Counting in Nagamese, Ao and Sümi

One to ten, side by side

Two rows stand out. Five is pungu in both languages — identical. Nine is teku in Ao and toku in Sümi, differing by a single vowel. These are cognates: forms inherited from a shared ancestor rather than borrowed.

Elsewhere the resemblance is fainter. Ao ana "two" and Sümi kini "two" do not look alike at first glance, though the -ni element is suggestive. Ao ka "one" and Sümi lakhi "one" look unrelated entirely. This mixture — some numerals obviously shared, others not — is exactly what you expect from languages that separated a long time ago and have been changing independently ever since.

Nagamese numbers work differently

Nagamese is not Sino-Tibetan — it is an Assamese-lexified creole — so its numerals come from a completely different source and look nothing like the Ao and Sümi ones.

In practice, spoken Nagamese draws on Assamese-derived numerals for small counts, and increasingly on English numerals for larger ones, prices and years. You will hear 400 rupees said exactly that way. This is not sloppiness: code-switching for numbers is a stable feature of the language, particularly in commerce and among younger urban speakers.

Quantity words carry more everyday weight in Nagamese than exact numerals do:

Where the numeral goes

Both Ao and Sümi put modifiers after the head noun, and numerals are modifiers. Ao senpungu is sen "money" + pungu "five" — "five rupees", noun first. Compare English, which puts the numeral first, and Nagamese, which follows the Indo-Aryan pattern of numeral before noun.

Ao also drops the plural marker -tem when a numeral greater than one is present — the count already conveys plurality, so marking it twice is redundant. Sümi behaves similarly, and lakhi "one" doubles as something close to an indefinite article: aki lakhi is "a house" as much as "one house".