Building a Tri-Language Kanji/Hanzi/Hanja Cognate Map

Core examples: 学校, 銀行, 医学, 法律, 経済, 政府, 社会, 電話, 研究, 文化, 権利, 安全.

A cognate list is not a map

A learner collects words:

学校 銀行 医学 法律 経済 政府 社会 電話 研究 文化 権利 安全

They notice that Chinese and Korean have related forms. This is useful, but a list alone quickly becomes messy. Which forms are simplified or traditional? What is the Japanese reading? What is the Mandarin pronunciation? What is the Korean Hangul form? Are the meanings identical? Is the word formal in one language and everyday in another? Is there a false-friend risk?

The key principle is:

A CJK cognate map must record differences as carefully as similarities.

The goal is not to prove the languages are “the same.” The goal is to turn shared character roots into a disciplined learning tool.

What a tri-language map needs

A useful entry should include:

1. Japanese form.
2. Japanese reading.
3. Mandarin form and pinyin.
4. Korean Hangul form.
5. Hanja if relevant.
6. English/semantic gloss.
7. Meaning overlap score.
8. Register/domain.
9. False-friend risk.
10. Example sentence in Japanese.
11. Notes on form differences.

A weak map says:

学校 = school = 학교

A strong map says:

Japanese: 学校, がっこう Korean: 학교, Hanja 學校 Mandarin: 学校/學校, xuéxiào Meaning: school; strong overlap Register: ordinary/institutional Japanese collocations: 学校に通う, 学校教育, 小学校

Form, sound, meaning, register

Every cognate has at least four layers.

A good cognate map prevents you from confusing these layers.

Domain clusters

Cognate mapping works best by domain.

School/education:

学校 大学 研究

Law/politics:

法律 政府 権利

Economy/business:

経済 銀行 会社

Culture/society:

社会 文化 安全

Domain grouping helps because cognates often cluster where Chinese-character vocabulary historically dominated: scholarship, law, religion, administration, science, and public life.

False-friend risk marking

Not every cognate is safe. Mark entries as:

• strong overlap,
• partial overlap,
• register mismatch,
• false friend,
• form mismatch,
• pronunciation trap,
• context-dependent.

Example:

手紙

This should be marked high-risk for Japanese-Mandarin comparison.

A cognate map without risk labels creates false confidence.

Example bank walkthrough

学校

Strong overlap across CJK.

Learner action: add readings and collocations.

銀行

Bank.

Learner action: strong institutional cognate; reading differs.

医学

Medicine as academic field.

Learner action: domain-specific; not every “medical” phrase uses 医学.

法律

Law.

Learner action: legal terms need jurisdiction notes.

経済

Economy.

Learner action: strong formal/news cognate.

政府

Government.

Learner action: strong political cognate.

社会

Society.

Learner action: modern concept word; note philosophical/policy uses.

電話

Telephone/phone.

Learner action: common cognate, but usage collocations differ.

研究

Research.

Learner action: Japanese 研究する, 研究者, 研究室.

文化

Culture.

Learner action: strong academic/public vocabulary.

権利

Rights.

Learner action: legal/political concept; jurisdiction matters.

安全

Safety.

Learner action: strong overlap, common in signs and policy.

Cognate-map template

For each entry, fill:

Japanese: Reading: Chinese simplified/traditional: Pinyin: Korean Hangul: Hanja: Meaning overlap: strong / partial / weak Domain: Register: False-friend risk: Japanese collocations: Example sentence: Notes:

Map design: do not make a flat spreadsheet

A good tri-language cognate map needs layers, not just columns.

Without register and example sentences, the map becomes a pretty trap.

Overlap score

Use a simple score:

• A: strong overlap — 学校, 文化, 政府.
• B: partial overlap — 先生, 料理, 安全 depending context.
• C: register mismatch — formal in one language, everyday in another.
• D: false friend — 手紙, 勉強 in Japanese-Mandarin comparison.
• E: high-stakes caution — legal/medical terms that may look shared but depend on system.

This lets the learner review dangerous items more often than easy ones.

Japanese-first output

Even in a tri-language map, the target article is Japanese. Every entry should end with Japanese usage:

Japanese collocation Japanese sentence Japanese register note

Otherwise the map becomes comparative linguistics without improving Japanese production.

A strong tool for this article would let users build structured cognate cards.

Suggested functions:

1. Cognate entry form.
2. Character-form comparison.
3. Japanese reading lookup.
4. Mandarin pinyin field.
5. Korean Hangul/Hanja field.
6. Overlap/risk labels.
7. Domain tags.
8. Example sentence storage.
9. Export to study deck.

Final rule

A cognate map is useful only if it maps differences.

Record forms, readings, meanings, register, domain, and false-friend risk. Shared characters are the start of comparison, not the end.

Cross-CJK learning rewards disciplined curiosity.