Note
This romanisation guide is not entirely finished and might change at any moment, be it partially or entirely.
Last update: 2026-01-21
Core Romanisation Style#↑
This style guide defines how to consistently romanise and style Mandarin Chinese titles (manhua) for metadata purposes. The goal is to provide a standard that is easy to read and type while preserving the tonal nature of the language through Hanyu Pinyin.
The Basis#↑
We follow standard Hanyu Pinyin with the following specific enhancements and deviations:
- Tonal Diacritics: We strictly use tone marks (diacritics) for all syllables to ensure accurate identification (e.g., Mā vs Má vs Mǎ vs Mà) ― major deviation.
- Loanword Restoration: Translate clear phonetic loanwords into their English/foreign spelling (e.g., 巴士 / Bāshì → "Bus", 粉丝 / Fěnsī → "Fans") instead of strict pinyin ― major deviation.
- The "Zhi" Separation: The character 之 (Zhī) is treated as a delimiter (colon) when separating a main title from a subtitle.
- Agglutination: We fuse compund words (polysyllabic words) according to standard orthography (e.g., Dàxué and not Dà Xué), but keep grammatical particles separate.
Capitalisation#↑
As we are romanising titles, we use the Chicago Headline Style:
- Capitalise all nouns, verbs, adjectives, adverbs, pronouns, and auxiliary verbs.
- Keep all particles (e.g., de, le, zhe, ma), prepositions (e.g., zài, cóng), and copulas (e.g., shì) in lowercase.
Word Spacing, Fusions, and Hyphenation#↑
Chinese characters (Hanzi) represent syllables, but Pinyin represents words (Ci). We standardise boundaries as follows:
- Compounds: Fuse standard dictionary compounds. If a word is a single concept composed of multiple characters, it is one word (e.g., Péngyǒu for Friend, not Péng Yǒu).
- Names: Surname First (Capitalised) + Given Name (Capitalised and Fused). Unlike Korean, Chinese given names are fused without a hyphen (e.g., Máo Zédōng).
- Erhua (R-coloring): The suffix 儿 (-r) is fused to the preceding syllable (e.g., Huār not Huā Er).
- Separation: Separate words that do not form a tight compound. Subject-Verb-Object structures are always separated.
- Apostrophes: Use an apostrophe strictly to disambiguate syllable boundaries when a syllable starting with a, o, or e follows another syllable (e.g., Xī'ān vs Xiān).
Tone Sandhi
We romanise using the original dictionary tone, not the sandhi (changed) tone.
- Nǐ hǎo (你好) not Ní hǎo (even though it is pronounced with a rising tone on the first syllable).
- Yī (一) and Bù (不) retain their dictionary tones (yī, bù) regardless of the tone of the following word, to ensure search consistency.
Actual Title Showcasing the Style#↑
重生之黑客系统的逆袭 ~在异世界当骇客,我的粉丝全是僵尸~
Chóngshēng: Hacker Xìtǒng de Nìxí ~Zài Yìshìjiè Dāng Hacker, Wǒ de Fans Quán shì Jiāngshī~
- Loanwords: 黑客 (Hēikè) / 骇客 (Hàikè) → Hacker. 粉丝 (Fěnsī) → Fans.
- Native Nouns: Xìtǒng (System), Yìshìjiè (Otherworld), Jiāngshī (Zombie) are kept in Pinyin as they are standard native/Sino compounds or not direct phonetic loans.
- The "Zhi" Rule: Chóngshēng (Rebirth) is followed by Zhi acting as a separator, so it becomes a colon.
- Particles: de (possessive) is separate and lowercase.
- Prepositions/Copula: zài (at/in) and shì (is) are separate and lowercase.
- Compounds: Nìxí (Counterattack) is fused.
Further Clarification#↑
Loanword Restoration Rule#↑
Chinese often absorbs foreign words phonetically. To improve readability, we restore these to their original foreign counterparts if they are direct phonetic transliterations.
- 巴士 (Bāshì) → Bus
- 沙发 (Shāfā) → Sofa
- 巧克力 (Qiǎokèlì) → Chocolate
- 朋克 (Péngkè) → Punk
- S级 (S-jí) → S-jí (or S-Rank if "Rank" is implied contextually).
Semantic Translations vs Phonetic Loans
We do not translate words that are semantic translations (translating the meaning rather than the sound).
- 电脑 (Diànnǎo - Electric Brain) → Diànnǎo (Keep Pinyin, do not write Computer).
- 手机 (Shǒujī - Hand Machine) → Shǒujī (Keep Pinyin, do not write Mobile).
