Phi practice Counting in Phi: the ternary numerals and the four natures

Reading 9 of 11

Part 8

Common errors

The numeral system's errors divide in two: forms the grammar forbids, and forms the grammar allows that say something you did not mean. Both kinds below, with the repair reasoning.

Error 1a remainder of three

Wrong:

*ta shao ta shao lipha powea.

Right:

wi shao lipha powea.

two three-group LIFE.CLF egg.

(Six eggs.)

If three remain, another group has formed. ta shao ta shao (3 + 3) is not a number any more than "twelve-teen" is; the composition always uses the largest unit first and never repeats a unit. Six is two threes: wi shao. The same error at the next rung (*wi phoi wi phoi for 36) repairs to ta lau ta phoi (27 + 9).

Error 2units out of order

Wrong:

*ta shao ta phoi miona.

Right:

ta phoi ta shao miona.

one nine-group one three-group person.

(Twelve people.)

Largest unit first, always: the ladder descends. The wrong form is not ambiguous (its parts still sum) but it breaks the reading rhythm every fluent listener leans on: the shape of a Phi numeral tells you its magnitude before its details, exactly as modifier-first tells you a sentence's frame before its content.

Error 3digit three

Wrong (for the intended meaning):

*shao lipha powea.

Intended as exactly three eggs. This form is grammatical, but it is Part 5's about: a bare scale unit gives an approximate magnitude around three. Exactly three is counted: ta shao lipha powea. The minimal pair matters in trade, recipes, and any context where the difference affects action; mixing them up is how a soup comes out wrong. Digit-then-unit for exact; bare unit for an estimate.

Error 4the classifier misfiled

Wrong (for the market's eggs):

wi themo powea.
two THING.CLF egg.

Grammatical, and for some eggs even right: the lexicon licenses themo for an egg counted as food or detached material. The market's eggs are meant for hatching, life in waiting, so themo misfiles them. Conversely, a branch in the woodpile has left lipha behind. The repair is not syntactic; it is the nature-now question from Part 3, asked of these eggs rather than eggs in general: what is this, now? When genuinely torn (a seed packet? cut flowers in water?), remember that classifiers are optional, and declining to classify is the honest form of not knowing: wi powea is always complete.

Error 5classifier disagreement across an operation

Wrong:

*wi himo phemi nela ta shao themo noru sholei.

You cannot add two guests to three bowls: addition requires commensurable quantities. Count the guests, count the bowls, and relate them with an ordinary sentence (theli phemi ta noru howela. — each guest receives a bowl). This same-kind rule applies to addition and subtraction; multiplication uses a counted quantity plus a bare factor, and division may relate a total to differently classified recipients or groups.

Error 6no operand boundary

Wrong (for 3 + 2):

*ta shao wi sholei. ta shao wi kelai.

Right:

ta shao nela wi sholei. ta shao wi kelai.

one three-group COORD two gather. one three-group two equals.

(Three plus two results in five.)

ta shao wi is the single numeral five, not an audible pair of operands. Rhythm, pause, and context are not grammatical boundaries. Arithmetic, equality, and magnitude comparison require nela between the two quantities.

Error 7ordering people

Grammatical, and against the grain:

nu ta himo miona.

ORD one HUM.CLF person.

(the first person)

Legal, and sometimes exactly right: someone is first in line for soup. The grammatical form does not distinguish a harmless queue from a consequential ranking; who set the order, by what criterion, and at what cost are the questions the ordinal leaves open. Order events, days, stories, and steps freely; order people where a line is truly just a line.

Error 8unsupported precision

Grammatical, but unsupported in the stated scene:

ta rei ta lau wi phoi wi silero.

One hundred twenty-eight stars, composed correctly, but the scene says nobody counted them. The problem is not precision itself; it is asserting evidence the speaker does not have. If only a rough scale is known, rei silero says so without claiming literal uncountability. Where someone truly counted (a scientific observation, a child's finished tally), the long exact form is the honest one, and its length is the price of knowing.