** Writing this page it felt like subject-verb-object over and over, so the first version of this section said so. Then it was measured, and the measurement disagrees twice. By sentence shape — every sentence reduced to its parts of speech — this text uses **52 distinct shapes**, more than any other text in the language; the next highest are story 2 and [text 22](text-22-agar-yamur-lai.md) with 27 each. Its three commonest shapes cover 33% of it, which makes it the eighth *least* repetitive of thirty texts. And 8% of adjacent sentence pairs share a shape, fifteenth of thirty: the recipe is 28% and the poem 25%, both for good reasons. *Four of those five figures were wrong until September 6, 2026.* The paragraph said 56 shapes, 31%, fifth of twenty-one and eleventh of twenty-one. Fifty-six is not a figure the repository's own shape scanner produces at any scope — it gives 52 over the poem's block and 53 over the whole file — and the other three went stale when [text 22](text-22-agar-yamur-lai.md) and [text 23](text-23-kim-paling-hayai.md) were written four days later. This is the section that exists because a claim about this text was measured and refuted; measuring it again is the same act, and it should not have taken four days. Every figure in it is derived now. So length did not flatten the language. It did the opposite — a long text visits more situations and reaches for more shapes. The drone was in the writing of it, not in the reading, and the honest record of that is this paragraph rather than the one it replaced. ## Gaps **No word for a song, and this page had been borrowing the verb.** The line read *Ta-ta kanta kanta eski* — glossed "they sing an old song" — using *kanta* once as the verb and once as a noun. [The dictionary glosses it *to sing*](../dictionary/dictionary.md) and nothing else, and [one root does one job](../CONTRIBUTING.md): the two roots that break that rule are named and this was quietly a third. The line says *Ta-ta kanta tena* now — they sing again — which loses the song and the age of it. Found September 6, 2026 by reducing every sentence in the material to its parts of speech and reading the shapes that occur once; the song is [on the list of words the writing has asked for](../dictionary/README.md#words-the-writing-has-asked-for). **I reached for a word that does not exist, twice, without meaning to.** Both drafts of this page wrote *lakin* — the candidate in [the briefing for "but"](../grammar/proposal-but.md) — once at *Pesa keci, lakin…* and once at *Kamra keci, lakin…*. [`check.py`](../check.py) refused both, correctly: a text invents nothing. That makes six pages that have wanted the word, and this is the first time it arrived in a draft by reflex rather than by argument. **No word for *then*, over eighty-eight sentences** ([the list](../dictionary/README.md#words-the-writing-has-asked-for))**.** A life is a sequence and the page can only set its parts side by side. Every paragraph break in the block is doing work the grammar cannot. **Nothing for growing old.** (Neither on [the list](../dictionary/README.md#words-the-writing-has-asked-for) nor in [the checklist](../dictionary/a1-checklist.md); it is recorded here as a judgement rather than a hole.) *eski* is old of things; a person is *no genc*, not young, which is what the text says. Whether that is a gap or a language choosing not to have one is a judgement, and it is recorded here rather than claimed as a hole. ## Roots used 122 of 300. --- # Agar yamur lai *If the rain comes.* The twenty-second text, and the first written **for subordination**. Of the fifteen rules a scanner can find in a sentence, it stood in the fewest texts — five, against place's nineteen — and the two rules that were thin before it, the adverb rule and commands, each got a page of their own and stopped being thin. This is the same remedy applied to the third. **The premise turned out to be an artefact, and this page is what exposed it.** Writing the four shapes out made it plain that one of them has no word in it, so the scan that called subordination thinnest had been counting only the marked half. Counting both, the rule stands in seven texts and *comparison* is the thinnest at six. The page is still worth its place — the eleven unmarked clauses in the material were being counted nowhere, and three of them are here — but the reason it was written did not survive being written. The rule has three marked shapes and one unmarked one. *porke*, *kab* and *agar* each open a clause; a clause that is simply the object of a verb opens with nothing at all. A whole story of deciding, waiting and changing one's mind is made of little else, which is why the subject is a journey that may not happen. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Pagi. Mi kara surat dari rafiki mi. Rafiki mi katab ta mau kan mi. Ta hidup in madina baid. Mi fikir safari muskil porke tarik cang. Mama tanya: "Yu saufa go kab?" Mi respon: "Mi saufa go besok agar yamur no lai." Kab yamur lai, tarik kotor aur tren stop. Kab sol upar, safari asan. Aksam. Angin kuat in bage. Mi lihat asman. Es badal cok upar. Mi fikir yamur saufa lai. Mi katab surat por rafiki mi: "Mi saufa lai agar sol upar. Mi saufa espera agar yamur kuat. Mi mau yu bil mi mau lai." Mi kirim surat kab mi finis. Rat. Mi lala porke din cang. Mi ingat rafiki mi. Pagi tena. Yamur no lai. Sol upar. Mi ambil kertas mi aur go tren. Mi senang porke safari start. ``` --- ## Line by line | Amadunia | English | |---|---| | Pagi. | Morning. | | Mi kara surat dari rafiki mi. | I read a letter from my friend. | | Rafiki mi katab ta mau kan mi. | My friend writes he wants to see me. | | Ta hidup in madina baid. | He lives in a far city. | | Mi fikir safari muskil porke tarik cang. | I think the journey is difficult because the road is long. | | Mama tanya: "Yu saufa go kab?" | Mother asks: "When will you go?" | | Mi respon: "Mi saufa go besok agar yamur no lai." | I answer: "I will go tomorrow if it does not rain." | | Kab yamur lai, tarik kotor aur tren stop. | When rain comes, the road is dirty and the trains stop. | | Kab sol upar, safari asan. | When the sun is up, the journey is easy. | | Aksam. | Evening. | | Angin kuat in bage. | The wind is strong in the garden. | | Mi lihat asman. | I look at the sky. | | Es badal cok upar. | There are many clouds above. | | Mi fikir yamur saufa lai. | I think rain will come. | | Mi katab surat por rafiki mi: | I write a letter to my friend: | | "Mi saufa lai agar sol upar. | "I will come if the sun is up. | | Mi saufa espera agar yamur kuat. | I will wait if the rain is strong. | | Mi mau yu bil mi mau lai." | I want you to know I want to come." | | Mi kirim surat kab mi finis. | I send the letter when I finish. | | Rat. | Night. | | Mi lala porke din cang. | I sleep because the day is long. | | Mi ingat rafiki mi. | I remember my friend. | | Pagi tena. | Morning again. | | Yamur no lai. Sol upar. | No rain comes. The sun is up. | | Mi ambil kertas mi aur go tren. | I take my papers and go to the train. | | Mi senang porke safari start. | I am happy because the journey begins. | ## What subordination did **Eleven of the twenty-six sentences carry a clause**, and between them they cover all four shapes: | | | |---|---| | porke tarik cang, porke din cang, porke safari start | a reason | | kab yamur lai, kab sol upar, kab mi finis | a time | | agar yamur no lai, agar sol upar, agar yamur kuat | a condition — and the first of the three says the rain does not come | | ta mau kan mi, yamur saufa lai, yu bil mi mau lai | an object, with nothing marking it | The last row is the one the scanner cannot see. A clause object opens with no word at all, so nothing in [`check.py`](../check.py) can count it — the detector looks for *porke*, *kab* and *agar*, and an unmarked clause is invisible to it by construction. This page uses three, and the table above is the only record of them. **A rule with no marker cannot be measured by a scanner that reads markers,** which is worth saying plainly on the page that was written to exercise it. *Mi mau yu bil mi mau lai* puts one clause inside another: the object of *bil* is itself a sentence, and *bil* is the object of *mau*. Nothing in the rule forbids the nesting and nothing in the corpus had tried it before. ## Gaps **"If it does not rain" needs a subject the language will not give it.** English says *it* rains; Amadunia has no dummy subject, so the sentence is *yamur no lai* — "rain does not come" — and the weather is always a thing that arrives. That works for rain and snow and wind, which are nouns here, and it has no way at all to say *it is cold today*: *barid* is an adjective with nothing to attach to. The line was written round rather than through, and the missing weather predicate is [on the list of words the writing has asked for](../dictionary/README.md#words-the-writing-has-asked-for). **No word for "then", again** ([the list](../dictionary/README.md#words-the-writing-has-asked-for))**.** The letter is written, then sent, then the night passes. Every one of those is a separate sentence in the order it happened, and the order on the page is doing the work a word would do. **"Whether" is not available and was not used.** The natural sentence at the window is *I do not know whether the rain will come*, which is an indirect question — [an open question](../grammar/questions.md), and not to be invented. The page says *Mi fikir yamur saufa lai*, "I think rain will come", which is a different sentence: a guess instead of an admission of not knowing. That is a loss the reader cannot see, which is the kind worth recording. ## Roots used 62 of 300. --- # Kim paling hayai *Who is the fastest.* The twenty-third text, and the first written **for comparison** — which became the thinnest rule the day before this page was written, when subordination turned out to have a second shape and stopped being it. Comparison stood in six texts of twenty-two, against place's twenty. It is the fourth rule to be given a page of its own, and the shortest argument for one: children rank things. *lebi*, *kurang* and *paling* say more, less and most; *dari* is "than" and *kadar* is "as ... as". A conversation about which animal is fastest uses all five without being made to. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Pagi. Anak-anak otur in bage. Ta-ta kan animal-animal. Anak kabir sema: "Farasi paling kabir." Anak keci respon: "No. Farasi kabir, aur pahar lebi kabir dari farasi." Kanis hayai. Kucing hayai kadar kanis. Farasi lebi hayai dari kanis. Tair paling hayai porke tair in udara. Samaki kurang kabir dari farasi. Samaki hidup in sungai. Sungai lebi baid dari bage. Anak keci sema: "Mi hayai kadar tair." Dugu ta sema: "Yu kurang hayai dari kucing." Anak keci kimbia. Kucing kimbia lebi hayai. Anak-anak kanta. Mualim lai aur tanya: "Kim paling hayai?" Anak-anak respon: "Tair!" Mualim sema: "Tair paling hayai in udara. Samaki paling hayai in akua." Ta sema kanis kurang hayai dari farasi. Ta sema kanis es rafiki lebi hao. Aksam. Anak-anak go dom. Anak keci fikir tair lebi hao dari kucing. Dugu ta fikir kucing lebi hao dari tair. Du anak senang. ``` --- ## Line by line | Amadunia | English | |---|---| | Pagi. | Morning. | | Anak-anak otur in bage. | The children sit in the garden. | | Ta-ta kan animal-animal. | They look at the animals. | | Anak kabir sema: "Farasi paling kabir." | The big child says: "The horse is the biggest." | | Anak keci respon: "No. | The small child answers: "No. | | Farasi kabir, aur pahar lebi kabir dari farasi." | The horse is big, and the mountain is bigger than the horse." | | Kanis hayai. | The dog is fast. | | Kucing hayai kadar kanis. | The cat is as fast as the dog. | | Farasi lebi hayai dari kanis. | The horse is faster than the dog. | | Tair paling hayai porke tair in udara. | The bird is the fastest because the bird is in the air. | | Samaki kurang kabir dari farasi. | The fish is less big than the horse. | | Samaki hidup in sungai. | The fish lives in the river. | | Sungai lebi baid dari bage. | The river is further than the garden. | | Anak keci sema: "Mi hayai kadar tair." | The small child says: "I am as fast as a bird." | | Dugu ta sema: "Yu kurang hayai dari kucing." | Her sibling says: "You are less fast than the cat." | | Anak keci kimbia. | The small child runs. | | Kucing kimbia lebi hayai. | The cat runs faster. | | Anak-anak kanta. | The children sing. | | Mualim lai aur tanya: "Kim paling hayai?" | The teacher comes and asks: "Who is the fastest?" | | Anak-anak respon: "Tair!" | The children answer: "The bird!" | | Mualim sema: "Tair paling hayai in udara. | The teacher says: "The bird is the fastest in the air. | | Samaki paling hayai in akua." | The fish is the fastest in the water." | | Ta sema kanis kurang hayai dari farasi. | She says the dog is less fast than the horse. | | Ta sema kanis es rafiki lebi hao. | She says the dog is a better friend. | | Aksam. | Evening. | | Anak-anak go dom. | The children go home. | | Anak keci fikir tair lebi hao dari kucing. | The small child thinks the bird is better than the cat. | | Dugu ta fikir kucing lebi hao dari tair. | Her sibling thinks the cat is better than the bird. | | Du anak senang. | Two children are happy. | ## What comparison did **Fourteen of the twenty-nine sentences compare something**, and between them they use every part of the rule: | | | |---|---| | lebi kabir, lebi hayai, lebi baid, lebi hao | more | | kurang kabir, kurang hayai | less | | paling kabir, paling hayai | most | | dari farasi, dari kanis, dari bage, dari kucing, dari tair | than | | kadar kanis, kadar tair | as ... as | *Kucing kimbia lebi hayai* is the one that puts a degree word on a verb rather than a noun — *lebi* in the adverb slot, scaling *hayai* where the adverb rule puts it. The rule page states that shape in one line and the corpus had barely used it. The last three sentences are deliberate. Two children compare the same two animals and reach opposite answers, and the language says both with the same four words in the same order. **Nothing in a comparison marks whose it is** — *lebi hao* is "better" and never "better to me" — so the two lines are not contradictory, they are two people each stating a fact the language has no way to qualify. ## Gaps **"The biggest of the three" cannot be said.** *paling* names a superlative without naming what it is superlative among, and [the rule page records the gap](../grammar/comparison.md): *paling kabir dari tri* is untested and was not tested here. The teacher's answer works around it by naming the place instead — *paling hayai in udara* — which is a different sentence and an honest one. **No word for "than" that is not also "from".** *dari* does both jobs, which is settled and deliberate — [the rule page argues it](../grammar/comparison.md) — but this page puts the two jobs one line apart: *Samaki hidup in sungai* and *Sungai lebi baid dari bage*. A reader meeting *dari* in the second sentence has to know the first word after it is a standard and not a source, and nothing but the *lebi* before it says so. **No comparative for a noun against a noun without an adjective.** *A dog is more of a friend than a cat* has no form: the page says *kanis es rafiki lebi hao*, "the dog is a better friend", which needs an adjective to lean on. [Recorded with the other gaps in comparison.md](../grammar/comparison.md). ## Roots used 50 of 300. --- # Mama mama mi *My mother's mother.* The twenty-fourth text, and the first **portrait of one person**. The genres so far are stories, arguments, dialogues, a poem, a letter, a recipe, a meal, a room, a life and a journey; nobody had described a face. It is also a test of something the gap list raises and does not settle. The dictionary has no word for a grandmother, and [the kinship set is fourteen of twenty-seven](../dictionary/a1-checklist.md) — but possession puts the owner straight after the thing owned, so *mama mama mi* is the mother of my mother and needs no root at all. This page is written to find out whether that holds for a whole text or only for a heading. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Mama mama mi hidup in dom keci. Dom ta in tarik. Sungai dekat. Rambut ta putih. Ain ta keci aur hao. Mano ta kuat porke ta suda hidup cok tahun. Cehra ta punya cok historia. Ta otur in kursi kab aksam lai. Ta kara libro eski. Ta no mau telefon. Mi lai. Ta sema: "Otur sini, anak." Ta beri ca garam aur pan dulce. Kita sema. Kita kanta. Ta ingat tempo baid. Ta sema: "Mi suda kan dunia." Mi dengar. Mi no sema. Ta suda hidup umur cang. Ta punya cok sabar aur cok kalima. Rat lai. Ta lala. Mi kan cehra ta. Mi senang aur mi sedih. ``` --- ## Line by line | Amadunia | English | |---|---| | Mama mama mi hidup in dom keci. | My mother's mother lives in a small house. | | Dom ta in tarik. Sungai dekat. | Her house is on a road. The river is near. | | Rambut ta putih. | Her hair is white. | | Ain ta keci aur hao. | Her eyes are small and kind. | | Mano ta kuat porke ta suda hidup cok tahun. | Her hands are strong because she has lived many years. | | Cehra ta punya cok historia. | Her face has much history. | | Ta otur in kursi kab aksam lai. | She sits in a chair when evening comes. | | Ta kara libro eski. | She reads an old book. | | Ta no mau telefon. | She does not want a telephone. | | Mi lai. | I come. | | Ta sema: "Otur sini, anak." | She says: "Sit here, child." | | Ta beri ca garam aur pan dulce. | She gives hot tea and sweet bread. | | Kita sema. Kita kanta. | We talk. We sing. | | Ta ingat tempo baid. | She remembers a far-off time. | | Ta sema: "Mi suda kan dunia." | She says: "I have seen the world." | | Mi dengar. Mi no sema. | I listen. I do not speak. | | Ta suda hidup umur cang. | She has lived a long life. | | Ta punya cok sabar aur cok kalima. | She has much patience and many words. | | Rat lai. Ta lala. | Night comes. She sleeps. | | Mi kan cehra ta. | I look at her face. | | Mi senang aur mi sedih. | I am happy and I am sad. | ## What the possession rule did **It gave the language a grandmother without a root.