Languages can be romanized in a number of ways, as shown here with Mandarin Chinese
Romanization or romanisation, in linguistics, is the conversion of writing from a different writing system to the Roman (Latin) script, or a system for doing so. Methods of romanization include transliteration, for representing written text, and transcription, for representing the spoken word, and combinations of both. Transcription methods can be subdivided into phonemic transcription, which records the phonemes or units of semantic meaning in speech, and more strict phonetic transcription, which records speech sounds with precision.
There are many consistent or standardized romanization systems. They can be classified by their characteristics. A particular system’s characteristics may make it better-suited for various, sometimes contradictory applications, including document retrieval, linguistic analysis, easy readability, faithful representation of pronunciation.
Source, or donor language – A system may be tailored to romanize text from a particular language, or a series of languages, or for any language in a particular writing system. A language-specific system typically preserves language features like pronunciation, while the general one may be better for cataloguing international texts.
Target, or receiver language – Most systems are intended for an audience that speaks or reads a particular language. (So-called international romanization systems for Cyrillic text are based on central-European alphabets like the Czech and Croatian alphabet.)
Simplicity – Since the basic Latin alphabet has a smaller number of letters than many other writing systems, digraphs, diacritics, or special characters must be used to represent them all in Latin script. This affects the ease of creation, digital storage and transmission, reproduction, and reading of the romanized text.
Reversibility – Whether or not the original can be restored from the converted text. Some reversible systems allow for an irreversible simplified version.
If the romanization attempts to transliterate the original script, the guiding principle is a one-to-one mapping of characters in the source language into the target script, with less emphasis on how the result sounds when pronounced according to the reader's language. For example, the Nihon-shiki romanization of Japanese allows the informed reader to reconstruct the original Japanese kana syllables with 100% accuracy, but requires additional knowledge for correct pronunciation.
Most romanizations are intended to enable the casual reader who is unfamiliar with the original script to pronounce the source language reasonably accurately. Such romanizations follow the principle of phonemic transcription and attempt to render the significant sounds (phonemes) of the original as faithfully as possible in the target language. The popular Hepburn romanization of Japanese is an example of a transcriptive romanization designed for English speakers.
A phonetic conversion goes one step further and attempts to depict all phones in the source language, sacrificing legibility if necessary by using characters or conventions not found in the target script. In practice such a representation almost never tries to represent every possible allophone—especially those that occur naturally due to coarticulation effects—and instead limits itself to the most significant allophonic distinctions. The International Phonetic Alphabet is the most common system of phonetic transcription.
For most language pairs, building a usable romanization involves tradeoffs between the two extremes. Pure transcriptions are generally not possible, as the source language usually contains sounds and distinctions not found in the target language, but which must be shown for the romanized form to be comprehensible. Furthermore, due to diachronic and synchronic variance no written language represents any spoken language with perfect accuracy and the vocal interpretation of a script may vary by a great degree among languages. In modern times the chain of transcription is usually spoken foreign language, written foreign language, written native language, spoken (read) native language. Reducing the number of those processes, i.e. removing one or both steps of writing, usually leads to more accurate oral articulations. In general, outside a limited audience of scholars, romanizations tend to lean more towards transcription. As an example, consider the Japanese martial art 柔術: the Nihon-shiki romanization zyûzyutu may allow someone who knows Japanese to reconstruct the kana syllables じゅうじゅつ, but most native English speakers or rather readers would find it easier to guess the pronunciation from the Hepburn version, jūjutsu.
The Brahmic family of abugidas is used for languages of the Indian subcontinent and south-east Asia. There is a long tradition in the west to study Sanskrit and other Indic texts in Latin transliteration. Various transliteration conventions have been used for Indic scripts since the time of Sir William Jones.