- Only restore words where the Chinese characters were chosen specifically to mimic the English/foreign sound (Phonetic Loans).
The Particle "De" (的 / 地 / 得)#↑
The natural tone particle de is the most common grammatical marker. It is always separated and lowercase.
- 红色跑车 (Hóngsè Pǎochē) → Hóngsè Pǎochē (Red Sports Car - no particle)
- 红色的跑车 (Hóngsè de Pǎochē) → Hóngsè de Pǎochē (Red-colored Sports Car)
- 慢慢地走 (Mànmàn de Zǒu) → Mànmàn de Zǒu (Walk slowly - adverbial marker)
- 跑得快 (Pǎo de Kuài) → Pǎo de Kuài (Run fast - complement marker)
Tones on Particles
While often pronounced neutral, for metadata consistency, if a particle has a dictionary tone variant, prefer the neutral (no mark) for grammatical particles listed in the table below to reduce keyboard friction, unless distinct emphasis is required.
Copula (Is/Am/Are)#↑
The verb shì (是) functions as the copula. While syntactically a verb, for Title Styling purposes to align with our Japanese/Korean guidelines, we treat it as a functional connector.
- Lowercase shì when it acts as a simple distinct equating verb in a title.
- Fuse and Capitalise if it is part of a lexicalised compound (e.g., Wèishénme - Why).
Hyphenation and Special Characters#↑
Personal Names#↑
Surname + Given Name
- 李小龙 → Lǐ Xiǎolóng
- 司马光 → Sīmǎ Guāng (Compound Surname Sima is fused).
Honorifics and Prefixes#↑
Chinese uses prefixes/suffixes for familiarity. Hyphenate these to clarify they are not part of the name proper.
- 老王 (Lǎo Wáng) → Lǎo-Wáng (Old Wang)
- 小猫 (Xiǎo Māo) → Xiǎo-Māo (Little Mao)
- 这里的黎明静悄悄 → Zhèlǐ de Límíng Jìngqiāoqiāo (No honorifics here, just standard fusion).
Numbers and Counters#↑
Use Arabic numerals. Hyphenate the counter.
- 第101次 → Dì 101-cì (The 101st Time)
- 3个人 → 3-gè Rén (3 People)
The Character "Zhī" (之)#↑
As a Separator: If it appears after the initial noun phrase to introduce the subtitle/description, replace it with a Colon (:).
- System Zhī Wáng → System: Wáng
As a Possessive: If it appears inside a phrase meaning "of", write as zhī (lowercase). - Sānfēn zhī Yī (One third) → Sānfēn zhī Yī
Detailed Tables#↑
Common Particles & Functional Words#↑
These words should generally be lowercase and separated.
| Pinyin | Chinese | Function | Notes |
|---|---|---|---|
| de | 的 / 地 / 得 | Possessive / Adverbial / Complement | Always neutral tone in this context. |
| le | 了 | Aspect particle (Completion) | |
| zhe | 着 | Aspect particle (Continuous) | |
| guò | 过 | Aspect particle (Experience) | Write with tone guò or neutral. |
| men | 们 | Plural marker | Usually fused to pronouns (Wǒmen), separate for nouns (Rén men). |
| ba | 吧 | Modal particle (Suggestion) | |
| ma | 吗 | Question particle | |
| ne | 呢 | Query particle | |
| bèi | 被 | Passive marker | |
| gěi | 给 | Preposition (To/For) | |
| zài | 在 | Preposition (At/In) | |
| hé / yǔ | 和 / 与 | Conjunction (And) | |
| shì | 是 | Copula (Is/Am/Are) | Lowercase in titles. |
| wéi | 为 | Copula (Act as/Become) | Lowercase in titles. |
Suffixes (Treated as Loanwords/Restoration)#↑
| Chinese | Pinyin Reading | Restoration | Condition |
|---|---|---|---|
| S级 | S-jí | S-Rank | If context implies ranking. |
| APP | A-P-P | App | |
| V | V | V |
LLM Ruleset and Prompts#↑
YAML Ruleset#↑
romanization_style:
name: "MangaBaka Chinese Romamization Style"
version: "1.0"
base_system: "Hanyu Pinyin (ISO 7098)"
tones:
usage: "obligatory"
type: "diacritics"
rule: "Use original dictionary tones (e.g., Nǐ hǎo), do not apply tone sandhi rules."
neutral_tone: "Do not mark neutral tones (e.g., de, le, zi)."
vowels: "Use ü for u-umlaut (e.g., nǚ)."
capitalization: >
Capitalize all Nouns, Verbs, Adjectives, Adverbs, Pronouns, and Auxiliaries.