** *mama mama mi* is two applications of one rule: *mama mi* is my mother, and *mama* in front of that phrase owns it. Nothing marks either step. The same shape gives a grandfather, and *dugu mama mi* an uncle or an aunt — which is [what the gap note already suspected](../dictionary/README.md#words-the-writing-has-asked-for) and what a whole page now shows. **And it holds for exactly one step of remove.** *mama mama mama mi* would be a great-grandmother, and it is unreadable rather than ungrammatical: three identical words in a row with nothing to group them. The rule does not forbid it and no reader could parse it, which is a different kind of limit from the ones this repository usually records — not a missing word but a shape that runs out. ## Gaps **Nothing can be near anything.** The first draft of the second line read *Dom ta in tarik dekat sungai* — her house is on a road near the river — and it is not a sentence: *dekat* is an adjective, adjectives follow their noun, and so *tarik dekat* is "a near road" with the river left dangling. The language has three prepositions, *in*, *dari* and *por*, and none of them is *beside*. The page says *Dom ta in tarik. Sungai dekat.* — two sentences, with the relation between them left to the reader. [Recorded with the other things the writing has asked for](../dictionary/README.md#words-the-writing-has-asked-for), and it is the first gap this repository has found in the [preposition set](../grammar/place.md) rather than in the vocabulary. **No word for an old person.** *eski* is old and is used here of a book. Using it of a person is grammatical and cold; the text says *ta suda hidup umur cang*, she has lived a long life, which is longer and kinder and not the same thing. *genc* has an opposite in [the dictionary](../dictionary/dictionary.md) only for objects. **No word for *kind*.** *Ain ta keci aur hao* is glossed "small and kind" because that is what the sentence means here, and *hao* is simply "good". The language cannot separate a good person from a kind one, and this page leans on context to do it. [Recorded with the other gaps](../dictionary/README.md#words-the-writing-has-asked-for). **A face cannot hold anything.** *Cehra ta punya cok historia* — her face has much history — is the nearest the language gets to a face that shows a life. *punya* is possession and nothing else; there is no *show*, no *carry*, no *tell* that takes a thing rather than a person, and [the verb list](../dictionary/dictionary.md) has no candidate to press into the job. **And the portrait has no word for what a portrait is about.** There is no *look* as a noun, no *age* apart from *umur* which is also *life*, and no way to say someone is beautiful in the sense a grandchild means. *sundar* is in [the dictionary](../dictionary/dictionary.md) and is not used here, because it would have been the wrong word. ## Roots used 59 of 300. --- # Tarik por hospital *The way to the hospital.* The twenty-fifth text, and the first written as **directions to a stranger**. The phrasebook has the fragments — *Hotel nali?*, *Go yamin* — and no page has ever put them together into somebody actually finding a place. It was written for two reasons. Commands stand in seven texts, level with comparison at the floor of the exercise table, and directions are made of them. And [text 24 found that nothing can be near anything](text-24-mama-mama-mi.md): the language has three prepositions and none of them is *beside*. A page about where things are is the hardest test that gap can be given, and a second witness is worth more than a second argument. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` — Pardon. Hospital nali? — Hospital baid. Yu anda o yu go in auto? — Mi anda. — Ok. Anda tarik ini. Go yamin in ponte. — Ponte cang? — Ya. Sungai sub. — Mi kan market kabir sasa. — Hao. Hospital baid dari market. — Berapa tempo? — Des momen agar yu anda hayai. — Posta nali? — No go por posta. Posta kiri, hospital yamin. — Mi no faham. Sema tena, plis. — Anda tarik cang. No go kiri. Go yamin. — Ok. Mersi! — Hospital kabir aur putih. Yu saufa kan ta. ``` --- ## Line by line | Amadunia | English | |---|---| | — Pardon. Hospital nali? | — Excuse me. Where is the hospital? | | — Hospital baid. | — The hospital is far. | | Yu anda o yu go in auto? | Are you walking or going by car? | | — Mi anda. | — I'm walking. | | — Ok. Anda tarik ini. | — Ok. Walk this road. | | Go yamin in ponte. | Turn right at the bridge. | | — Ponte cang? | — Is the bridge long? | | — Ya. Sungai sub. | — Yes. The river is below. | | — Mi kan market kabir sasa. | — I see a big market now. | | — Hao. Hospital baid dari market. | — Good. The hospital is far from the market. | | — Berapa tempo? | — How much time? | | — Des momen agar yu anda hayai. | — Ten moments if you walk fast. | | — Posta nali? | — Where is the post? | | — No go por posta. | — Do not go to the post. | | Posta kiri, hospital yamin. | The post is left, the hospital right. | | — Mi no faham. Sema tena, plis. | — I don't understand. Say it again, please. | | — Anda tarik cang. No go kiri. Go yamin. | — Walk the long road. Do not go left. Go right. | | — Ok. Mersi! | — Ok. Thank you! | | — Hospital kabir aur putih. | — The hospital is big and white. | | Yu saufa kan ta. | You will see it. | ## What directions are made of **Seven of the twenty-eight sentences are commands**, and three of the seven are negative. *No go kiri* and *No go por posta* carry as much of this page as *Go yamin* does — a direction is as often about where not to turn — and no other text has needed the negative command more than once. The rest is the question rule doing the other half. *Hospital nali?*, *Posta nali?*, *Berapa tempo?* — the question word stands where the answer will stand, so the answer is the question with one word swapped, and a stranger who learns the question has learned the answer. ## Gaps **Nothing is beside anything, for the second time.** *Sungai sub* says the river is below and stops there; the sentence a speaker wants is *the river is below the bridge*, and *sub* cannot take a noun any more than *dekat* could in [text 24](text-24-mama-mama-mi.md). This page had to put the bridge in one sentence and the river in the next and let the reader join them. [The rule page records it](../grammar/place.md) as a missing slot rather than a missing word, and this is the second text to run into it in two days — the first by accident, this one on purpose. **There is no word for a corner, a turn or a crossing.** *Go yamin in ponte* is "go right at the bridge", and a bridge is the only landmark this page can use because the language has one — [the house and the city are the two thinnest domains on the checklist](../dictionary/a1-checklist.md). Directions in any city are given by corners. **No word for *far* that is not also a judgement.** *baid* is far, and *Hospital baid* is the first thing the stranger is told; *ten moments* has to do the work of saying how far, because a distance the language can name does not exist. [Recorded with the other things the writing has asked for](../dictionary/README.md#words-the-writing-has-asked-for). **And *ten moments* is not a time.** *momen* is a moment and *des momen* is ten of them, which is a number of an unmeasured thing. [The clock has no word](../dictionary/README.md#words-the-writing-has-asked-for) and neither has the minute, so the only honest answer to *how much time* is a guess dressed as a count. ## Roots used 45 of 300. --- # Dom eski *The old house.* The twenty-sixth text, and the first **memory**. Everything in it happened, and the language has one way to say so: *suda* in front of the verb, the same mark for a thing that happened once and a thing that went on for years. It was written to find out what that costs. A memory is made of the difference between what was and what is, and Amadunia can say both — but it cannot say that one thing had already finished when another began, which is the shape most of a memory takes. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Mi ingat dom eski. Dom itu suda es dom familia mi. Dom itu lebi keci dari dom mi sasa. Bage itu lebi kabir. Pohon cang suda hidup in bage. Anak-anak suda kimbia in bage. Mama suda tabak corba. Papa suda kanta. Kamra mi keci aur barid. Uan pencere. Uan lampu. Mi suda lala kab luma suda mati. Sasa dom itu no es dom mi. Insan-insan baru hidup situ. Pohon no situ. Mi suda go tarik itu. Mi suda kan pencere. Mi no suda kan anak-anak. Dom mi baru lebi kabir aur lebi safi. Mi senang in dom baru. Aur mi sedih kab mi ingat dom eski. ``` --- ## Line by line | Amadunia | English | |---|---| | Mi ingat dom eski. | I remember the old house. | | Dom itu suda es dom familia mi. | That house was my family's house. | | Dom itu lebi keci dari dom mi sasa. | That house is smaller than my house now. | | Bage itu lebi kabir. | That garden is bigger. | | Pohon cang suda hidup in bage. | A tall tree lived in the garden. | | Anak-anak suda kimbia in bage. | The children ran in the garden. | | Mama suda tabak corba. | Mother cooked soup. | | Papa suda kanta. | Father sang. | | Kamra mi keci aur barid. | My room is small and cold. | | Uan pencere. Uan lampu. | One window. One lamp. | | Mi suda lala kab luma suda mati. | I slept when the light had died. | | Sasa dom itu no es dom mi. | Now that house is not my house. | | Insan-insan baru hidup situ. | New people live there. | | Pohon no situ. | The tree is not there. | | Mi suda go tarik itu. | I went to that road. | | Mi suda kan pencere. | I saw the window. | | Mi no suda kan anak-anak. | I did not see the children. | | Dom mi baru lebi kabir aur lebi safi. | My new house is bigger and cleaner. | | Mi senang in dom baru. | I am happy in the new house. | | Aur mi sedih kab mi ingat dom eski. | And I am sad when I remember the old house. | ## What one past marker does *suda* is in nine of these twenty sentences and it is the same word every time. *Papa suda kanta* is one evening; *Pohon cang suda hidup in bage* is thirty years. The language does not distinguish them and this page does not try: the difference is carried by what is being talked about, which is how Chinese, Indonesian and Malay carry it too. **The comparisons are what make it a memory rather than a list.** *Dom itu lebi keci dari dom mi sasa* puts then beside now in one sentence — and the tense marker is on neither half, because a comparison of an adjective has no verb to mark. The old house is smaller *now*, in the remembering, which is not quite what the sentence means and is exactly what it says. ## Gaps **Nothing can have already happened.** *Mi suda lala kab luma suda mati* is "I slept when the light died", and the sentence wants *had died* — one past inside another. Both verbs take the same *suda* and nothing orders them. [The tense page](../grammar/tense.md) settles that the verb never changes, so this is not a missing word but the price of that decision, and this is the first page to pay it in full. **No *before*, no *after*, no *until*.** Each is [open](../grammar/README.md) and none was invented here. A memory is mostly made of those words, and the page is written round them: every sentence stands on its own and the order they are printed in does the work. **No *ago*, and no way to say how long.** *tahun ini* is this year and *kemarin* is yesterday, and between them there is nothing — no *ten years ago*, no *for a long time*. The tree lived in the garden and the page cannot say for how long, which is the sentence a memory most wants. [Recorded with the other things the writing has asked for](../dictionary/README.md#words-the-writing-has-asked-for). **An adjective predicate has no granted way into the past.** *Kamra mi keci* is "my room is small", and the sentence this page wanted is *my room was small*. *suda* goes in front of a verb and an adjective predicate has none; *es* before the adjective is [refused by the copula rule](../grammar/copula.md) and by the checker. So the room is described in the present inside a page that is entirely past, and the reader carries the tense across from the sentences around it. **The course has been doing it since Lesson 11 anyway.** Written the day after this page, and it corrects what this page first said: *Dom suda kabir* — the house was big — puts the marker straight in front of the adjective, where the verb would be, and [Lesson 11](../lessons/lesson-11-being.md) teaches it, twice more in the future. Five sentences in the material use the shape and no rule grants it. So the language does have a way and this page did not use it, which is the right way round: [an open shape is not for new writing until it is settled](../grammar/tense.md), and finding it in the lessons is worth more than inventing it here would have been. ## Roots used 45 of 300. --- # Kalima por anak-anak *Words for the children.* The twenty-seventh text, and the first **notice** — a page written to be pinned on a wall and read by everybody, rather than said by somebody to somebody. It was drafted claiming to be the only page in the corpus with nobody speaking and nobody addressed, and that was not measured and is not true. **Four texts use no *mi* at all**: this one, [the recipe](text-8-kaifa-suru-ca.md), and the portraits in [text 6](text-6-seti-din.md) and [text 11](text-11-anak-katab.md). The recipe is the nearest thing to a notice already here, and what separates them is who *yu* is — the recipe's *yu* is one cook following one method, and a notice's *yu* is whoever happens to be standing in front of the wall. The language spells those the same way, which is the finding this page was written to reach and not the one it set out for. A notice is made of commands and of the word *everyone*, and the language has the first and not the second. This page was written to find out how far the first alone gets. Written at 300 roots, with no grammar beyond Lesson 26. --- ``` Kalima por anak-anak in skola. Lai in pagi. Buka pintu. Buka libro yu. Kara. Dengar mualim. No kimbia in kamra. Kimbia in bage. No kula in kamra. Kula in bage. Leta libro aur kertas. Leta akua. No leta kanis. Agar yu no faham, tanya. Agar yu bimar, sema mualim. Gasil mano yu. Gasil piring yu. Funga pintu kab yu go. Skola buka in pagi. Skola funga in aksam. Kita studi una. ``` --- ## Line by line | Amadunia | English | |---|---| | Kalima por anak-anak in skola. | Words for the children at school. | | Lai in pagi. Buka pintu. | Come in the morning. Open the door. | | Buka libro yu. Kara. | Open your book. Read. | | Dengar mualim. | Listen to the teacher. | | No kimbia in kamra. | Do not run in the room. | | Kimbia in bage. | Run in the garden. | | No kula in kamra. | Do not eat in the room. | | Kula in bage. | Eat in the garden. | | Leta libro aur kertas. | Bring a book and paper. | | Leta akua. | Bring water. | | No leta kanis. | Do not bring a dog. | | Agar yu no faham, tanya. | If you do not understand, ask. | | Agar yu bimar, sema mualim. | If you are sick, tell the teacher. | | Gasil mano yu. Gasil piring yu. | Wash your hands. Wash your plate. | | Funga pintu kab yu go. | Close the door when you go. | | Skola buka in pagi. | The school opens in the morning. | | Skola funga in aksam. | The school closes in the evening. | | Kita studi una. | We learn together. | ## What a notice is made of **Fifteen of the twenty-one sentences are commands**, and three of those are negative. A notice is the genre the imperative was granted for, and it did not exist when the imperative was granted: [the briefing](../grammar/proposal-sentence-types.md) counted seven commands in the whole material and one text written deliberately without any. The other six are the title, three statements about the school itself, and the two conditionals. **Both conditionals speak to one reader**, which no command here does: *Agar yu no faham, tanya* and *Agar yu bimar, sema mualim* single out the child who does not understand and the child who is ill, and each is fronted with a comma, which is the one fronting the rules grant. ## Gaps **There is no word for everyone, and a notice is addressed to everyone.** The page says *yu* throughout — the same *yu* that means one person — and only the title, *kalima por anak-anak*, says who is meant. *all*, *some* and *none* are [an open question](../grammar/plural.md) and none was invented here; without them a public notice cannot say it is public. **No word for late or early.