ISO 15919 (2001): A standard transliteration convention was codified in the ISO 15919 standard. It uses diacritics to map the much larger set of Brahmic consonants and vowels to the Latin script. See also Transliteration of Indic scripts: how to use ISO 15919. The Devanagari-specific portion is very similar to the academic standard, IAST: "International Alphabet of Sanskrit Transliteration", and to the United States Library of Congress standard, ALA-LC, although there are a few differences
In Pakistan: Standard (Saaf or Khaalis) Urdu is the "high" variety, whereas Hindustani is the "low" variety used by the masses (called Urdu, written in nastaʿlīq script).
In India, both Standard (Shuddh) Hindi and Standard (Saaf or Khaalis) Urdu are the "H" varieties (written in devanagari and nastaʿlīq respectively), whereas Hindustani is the "L" variety used by the masses and written in either devanagari or nastaʿlīq (and called 'Hindi' or 'Urdu' respectively).
The digraphia renders any work in either script largely inaccessible to users of the other script, though otherwise Hindustani is a perfectly mutually intelligible language, essentially meaning that any kind of text-based open source collaboration is impossible among devanagari and nastaʿlīq readers.
Initiated in 2011, the Hamari Boli Initiative is a full-scale open-source language planning initiative aimed at Hindustani script, style, status & lexical reform and modernization. One of primary stated objectives of Hamari Boli is to relieve Hindustani of the crippling devanagari–nastaʿlīq digraphia by way of romanization.
Romanization of the Sinitic languages, particularly Mandarin, has proved a very difficult problem, although the issue is further complicated by political considerations. Because of this, many romanization tables contain Chinese characters plus one or more romanizations or Zhuyin.
Hanyu Pinyin (1958): In mainland China, Hanyu Pinyin has been used officially to romanize Mandarin for decades, primarily as a linguistic tool for teaching the standardized language. The system is also used in other Chinese-speaking areas such as Singapore and parts of Taiwan, and has been adopted by much of the international community as a standard for writing Chinese words and names in the Latin script. The value of Hanyu Pinyin in education in China lies in the fact that China, like any other populated area with comparable area and population, has numerous distinct dialects, though there is just one common written language and one common standardized spoken form. (These comments apply to Romanization in general)
Pe̍h-ōe-jī (POJ), once the de facto official script of the Presbyterian Church in Taiwan (since the late 19th century). Technically this represented a largely phonemic transcription system, as Min Nan was not commonly written in Chinese.
While romanization has taken various and at times seemingly unstructured forms, some sets of rules do exist:
McCune–Reischauer (MR; 1937?), the first transcription to gain some acceptance. A slightly changed version of MR was the official system for Korean in South Korea from 1984 to 2000, and yet a different modification is still the official system in North Korea. Uses breves, apostrophes and diereses, the latter two indicating orthographic syllable boundaries in cases that would otherwise be ambiguous. What is called MR may in many cases be any of a number of systems that differ from each other and from the original MR mostly in whether word endings are separated from the stem by a space, a hyphen or – according to McCune's and Reischauer's system – not at all; and if a hyphen or space is used, whether sound change is reflected in a stem's last and an ending's first consonant letter (e.g. pur-i vs. pul-i). Although mostly irrelevant when transcribing uninflected words, these aberrations are so widespread that any mention of "McCune-Reischauer romanization" may not necessarily refer to the original system as published in the 1930s.
There is, for example, the ALA-LC / U.S. Library of Congress system, based on MR but with some deviations. Word division is addressed in detail, with a generous use of spaces to separate word endings from stems that is not seen in MR. Syllables of given names are always separated with a hyphen, which is expressly never done by MR. Sound changes are ignored more often than in MR. Distinguishes between ‘ and ’.
Several problems with MR led to the development of the newer systems:
Yale (1942): This system has become the established standard romanization for Korean among linguists. Vowel length in old or dialectal pronunciation is indicated by a macron. In cases that would otherwise be ambiguous, orthographic syllable boundaries are indicated with a period. Indicates disappearance of consonants.