Lowercase grammatical particles, prepositions, and copulas (shì, wéi).
word_segmentation:
compounds: "Fuse all lexicalized dictionary compounds (e.g., Dàxué, Shénme)."
phrases: "Separate words that are syntactically distinct (Subject Verb Object)."
names: "Surname (Cap) + GivenName (Cap+Fused). Example: Lǐ Xiǎolóng."
erhua: "Fuse the 'r' suffix to the stem (e.g., Huār)."
special_characters:
zhi_rule: >
If '之' functions as a subtitle delimiter, romanize as a colon (:).
If '之' functions as a possessive/grammatical particle, romanize as 'zhī' (lowercase).
apostrophe: "Use strictly before a, o, e syllables that follow another syllable (e.g., Xī'ān)."
loanwords: >
Restore English spelling ONLY if the Chinese word is a direct phonetic transliteration.
Examples: Bāshì -> Bus, Fěnsī -> Fans, Jiákè -> Jacket.
Do not restore semantic translations (e.g., Shǒujī stays Shǒujī, not Mobile).
particles: >
Always separate and lowercase: de, le, zhe, guo, ba, ma, ne, bei, gei, zai, he, yu.
Copulas (shi, wei) are lowercase.
hyphens: >
Use hyphens for:
- Honorific prefixes/suffixes (Lǎo-Wáng, Xiǎo-Lǐ).
- Arabic numerals + Counters (10-gè, Dì-1-cì).
example_validation:
input: "重生之在大巴上捡到了神级系统"
output: "Chóngshēng: Zài Bus shàng Jiǎndàole Shénjí Xìtǒng"
explanation: "Zhī becomes colon. Dàbā restored to Bus. Zài/shàng are prepositions/locatives (lowercase/separate). Shénjí (God-level) is fused."
Old Guidelines (Kept temporarily for internal reference. DO NOT USE)#↑
Overview#↑
This public guide explains how Mandarin Chinese titles are romanised and styled for display on MangaBaka. The goal is readable, frontend-friendly Pinyin with clear editorial rules that keep titles consistent across the site.
Canonical system: Hanyu Pinyin with tone marks for display, with editorial fallbacks for clarity.
Title field priorities#↑
- Native Title: the original Chinese characters as shown on the cover or source (Simplified or Traditional).
- Romanized Title: the Pinyin shown to users (with tone marks where applicable).
Pinyin display rules#↑
- Use diacritics (tone marks) for display: e.g., 鹤报刀法,我一刀一个 → Hébào Dāofǎ, Wǒ Yì Dāo Yí Gè.
- Use ü for the vowel ü in display (e.g., nǚ).
- Use an apostrophe to disambiguate syllable boundaries when necessary (Xi'an).
- Capitalize content words in title-style capitalization; keep grammatical particles (de, di, le, zhe, ba, ne, ma, ye) lowercased when they are functional.
之 as subtitle divider#↑
When 之 is used as a subtitle divider, it is romanised as a colon (':'), but when it is used as a possessive, it has to be romanised as zhī.
Loanword Rule#↑
- When a Chinese title clearly uses a loanword and there is a common real-world/English term, prefer the real-world term in the romanisation or as the visible English phrase.
Example: if the title references a specific English concept such as "Rhythm Game," present that term visibly while also keeping pinyin for the native characters.
- Maintain the original Chinese title in the native field.
Names & capitalization#↑
- Follow Chinese name order (surname first) for metadata unless an established English name exists. In that case, include the common English form as an alternative display title.
- Capitalize main content words in display pinyin; keep small grammatical particles lowercased.
Numbers, dates, counters#↑
- Use Arabic numerals for clarity and sorting: 第四部 → Di 4-bù (display may include tone marks where useful).
- Hyphens for counters are acceptable to align with cross-language conventions: Di 3-bù: ....
Quotation & punctuation#↑
- Decorative Chinese quotation marks in the native title are retained in the native field. For pinyin display, use standard ASCII quotes.
- Use apostrophes in pinyin display where they clarify syllable boundaries (e.g., Xī'ān).
Examples#↑
- Native: 鹤报刀法,我一刀一个
Romanization: Hébào Dāofǎ, Wǒ Yì Dāo Yí Gè
- Native: 西安事变
Romanization: Xī'ān Shìbiàn
- Native: 女
Romanization: nǚ
Notes for editors#↑
- Do not adopt publisher romanisation as the default; publishers often use inconsistent romanisations.