** The first line wants *do not come late* and says *come in the morning*, which is a different instruction: it tells a punctual child something they already know instead of telling a late one anything. *late* and *early* are both [missing from the checklist's time domain](../dictionary/a1-checklist.md). **A notice wants a numbered list and cannot have one.** *First, second, third* have no form — [ordinals are open](../grammar/numbers.md) — so the instructions stand in the order they are printed and nothing says that the order matters. For *come in the morning* before *close the door when you go* it does. **And there is no word for a rule.** *legis* is law, which is a courtroom, and *kalima* is a word; neither is what a notice is made of, and *a rule* now stands on [the gap list](../dictionary/README.md) with the rest. The title says *words for the children* because *rules for the children* cannot be said, and a notice that cannot name what it is is relying on the wall it is pinned to. ## Roots used 39 of 300. --- # Din mi *My day.* The twenty-eighth text, and the new lowest rung on [the reading ladder](../lessons/reading-ladder.md): it uses **only the roots the lessons have taught by Lesson 06**, and no rule beyond that lesson. Until it was written the ladder began at Lesson 10, which meant nine lessons of teaching before there was anything to read. This page was written for those nine. Six lessons in, there is no *aur*, no *no*, no question word, no *es*, no preposition, no *ini* or *itu*, no adverb, no comparison and no clause inside a clause. What is there is a subject, a verb, an object, an adjective after its noun, an adjective standing as the whole predicate, the plural, possession, the two tense markers and the numbers. Everything below is made of that. --- ``` Sol lai. Pagi hao. Mama mi lai. Mama mi beri ca. Mi nomu ca. Ca hao. Papa mi kula pan. Anak keci nomu akua. Sol kabir. Din kabir. Mi go. Rafiki mi go. Kita kan dunia. Dunia kabir. Rafiki mi sema. Mi sema. Aksam lai. Sol keci. Papa mi lai dom. Kita kula pan. Kita nomu ca. Rat lai. Dom mi keci. Anak-anak keci. Mi ama anak-anak. Mi ama mama mi. Mi ama papa mi. Mi ama rafiki-rafiki mi. Sol suda go. Sol saufa lai besok. Din hao. ``` --- ## Line by line | Amadunia | English | |---|---| | Sol lai. Pagi hao. | The sun comes. The morning is good. | | Mama mi lai. Mama mi beri ca. | My mother comes. My mother gives tea. | | Mi nomu ca. Ca hao. | I drink tea. The tea is good. | | Papa mi kula pan. | My father eats bread. | | Anak keci nomu akua. | The small child drinks water. | | Sol kabir. Din kabir. | The sun is big. The day is big. | | Mi go. Rafiki mi go. | I go. My friend goes. | | Kita kan dunia. Dunia kabir. | We see the world. The world is big. | | Rafiki mi sema. Mi sema. | My friend speaks. I speak. | | Aksam lai. Sol keci. | Evening comes. The sun is small. | | Papa mi lai dom. | My father comes home. | | Kita kula pan. Kita nomu ca. | We eat bread. We drink tea. | | Rat lai. Dom mi keci. | Night comes. My house is small. | | Anak-anak keci. Mi ama anak-anak. | The children are small. I love the children. | | Mi ama mama mi. Mi ama papa mi. | I love my mother. I love my father. | | Mi ama rafiki-rafiki mi. | I love my friends. | | Sol suda go. Sol saufa lai besok. | The sun has gone. The sun will come tomorrow. | | Din hao. | The day is good. | ## What six lessons buy Every sentence here is two or three words, and that is not a style. **No sentence can be joined to another**, because *aur* arrives in Lesson 14, so *Mi nomu ca. Ca hao.* is two sentences where one was meant. **Nothing can be denied**, because *no* before a predicate is Lesson 14 as well: the day in this text is good throughout, and it could not have been otherwise. **Nothing can be asked**, because the question words arrive in Lesson 07. And **nothing is anywhere**: *in*, *dari* and *por* are Lesson 15, so the father comes home and the family is never *in* the house. What is left turns out to be a day. The three time words — *pagi*, *aksam*, *rat* — carry the whole shape of it without a preposition or a marker between them, standing as the subject of a sentence of their own: *Aksam lai*, evening comes. **The sun does the work a conjunction would.** *Sol lai* opens it, *Sol kabir* is the middle of the day, *Sol keci* is the evening and *Sol suda go* is the end. A reader six lessons in has no word that means *then*, so the sun is the only thing that can say it. ## Gaps **Five adjectives, and the tea cannot be hot.** Everything describable by Lesson 06 is describable as *hao*, *mal*, *kabir*, *keci* or *cok* — good, bad, big, small, many. *garam*, hot, is Lesson 08. The first draft of this page wrote *Pagi luma* for *the morning is bright* three times, and [`check.py`](../check.py) rejected all three: *luma* is the noun **light**, so the line said *morning light* and had no predicate in it at all. There is no adjective for bright at any lesson, so the line is *Pagi hao* now — which is [the greeting](../phrasebook.md), and the next gap is about that. **A noun cannot be a predicate here.** *My mother is a doctor* needs *es*, and [*es* arrives in Lesson 11](../lessons/lesson-11-being.md) — the word is met earlier and the sentence pattern is not. The one thing a person most wants to say when introducing a family is the one thing this rung cannot say. **And *Pagi hao* is both a sentence and a greeting.** *The morning is good* and *Good morning!* are the same three syllables with nothing to tell them apart; [copula.md records it](../grammar/copula.md) as a decision the language has not made. A page at this rung cannot write around it, because the only other bright-morning adjective is *kabir*. ## Roots used 30 of 300. --- # Yu mau ke? *What do you want?* The twenty-ninth text, and the second rung on [the reading ladder](../lessons/reading-ladder.md): **only the roots the lessons have taught by Lesson 08**, and no rule beyond Lesson 07. [Text 28](text-28-din-mi.md) opened the ladder at Lesson 06 with a day that nobody speaks in. This is the other thing eight lessons buy: a conversation. Lesson 07 gives the question words, and almost every line here is a question or an answer to one. Eight lessons in there is still no *aur*, no *no* before a predicate, no *es*, no *ini* or *itu*, no preposition and no verb chain. A question, a possessor and the two tense markers are the whole apparatus. --- ``` Mama mi go market. Mi go market. — Hi! Yu mau ke? — Mi mau pan. — Berapa pan? — Du pan. — Des pesa. — Mi beri des pesa. Mersi! — Yu mau ca? — Ya. Ca garam? — Ya, ca garam. — Mi mau ca. Mi mau akua barid. Insan lai. Insan kan foto mi. — Kim? — Papa mi. — Ta lai kab? — Ta saufa lai aksam. — Yu go nali? — Mi go dom. Mama mi go dom. — Bai! — Bai! Mi kula pan. Mi nomu ca. Mama mi nomu akua. Din hao. ``` --- ## Line by line | Amadunia | English | |---|---| | Mama mi go market. Mi go market. | My mother goes to the market. I go to the market. | | — Hi! Yu mau ke? | — Hello! What do you want? | | — Mi mau pan. | — I want bread. | | — Berapa pan? | — How many breads? | | — Du pan. | — Two breads. | | — Des pesa. | — Ten money. | | — Mi beri des pesa. Mersi! | — I give ten money. Thank you! | | — Yu mau ca? | — Do you want tea? | | — Ya. Ca garam? | — Yes. Is the tea hot? | | — Ya, ca garam. | — Yes, the tea is hot. | | — Mi mau ca. Mi mau akua barid. | — I want tea. I want cold water. | | Insan lai. Insan kan foto mi. | A person comes. The person looks at my photo. | | — Kim? | — Who? | | — Papa mi. | — My father. | | — Ta lai kab? | — When does he come? | | — Ta saufa lai aksam. | — He will come in the evening. | | — Yu go nali? | — Where are you going? | | — Mi go dom. Mama mi go dom. | — I am going home. My mother is going home. | | — Bai! — Bai! | — Goodbye! — Goodbye! | | Mi kula pan. Mi nomu ca. | I eat bread. I drink tea. | | Mama mi nomu akua. | My mother drinks water. | | Din hao. | The day is good. | ## Two kinds of question, and both are here *Ke*, *berapa*, *kim*, *kab* and *nali* stand exactly where the answer will stand — **Yu mau ke?** is answered **Mi mau pan**, with the word swapped and nothing else moved. That is the Lesson 07 rule and it is the reason this page can be a conversation at all. *Ca garam?* has no question word in it. It is a statement asked on a rising tone, and the lessons have been doing that since Lesson 01 — which is why the reading ladder counts the two separately, and why this text is dated by its words rather than by its questions. **A fragment answers most of them.** *Du pan*, *Des pesa*, *Papa mi*, *Kim* — four turns with no verb in them. Nothing was added to allow that; a phrase may stand alone [when the question makes it clear](../grammar/sentence-types.md). ## Gaps **A price cannot be named, only counted.** *Des pesa* is "ten money", and the shopkeeper who says it is doing arithmetic out loud. *harga*, price, arrives in Lesson 21 — thirteen lessons after a learner can walk into a market and ask for bread. [Text 16](text-16-berapa-harga.md) records the same gap at 300 roots for a different reason: a price cannot be named *cheap* or *dear* either. **Nothing can be asked with *kaifa*.** *How* is in [Lesson 07](../lessons/lesson-07-questions.md) with the other five question words and there is nothing here to ask it about: it wants a verb of making or doing, and *suru*, to make, arrives in Lesson 22. The word is taught nine lessons before the first sentence that can use it. **Two turns cannot be joined and none can be denied.** *aur* and predicate *no* are both [Lesson 14](../lessons/lesson-14-not-and.md), so the shopkeeper cannot say *I have bread and tea* and cannot say *there is no tea*. Every answer on this page is yes. **And nobody can be asked who they are.** *Kim yu?* would need *es* to be answered — *Mi es doktor* — and [the copula is Lesson 11](../grammar/copula.md). *Kim* here can only ask about a person already in the room, which is why it asks about a photograph. ## Roots used 37 of 300. --- # Katab hao *Write well.* The thirtieth text, and the fourth rung on [the reading ladder](../lessons/reading-ladder.md): **only the roots the lessons have taught by Lesson 12**, and no rule beyond that lesson. The ladder ran 06, 08, 10 and then nothing until 13. This page was written for the two lessons in that gap, and for what Lesson 12 is actually about: an adjective straight after the verb describes the verb. *Katab hao* is **write well**, and *katab hao* is the whole point of the page. Twelve lessons in there is no *aur*, no *no* before a predicate, no *ini* or *itu*, no preposition, no verb chain and no comparison. What there is that the earlier rungs did not have: the command, the copula, and the adverb. --- ``` Anak mi keci. Mano ta keci. Otur sini. Ambil mano mi. Kan mano mi. Mi katab. Katab keci. Katab hao. Anak mi katab hayai. Mi sema: "Katab hao." Ta katab hao. Ta katab nama mama. Ta katab nama papa. Ta katab nama ta. Ta kara sasa. Ta kara hayai. Ta kara hao. Anak mi es rafiki mi. Ta saufa katab hao. ``` --- ## Line by line | Amadunia | English | |---|---| | Anak mi keci. Mano ta keci. | My child is small. Her hand is small. | | Otur sini. Ambil mano mi. | Sit here. Take my hand. | | Kan mano mi. Mi katab. | Look at my hand. I write. | | Katab keci. Katab hao. | Write small. Write well. | | Anak mi katab hayai. | My child writes fast. | | Mi sema: "Katab hao." | I say: "Write well." | | Ta katab hao. | She writes well. | | Ta katab nama mama. | She writes mother's name. | | Ta katab nama papa. | She writes father's name. | | Ta katab nama ta. | She writes her name. | | Ta kara sasa. Ta kara hayai. | She reads now. She reads fast. | | Ta kara hao. | She reads well. | | Anak mi es rafiki mi. | My child is my friend. | | Ta saufa katab hao. | She will write well. | ## The one rule this page is for **An adjective after the verb describes the verb; an adjective after a noun describes the noun.** Nothing is added and nothing changes shape, so the same two words mean two things depending on which side of the verb they fall: | | | |---|---| | Anak mi katab hayai. | My child writes fast. | | Anak hayai katab. | The fast child writes. | Both are on this page's own terms — the second is not written above, because the child in it is not a fast child. That is the whole of Lesson 12, and a text is where a rule stops being a table. **The command is what makes it a lesson rather than a description.** **Five of the nineteen sentences are commands**, and zero of those are negative — *Otur sini*, *Ambil mano mi*, *Kan mano mi*, *Katab keci*, *Katab hao*. That is what teaching a child to hold a pen sounds like in any language, minus the half of it that says *don't*. ## Gaps **There is nothing to write on and nothing to read.** *kertas*, paper, arrives in Lesson 24 and *libro*, book, in Lesson 21, so a page about learning to write has no pen, no paper and no book in it. The child writes and reads, and the objects of both verbs are missing from the language a learner has at this point. [The reading ladder](../lessons/reading-ladder.md) says what is readable; it does not say what is writable, and this is the difference. **And *Ta katab nama ta* cannot say whose name it is.** *nama ta* is "her name", and nothing marks whether that is her own name or somebody else's — a reflexive is [an open question](../grammar/pronouns.md) and none was invented here. The line is meant as her own, and the page cannot say so. **Nothing can be joined and nothing denied.** *aur* and predicate *no* are both [Lesson 14](../lessons/lesson-14-not-and.md), which is why *Katab keci* and *Katab hao* stand as two orders where a parent would say one. The same gap [text 29](text-29-yu-mau-ke.md) recorded at Lesson 08 is still open four lessons later, and it is the longest-running gap on the low rungs. **And a child cannot want to do anything.** The first line of this page read *Ta mau katab*, she wants to write, until [`check.py`](../check.py) refused it: two verbs in a row is a verb chain and [the chain is Lesson 17](../grammar/verb-chains.md). *mau* takes a noun at this rung — [text 29](text-29-yu-mau-ke.md) says *Mi mau ca* five lessons earlier — so wanting a thing is sayable and wanting to do a thing is not. The line is *Mano ta keci* now, her hand is small, which is a fact about the child rather than a wish of hers. ## Roots used 21 of 300. --- # Amadunia phrasebook Everything here obeys the rules in [grammar/](grammar/) and uses only words in [the dictionary](dictionary/dictionary.md). [`check.py`](check.py) tests every line on this page along with the lessons. [Lesson 1](lessons/lesson-01-greetings.md) gets you five of the eight lines in the first section — *Hi!*, *Salam!*, *Ok?*, *Ya, ok.* and *Bai!*. The two greetings for a time of day wait for Lesson 03, and the last line, "I'll come again", waits for **Lesson 16**, which is where *tena* is taught. The rest of the page draws on the whole course; its last word arrives in [Lesson 25](lessons/lesson-25-doing-and-being.md). Read it for what you can use now rather than as a test of what you have finished. That paragraph said "if you have read Lesson 1 you can already say the first section" until September 6, 2026. Six of the seven roots in that section are taught after Lesson 01 and one of them fifteen lessons after it, so the claim was out by more than half the course. Both figures are computed from the lessons now, the way the [reading ladder](lessons/reading-ladder.md) is. ## Saying it A phrasebook is for saying things to somebody, so three facts before the phrases. All three were settled rather than guessed, and none has an exception. **Every letter has one sound and keeps it.** *c* is the *ch* of *chai*; *g* is always hard; *s* is always the *s* of *sun*, never a *z*. Nothing is silent. [The full table](grammar/pronunciation.md). **The beat is on the second-to-last syllable, always.** *SA-lam*, *PA-gi*, *AK-sam*, *a-ma-DU-nia*. A one-syllable word — *hi*, *ok*, *ya*, *bai* — has nowhere else to put it. A vowel pair counts as one syllable, so *bai* is one beat and *dunia* is two. [The rule](grammar/stress.md). **Unstressed vowels do not change.