Two ways to say each number
Japanese has two full sets of number words. One is native to Japanese. The other was borrowed from Chinese many centuries ago and adapted to Japanese sounds. Writers call the native reading “kun’yomi” and the Chinese-derived reading “on’yomi”, and for numbers the second is often called the Sino-Japanese set. The written digits and kanji are the same for both, so the same symbol can be read in either way, and context decides.
| Number | Sino-Japanese | Native |
|---|---|---|
| 1 | ichi | hitotsu |
| 2 | ni | futatsu |
| 3 | san | mittsu |
| 4 | shi, yon | yottsu |
| 5 | go | itsutsu |
| 6 | roku | muttsu |
| 7 | shichi, nana | nanatsu |
| 8 | hachi | yattsu |
| 9 | kyu, ku | kokonotsu |
| 10 | ju | to |
Where the Chinese-derived set comes from
The Sino-Japanese words are borrowed from Chinese pronunciations that reached Japan over several periods. Linguists distinguish layers that entered at different times, and the forms were adjusted to fit Japanese syllables. They are loosely related to modern Chinese numerals, but the sounds have diverged on both sides, which is why “san” resembles Mandarin “san” but “shi” for four only faintly recalls Mandarin “si”. The close resemblance to Korean Sino-Korean numerals comes from the same Chinese source.
The native series is the older Japanese vocabulary. It is related to the ordinary Japanese way of counting objects up to ten. The pattern in the native forms is interesting: several of them seem to be related in pairs, such as “hitotsu” and “futatsu”, and some scholars have proposed that forms like those for one and two, three and six, or four and eight were related by sound changes. These proposals are speculative and are not generally accepted as settled.
When each set is used
In general, the Sino-Japanese set is used for abstract counting and for most practical purposes:
- arithmetic, phone numbers, prices and larger numbers
- dates, months and times of day
- numbers combined with most counters, such as “san-bon” for three long objects, and the numeral classifiers that follow
The native set is used for counting small groups of objects, usually up to ten, when no classifier is attached: “hitotsu, futatsu, mittsu”. It also appears in the ages of small children and in set expressions. Beyond ten, the native series is not normally used, and the Sino-Japanese forms take over.
Dates show a mixture. The days of the month from the second to the tenth use irregular forms based on the native series, such as “futsuka” and “mikka”, while most of the later days use Sino-Japanese numbers with “nichi”, with a few exceptions such as “hatsuka” for the twentieth. Month names, on the other hand, are straightforward combinations of a Sino-Japanese numeral with “gatsu”.
People are counted with an irregular native pair for the first two, “hitori” for one person and “futari” for two. From three people, the counter “nin” is used with Sino-Japanese numerals: “sannin”, “yonin”.
Why four, seven and nine have alternate readings
Four, seven and nine are the numbers where learners meet two choices even inside the Sino-Japanese set.
Four is “shi” or “yon”. “Shi” is the Chinese-derived reading, and it sounds the same as the word for death (written with a different character). Because of this, “yon” is often preferred, particularly in daily speech. The time “four o’clock” is “yoji” and the age “four years old” is “yon-sai”, while “April” is “shigatsu”, showing that conventions vary by expression. The avoidance of four is part of a wider East Asian pattern described as tetraphobia in East Asia, and the sound-based associations behind it are part of the wider topic of Chinese number homophones. It is a linguistic and cultural tendency, and the degree of avoidance varies between people.
Seven is “shichi” or “nana”. “Shichi” is the Sino-Japanese form, but it sounds similar to “ichi” (one) and “hachi” (eight) over a poor connection or in noisy conditions, so “nana” is often used to avoid confusion. Both are used. The month of July is “shichigatsu” and seven o’clock is “shichiji”, while “nana” is common in phone numbers and counting out loud.
Nine is “kyu” or “ku”. “Ku” is the word that can resemble the word for suffering or hardship, written with another character, and so “kyu” is often used. September, though, is “kugatsu” and nine o’clock is “kuji”, which again shows that no single rule covers everything.
The alternate readings, then, have two kinds of motive: avoiding possible unlucky associations, and avoiding confusion between similar-sounding numbers. How strongly speakers act on the first motive differs; many use the “ordinary” reading without a thought.
Why the system persists
Having two sets might seem inefficient, but it works because the sets divide up the work. The Sino-Japanese numbers are the base of large-number arithmetic and the numerals used with counters, so they are used constantly. The native numbers survive in the small-count, child-level, everyday contexts. As a result, Japanese speakers rarely need to think about which set to choose; it is a habit built through use.
The pattern is a clear example of the borrowing described in the broader story of number words: a second layer of vocabulary arrives with writing, trade and learning, and it settles into specific roles beside the older layer without replacing it. Korean is similar, with its own native and Sino-Korean numerals. Both languages also took over the Chinese way of writing numbers, which linked to calculating tools such as the abacus; the Japanese version is the soroban, a chapter in abacus history.
Summary
Japanese counts in two ways because it has two layers of number words, the native one and the one brought from China. The Sino-Japanese layer carries most arithmetic and compound counting, while the native layer survives for small counts. Four, seven and nine have alternates because of possible unlucky associations and the wish to be heard clearly, not because of any inherent power in the numbers.