Revised Romanization of Korean (RR; 2000): Includes rules both for transcription and for transliteration. South Korea now officially uses this system that was approved in 2000. Road signs and textbooks were required to follow these rules as soon as possible, at a cost estimated by the government to be at least US$20 million. All road signs, names of railway and subway stations on line maps and signs etc. have been changed. The change has been either ignored or grandfathered in some cases, notably the romanization of names and existing companies. RR is generally similar to MR, but uses no diacritics or apostrophes, and uses distinct letters for ㅌ/ㄷ (t/d), ㅋ/ㄱ (k/g), ㅊ/ㅈ (ch/j) and ㅍ/ㅂ (p/b). In cases of ambiguity, orthographic syllable boundaries were intended to be indicated with a hyphen, but this is inconsistently applied in practice.
ISO/TR 11941 (1996): This actually is two different standards under one name: one for North Korea (DPRK) and the other for South Korea (ROK). The initial submission to the ISO was based heavily on Yale and was a joint effort between both states, but they could not agree on the final draft.
Lukoff romanization, developed 1945–47 for his Spoken Korean coursebooks
When Spain colonised the Philippines in the late 16th century, the numerous languages of the Philippines were written in various scripts, such as Baybayin. These were initially promoted by the colonists but later replaced by Spanish transcriptions, which are still evident in place names and surnames. Letters such as C, Ll, and Ñ were considered Hispanic additions and removed in the Abakada, an attempt at a more indigenous alphabet devised by Lope K. Santos in 1940. These were eventually superseded by the Pilipino Alphabet and by the 28-letter, Modern Filipino Alphabet, which adds Ñ and the native Ng to the standard, 26-letter Latin alphabet.
While the Spanish language itself uses a very phonemic spelling, the romanised spelling created for Philippine languages is even more so. For example, the Spanish caballo ([kaˈβa.ʎo], "horse"), the same word in Tagalog is kabayo (demonstrating yeismo in the pronunciation of the Spanish "Ll" digraph).
A system based on scientific transliteration and ISO/R 9:1968 was considered official in Bulgaria since the 1970s. Since the late 1990s, Bulgarian authorities have switched to the so-called Streamlined System avoiding the use of diacritics and optimized for compatibility with English. This system became mandatory for public use with a law passed in 2009. Where the old system uses <č,š,ž,št,c,j,ă>, the new system uses <ch,sh,zh,sht,ts,y,a>.
The new Bulgarian system was endorsed for official use also by UN in 2012, and by BGN and PCGN in 2013.
There is no single universally accepted system of writing Russian using the Latin script—in fact there are a huge number of such systems: some are adjusted for a particular target language (e.g. German or French), some are designed as a librarian's transliteration, some are prescribed for Russian travellers' passports; the transcription of some names is purely traditional. All this has resulted in great reduplication of names. E.g. the name of the Russian composer Tchaikovsky may also be written as Tchaykovsky, Tchajkovskij, Tchaikowski, Tschaikowski, Czajkowski, Čajkovskij, Čajkovski, Chajkovskij, Çaykovski, Chaykovsky, Chaykovskiy, Chaikovski, Tshaikovski, Tšaikovski, Tsjajkovskij etc. Systems include:
BGN/PCGN (1947): Transliteration system (United States Board on Geographic Names & Permanent Committee on Geographical Names for British Official Use).
GOST 16876-71 (1971): A now defunct Soviet transliteration standard. Replaced by GOST 7.79, which is an ISO 9 equivalent.
Ukrainian personal names are usually transcribed phonetically; see the main article section Conventional romanization of proper names. The Ukrainian National system is used for geographic names in Ukraine.
The chart below shows the most common phonemic transcription romanization used for several different alphabets. While it is sufficient for many casual users, there are multiple alternatives used for each alphabet, and many exceptions. For details, consult each of the language sections above. (Hangul characters are broken down into jamo components.)
^The News International - Dec 29, 2011 -- "Hamari Boli (our language) is perhaps one of the very first serious undertakings to explore, develop and encourage the growth of Roman script in the use of Urdu/Hindi language"