** The *a* of *salam* is the same *a* in both syllables. This is how Indonesian, Swahili and Japanese work, and not how English does — do not let an unstressed vowel slide toward *uh*. --- ## Hello and goodbye | | | |---|---| | Hi! | Hello! | | Salam! | Hello — the peaceful one. | | Ok? | Are you all right? | | Ya, ok. | Yes, fine. | | Pagi hao! | Good morning! | | Aksam hao! | Good evening! | | Bai! | Goodbye! | | Mi saufa lai tena. | I'll come again. | ## Meeting someone | | | |---|---| | Nama yu es ke? | What is your name? | | Nama mi es Sol. | My name is Sol. | | Yu es kim? | Who are you? | | Mi es rafiki Luma. | I am Luma's friend. | | Yu dari nali? | Where are you from? | | Mi dari negara baid. | I am from a far country. | | Mi baru in madina ini. | I am new in this city. | | Mi studi Amadunia. | I am learning Amadunia. | | Mi no faham. Sema tena, plis. | I don't understand. Say it again, please. | | Yu sema hao! | You speak well! | ## Please, thank you, sorry | | | |---|---| | Plis. | Please. | | Mersi! | Thank you! | | Pardon! | Sorry. / Excuse me. | | Ok. No es problema. | Ok. There's no problem. | | Yu hao! | You are good! | ## Asking the way | | | |---|---| | Pardon! Yu punya tempo? | Excuse me! Do you have a moment? | | Hotel nali? | Where is the hotel? | | Market nali? | Where is the market? | | Anda tarik ini. | Walk this road. | | Go yamin. | Go right. | | Go kiri. | Go left. | | Situ, in ponte. | There, at the bridge. | | Tarik cang o duan? | Is the road long or short? | | Berapa hora? | How many hours? | | Mi anda o mi go in tren? | Do I walk or go by train? | ## Buying | | | |---|---| | Berapa harga? | How much? | | Harga kabir! | The price is big! — the nearest to *that's expensive*; [there is no word for it](dictionary/README.md#words-the-writing-has-asked-for) | | Mi mau kupi ini. | I want to buy this. | | Mi mau ini. Mi no mau itu. | I want this. I do not want that. | | Es kamisa lebi keci? | Is there a smaller shirt? | | Mi punya des pesa. | I have ten money. — *pesa* is the substance, not a coin; [the gap is recorded](dictionary/README.md#words-the-writing-has-asked-for) | | Ok, mi kupi. | Ok, I'll buy it. | | Mersi! Yu hao. | Thank you! You are good. | ## Eating and drinking | | | |---|---| | Mi mau kula. | I want to eat. | | Es restoran sini? | Is there a restaurant here? | | Corba, plis. | Soup, please. | | Mi mau ca garam. | I want hot tea. | | Akua barid, plis. | Cold water, please. | | Ini hao! | This is good! | | Mi no mau lebi, mersi. | I don't want more, thank you. | | Bas, mersi. | That's enough, thanks. | ## Time | | | |---|---| | Sasa. | Now. | | Din ini. | Today. | | Rat ini. | Tonight. | | Besok. | Tomorrow. | | Kemarin. | Yesterday. | | Mi lai kab? | When shall I come? | | Mi lai in pagi. | I'll come in the morning. | | Mi no punya tempo. | I have no time. | ## If something is wrong | | | |---|---| | Madad! | Help! | | Mi bimar. | I am sick. | | Sar mi garam. | My head is hot. — the nearest the language gets to *my head hurts*; [pain has no word](dictionary/README.md#words-the-writing-has-asked-for) | | Mi mau doktor. | I want a doctor. | | Hospital nali? | Where is the hospital? | | Mi suda unut anahtar mi. | I have forgotten my key. | | Mi no naiti dom mi. | I can't find my house. | | Es bahaya? | Is it dangerous? | | Polisi nali? | Where are the police? | ## Small talk | | | |---|---| | Din ini hao. | It's a good day. | | Yamur lai. | It's raining. | | Angin barid. | The wind is cold. | | Asman asul. | The sky is blue. | | Madina ini sundar. | This city is beautiful. | | Yu punya familia? | Do you have a family? | | Mi punya du anak. | I have two children. | | Mi rabota in skola. | I work at a school. | | Mi ama musik. | I love music. | ## Twenty words that will get you a long way *hi · salam · plis · mersi · pardon · ya · no · ok · bas · madad · mi · yu · es · mau · go · lai · nali · berapa · harga · bai* With these and a pointing finger you can cross a city. ## What this page could not say A phrasebook is a hard test, because it is made of the things people actually say rather than the things a grammar has got round to. Four gaps showed up immediately, and a fifth was found later by reading this page against the lessons: - **No word for "very".** *Thank you very much* and *this is very good* both had to lose the "very". *cok* means "much" and whether it can carry "very" is [undecided](grammar/adverbs.md). - ~~**Almost every line is a fragment.**~~ *Sasa.* *Besok.* *Corba, plis.* *Madad!* That is what a phrasebook is, and as of September 3, 2026 [the grammar covers it](grammar/sentence-types.md): any constituent may stand alone as an utterance. - ~~**Half the directions are commands**~~ — *Anda tarik ini*, *Go yamin* — and [they are now granted](grammar/sentence-types.md): a verb with no subject is a command. Neither line changed; the rule caught up with them. - **The clock cannot be asked.** *Berapa hora* is "how many hours", not "what time"; the language has no word for a clock or an hour of the day. - **Pain cannot be said.** *Sar mi garam* is "my head is hot", which is the nearest a traveller gets to *my head hurts*. This page glossed it as *my head hurts* until September 4, 2026 — translating around a gap the repository had already recorded, which is the one thing a phrasebook must not do. The gloss says what the words say now, and [the missing word is on the list](dictionary/README.md#words-the-writing-has-asked-for). Nothing was invented to paper over any of them. Four checks in [`check.py`](check.py) also caught four copula slips in the first draft of this page — *es* in front of an adjective, the same mistake the lessons made thirteen times. --- *Mi ama dunia. Yu ama dunia. Kita ama dunia.* --- # The A1 checklist Every gap in [the list the writing has asked for](README.md#words-the-writing-has-asked-for) was found by trying to say something and failing. That method has a blind spot, and [text 19](../texts/text-19-kamra-mi.md) found it: a wall and a floor are exactly what a thematic wordlist would have contained, and eighteen texts went past them because none had described a room. So this page is the other method, kept deliberately separate. It is a hand-made list of ordinary A1 concepts by domain — the judgement is which concepts belong on it — and [`check.py`](../check.py) looks each one up in [the English index](index-english.md) and counts what is missing. **It proposes nothing.** A concept missing here is not a word owed; several are missing on purpose, and *all*, *some*, *none* and *every* are [an open question](../grammar/plural.md) rather than an oversight. **One word appears twice, and it should.** *clean* is in **actions**, where it is the verb, and in **qualities**, where it is what a washed room is. They are two concepts and the list counts them as two. The English index answers *clean* with *safi*, the quality, and until September 5, 2026 that one entry was marking both of them present — so the verb, which the language does not have, was counted as had. An action now counts only if the index holds it as **to X**. *light* is the same shape and comes out right by luck: the index has both *light* and *light in weight*, and the quality is really there. **Four of the missing are a part of speech, not a concept.** The list counts a concept as had when the index answers it, and these four the language already has under another word class: | Missing | What the dictionary holds | | |---|---|---| | **work**, the action | *rabota*, the noun *work* | and whether one root may hold both jobs is [the open question](../grammar/proposal-two-jobs.md) — so it is [set aside from the count entirely](proposal-a2.md), rather than counted either way | | **clean**, the action | *safi*, the quality *clean* | the pair this page was corrected for on September 5, 2026 | | **free** | *uhuru*, *freedom* | | | **angry** | *gusa*, *anger* | | That is the same shape as *quiet* on [the other list](README.md#words-the-writing-has-asked-for), where *sukut* is the noun *silence* and no adjective stands beside it. The two pages found the same thing by two methods and neither named it as a kind. It is worth a name because it is a cheaper gap than the others here: nobody has to decide what the idea is, only which word class should carry it — and for *work* that decision is already open under a different heading. **This still proposes nothing.** Noticed September 6, 2026. **Thirteen of the missing are kinship, and five of them are not missing.** *people* is the largest block of absences on this list, and [text 24](../texts/text-24-mama-mama-mi.md) showed why the block overstates the hole: possession puts the owner straight after the thing owned and applies twice without a marker, so a grandmother is the mother of my mother and needs no root. Measured on September 6, 2026, the thirteen divide three ways. | | | | |---|---|---| | **Composed already** | grandmother, grandfather, cousin | the mother of my mother, the father of my father, the child of a parent's sibling — settled possession and nothing else | | **Refused by design** | man, woman, son, daughter, uncle, aunt | the language has no gender anywhere: *ta* is he, she and it, *dugu* is a sibling of either, *anak* a child of either. A parent's sibling composes; *uncle* as against *aunt* does not, and a root for either would be the first gendered word in the dictionary | | **Genuinely absent** | husband, wife, baby, neighbour | four roots. *baby* comes close — *anak keci* is a small child — and *neighbour* is blocked twice: it needs a word, and it needs the [preposition that does not exist](../grammar/place.md), since nobody can live *near* anything | So the domain is thinner on paper than in use, and the split matters more than the total. **Three of the thirteen are already sayable**, at a cost in length — a cousin is the child of the sibling of my mother, four words, and length is a real price but a different one from silence. **Six are not gaps at all** but the no-gender rule doing what it was chosen to do, and counting them as missing vocabulary reads a decision as an absence. **Four are the real hole.** This proposes nothing; the [A2 briefing](proposal-a2.md) is where the question belongs, and it is a narrower question than thirteen. **And kinship is not the only place composition hides a concept.** Swept on September 6, 2026: every two-word Amadunia phrase in the lessons and the phrasebook that is glossed by a single English word. There are exactly two, both in [the phrasebook](../phrasebook.md), and both are the demonstrative rather than possession — *din ini* for **today** and *rat ini* for **tonight**, this day and this night. **One of the two is on this list as missing.** *today* stands in the **time** domain and the index has no entry for it, because the index holds roots and *din ini* is two of them. So the same thing is true of it as of a grandmother: the language says it, the count cannot see it, and the count is not wrong — it is measuring roots, which is what it says it measures. *tonight* is not on the list at all, which is the checklist's own judgement about what an A1 learner needs and not an oversight. Two rules compose, then, and only two: possession and the demonstrative. Nothing else joins — [no compound word has ever been formed](../grammar/word-formation.md) — so the list of concepts the language reaches without a root is short and can be written out in full, which is what these two paragraphs do. ## the house wall, floor, roof, kitchen, bathroom, stairs, garden, bed, door, window, table, chair, lamp, key, soap, plate, glass, knife, spoon, cup, clock, mirror, towel, box, bag ## the body head, hair, face, eye, ear, nose, mouth, tooth, hand, arm, leg, foot, finger, back, stomach, heart, skin, blood, bone, neck ## food bread, rice, meat, fish, egg, milk, water, tea, coffee, sugar, salt, oil, fruit, vegetable, soup, apple, cheese, butter, potato, onion, banana ## clothing shirt, shoe, trousers, coat, hat, dress, sock, button, cloth ## people person, child, mother, father, sibling, friend, teacher, doctor, farmer, merchant, driver, police, man, woman, baby, neighbour, name, family, uncle, aunt, grandmother, grandfather, cousin, son, daughter, husband, wife ## time day, night, morning, evening, week, month, year, hour, minute, moment, today, tomorrow, yesterday, now, season, late, early, Monday, Sunday, January, summer, winter ## weather rain, snow, wind, sun, cloud, sky, hot, cold, storm, ice, north, south, east, west ## actions eat, drink, sleep, walk, run, go, come, see, hear, speak, read, write, give, take, buy, sell, open, close, wash, cook, work, play, sit, stand, wait, find, lose, carry, put, begin, stop, help, learn, think, know, want, live, die, love, laugh, cry, sing, dance, jump, fall, push, pull, break, build, send, meet, ask, answer, forget, remember, try, feel, hold, throw, clean ## qualities big, small, long, short, new, old, good, bad, fast, slow, clean, dirty, full, empty, strong, weak, easy, difficult, near, far, heavy, light, beautiful, young, happy, sad, angry, tired, hungry, thirsty, sick, free, true, wrong, cheap, expensive, dark, bright, soft, hard, wet, dry, fluent, brown, grey ## basics yes, no, and, or, but, because, if, when, where, who, what, how, why, this, that, here, there, all, some, none, every, many, few, again, also, only, very, then, never, always, sometimes --- # Where the words come from Design rule 4 says **no language family dominates**. That had been balanced batch by batch and never measured over the whole dictionary. This is the measurement, regenerated from [the dictionary](dictionary.md) itself. Two numbers are worth separating. **Origin** is the family a root came from — the first source its entry names. **Reach** is every family whose own vocabulary names that root, which is what decides whether a stranger recognises it. ## All 300 roots | Family | Origin | | Reach | | |---|---|---|---|---| | Austronesian | 67 | 22.3% | 115 | 38.3% | | Turkic | 23 | 7.7% | 94 | 31.3% | | Latin/Romance | 75 | 25.0% | 90 | 30.0% | | Indo-Aryan | 22 | 7.3% | 83 | 27.7% | | Semitic | 49 | 16.3% | 68 | 22.7% | | Niger-Congo | 14 | 4.7% | 64 | 21.3% | | Slavic | 5 | 1.7% | 48 | 16.0% | | Germanic | 13 | 4.3% | 46 | 15.3% | | Iranian | 9 | 3.0% | 36 | 12.0% | | Greek | 8 | 2.7% | 17 | 5.7% | | Sino-Tibetan | 11 | 3.7% | 12 | 4.0% | | Japonic | 4 | 1.3% | 8 | 2.7% | | Koreanic | 0 | 0.0% | 2 | 0.7% | | Austroasiatic | 0 | 0.0% | 1 | 0.3% | | Celtic | 0 | 0.0% | 1 | 0.3% | | Tai-Kadai | 0 | 0.0% | 1 | 0.3% | Every root now names at least one language, so no row is left over. Six once were: *bai, foto, hi, hotel, ok, taksi* said only "already-global" and were counted as belonging to no family. They belong to English, French and Greek, and the label was hiding it: counted as belonging to nobody, they made the European share look smaller than it is. **This table is regenerated by [`check.py`](../check.py) and compared against what is written here,** so it cannot drift from the dictionary again. It did: the numbers above replace a set that could not be reproduced by the method this page states. ## What it shows **Against Esperanto.** The README's case is that Esperanto's roots are about 95% European. Esperanto's figure is about root origins, so this one is too — the family a root came from, counted once. Taking Latin, Romance, Germanic, Slavic and Greek together, Amadunia's European roots are **101 of 300, or 34%**. Indo-Aryan and Iranian are Indo-European too but they are not European — Hindi, Urdu, Bengali and Persian are spoken by well over a billion people in Asia, and counting them as European would be the kind of sleight the project exists to avoid. **By origin there is a largest bloc.** Latin/Romance is 25.0%, ahead of Austronesian at 22.3% and Semitic at 16.3%. Nothing dominates, but nothing is level either. **By reach it is much flatter.** The top six run from 38.3% down to 21.3%; by origin the top six run from 25.0% down to 4.7%. **Six families are named in a fifth or more of the dictionary** — Austronesian, Turkic, Latin/Romance, Indo-Aryan, Semitic, Niger-Congo. This paragraph said those six sat "between 20% and 29%" until September 5, 2026, and three of them do not: Austronesian is 38.3%, Turkic 31.3% and Latin/Romance 30.0%. The figure it was reaching for is the flatness, and the drop from first to sixth says it. This is the more meaningful figure, and the gap between the two columns is the point: so many roots are shared across families that asking where a word *came from* overstates the imbalance. **Forty roots belong to five traditions or more**, and one — *sabun* — belongs to seven: Austronesian, Indo-Aryan, Latin/Romance, Niger-Congo, Semitic, Slavic and Turkic. Fourteen more name six, among them *insan*, *mama*, *papa*, *ca*, *mersi* and *musik*. That sentence read "*sabun*, *kertas*, *kalima*, *insan*, *safari* and *kursi* each belong to five or six traditions at once" until September 6, 2026, and three of its six examples were outside the range it claimed — *sabun* names seven, *kertas* four and *safari* three. *safari* was the weakest example on the page and it was carrying an argument about breadth: it names Swahili, Arabic and German, which is a real spread and not the one the sentence promised. The examples are read off the dictionary now, so an etymology edited tomorrow moves them. ## The weak spot **Sino-Tibetan reaches 4.0% of the dictionary and Japonic 2.7%.** Between them that is roughly a billion and a half speakers, and the language draws almost nothing from them. The cause was recorded as structural: Chinese and Japanese roots are one or two short syllables — *hao*, *kan*, *lai*, *cang*, *duan*, *yuki*, *nomu*, *suru* — every one taken while the language was small, and with the two- and three-letter space now full, no new short root can be added without a collision. **That explanation was checked on September 3, 2026 and is wrong.