Korean vs Japanese

Structurally alike to a striking degree, yet with almost no shared native vocabulary and no accepted family relationship.

Last updated 2026-08-01

Are Korean and Japanese similar?

Structurally alike to a striking degree, yet with almost no shared native vocabulary and no accepted family relationship.

The grammatical parallels are real: both are subject-object-verb, both mark grammatical roles with particles attached to nouns, and both have elaborate honorific systems that change verb forms according to social relationship. Learners of one often find the other's structure familiar.

Vocabulary tells a different story. Native Korean and Japanese words are largely unrelated. What overlap exists comes from both having borrowed extensively from Chinese, so many formal and technical terms are recognisably cognate โ€” while everyday words are not.

The writing systems are unrelated in design. Korean uses Hangul, a featural alphabet created in the 15th century and widely regarded as one of the most systematic writing systems ever devised. Japanese combines two syllabaries with Chinese characters.

Whether the two are genealogically related remains unsettled; most linguists treat Korean as an isolate and Japanese as Japonic, with the similarities attributed to long contact rather than shared descent.

Side by side

KoreanJapanese
Writing systemHangulJapanese kana and kanji
Reading directionLeft to rightLeft to right
Language familyUnclassifiedJaponic
Word ordersubject-object-verb (SOV)subject-object-verb (SOV)
SpeakersApproximately 81 million speakers worldwide.Approximately 128 million speakers, predominantly in Japan.
  • โœ— Different writing systems โ€” neither can be read without learning the other's script.
  • โœ— Different language families โ€” few cognates to rely on.
  • โœ“ Same word order โ€” sentences keep their shape between them.

The same phrases in both

Nothing shows how close two languages are like seeing identical meaning rendered in each. These are independent translations of the same English source.

Greetings

KoreanJapanese
์•ˆ๋…•ํ•˜์„ธ์š”ใ“ใ‚“ใซใกใฏ
์ข‹์€ ์•„์นจใŠใฏใ‚ˆใ†ใ”ใ–ใ„ใพใ™
์ž˜์ž์š”ใŠใ‚„ใ™ใฟ
์•ˆ๋…•ใ•ใ‚ˆใ†ใชใ‚‰
๋‚˜์ค‘์— ๋ด์š”ใพใŸไผšใ„ใพใ—ใ‚‡ใ†
๋‚ด์ผ ๋ด์š”ใพใŸๆ˜Žๆ—ฅ
ํ™˜์˜ํ•ฉ๋‹ˆ๋‹คใ‚ˆใ†ใ“ใ
์ž˜ ์ง€๋‚ด์„ธ์š”?่ชฟๅญใฏใฉใ†ใงใ™ใ‹๏ผŸ
๋‚œ ๊ดœ์ฐฎ์•„์š”, ๊ณ ๋งˆ์›Œ์š”็งใฏๅ…ƒๆฐ—ใงใ™ใ€ใ‚ใ‚ŠใŒใจใ†
์ข‹์€ ํ•˜๋ฃจ ๋ณด๋‚ด์„ธ์š”่‰ฏใ„ไธ€ๆ—ฅใ‚’ใŠ้Žใ”ใ—ใใ ใ•ใ„

Courtesy

KoreanJapanese
์ œ๋ฐœใŠ้ก˜ใ„ใ—ใพใ™
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹คใ‚ใ‚ŠใŒใจใ†
์ •๋ง ๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹คใ‚ใ‚ŠใŒใจใ†ใ”ใ–ใ„ใพใ—ใŸ
์ฒœ๋งŒ์—์š”ใฉใ†ใ„ใŸใ—ใพใ—ใฆ
์˜ˆใฏใ„
์•„๋‹ˆ์š”ใ„ใ„ใˆ
์‹ค๋ก€ํ•ฉ๋‹ˆ๋‹คใ™ใฟใพใ›ใ‚“
๋ฏธ์•ˆํ•ด์š”ใ”ใ‚ใ‚“ใชใ•ใ„
๋ฌธ์ œ์—†์–ดๅ•้กŒใ‚ใ‚Šใพใ›ใ‚“
๋ฌผ๋ก ์ด๋‹คใ‚‚ใกใ‚ใ‚“

Introductions

KoreanJapanese
๋‹น์‹ ์˜ ์ด๋ฆ„์€ ๋ฌด์—‡์ž…๋‹ˆ๊นŒ?ใ‚ใชใŸใฎๅๅ‰ใฏไฝ•ใงใ™ใ‹๏ผŸ
๋งŒ๋‚˜์„œ ๋ฐ˜๊ฐ€์›Œ์š”ใ‚ˆใ‚ใ—ใใŠ้ก˜ใ„ใ—ใพใ™
๋‹น์‹ ์€ ์–ด๋””์„œ ์™”๋‚˜์š”?ใฉใ“ใฎๅ‡บ่บซใงใ™ใ‹๏ผŸ
๋ช‡ ์‚ด์ด์—์š”?ไฝ•ๆญณใงใ™ใ‹๏ผŸ
๋‚˜๋Š” ํ•™์ƒ์ด๋‹ค็งใฏๅญฆ็”Ÿใงใ™
๋‚˜๋Š” ํœด์ผ์— ์—ฌ๊ธฐ ์žˆ์–ด์š”็งใฏไผ‘ๆš‡ใงใ“ใ“ใซใ„ใพใ™
๋‚˜๋Š” ์ผ ๋•Œ๋ฌธ์— ์—ฌ๊ธฐ์— ์™”์–ด์š”็งใฏไป•ไบ‹ใฎใŸใ‚ใซใ“ใ“ใซใ„ใพใ™
์ด ์‚ฌ๋žŒ์€ ๋‚ด ์นœ๊ตฌ์•ผใ“ใ‚Œใฏ็งใฎๅ‹้”ใงใ™

Numbers

KoreanJapanese
ํ•˜๋‚˜1 ใค
2๊ฐœ2
์„ธ3
4๊ฐœ4
5๊ฐœ5
์—ฌ์„ฏ6
์„ธ๋ธใ‚ปใƒ–ใƒณ
์—ฌ๋Ÿ8
๋‚˜์ธใƒŠใ‚คใƒณ
1010

See all 116 phrases and translate Korean to Japanese โ†’

In practice

Grammar knowledge transfers unusually well between them. Vocabulary does not, outside Chinese-derived terms.

Other comparisons

Method and limitations

The structural rows โ€” writing system, family, word order, speaker numbers โ€” are taken from each language's own entry in our language data. The assessment of mutual intelligibility is editorial: it reflects the general consensus, and intelligibility always varies by speaker, dialect and exposure.

The phrase comparison uses the same 116 seed phrases translated into every language we support, so the two columns are independent translations of identical English source text rather than one translated through the other. Those translations are machine generated; the sample skips any row where one side repeats an earlier row, since a collapsed translation shows nothing useful.