Korean Romanization Systems and Why They Frustrate Everyone

Core examples: 부산/Busan/Pusan; 김포/Gimpo/Kimpo; 대구/Daegu; 서울/Seoul; 박/Park; 이/Lee/Yi.

Romanization is a tool, not Korean

Korean learners often meet the same place or name in several spellings: Busan and Pusan, Gimpo and Kimpo, Daegu and Taegu, Seoul and Sŏul, Lee and Yi, Park and Bak. It can feel as if romanization is random.

It is not random. It is worse than random: it is historical.

Different spellings come from different systems, different periods, different institutions, and different personal-name practices. Some romanizations are designed for road signs and public use. Some are designed for linguistic analysis. Some preserve older spellings. Some are personal identity choices. Some are English-friendly compromises that no longer match a government system neatly.

The most important rule is this:

Romanization is not Hangul. It is an approximation made for a particular purpose.

A romanized spelling can help you search, travel, or identify a name. It should not replace Hangul in your learning.

Revised Romanization: the public default in South Korea

South Korea’s current public romanization standard is Revised Romanization. It generally avoids diacritics and special marks, which makes it easier for signs, maps, web addresses, and databases. Under this system, 부산 is Busan, 김포 is Gimpo, 대구 is Daegu, and 한글 is Hangeul.

The system is based on standard pronunciation rather than a simple one-letter-to-one-letter spelling code. That is a strength for public reading, but it also means learners should not reverse-engineer Hangul from romanization too confidently.

For example, the letter sequence g in a romanized Korean word does not mean English g in every environment. It is part of a Korean-specific mapping. Likewise eo represents ㅓ, not the English sequence in leopard or people. The romanization is a convention, not English pronunciation.

What the public standard actually asks you to do

The current South Korean public system has several practical consequences for learners. First, it is tied to standard pronunciation, not to a naive one-symbol-per-Hangul-letter code. Second, it avoids non-Roman symbols where possible, which is one reason public spellings like Busan and Daegu are easier to type in ordinary databases than mark-heavy systems. Third, vowel spellings such as eo, eu, oe, wi, and ui are conventions. They are not instructions to pronounce the letters as English.

Personal names require extra caution. Public romanization rules give a framework for names, including surname–given-name order and optional hyphenation inside given names, but actual passport and public identity spellings often preserve family or personal choices. That is why Lee, Park, Choi, Chung, Jung, and Kim must be treated as identity spellings, not simply as learner exercises.

McCune-Reischauer: older, precise, and mark-heavy

McCune-Reischauer is an older romanization system that many readers still encounter in academic books, older maps, library records, and historical materials. It uses breves and apostrophes to mark distinctions that Revised Romanization handles differently.

That is why 서울 may appear as Sŏul, and 부산 as Pusan in older materials. The spelling Pusan is not a typo. It reflects a system and a period.

McCune-Reischauer can be useful for specialists, but it frustrates general users because marks are often dropped in plain text. Once the breves and apostrophes disappear, distinctions become harder to recover. A carefully written form and a lazy ASCII-only form are not equivalent.

Yale romanization: useful for linguists, strange for travelers

Yale romanization is common in linguistic work because it represents Korean morphology and sound structure in a way that is useful for analysis. It is not designed to be a tourism spelling system.

A learner who sees Yale forms in a grammar paper should not expect them to match subway signs. Yale answers a different question: how can we represent Korean forms systematically for analysis? Public romanization answers a practical signage question. Passport names answer an identity and administrative question.

Do not force one system to do another system’s job.

Passport names are their own problem

Personal names are the most frustrating area because they are not governed only by romanization tables. Families often preserve spellings that are familiar, prestigious, or long established: Lee for 이, Park for 박, Kim for 김, Choi for 최, Chung or Jung for 정, and so on.

Under a strict system, 박 might be rendered differently from the familiar Park. But a person’s passport spelling is part of their identity in international contexts. Do not “correct” it just because you know another romanization system.

For names, use this rule:

If identity matters, use the person’s own official spelling.

You can note the Hangul form separately: 박지훈 / Park Ji-hoon, 이서준 / Lee Seo-jun, 김민지 / Kim Min-ji. But the romanized form itself may follow personal, family, or institutional practice.

Why reverse conversion is dangerous

Romanization loses information. That is true even in the best systems.

If you see Kim, the Hangul is probably 김, but you still need context. If you see Lee, the Hangul may be 이, but the romanized spelling does not show the initial silent ㅇ. If you see Park, the Hangul is usually 박, but the spelling contains an English r that does not correspond to ㄹ in the Korean form.

Place names are usually safer because they have official forms. Personal names are not. Brand names are even less predictable.

Romanization also hides spacing and morphology. Hanguk University might represent 한국대학교, 한국 대학, or a stylized institutional English name. You need the Korean source.

Romanization and pronunciation traps

English speakers often read romanization through English spelling habits. This is why eo, eu, ae, oe, and ui cause trouble.

The romanized spelling is only a cue. Once you know Hangul, return to Hangul.

Where romanization appears in real life

Romanization is not just for learners. It appears on subway signs, airport displays, maps, passports, business cards, academic citations, tourism sites, menus, product labels, and search results. Each domain has its own pressures.

A subway sign needs consistency and quick recognition. A passport needs legal identity continuity. An academic citation may preserve a published spelling. A restaurant may use an English-friendly brand spelling. A map may contain legacy data.

When spellings disagree, ask what domain produced them.

A romanization workflow

Use this routine when you meet a romanized Korean form:

1. Identify the domain: map, passport, academic text, sign, brand, or learner material.
2. Decide whether it is likely Revised Romanization, McCune-Reischauer, Yale, or a personal spelling.
3. Convert back to Hangul cautiously, not automatically.
4. Search or verify the official Hangul form when accuracy matters.
5. Once you know the Hangul, study and remember the Hangul.
6. Keep the romanized form only for the domain where it matters.

Mini practice: do not treat spellings as random

Suggested functions:

1. Input Hangul: output Revised Romanization, McCune-Reischauer, and Yale.
2. System labels: explain which system is useful for signs, scholarship, or linguistic analysis.
3. Name warning: flag surnames and personal names as identity-sensitive.
4. Reverse caution mode: show multiple possible Hangul forms where romanization is ambiguous.
5. Map mode: compare older and newer spellings for place names.

Final rule

Romanization helps you get to Korean. It should not become the Korean you learn.

Use romanization to search, travel, cite, and compare systems. Use Hangul to read, pronounce, spell, and understand Korean itself.