** It is true of the words already taken and not of the languages they came from. Modern Mandarin vocabulary is overwhelmingly two syllables, and so is much of Japanese, which is exactly the length the four-letter rule wants. Ten ordinary words were put through every rule in [`check.py`](../check.py) — alphabet, consonant runs, vowel pairs, minimal pairs, length — and seven passed: | Source | Shape | Verdict | |---|---|---| | 面包 *miànbāo*, bread | *mianbao* | passes | | 老师 *lǎoshī*, teacher | *laosi* | passes | | 电话 *diànhuà*, telephone | *dianua* | passes | | 友達 *tomodachi*, friend | *tomodaci* | passes — *c* is exactly that *ch* | | 元気 *genki*, well, lively | *genki* | passes | | 学校 *gakkō*, school | *gako* | passes | | 食べ物 *tabemono*, food | *tabemono* | passes | | 朋友 *péngyǒu*, friend | *pengyu* | rejected — *ngy*, three consonants | | 时间 *shíjiān*, time | *sicien* | rejected — *ie* is not an attested pair | | 先生 *sensei*, teacher | *sensei* | rejected — *ei* is not an attested pair | None of these has been added; the dictionary is closed at 300 until A2. They are here as evidence about the cause. The thinness is not a wall the phonology built — it is which words the borrowing reached for. It went to the classical monosyllables and stopped. No root comes from Korean or from a Dravidian language. **Nine sorts in the checker had the same shape, and eight of them could not have shown it.** After the balance table was fixed, every ordering in [`check.py`](../check.py) that decides what a file should say was given a tiebreak, and the file now reads itself to refuse a single-term sort key. The other eight were safe only because their values happened not to tie — which is the same protection this table had until three families reached one root each. **And the three rows arrived in a random order.** The table is sorted by reach, and the three new families reach one root each, so the sort had nothing to separate them — it kept the order they happened to be inserted in, which comes from iterating a set of family names, which depends on the interpreter's hash seed. One run of the checker called the table right and the next called it wrong, on a file nobody had touched. Ties are broken by name from September 6, 2026, and the row order is a fact about the dictionary rather than about the process that read it. Three families were invisible to this table until September 6, 2026, and the reason is that the map from a language to its family did not hold them: *negara* names **Thai** and **Khmer**, *du* names **Welsh**, and Tai-Kadai, Austroasiatic and Celtic had nowhere to go. The reach column dropped them in silence — no row, no warning, and the roots counted for every other family they named. Three rows of one now, and *du* moved from five families to six with them. Korean is not quite absent — *ok* and *pan* name it among their sources, which is the 0.7% of reach in the table — but Tamil and Telugu appear nowhere in the dictionary, in either column. This line read "Korean and Dravidian have no roots at all" until September 6, 2026, which the table three paragraphs above it had already contradicted. Fixing it would mean taking two-syllable compounds from those languages rather than single words. That is a founder decision and is not recorded as an open question yet, because it is not a gap in the grammar — it is a gap in the vocabulary policy. --- *[`check.py`](../check.py) recomputes these figures from the dictionary on every run and fails if this page has gone stale.* --- # Amadunia — Dictionary **300 roots — the A1 target, reached September 3, 2026.** Target for A1: 300. | Word | Meaning | Source languages | |---|---|---| | **Greetings and basics** | | | | bai | goodbye | English *bye* — borrowed as Mandarin 拜拜 *bàibài*, Japanese *baibai*, Hindi/Urdu *bāy*, Indonesian *bai* — already-global | | bas | enough, stop | Hindi/Urdu *bas*, Persian *bas*, Swahili *basi*, Turkish *bas* | | hi | hello | English *hi* — travels with English into casual speech nearly everywhere, and the shape never changes — already-global | | mersi | thank you | French *merci* — adopted in Persian, Turkish, Romanian, Bulgarian, Greek, Egyptian Arabic | | no | no; **not** — before the predicate, see [grammar/negation.md](../grammar/negation.md) | the *na/ne* root: Spanish/Italian *no*, English *no/not*, Latin *non*, Russian *ne*, Polish *nie*, Sanskrit/Hindi *na*, Persian *na-* | | ok | ok | English *OK* — Japanese *ōkē*, Korean *okei*, Turkish *okey*, Russian *okey*, Spanish/Portuguese *ok*; among the most widely borrowed words on Earth — already-global | | pardon | sorry, excuse me | French *pardon* — adopted in Turkish, Russian, Persian, Arabic dialects; Spanish *perdón* | | plis | please | English *please*, already-global | | salam | peaceful greeting | Arabic, Hebrew, Persian, Turkish, Urdu, Swahili | | ya | yes | Indonesian/Malay *ya*, Javanese *ya*; German/Dutch/Swedish/Norwegian *ja* | | **People** | | | | anak | child | Indonesian/Malay *anak*, Tagalog *anak*, Javanese *anak* | | dugu | sibling — no gender | Swahili *ndugu* | | familia | family | Latin *familia*, Spanish *familia*, Swahili *familia*, Polish *familia*, Russian *familiya* | | insan | person, human | Arabic *insān* — adopted in Turkish, Persian, Urdu, Indonesian, Swahili *insani* | | kisan | farmer | Hindi/Urdu *kisān*, Nepali *kisan*, Bengali *kisan* | | kita | we, including you — see [grammar/pronouns.md](../grammar/pronouns.md) | Indonesian, Malay, Tagalog | | mama | mother | a nursery word shared across nearly every family on Earth: Mandarin *māma*, Swahili *mama*, Spanish *mamá*, Russian *mama*, Arabic *māmā*, Hindi *mā̃* | | mi | I, me | Spanish/Italian *mi*, English *me*, Swahili *mimi* | | mualim | teacher | Arabic *muʿallim* (معلم), Swahili *mwalimu*, Turkish *muallim*, Urdu *muallim*, Indonesian *muallim* | | nama | name | Sanskrit/Hindi *naam*, Indonesian *nama*, Japanese *namae*, English *name* | | nasab | relative, kin | Arabic *nasab* (lineage), Turkish *nesep*, Urdu/Hindi *nasab*, Swahili *nasaba* | | papa | father | the nursery pair of *mama*, as widespread: Mandarin *bàba*, Swahili *baba*, Turkish *baba*, Spanish *papá*, Hindi *pāpā*, Russian *papa* | | polisi | police | Greek *polis* → global: Swahili *polisi*, Indonesian *polisi*, Turkish *polis*, Russian *politsiya* | | rafiki | friend | Swahili *rafiki*, Arabic *rafīq*, Persian *rafiq* | | sofer | driver | French *chauffeur* → Turkish *şoför*, Russian *shofyor*, Indonesian *sopir*, Arabic *shufīr* | | ta | he, she, it — no gender | Chinese *tā* 他/她/它, one sound for all three | | tacir | merchant | Arabic *tājir* (تاجر), Turkish *tacir*, Urdu *tājir*, Swahili *tajiri* | | yu | you — one form for everyone | English *you* | | **Body** | | | | ain | eye | Arabic *ʿayn* (عين), Hebrew *ayin* | | burun | nose | Turkish *burun*, Azeri *burun*, Turkmen *burun* | | cehra | face | Hindi/Urdu *chehra*, Persian *chehreh*, Nepali *cehara* | | dente | tooth | Latin *dentem* → Italian *dente*, Spanish *diente*, Portuguese *dente*, French *dent*, English *dental* | | kaki | foot, leg | Indonesian/Malay/Tagalog *kaki* | | kalb | heart | Arabic *qalb* (قلب), Turkish *kalp*, Urdu *qalb*, Persian *qalb* | | korpo | body | Latin *corpus* — global through *corps*, *corporate*, Spanish *cuerpo*, Italian *corpo* | | kulit | skin | Indonesian/Malay *kulit* | | mano | hand | Spanish/Italian *mano*, Latin *manus*, Portuguese *mão*, French *main* | | mulut | mouth | Indonesian/Malay *mulut* | | rambut | hair | Indonesian/Malay *rambut* | | sar | head | Persian *sar*, Hindi/Urdu *sir/sar*, Sanskrit *śiras* | | telinga | ear | Indonesian/Malay *telinga*, Javanese *talingan* | | **Feelings** | | | | gusa | anger | Hindi/Urdu *gussa*, Persian *ghosse*, Nepali *gussa* | | korku | fear | Turkish *korku*, Azeri *qorxu* | | sabar | patience | Arabic *ṣabr* (صبر), Turkish *sabır*, Urdu/Hindi *sabr*, Indonesian *sabar*, Swahili *subira* | | **Food and water** | | | | akua | water | Latin *aqua*, Spanish *agua*, Portuguese *água*, Italian *acqua* | | ca | tea | Chinese *chá*, Hindi *chai*, Turkish *çay*, Swahili *chai*, Russian *chay*, Persian *chai* | | corba | soup | Persian *shorbā* → Turkish *çorba*, Arabic *shurba*, Urdu *shorba*, Bulgarian/Serbian *čorba* | | fruta | fruit | Latin *fructus*, Spanish/Portuguese *fruta*, Russian *frukt*, English *fruit*, German *Frucht* | | halib | milk | Arabic *ḥalīb* (حليب), Swahili *haliba*, Maltese *ħalib* | | kafe | coffee | Arabic *qahwa* → Turkish *kahve* → global: *café*, *coffee*, *Kaffee*, *kofe*, Swahili *kahawa* | | legum | vegetable | Latin *legumen* → French *légume*, Spanish *legumbre*, English *legume* | | minyak | oil | Indonesian/Malay *minyak* | | namak | salt | Hindi/Urdu *namak*, Persian *namak*, Bengali *nun* | | nyama | meat | Swahili *nyama*, and Bantu-wide | | pan | bread, food | Spanish/Portuguese *pan/pão*, Japanese *pan*, Korean *ppang* | | risi | rice | Greek *óryza* → Arabic *aruzz* → Italian *riso*, French *riz*, English *rice*, Russian *ris* | | sukar | sugar | Sanskrit *śarkarā* → Arabic *sukkar* → Turkish *şeker*, Swahili *sukari*, Russian *sakhar*, English *sugar* — one of the most travelled words on Earth | | telur | egg | Indonesian/Malay *telur*, Javanese *telur* | | **Health** | | | | bimar | sick | Persian *bīmār*, Urdu/Hindi *bīmār*, Bengali *bimar* | | ilac | medicine | Turkish *ilaç*, Azeri *ilaç* — from Arabic *ʿilāj* (treatment) | | **Clothing** | | | | kamisa | shirt | Latin *camisia* → Spanish/Portuguese *camisa*, Arabic *qamīṣ*, Urdu/Hindi *kameez*, Swahili *kamisi* | | sapatu | shoe | Portuguese *sapato*, Spanish *zapato*, Indonesian/Malay *sepatu*, Tagalog *sapatos* | | **Home and world** | | | | anahtar | key | Turkish *anahtar*, Azeri *açar* aside; from Greek *anoiktēri* | | ates | fire | Turkish *ateş*, Azeri *ateş*, Persian *ātash* | | bage | garden | Persian *bāgh*, Urdu/Hindi *bāgh*, Turkish *bağ* | | bahari | sea | Arabic *baḥr* (بحر), Swahili *bahari*, Turkish *bahriye*, Indonesian *bahari* | | bulan | moon | Indonesian/Malay *bulan*, Javanese *wulan*, Tagalog *buwan* | | dom | house, home | Russian/Polish *dom*, Czech *dům*, Latin *domus* — global through *domestic*, *domicile*. Replaced *ruma* on September 3, 2026 | | dunia | world | Arabic, Hindi, Swahili, Indonesian, Turkish | | gelas | glass, cup | Dutch *glas* → Indonesian/Malay *gelas*, Swahili *glasi*; English *glass*, German *Glas* | | hutan | forest | Indonesian/Malay *hutan* | | kamra | room | Hindi/Urdu *kamra*, Nepali *kamra*, Bengali *kamra* | | lampu | lamp | Indonesian/Malay *lampu* — global: *lamp*, *lampe*, *lampa*, *lampara* | | madina | city | Arabic *madīna* (مدينة), Turkish *medine*, Urdu *madīna*, Swahili *mji/madina* | | negara | country | Sanskrit *nagara* → Indonesian/Malay *negara*, Thai *nakhon*, Khmer, Javanese | | pahar | mountain | Hindi/Urdu *pahāṛ*, Nepali *pahad* | | pencere | window | Turkish *pencere*, Persian *panjare*, Urdu *panjra* | | petra | stone | Greek *pétra* → Latin, Spanish *piedra*, Italian *pietra* — global through *petroleum*, *petrify* | | pintu | door | Indonesian/Malay *pintu*, Tagalog *pinto* | | piring | plate | Indonesian/Malay *piring* | | ponte | bridge | Latin *pontem* → Italian/Portuguese *ponte*, Spanish *puente*, French *pont*, Romanian *punte* | | sabun | soap | Arabic *ṣābūn* → Turkish *sabun*, Swahili *sabuni*, Indonesian *sabun*, Hindi *sābun*, Russian... Spanish *jabón*, French *savon* | | sikin | knife | Arabic *sikkīn* (سكين), Urdu *sikkīn*, Swahili *kisu* aside; Maltese *sikkina* | | sol | sun | Latin/Spanish/Portuguese *sol* | | stela | star | Latin *stella* → Italian *stella*, Spanish *estrella*, Portuguese *estrela* — global through *stellar*, *constellation* | | sungai | river | Indonesian/Malay *sungai*, Javanese *kali/sungai* | | tabla | table | Latin *tabula* → Spanish *tabla*, English *table*, Russian *tablitsa* | | tanah | earth, ground | Indonesian/Malay *tanah*, Javanese *tanah* | | udara | air | Indonesian/Malay *udara*, from Sanskrit *udāra* | | yatak | bed | Turkish *yatak*, Azeri *yataq* | | **Weather** | | | | angin | wind | Indonesian/Malay *angin*, Javanese *angin* | | asman | sky | Persian *āsmān*, Urdu/Hindi *āsmān*, Turkish *asuman* | | badal | cloud | Hindi/Urdu *bādal*, Bengali *badol*, Nepali *badal* | | yamur | rain | Turkish *yağmur*, Azeri *yağmur*, Turkmen *ýagmyr* | | yuki | snow | Japanese *yuki* (雪) | | **Animals and plants** | | | | animal | animal | Latin *animal* — global: *animal*, *animale*, *Tier* aside, Indonesian *animalia* | | farasi | horse | Arabic *faras* (فرس), Swahili *farasi*, Turkish *feres* | | flor | flower | Latin *flōrem* → Spanish/Portuguese *flor*, Italian *fiore*, French *fleur*, English *flora* | | kanis | dog | Latin *canis* — global through *canine*, Italian *cane*, Spanish *can*, French *chien* | | kucing | cat | Indonesian/Malay *kucing* | | pohon | tree | Indonesian/Malay *pohon* | | samaki | fish | Swahili *samaki*, Arabic *samak* (سمك) | | tair | bird | Arabic *ṭayr* (طير), Urdu *tair*, Persian *tayr* | | **Time** | | | | aksam | evening | Turkish *akşam*, Azeri *axşam*, Turkmen *agşam* | | besok | tomorrow | Indonesian/Malay *besok*, Javanese *sesuk* | | din | day | Hindi/Urdu *din*, Bengali *din*, Persian *din* | | hafta | week | Persian *hafta*, Turkish *hafta*, Hindi/Urdu *hafta*, Arabic dialects | | hora | hour | Spanish/Portuguese *hora*, Italian *ora*, Latin *hora*, Greek *hōra* — global through *hour* | | kemarin | yesterday | Indonesian/Malay *kemarin* | | mausim | season | Arabic *mawsim* (موسم) → Indonesian *musim*, Turkish *mevsim*, Urdu *mausam*, Swahili *msimu* — the source of *monsoon* | | mensis | month | Latin *mēnsis* → Spanish *mes*, Italian *mese*, Portuguese *mês*, French *mois* — global through *bimonthly*, *menstrual* | | momen | moment | Latin *momentum* — English *moment*, Spanish *momento*, German *Moment*, Russian *moment*, Indonesian *momen* | | pagi | morning | Indonesian/Malay *pagi* | | rat | night | Hindi/Urdu *raat*, Bengali *raat*, Sanskrit *rātri* | | sasa | now | Swahili *sasa* | | tahun | year | Indonesian/Malay *tahun*, Javanese *taun* | | tempo | time | Latin *tempus* → Italian/Portuguese *tempo*, Spanish *tiempo*, Indonesian *tempo*, Russian *temp* | | **This and that** | | | | ini | this — after the noun, last in the phrase | Indonesian/Malay *ini*, Tagalog *ito* | | itu | that | Indonesian/Malay *itu* | | **Place** | | | | kiri | left | Indonesian/Malay *kiri* | | sini | here | Indonesian/Malay *sini* | | situ | there | Indonesian/Malay *situ* — the pair of *sini* | | sub | down, below | Latin *sub* — global as the prefix in *submarine*, *subway*, Spanish/Italian *sotto/sub*, Turkish *sub-* in loanwords | | tempat | place | Indonesian/Malay *tempat* | | upar | up, above | Hindi/Urdu *ūpar*, Bengali *upore*, Nepali *upar* | | yamin | right | Arabic *yamīn* (يمين) — the source of *Yemen*, the land on the right hand | | **Actions** | | | | ama | to love | Latin | | ambil | to take | Indonesian/Malay *ambil* | | anda | to walk | Spanish/Portuguese *andar*, Italian *andare* | | beca | to sell | Hindi/Urdu *bechnā*, Bengali *becha* | | beri | to give | Indonesian/Malay *beri*, Turkish *vermek* | | bil | to know | Turkish *bilmek*, Azeri/Uzbek/Kazakh *bil-* | | bisa | can, to be able — before another verb, see [grammar/verb-chains.md](../grammar/verb-chains.md) | Indonesian/Malay *bisa* | | buka | to open | Indonesian/Malay *buka*, Javanese *bukak* | | dengar | to hear | Indonesian/Malay *dengar* | | espera | to wait | Spanish/Portuguese *esperar*, Italian *sperare* — the root of *Esperanto*, "one who hopes" | | faham | to understand | Arabic *fahima*, Indonesian/Malay *paham*, Swahili *fahamu*, Urdu *fahm*, Turkish *fehim* | | fikir | to think | Arabic *fikr*, Turkish *fikir*, Urdu/Hindi *fikr*, Swahili *fikiri*, Indonesian *pikir* | | finis | to finish | Latin *finis* — global: *finish*, *final*, *finale*, *fin*, Turkish *final* | | funga | to close | Swahili *funga* | | gasil | to wash | Arabic *ghasala* (غسل), Urdu *ghusl*, Swahili *kuosha* aside; Turkish *gasil* | | go | to go | English *go*, already-global | | hidup | to live | Indonesian/Malay *hidup* | | ingat | to remember | Indonesian/Malay *ingat* | | kan | to see | Chinese *kàn* (看) | | kanta | to sing | Latin *cantāre*, Spanish *cantar*, Italian *cantare*, Portuguese *cantar*, French *chanter* | | kara | to read | Arabic *qaraʾa* (قرأ) — the root of *Quran*, "that which is read" | | katab | to write | Arabic *kataba* (كتب) — the root of *kitab* (book), Turkish *katip*, Swahili *kitabu* | | kimbia | to run | Swahili *kimbia* | | kirim | to send | Indonesian/Malay *kirim* | | kula | to eat | Swahili *kula*, Arabic *akala* | | kupi | to buy | Russian *kupit'*, Bulgarian/Serbian/Croatian *kupi*, Czech *koupit* | | lai | to come | Chinese *lái* (来) | | lala | to sleep | Swahili *kulala* | | lasim | must, to have to | Arabic *lāzim*, Turkish *lazım*, Persian *lāzem*, Urdu/Hindi *lāzim*, Swahili *lazima*, Indonesian *lazim* — written with *s*, the alphabet having no *z* | | leta | to bring | Swahili *leta* | | lihat | to look, to watch | Indonesian/Malay *lihat* | | madad | help; to help | Arabic *madad*, Hindi/Urdu *madad*, Persian *madad*, Turkish *medet* | | mati | to die | Indonesian/Malay *mati*, Tagalog *matay* | | mau | to want | Indonesian/Malay *mau* | | naiti | to find | Russian *naiti*, Polish *najść*, Ukrainian *znaity* | | nomu | to drink | Japanese *nomu* (飲む) | | otur | to sit | Turkish *otur-*, Azeri *otur-* | | pone | to put | Latin *pōnere* → Spanish *poner*, Italian *porre*, Portuguese *pôr* — global through *position*, *component* | | proba | to try | Latin *probāre* — global: *probe*, *prueba*, *prova*, *proba*, Russian *proba* | | punya | to have | Indonesian/Malay *punya* | | respon | to answer | Latin *respondēre* — global: *respond*, *responder*, *rispondere*, Indonesian *respon* | | sema | to say, to speak | Swahili *sema* | | senti | to feel | Latin *sentīre* → Spanish/Portuguese/French *sentir*, Italian *sentire* — global through *sentiment*, *consent*, *resent* | | start | to begin | English *start* — global: German *starten*, Russian *start*, Turkish/Indonesian *start* | | stop | to stop | English *stop* — global: *stopp*, *stop*, *stoppen*, Indonesian *stop* | | studi | to learn, to study | Latin *studere* — global: English *study*, Spanish *estudiar*, German *Studium*, Russian *studirovat'*, Indonesian *studi* | | suru | to do, to make | Japanese *suru* (する) | | tabak | to cook | Arabic *ṭabakha* (طبخ), Turkish *tabh*, Urdu *tabakh* | | tanya | to ask | Indonesian/Malay *tanya* | | tena | again | Swahili *tena* | | unut | to forget | Turkish *unut-*, Azeri *unut-*, Turkmen *unut-* | | **Play** | | | | bola | ball | Portuguese/Spanish *bola*, Indonesian/Malay *bola*, Tagalog *bola*, Swahili *boli* | | oyun | game | Turkish *oyun*, Azeri *oyun*, Turkmen *oýun* | | **Town and money** | | | | auto | car | German *Auto*, Spanish *auto*, Turkish/Indonesian *oto*, Russian *avto* — global | | banka | bank | Italian *banca* — English *bank*, Spanish *banco*, Turkish *banka*, Russian *bank*, Indonesian *bank*, Swahili *benki* | | harga | price | Indonesian/Malay *harga*, from Sanskrit *argha* | | hospital | hospital | Latin *hospitale* — global: *hospital*, *hôpital*, *hospitalis*, Indonesian *hospital* | | kertas | paper | Greek *chártēs* → Arabic *qirṭās* → Indonesian/Malay *kertas*, Swahili *karatasi* | | kursi | chair | Arabic *kursī*, Turkish *kürsü*, Hindi/Urdu *kursī*, Indonesian *kursi*, Swahili *kiti* | | market | market, shop | already-global — English, Turkish, Russian, Arabic dialects, Indonesian | | mesin | machine | Latin *machina* — global: *machine*, *máquina*, *Maschine*, *mashina*, Indonesian *mesin* | | pesa | money | Swahili *pesa*, Hindi/Urdu *paisa*, Spanish/Tagalog *peso/piso* | | posta | mail | Latin *posita* — global: *post*, *posta*, *pochta*, *pos*, Turkish *posta*, Swahili *posta* | | rabota | work | Russian *rabota*, Polish/Czech *robota* — source of the global word *robot* | | regalo | gift | Spanish/Italian *regalo*, Portuguese *regalo* | | restoran | restaurant | French *restaurant* — global: Turkish/Indonesian/Russian/Polish *restoran* | | skola | school | Latin *schola* — global: Russian *shkola*, Polish *szkoła*, Indonesian *sekolah*, Swahili *shule*, German *Schule*, English *school* | | tarik | road, way | Arabic *ṭarīq* (طريق), Turkish *tarik*, Urdu *tarīq*, Swahili *tariki* | | tren | train | Spanish/Turkish *tren*, French *train*, English *train*, Russian *tren* — global | | **Colours** | | | | asfar | yellow | Arabic *aṣfar* (أصفر), Swahili *safari* (a journey, from the same root's travel sense) aside; Urdu *asfar* | | asul | blue | Arabic *lāzaward* → Spanish/Portuguese *azul*, Italian *azzurro*, English *azure* | | merah | red | Indonesian/Malay *merah* | | putih | white | Indonesian/Malay *putih*, Tagalog *puti* | | siya | black | Persian *siyāh*, Turkish *siyah*, Urdu/Hindi *siyāh*, Bengali *shyam* | | yesil | green | Turkish *yeşil*, Azeri *yaşıl*, Turkmen *ýaşyl* — written with *s*, the alphabet having no *sh* | | **Qualities and ideas** | | | | amani | peace | Arabic *amān* (أمان), Swahili *amani*, Turkish *aman*, Urdu *amān* | | amik | deep | Arabic *ʿamīq* (عميق), Urdu *amīq* | | arte | art | Latin *artem* — global: *art*, *arte*, *arta*, Russian *art* | | asan | easy | Hindi/Urdu *āsān*, Persian *āsān*, Bengali *ashan* | | bahaya | danger | Indonesian/Malay *bahaya*, from Sanskrit *bhaya* | | baid | far | Arabic *baʿīd* (بعيد) | | barid | cold | Arabic *bārid*, Swahili *baridi* | | baru | new | Indonesian/Malay *baru* | | benar | true | Indonesian/Malay *benar* | | cang | long | Chinese *cháng* (长) | | cok | many, much | Turkish *çok*, Azeri *çox* | | daima | always | Arabic *dāʾiman* (دائما), Swahili *daima*, Turkish *daima*, Urdu *dāʾim* | | dekat | near | Indonesian/Malay *dekat* | | duan | short | Chinese *duǎn* (短) | | dulce | sweet | Latin *dulcis* → Spanish/Portuguese *dulce/doce*, Italian *dolce*, French *doux* | | eski | old | Turkish *eski*, Azeri *əski* | | garam | hot | Hindi/Urdu *garam*, Persian *garm*, Bengali *gorom* | | genc | young | Turkish *genç*, Azeri *gənc* | | golos | voice | Russian *golos*, Polish *głos*, Czech *hlas*, Serbian *glas* | | grupo | group | Latin *gruppo* — global: *group*, *grupo*, *Gruppe*, *gruppa*, Indonesian *grup* | | habari | news | Arabic *khabar* (خبر) → Swahili *habari*, Turkish *haber*, Urdu *khabar*, Indonesian *kabar* | | hafif | light in weight | Arabic *khafīf* (خفيف), Turkish *hafif*, Urdu *hafīf* | | hao | good | Chinese *hǎo* (好) | | hayai | fast | Japanese *hayai* (速い) | | historia | history | Greek *historía* — global: *history*, *historia*, *Geschichte* aside, *istoriya*, Indonesian *historia* | | ide | idea | Greek *idéa* — global: *idea*, *idée*, *Idee*, *ideya*, Indonesian *ide* | | kabir | big | Arabic *kabīr*, Swahili *-kubwa*, Turkish *kebir* | | kadang | sometimes | Indonesian/Malay *kadang* | | kadar | as ... as | Arabic *qadar*, Turkish *kadar*, Urdu/Hindi *qadar*, Persian *qadr* | | kalima | word | Arabic *kalima* (كلمة), Turkish *kelime*, Urdu *kalima*, Swahili *kalima*, Indonesian *kalimat* | | keci | small | Indonesian/Malay *kecil*, Turkish *küçük* | | kosong | empty | Indonesian/Malay *kosong* | | kotor | dirty | Indonesian/Malay *kotor* | | kuat | strong | Indonesian/Malay *kuat* | | kultura | culture | Latin *cultura* — English *culture*, Spanish/Italian *cultura*, German *Kultur*, Russian *kultura*, Turkish *kültür* | | kurang | less | Indonesian/Malay *kurang* | | lebi | more | Indonesian/Malay *lebih* | | legis | law | Latin *lēgis* — English *legal*, Spanish *ley*, Italian *legge*, French *loi* | | lingua | language | Latin *lingua* → Italian/Portuguese *lingua*, Spanish *lengua*, French *langue* — global through *linguistics*, *lingua franca* | | luma | light | Latin *lūmen* → Spanish *lumbre*, Italian/Portuguese *lume* — global through *luminous*, *illuminate* | | mal | bad | Latin *malus*, Spanish/French *mal*, already-global via *malaria* | | masal | story | Turkish *masal*, from Arabic *mathal*; Azeri *masal* | | mimpi | dream | Indonesian/Malay *mimpi* | | muskil | difficult | Arabic *mushkil* (مشكل), Turkish *müşkül*, Urdu/Hindi *mushkil*, Persian *moshkel* — written with *s*, the alphabet having no *sh* | | natura | nature | Latin *natura* — English *nature*, Spanish/Italian *natura*, German *Natur*, Russian *natura* | | numero | number | Latin *numerus* — English *number*, Spanish *número*, Italian *numero*, German *Nummer*, Russian *nomer*, Turkish *numara* | | paling | most — the superlative, before the adjective | Indonesian/Malay *paling* | | plen | full | Latin *plēnus*, Spanish *lleno*, French *plein*, English *plenty* | | problema | problem | Greek *próblēma* — English *problem*, Spanish/Italian *problema*, German *Problem*, Russian *problema*, Turkish *problem*, Indonesian *problema* | | safari | journey | Arabic *safar* (سفر) → Swahili *safari*, and global in English | | safi | clean | Arabic *ṣāfī* (صافي), Swahili *safi*, Hindi/Urdu *sāf*, Turkish *saf* | | sakil | heavy | Arabic *thaqīl* (ثقيل), Turkish *sakil*, Urdu *saqīl* | | sansi | chance | French *chance* → Turkish *şans*, Russian *shans*, Polish *szansa* | | sedih | sad | Indonesian/Malay *sedih* | | senang | happy | Indonesian/Malay *senang* | | sukut | silence | Arabic *sukūt* (سكوت), Turkish *sükût*, Urdu *sukūt* | | sundar | beautiful | Hindi/Urdu *sundar*, Sanskrit *sundara*, Nepali *sundar* | | surat | letter | Indonesian/Malay *surat*, from Arabic *sūra* (a written passage) | | uhuru | freedom | Swahili *uhuru* | | umid | hope | Persian *omīd*, Urdu *ummīd*, Hindi *ummeed*, Turkish *ümit* | | umur | life, age | Arabic *ʿumr* (عمر), Turkish *ömür*, Urdu *umr*, Indonesian *umur*, Swahili *umri* | | una | together; *una* + noun = with — see [grammar/place.md](../grammar/place.md) | Latin *ūnā*, "together" | | yalan | lie, falsehood | Turkish *yalan*, Azeri *yalan*, Turkmen *ýalan* | | yanlis | wrong | Turkish *yanlış*, Azeri *yanlış* — written with *s*, the alphabet having no *sh* | | **Question words** | | | | berapa | how many, how much | Indonesian/Malay *berapa* | | kab | when | Hindi/Urdu *kab*, Bengali *kobe* | | kaifa | how | Arabic *kayfa* (كيف) | | ke | what | Spanish *qué*, Portuguese *que*, Italian *che*, French *que* | | kim | who | Turkish, Azeri, Uzbek, Kazakh *kim* | | nali | where | Chinese *nǎlǐ* (哪里) | | porke | why; because | Spanish *por qué*, Italian *perché*, French *pourquoi* — "for what" | | **Prepositions** — before the noun | | | | dari | from | Indonesian/Malay *dari* | | in | at, in, on | Latin, English, German, Italian, Dutch *in* | | por | to, for | Spanish/Portuguese *por*, Italian *per*, French *pour*, Latin *pro* | | **Grammar particles** | | | | agar | if | Persian/Urdu *agar*, Hindi *agar*, Turkish *eğer* | | aur | and | Hindi/Urdu *aur* | | es | is, am, are — before a noun predicate only; see [grammar/copula.md](../grammar/copula.md) | Spanish *es*, Latin *est*, Italian *è*, French *est*, German *ist*, English *is*, Russian *est'*, Hindi *hai*, Persian *ast* | | o | or | Spanish/Italian *o*, Portuguese *ou*, Latin *aut* — the language's first one-letter word | | saufa | future marker (before the verb) | Arabic *sawfa* (سوف) | | suda | past marker (before the verb) | Indonesian/Malay *sudah* | | **Already-global loans** | | | | doktor | doctor | Turkish, Russian, Indonesian, Hindi, Swahili *daktari* | | foto | photo | Greek *phōs, phōtos* (light) via *photograph* — the clipped *foto* is ordinary in German, Dutch, Spanish, Portuguese, Italian, Polish, Turkish, Indonesian, Russian *foto* — already-global | | hotel | hotel | French *hôtel* — English *hotel*, Spanish/Portuguese *hotel*, Indonesian *hotel*, Turkish *otel*, Russian *otel*, Hindi *hoṭal* — already-global | | libro | book | Spanish/Italian *libro*, Portuguese *livro*, French *livre* — global through *library* | | musik | music | already-global — German *Musik*, Turkish *müzik*, Russian *muzyka*, Indonesian *musik*, Arabic *mūsīqā*, Swahili *muziki* | | taksi | taxi | French *taxi*, from *taximètre* — Turkish *taksi*, Indonesian *taksi*, Russian *taksi*, Swahili *teksi* — already-global | | telefon | telephone | already-global — Turkish, German, Russian, Polish, Indonesian *telepon*, Arabic *tilifūn* | | **Numbers** — in numerical order, not alphabetical | | | | uan | 1 | Latin *ūnus* → Spanish/Italian *uno*, French *un*, Portuguese *um*; English *one* | | du | 2 | Latin *duo*, Greek *dýo*, Russian *dva*, Hindi/Urdu *do*, Persian *do*, Welsh *dau* — global through *duo*, *dual* | | tri | 3 | Latin *trēs*, Greek *treîs*, Sanskrit *tri*, Russian *tri*, Hindi *tīn* — global through *tri-* | | pat | 4 | Indonesian/Malay *empat*, Tagalog *apat*, Javanese *papat* | | fai | 5 | the Germanic five — English *five*, German *fünf*, Dutch *vijf*, Swedish *fem* — from the root that also gave Latin *quīnque* and Greek *pénte* | | sis | 6 | Latin *sex* → Spanish *seis*, French/English *six*, German *sechs*; Greek *héx*, Persian *shesh* | | seti | 7 | Latin *septem* → Spanish *siete*, Italian *sette*, French *sept*; Sanskrit *sapta*, Hindi *sāt*, Greek *heptá*, English *seven* | | ba | 8 | Chinese | | nau | 9 | Hindi/Urdu *nau*, Bengali *noy*, Persian *noh* — the same root as Latin *novem*, German *neun*, English *nine* | | des | 10 | Hindi/Urdu *das*, Sanskrit *daśa*; Latin *decem* → Spanish *diez*, Italian *dieci*, French *dix*; Greek *déka* | | sen | 100 | Latin *centum* → French *cent*, Spanish *cien*, Italian *cento*; Russian *sto*, Sanskrit *śatam*, Hindi *sau*, Persian *sad* | | mila | 1000 | Latin *mille* → Spanish *mil*, Italian *mila*, French *mille*, Turkish *milyon* | **Every root is sourced.** The dictionary contains no blank etymology. ## Counting to a hundred These are **not roots** and are not counted in the 300 — a figure that read *the 113* until September 5, 2026, left behind by every batch since September 2. Every one is built from the roots above by the rule in [grammar/numbers.md](../grammar/numbers.md) — the digit, then *des*. Learn eleven words and you can count to a hundred. | | Amadunia | built from | |---|---|---| | 10 | des | *(root)* | | 20 | du-des | du + des | | 30 | tri-des | tri + des | | 40 | pat-des | pat + des | | 50 | fai-des | fai + des | | 60 | sis-des | sis + des | | 70 | seti-des | seti + des | | 80 | ba-des | ba + des | | 90 | nau-des | nau + des | | 100 | sen | *(root)* | The teens work the other way round — *des* first, then the digit: 11 = *des-uan*, 12 = *des-du*, 19 = *des-nau*. Anything in between joins the two: 21 = *du-des-uan*, 99 = *nau-des-nau*, 345 = *tri-sen pat-des fai*. There is not one irregular number anywhere in the system. ## A collision, resolved *luma* (light) and *ruma* (house) differed only in *l* versus *r* — the hardest contrast on Earth for over a billion speakers, and the ground on which *beli* and *cala* had been rejected. *luma* is the founder's word; *ruma* was the first batch's, added before the rule existed. On September 3, 2026 *ruma* became **dom**, and the dictionary now contains no l/r pair at all. ## Blocked candidates Words with excellent global reach that collide with vocabulary already settled. These need a founder decision before they can be added. | Candidate | Meaning | Reach | Problem | |---|---|---|---| | kitab | book | Arabic, Turkish, Hindi, Urdu, Swahili, Indonesian | near-minimal pair with *kita* (we) | | hi | sun | Japanese *hi* | *hi* is already **hello**; *sol* was used instead | --- # How often each root is actually used Generated from the material a learner reads — the twenty-six lessons, the thirty texts and the phrasebook — counting every word inside an Amadunia sentence and ignoring the vocabulary tables, which teach a word rather than use it. **6632 words of running Amadunia**, and 291 of the 300 roots appear in one. The other nine appear in no sentence of more than one word anywhere: *bai*, *bas*, *hi*, *mersi*, *ok*, *pardon*, *plis*, *ya* — and *madad*, which is [held back on purpose](../grammar/verb-chains.md) until its class is decided. The eight are interjections and are used alone, which the language grants: [a single constituent may stand as an utterance](../grammar/sentence-types.md). This page counts runs of two words or more, so a one-word utterance is invisible to it, and until September 6, 2026 the count read *every one of the 300* because [the phrasebook's list of twenty words](../phrasebook.md) — a row of vocabulary separated by dots — was being read as two ten-word sentences. Excluding it took twenty words off the total and left the eight with nothing, which is the truth about them and not a gap. [`check.py`](../check.py) regenerates this page and fails if it disagrees, so the numbers follow the corpus rather than the other way round. ## The shape of it | | Share of all running words | |---|---| | first 10 | 31% | | first 25 | 50% | | first 50 | 65% | | first 100 | 81% | | first 150 | 90% | | first 200 | 95% | | first 300 | 100% | Ten roots carry a third of everything written. That is steeper than a natural language and it is what a small dictionary looks like from the inside: the grammar words — *mi*, *yu*, *ta*, *es*, *in*, *ini*, *suda*, *no*, *aur* — do most of the work, and the nouns take turns. For a learner the practical reading is the second row: **twenty-five roots cover half of everything.** They are the first twenty-five below. ## The forty commonest | | Root | Meaning | Uses | Share | |---|---|---|---|---| | 1 | *mi* | I, me | 664 | 10.0% | | 2 | *yu* | you — one form for everyone | 240 | 3.6% | | 3 | *ta* | he, she, it — no gender | 215 | 3.2% | | 4 | *in* | at, in, on | 195 | 2.9% | | 5 | *anak* | child | 151 | 2.3% | | 6 | *es* | is, am, are — before a noun predicate only; see [grammar/copula.md](../grammar/copula.md) | 134 | 2.0% | | 7 | *lai* | to come | 123 | 1.9% | | 8 | *dom* | house, home | 120 | 1.8% | | 9 | *go* | to go | 116 | 1.7% | | 10 | *no* | no; **not** — before the predicate, see [grammar/negation.md](../grammar/negation.md) | 115 | 1.7% | | 11 | *ini* | this — after the noun, last in the phrase | 109 | 1.6% | | 12 | *suda* | past marker (before the verb) | 107 | 1.6% | | 13 | *aur* | and | 104 | 1.6% | | 14 | *hao* | good | 101 | 1.5% | | 15 | *mau* | to want | 99 | 1.5% | | 16 | *ca* | tea | 88 | 1.3% | | 17 | *sema* | to say, to speak | 82 | 1.2% | | 18 | *kita* | we, including you — see [grammar/pronouns.md](../grammar/pronouns.md) | 73 | 1.1% | | 19 | *keci* | small | 72 | 1.1% | | 20 | *kula* | to eat | 72 | 1.1% | | 21 | *rafiki* | friend | 70 | 1.1% | | 22 | *kabir* | big | 69 | 1.0% | | 23 | *saufa* | future marker (before the verb) | 58 | 0.9% | | 24 | *sol* | sun | 58 | 0.9% | | 25 | *dari* | from | 56 | 0.8% | | 26 | *insan* | person, human | 55 | 0.8% | | 27 | *kan* | to see | 55 | 0.8% | | 28 | *punya* | to have | 54 | 0.8% | | 29 | *pan* | bread, food | 50 | 0.8% | | 30 | *libro* | book | 47 | 0.7% | | 31 | *din* | day | 46 | 0.7% | | 32 | *lebi* | more | 45 | 0.7% | | 33 | *tarik* | road, way | 44 | 0.7% | | 34 | *market* | market, shop | 42 | 0.6% | | 35 | *akua* | water | 41 | 0.6% | | 36 | *sasa* | now | 41 | 0.6% | | 37 | *sini* | here | 40 | 0.6% | | 38 | *itu* | that | 39 | 0.6% | | 39 | *por* | to, for | 39 | 0.6% | | 40 | *mama* | mother | 38 | 0.6% | ## Used exactly once 1 roots appear in a single sentence in the whole corpus: *salam* The list read nine until September 6, 2026 — *bai*, *bas*, *hi*, *madad*, *mersi*, *ok*, *pardon*, *plis*, *ya* — and every one of those nine has moved to zero rather than one, because their single appearance was the twenty-word list this page had been counting as two sentences. Nothing about them changed; what changed is that they are no longer credited with a sentence nobody wrote. *salam* is the only root left that is used once and used in a sentence. It was 37 roots long two texts ago. [Text 9](../texts/text-9-pagi-in-madina.md) was written from it and took eighteen, one of which needed a second attempt because a word alone between two commas is named rather than used. What survived that was all of one kind — peace, art, history, law, nature, number, chance, to die, to try, a thousand — and eleven texts about tea, hospitals, letters and errands had not reached a single one. Abstract words need an argument, not a scene, so [text 10](../texts/text-10-mila-tahun.md) was written as one, and took all ten. --- # English → Amadunia An index into [the dictionary](dictionary.md), which runs the other way. Use the dictionary to read Amadunia and this to write it. Derived mechanically from the dictionary, so it cannot drift out of step with it. Where one English word has two entries here, the dictionary entry says which is which. | English | Amadunia | |---|---| | above | upar | | again | tena | | age | umur | | air | udara | | always | daima | | am | es | | and | aur | | anger | gusa | | animal | animal | | are | es | | art | arte | | as ... as | kadar | | at | in | | bad | mal | | ball | bola | | bank | banka | | beautiful | sundar | | because | porke | | bed | yatak | | below | sub | | big | kabir | | bird | tair | | black | siya | | blue | asul | | body | korpo | | book | libro | | bread | pan | | bridge | ponte | | can | bisa | | car | auto | | cat | kucing | | chair | kursi | | chance | sansi | | child | anak | | city | madina | | clean | safi | | cloud | badal | | coffee | kafe | | cold | barid | | country | negara | | culture | kultura | | cup | gelas | | danger | bahaya | | day | din | | deep | amik | | difficult | muskil | | dirty | kotor | | doctor | doktor | | dog | kanis | | door | pintu | | down | sub | | dream | mimpi | | driver | sofer | | ear | telinga | | earth | tanah | | easy | asan | | egg | telur | | empty | kosong | | enough | bas | | evening | aksam | | excuse me | pardon | | eye | ain | | face | cehra | | falsehood | yalan | | family | familia | | far | baid | | farmer | kisan | | fast | hayai | | father | papa | | fear | korku | | fire | ates | | fish | samaki | | flower | flor | | food | pan | | foot | kaki | | for | por | | forest | hutan | | freedom | uhuru | | friend | rafiki | | from | dari | | fruit | fruta | | full | plen | | future marker | saufa | | game | oyun | | garden | bage | | gift | regalo | | glass | gelas | | good | hao | | goodbye | bai | | green | yesil | | ground | tanah | | group | grupo | | hair | rambut | | hand | mano | | happy | senang | | he | ta | | head | sar | | heart | kalb | | heavy | sakil | | hello | hi | | help | madad | | here | sini | | history | historia | | home | dom | | hope | umid | | horse | farasi | | hospital | hospital | | hot | garam | | hotel | hotel | | hour | hora | | house | dom | | how | kaifa | | how many | berapa | | how much | berapa | | human | insan | | I | mi | | idea | ide | | if | agar | | in | in | | is | es | | it | ta | | journey | safari | | key | anahtar | | kin | nasab | | knife | sikin | | lamp | lampu | | language | lingua | | law | legis | | left | kiri | | leg | kaki | | less | kurang | | letter | surat | | lie | yalan | | life | umur | | light | luma | | light in weight | hafif | | long | cang | | machine | mesin | | mail | posta | | many | cok | | market | market | | me | mi | | meat | nyama | | medicine | ilac | | merchant | tacir | | milk | halib | | moment | momen | | money | pesa | | month | mensis | | moon | bulan | | more | lebi | | morning | pagi | | most | paling | | mother | mama | | mountain | pahar | | mouth | mulut | | much | cok | | music | musik | | must | lasim | | name | nama | | nature | natura | | near | dekat | | new | baru | | news | habari | | night | rat | | no | no | | nose | burun | | not | no | | now | sasa | | number | numero | | oil | minyak | | ok | ok | | old | eski | | on | in | | or | o | | paper | kertas | | past marker | suda | | patience | sabar | | peace | amani | | peaceful greeting | salam | | person | insan | | photo | foto | | place | tempat | | plate | piring | | please | plis | | police | polisi | | price | harga | | problem | problema | | rain | yamur | | red | merah | | relative | nasab | | restaurant | restoran | | rice | risi | | right | yamin | | river | sungai | | road | tarik | | room | kamra | | sad | sedih | | salt | namak | | school | skola | | sea | bahari | | season | mausim | | she | ta | | shirt | kamisa | | shoe | sapatu | | shop | market | | short | duan | | sibling | dugu | | sick | bimar | | silence | sukut | | skin | kulit | | sky | asman | | small | keci | | snow | yuki | | soap | sabun | | sometimes | kadang | | sorry | pardon | | soup | corba | | star | stela | | stone | petra | | stop | bas | | story | masal | | strong | kuat | | sugar | sukar | | sun | sol | | sweet | dulce | | table | tabla | | taxi | taksi | | tea | ca | | teacher | mualim | | telephone | telefon | | thank you | mersi | | that | itu | | there | situ | | this | ini | | time | tempo | | to | por | | to answer | respon | | to ask | tanya | | to be able | bisa | | to begin | start | | to bring | leta | | to buy | kupi | | to close | funga | | to come | lai | | to cook | tabak | | to die | mati | | to do | suru | | to drink | nomu | | to eat | kula | | to feel | senti | | to find | naiti | | to finish | finis | | to forget | unut | | to give | beri | | to go | go | | to have | punya | | to have to | lasim | | to hear | dengar | | to help | madad | | to know | bil | | to learn | studi | | to live | hidup | | to look | lihat | | to love | ama | | to make | suru | | to open | buka | | to put | pone | | to read | kara | | to remember | ingat | | to run | kimbia | | to say | sema | | to see | kan | | to sell | beca | | to send | kirim | | to sing | kanta | | to sit | otur | | to sleep | lala | | to speak | sema | | to stop | stop | | to study | studi | | to take | ambil | | to think | fikir | | to try | proba | | to understand | faham | | to wait | espera | | to walk | anda | | to want | mau | | to wash | gasil | | to watch | lihat | | to write | katab | | together | una | | tomorrow | besok | | tooth | dente | | train | tren | | tree | pohon | | true | benar | | up | upar | | vegetable | legum | | voice | golos | | water | akua | | way | tarik | | we | kita | | week | hafta | | what | ke | | when | kab | | where | nali | | white | putih | | who | kim | | why | porke | | wind | angin | | window | pencere | | word | kalima | | work | rabota | | world | dunia | | wrong | yanlis | | year | tahun | | yellow | asfar | | yes | ya | | yesterday | kemarin | | you | yu | | young | genc | ## Numbers | | Amadunia | |---|---| | 1 | uan | | 2 | du | | 3 | tri | | 4 | pat | | 5 | fai | | 6 | sis | | 7 | seti | | 8 | ba | | 9 | nau | | 10 | des | | 100 | sen | | 1000 | mila | Everything between is built by rule — see [grammar/numbers.md](../grammar/numbers.md). --- # A briefing for A2 **Nothing here is adopted.** The dictionary is closed at 300 roots. This page sets out what has been measured about the next 300 so that the decisions are made on evidence rather than on the order a wordlist happens to be in. The standing plan is A2 at around 600 roots — double A1. That number is not what this page is about. What the next 300 should be *made of* is. --- ## 1. The thin families are thin for the wrong reason The record said Chinese and Japanese contribute almost nothing because their words are short and the two- and three-letter space is full. That is true of the fifteen words already taken — *hao*, *kan*, *lai*, *cang*, *yuki*, *nomu*, *suru* and eight more — and false about the languages. That count read **eleven** until September 6, 2026, which is the Sino-Tibetan figure alone; the sentence is about Chinese *and* Japanese, and the seven words it names are four of one and three of the other. Eleven plus four is what the record should have said. Root length, measured across all 300: | Family | 2-3 letters | 4 | 5+ | Average | |---|---|---|---|---| | Austronesian | 7 | 24 | 84 | 5.1 | | Turkic | 7 | 22 | 65 | 4.9 | | Latin/Romance | 23 | 20 | 47 | 4.4 | | Indo-Aryan | 15 | 18 | 50 | 4.5 | | Semitic | 1 | 13 | 54 | 5.0 | | Sino-Tibetan | 7 | 5 | 0 | 3.2 | | Japonic | 3 | 4 | 1 | 3.6 | Three of those rows were stale until September 5, 2026 — Austronesian's 5+ column read 81 against 84, Turkic's 61 against 65, Latin/Romance's 46 against 47 — and two averages with them. The rows counted roots by **reach**, so a root added anywhere moves several of them at once, and nothing was recounting. The table is generated now. Sino-Tibetan is the only family in the dictionary with no root of five letters or more. But modern Mandarin vocabulary is overwhelmingly two syllables, and so is much of Japanese — which is the length the four-letter rule wants. Ten ordinary words run through every rule in [`check.py`](../check.py): seven passed. *mianbao*, *laosi*, *dianua*, *tomodaci*, *genki*, *gako*, *tabemono*. [The working is in balance.md](balance.md). **So the door is open.** The thinness is not something the phonology did. It is that the borrowing reached for the classical monosyllables and stopped. ## 2. What a balance floor would cost If A2 is to fix the imbalance rather than inherit it, the cost is countable. | Family | Reach now | For 10% of 600 | Roots needed | |---|---|---|---| | Sino-Tibetan | 12 (4.0%) | 60 | **+48** | | Japonic | 8 (2.7%) | 60 | **+52** | Together that is up to 100 of the 300 new roots — **a third of the entire A2 batch** — before a single word is chosen for what it means. The real figure is lower, because one word can reach several families at once and these count each separately, but the order of magnitude is right. That is the trade-off in one line: **A2 can be balanced, or it can be chosen freely by meaning, and it cannot be both.** This is a decision, not a calculation. ## 3. What the writing has asked for Sixteen gaps are on the list: thirteen found by trying to write something and failing, one by a sweep of sentence shapes, and two found a third way — by translating a sentence somebody asked for, and by somebody asking about kinship. [They are recorded with the sentence that stopped in each case](README.md#words-the-writing-has-asked-for). **The first 7, in the order they were found:** pain, love as a noun, to miss someone, the clock, a word for *then*, standing up, a coin as against an amount. That enumeration was printed with no count in front of it until September 6, 2026, directly under the word *sixteen*, so seven items read as the whole list. Three of those seven are about how a person feels or where a person hurts. No thematic wordlist predicted that, and it is the strongest evidence available about where 300 roots are actually thin. Two more were claimed and withdrawn on checking, which is why the list is audited rather than appended to. This section's heading said **thirty-four** until September 5, 2026 and the paragraph beneath it said two, which is what the checker had been holding all along. A heading is the one line on a page that nobody rereads. ## 3b. Two roots have never been used Being taught is not being used. Counted on September 4, 2026: **55 of the 300 roots appeared in no text at all** — met once in a lesson's word table, drilled in that lesson, and never written again. What they were is the finding. Not obscure nouns but the joints of the language: *o* (or), *kadar* (as … as), *kurang* (less), *lasim* (must), *berapa* (how many), *kaifa* (how), *fikir* (to think). A rule can be exercised by a text while the word that carries it is not. Three texts were written for them: [16](../texts/text-16-berapa-harga.md) took twenty-one, [17](../texts/text-17-tren-aur-farasi.md) twenty-seven, and [18](../texts/text-18-gusa-in-tarik.md) the last five that could be written. **Two are still unused**, and `check.py` recounts the number so this paragraph cannot drift: *kadang*, *madad*. **Neither can be written at all**, which is why they are the two that are left. *madad*'s class is undecided, so any sentence using it would answer [the open question](../grammar/verb-chains.md) by accident — the question has no evidence because the word has no use, and it can have no use until the question is answered. *kadang* is the same shape: where a frequency adverb stands is [open](../grammar/adverbs.md), so the word cannot appear in a sentence without settling it. The open questions are not only holes in what can be said. They are holes in what can be *practised*, and after two texts written to sweep the vocabulary they are what is left. ## 3c. The other method: a checklist, and what it says Every gap in section 3 was found by writing. [Text 19](../texts/text-19-kamra-mi.md) showed the blind spot in that method — a wall and a floor are exactly what a thematic list would have contained, and nineteen texts went past them. So the other method was run once, kept separate, and written down: [a hand-made list of 274 ordinary A1 concepts by domain](a1-checklist.md), each looked up in the English index by [`check.py`](../check.py). **167 of 273 are present. A hundred and six are not.** Both halves of that sentence are checked now. The second was not, until September 6, 2026: the checker's number parser stopped at ninety-nine, so a figure written out as *a hundred and six* was invisible to it, and this is the count the whole page turns on. A parser that cannot spell a number cannot test a sentence that spells it — and the shape of that failure is the one this repository keeps finding, a map returning nothing for what it does not know rather than saying it does not know. | Domain | Present | Missing | |---|---|---| | the house | 13 of 25 | wall, floor, roof, kitchen, bathroom, stairs, spoon, clock, mirror, towel, box, bag | | the body | 13 of 20 | arm, finger, back, stomach, blood, bone, neck | | food | 15 of 21 | apple, cheese, butter, potato, onion, banana | | clothing | 2 of 9 | trousers, coat, hat, dress, sock, button, cloth | | people | 14 of 27 | man, woman, baby, neighbour, uncle, aunt, grandmother, grandfather, cousin, son, daughter, husband, wife | | time | 13 of 22 | minute, today, late, early, Monday, Sunday, January, summer, winter | | weather | 8 of 14 | storm, ice, north, south, east, west | | actions | 42 of 59 | play, stand, lose, carry, laugh, cry, dance, jump, fall, push, pull, break, build, meet, hold, throw, clean | | qualities | 27 of 45 | slow, weak, angry, tired, hungry, thirsty, free, cheap, expensive, dark, bright, soft, hard, wet, dry, fluent, brown, grey | | basics | 20 of 31 | but, all, some, none, every, few, also, only, very, then, never | **One of those was counted present by an adjective.** *clean* stands in the checklist twice — in **actions**, where it is the verb, and in **qualities**, where it is what a washed room is — and the English index answers *clean* with *safi*, which is the quality. The language has *gasil*, to wash, and no verb for cleaning. From September 5, 2026 a concept in **actions** counts as present only if the index holds it as **to X**, which is what an action is. **And *work* is set aside from both sides of the fraction.** It is the only other action the index answers with a bare entry, and the reason is [*rabota*'s undecided class](../grammar/proposal-two-jobs.md). It was exempt from the rule and counted **present** until September 6, 2026 — which is the answer the exemption was written to avoid giving, because counting it present asserts that *rabota* is the verb and counting it missing asserts that it is not. Neither is neutral. It is now left out of the numerator and the denominator alike: 273 concepts counted, the 274th named and waiting. The totals moved by one each way, and the page and [the checklist](a1-checklist.md) now say the same thing about the same word, which they did not. ### The sets a language has to close Kinship was not an isolated hole. Checked on September 4, 2026, four more closed sets are missing entire, and none was on the checklist either: | Set | In the dictionary | |---|---| | the compass — north, south, east, west | **nothing**; the language has left, right, up, down | | the days of the week | **nothing**; it has *hafta*, week, and *din*, day | | the months | **nothing**; it has *mensis* | | the seasons | **nothing**; it has *mausim* | | the colours | six — *merah*, *asul*, *siya*, *putih*, *asfar*, *yesil*; no brown, grey, orange, purple or pink | **Three of the five may not need roots at all**, and that is the interesting part. Chinese names its weekdays by number — 星期一 is week-one — and its months the same way, and a language built on economy should probably do the same: *din du* for the second day, *mensis tri* for the third month. **But it cannot, yet.** Numbering a day or a month is naming an ordinal — first, second, third — and [ordinals have no form](../grammar/numbers.md). That open question, filed since text 6 wanted "on the eighth day", turns out to block two whole vocabulary sets as well as a sentence. It is the first case where answering one open question would remove the need for **nineteen** roots: seven weekday names and twelve month names. That figure read "a dozen" until September 5, 2026, while [numbers.md](../grammar/numbers.md) and [the ordinals briefing](../grammar/proposal-ordinals.md) both said nineteen — and the briefing cited **this page** for it. Three pages, two numbers, and the one being quoted was the odd one out. The seasons are the set left out of all three: four more that a number could name, and *mausim tri* for autumn is not what anybody calls it, so they are counted as roots the language would still need. The compass is the opposite: four directions that no number can build, and no root for any of them. **Kinship's absences are thirteen of twenty-seven, and four of the thirteen are the hole.** The *people* domain is the largest block of absences on the checklist, and [the checklist now splits it](a1-checklist.md): three of the thirteen are already sayable by possession alone — a grandmother is the mother of my mother — six are the no-gender rule doing what it was chosen to do rather than vocabulary that is missing, and four are roots the language does not have. Reading a settled decision as an absence is what the block was doing, and it made the domain look three times emptier than it is. **Clothing is two of nine** — a shirt and a shoe — and is the thinnest domain in the language by a distance; nothing composes a coat. **Kinship is fourteen of twenty-seven present**, which is the same measurement as the paragraph above and not a second one: fourteen present, thirteen absent, and only four of those thirteen a root the language would have to add. This page said *fourteen of twenty-seven* and *thirteen of twenty-seven* four paragraphs apart, as though they were the same count disagreeing with itself, until September 6, 2026. The founder noticed the domain on September 4, 2026 by asking how to say *uncle*; the checklist had listed eighteen people-words and none of the nine he asked after, which is the second time a hole in the checklist has been found by somebody simply wanting to say something. The everyday adjectives are close behind: there is no *slow*, no *angry*, no *tired*, no *hungry*, no *dark* and no *wet*, so a person can be sick or sad but not tired, and a night can be cold but not dark. **The two methods agree where they overlap.** Seven of the thirteen gaps found by writing — *slowly*, *to stand up*, *cheap*, *the clock*, *then*, *but* and *a wall* — are on this list too, which is the check that matters: the checklist reproduces what writing found and adds **99** more. What it cannot do is say which of the **106** matter, and that is exactly what writing does say. Neither list replaces the other and this page keeps both. Those two figures read "about eighty" and "the eighty-four" until September 5, 2026, and the overlap read five rather than seven — *then*, *but* and *a wall* were missing from it and *fluent* was in it, which is one of the two gaps that came from a question rather than from writing. All three numbers are the checklist's own arithmetic and are checked as such now. ## 4. What A2 content is blocked on Three decisions are waiting, and lesson and text material for A2 will hit them immediately. Two more were settled on September 3, 2026 and no longer block anything: - ~~stress~~ — [settled](../grammar/stress.md): the second-to-last syllable - ~~the imperative~~ — [settled](../grammar/sentence-types.md): a verb with no subject - **a mark for a name** ([briefing](../grammar/proposal-names.md)) - **"want to be" plus an adjective** ([briefing](../grammar/proposal-modal-adjective.md)) - ***rabota*'s class** — one word, [twenty places](../grammar/proposal-two-jobs.md) And on the open list, *before*, *after*, *until* and the ordinals were each hit by real writing: [the letter](../texts/text-7-surat-por-mama.md) was arranged into short sentences to avoid the first three, and [text 6](../texts/text-6-seti-din.md) could not say "on the eighth day". ## 5. The vocabulary is thin in the wrong places A1 is a level about concrete life — food, clothes, the body, the house, the weather, shopping, health. Counted by the dictionary's own thematic groups, this one is not shaped like that. | Theme | Roots | |---|---| | Qualities and ideas | 64 | | Actions | 51 | | Home and world | 28 | | People | 18 | | Town and money | 16 | | Food and water | 14 | | Time | 14 | | Body | 13 | | Numbers | 12 | | Greetings and basics | 10 | | Animals and plants | 8 | | Place | 7 | | Question words | 7 | | Already-global loans | 7 | | Colours | 6 | | Grammar particles | 6 | | Weather | 5 | | Feelings | 3 | | Prepositions | 3 | | Health | 2 | | Clothing | 2 | | This and that | 2 | | Play | 2 | **64 of the 300 roots — 21% — are qualities and ideas.** Health has two words, *bimar* and *ilac*; there is no pain and no fever. Clothing has two, a shirt and a shoe. Play has two. Feelings have three. Food has fourteen. That is the shape of a dictionary built by writing texts about ideas, which is what happened, and it is the opposite of what a beginner needs first. **The next roots should be concrete**, and those first seven — pain, love as a noun, to miss, the clock, *then*, to stand up, a coin — all sit in exactly those thin groups. ### Which raises the number itself The front page has said 300 is enough for A1 because things can be described with what is already there. The stronger version of that claim — that compounding does the rest — [was checked and is false](../grammar/word-formation.md). Estimates for A1 in a natural language run to about 500 words, and Basic English covers daily life in 850. So **300 looks like an A1 core rather than a finished A1**, and the honest reading is that A1 completes somewhere near 500 with the extra two hundred spent on concrete vocabulary. That would move A2 as well. It is a founder's decision and this page does not make it; it records that the number now rests on a weaker argument than the one it was chosen under. ## What is measured and what is judgement Measured: the length table, the seven passing shapes, the +48 and +52, the gaps and where they were found, and the theme table — which `check.py` regenerates from the dictionary, so it cannot drift. Judgement, and the founder's: whether to set a family floor and where, whether the feelings are a deliberate theme for A2 or a coincidence of the data points, **whether A1 finishes at 300 or nearer 500**, and whether 600 is still the right number. Note that the reason once given for a small dictionary does not hold: *mesin ambil foto korpo* says X-ray with no new root, but it is a sentence, not a compound, and [the language forms no compound words at all](../grammar/word-formation.md). --- # Dictionary One row per word: **word | meaning | source languages**. The *source languages* column records where a word's sound and sense come from. Amadunia aims for global balance, so this column is how that balance is checked — it should not fill up with one family. Every one of the 300 roots now carries a sourced etymology. The last fourteen blanks — the founder's first words and the numbers — were filled on September 3, 2026. Words are grouped by theme and alphabetised within each group — except the numbers, which run in numerical order. For the other direction, see [English → Amadunia](index-english.md). It is derived from this file mechanically, so the two cannot drift apart. For which families the roots come from, and how far each one reaches, see [Where the words come from](balance.md). For which roots actually carry the language — ten of them are a third of everything written — see [how often each root is used](frequency.md). For what the next three hundred should be made of — the measurements, and the decisions they do not make — see [A briefing for A2](proposal-a2.md). For tools rather than people: [dictionary.json](dictionary.json) and [dictionary.csv](dictionary.csv) carry the same 300 entries, generated from this file. The CSV imports straight into a flashcard deck. Edit the markdown, never the derived files — [`check.py`](../check.py) regenerates both and fails if they disagree. ## Words the writing has asked for These are not suggestions. Each one is a place where an original text or the phrasebook tried to say something ordinary and stopped, and the sentence that stopped is on record. They are the first candidates for A2, because a gap found by writing is worth more than a gap found by reading a wordlist. | Missing | Found by | The sentence that could not be finished | |---|---|---| | pain, to hurt | [text 6](../texts/text-6-seti-din.md) | a child ill for seven days never says where it hurts; *mal*, bad, is not the same thing | | love, as a noun | [text 7](../texts/text-7-surat-por-mama.md) | a letter cannot be signed *with love* — *ama* is the verb, and one root does one job | | to miss someone | [text 7](../texts/text-7-surat-por-mama.md) | *Mi sedih porke yu-yu baid* names the cause instead of the feeling | | a clock, an hour of the day | [the phrasebook](../phrasebook.md) | *Berapa hora* asks "how many hours", not "what time" | | then, next | [text 8](../texts/text-8-kaifa-suru-ca.md) | a recipe's steps can only be set side by side; *tena* is "again", not "next" | | to get up, to stand | [text 9](../texts/text-9-pagi-in-madina.md) | *Mi go dari yatak*, I go from the bed, paraphrases a position change; *otur* is to sit and has no opposite | | a coin, as against an amount | [text 9](../texts/text-9-pagi-in-madina.md) | *Mi beri fai pesa* is "I give five money" | | slowly | [text 11](../texts/text-11-anak-katab.md) | *hayai* is fast and has no opposite, so a child writing *slowly* had to be written as writing *badly* | | but | [Lesson 12](../lessons/lesson-12-colors-directions.md) | *— No, eski. Ma hao!* — "no, old, but good". *ma* was invented on the spot and stood in the lesson unnoticed until September 4; the line now reads *No, eski. Hao!* and the contrast is gone | | quiet, silent — the adjective | [text 13](../texts/text-13-kula-una.md), [text 21](../texts/text-21-uan-umur.md) | *Dom sukut*, the house is silent, three times in two texts — and *sukut* is the **noun** silence, so the phrase reads "the house's silence". The lines are *Es sukut in dom*, there is silence in the house, which is the existential doing the work an adjective would. Found September 5, 2026. | | cheap, dear | [text 16](../texts/text-16-berapa-harga.md) | a whole bargaining scene runs on *kabir* and *kurang* — big and less — and says *this price is as big as the price in the new city* where a person would say *this is dear* | | a wall, a floor | [text 19](../texts/text-19-kamra-mi.md) | the dictionary has a house, a room, a door, a window, a bed, a table, a chair and a lamp, and not one surface to put them against; a description of a room can only list its contents | | a song | [text 21](../texts/text-21-uan-umur.md) | *Ta-ta kanta kanta eski* — they sing an old song — used *kanta* as the verb and again as a noun, and the dictionary glosses it *to sing* and nothing else. The line is *Ta-ta kanta tena* now, they sing again, and the song is gone. Found September 6, 2026 by a shape sweep rather than by writing, which makes it the one entry here found by neither method the page describes | | a rule | [text 27](../texts/text-27-kalima-por-anak-anak.md) | a notice pinned on a school wall cannot say what it is. *legis* is law, which is a courtroom, and *kalima* is a word, so the title reads *words for the children* where *rules for the children* was meant | | fluent, fluently | [a sentence someone asked for](#a-third-way-of-finding-one) | *Mi parolas Esperanton tre flue* came out as *Mi sema hao Amadunia* — I speak Amadunia well. *hao* carries most of it, which is why this one is weaker than the others on the list | | uncle, aunt, grandmother, cousin | [a question someone asked](#a-third-way-of-finding-one) | the whole kinship set is six words — *mama*, *papa*, *anak*, *dugu*, *familia*, *nasab*. *dugu mama* is the mother's sibling and works today; whether the language wants its own root is A2's call | ### A third way of finding one Of the sixteen, fourteen name a text or a lesson as the place they were found — thirteen because the writing stopped there, and one, *a song*, because a sweep of every sentence shape in the material turned up a line using one root for two jobs. The fifteenth and sixteenth came from neither writing nor a wordlist: the founder asked how Amadunia says the Esperanto sentence *Mi parolas Esperanton tre flue* — "I speak Esperanto very fluently" — on September 4, 2026. The answer is *Mi sema hao Amadunia*, and two things fell out of the difference. **No word for *fluently*.** *hao*, well, carries most of it, which is why this entry is marked weaker than the others: nothing is unsayable, only less exact. That is worth recording honestly rather than counting it as a hole the size of *pain*. **And *very* was wanted for the second time**, after [the phrasebook](../phrasebook.md). It is [an open question](../grammar/adverbs.md), not a missing word, and this is now its second site. The method is worth naming because [texts/](../texts/) refuses translation as a genre — a translated book needs either a copyright or a vocabulary the language does not have. **One sentence is not a genre.** Asking how the language says a thing somebody actually said is a cheap probe, and it found something two other methods had missed. --- The sixteen divide three ways, and the division is the argument for finding gaps by writing. **Three of the sixteen are about how a person feels or where a person hurts** — pain, love as a noun, and missing someone. That is where 300 roots turn out to be thinnest, and it is not what a thematic wordlist would have predicted. **Seven of the sixteen are joints rather than things:** a point in the day, the word that puts one step after another, the opposite of sitting down, the difference between a coin and an amount, the opposite of *fast*, the word that contrasts two clauses, and the pair that prices a thing. A wordlist names things; only a sentence notices that it cannot get from one step to the next. **Six of the sixteen are neither**, and they are the ones this page argues about below: a wall and a floor, the kinship set, *fluent*, *a rule*, *quiet* — which is not a missing concept at all but a missing part of speech, since *sukut* is the noun *silence* and there is no adjective beside it — and *a song*, which is the same shape the other way round: *kanta* is the verb and there is no noun. Until September 6, 2026 those two paragraphs were one sentence printed twice, back to back, introducing the same seven items as *the rest* and then as *most of them*. Neither quantifier was right: the rest of fourteen is eleven, and seven of fourteen is exactly half. The list was correct both times and the sentence around it was wrong both times, in two different directions. This page used to put a number on that — "ten of the twelve are not nouns a category would have suggested" — by comparing the list against [the A1 checklist](a1-checklist.md). **That comparison is not the test it looks like, and the number is withdrawn.** The checklist was written later and by the same hand, and several of these concepts are on it *because* the writing had already found them. Counting the overlap of two lists one person wrote in one week measures the overlap, not the method. What survives without circularity is the exceptions, and they are worth the correction. **A wall and a floor** are exactly what a wordlist headed *the house* would have contained, and eighteen texts went past them without noticing, because no one had described a room until [text 19](../texts/text-19-kamra-mi.md); writing found it in the end, a list would have found it sooner. **Kinship** is the same shape and larger: uncle, aunt, grandmother and cousin sit in the checklist's *people* domain and no text had ever needed them. **A rule** is the third of that kind and the plainest: any wordlist with a *school* heading has it, and twenty-six texts went past it because none of them was a notice. And **fluent** came from neither method but from a single sentence asked in another language, which is the third way described above. The first eight were re-checked against the dictionary on September 3, 2026, and two entries did not survive it. Comparing one duration with another is granted by [comparison.md](../grammar/comparison.md) under a heading called *With verbs*. A word for *forever* may already be *daima*, always — whether it covers both is now [an open question](../grammar/adverbs.md) rather than a missing word. Both claims were made without reading the page that would refute them, which is why the list is now audited rather than appended to.