REPORT ON THE CURRENT STATUS OF
UNITED NATIONS ROMANIZATION SYSTEMS
FOR GEOGRAPHICAL NAMES

Compiled by the UNGEGN Working Group on Romanization Systems
Version 2.2. January 2003

Report Contents

INTRODUCTION

The report is an updated version of a document presented to the 20th Session of the United Nations Group of Experts on Geographical Names (UNGEGN, New York, 1728 January 2000, Working Paper No. 34 and 34/Corr.). The overall aim of the report is to clarify the present status and the degree of implementation of the United Nations romanization systems used for converting geographical names from other writing systems than Roman. Since 1967, the United Nations conferences on the standardization of geographical names have adopted resolutions recommending romanization systems that cover 28 languages and/or scripts: Amharic, Arabic, Assamese, Bengali, Bulgarian, Chinese, Greek, Gujarati, Hebrew, Hindi, Kannada, Khmer, Macedonian Cyrillic, Malayalam, Marathi, Mongolian (in China), Nepali, Oriya, Persian, Punjabi, Russian, Serbian, Tamil, Telugu, Thai, Tibetan, Uighur, Urdu. Not all of them have been implemented on equal level, both nationally and internationally.

It has been agreed from the beginning that one of the general aims of international standardization is "to arrive at an agreement on a single romanization system, based on scientific principles, from each non-Roman alphabet or script, for international application" (Resolution 9 of the first UN Conference on the Standardization of Geographical Names, 1967). The Fourth United Nations Conference on the Standardization of Geographical Names (1982) also adopted Resolution 15 which called for guidelines governing the consideration of romanization systems. The resolution stated that

  1. the new romanization systems for international use should be considered only on condition that the sponsoring nations implement such systems on their cartographical products;
  2. States should refrain from revising systems previously adopted for international use.

In the reports of the UNGEGN Working Group on Romanization Systems and elsewhere it has been emphasized that the process of adoption of a romanization system by the UN involves the following prerequisites.

  1. Sufficient time is required for appropriate consultations and an expression of all views on technical matters between the sponsoring country and the WG members.
  2. Specifically, the group considers whether the romanization system is based on sound scientific principles, the system's degree of reversibility, and the extent of its implementation on cartographic products (maps and charts) by the proposing country.

There are many aspects that one should consider when adopting new romanization systems for international use. These concern both the nature of the system itself, practical considerations and expectations for its future usage. The romanization systems would benefit from the following.

  1. The system should be reversible, i.e., it should be possible to reconstruct the original non-Roman script form on the basis of romanization.
  2. The system should be as simple and clear to use as possible. For example, if a character table is not adequate, the notes to the table should clarify all aspects of usage. It would not be to the advantage of the system if dictionaries and other sources would need to be consulted to obtain a correct romanization of names, or if the rules would allow for variations in the romanization of the same original script form.
  3. The romanized name forms should be as easy as possible to write, read, memorize and also store electronically. That would mean minimal use of diacritical marks, avoiding difficult and unusual character sequences, etc. A systematic representation of phonological features is also to be recommended.

If the above conditions should be irreconcilable as is sometimes the case, it is the practical aspects that are the most decisive. For writing systems which would not allow for practicable and easy-to-use reversible romanization systems, it is often necessary to adopt two different romanization methods. The transcription method is the one recommended for wide usage, including the romanization of geographical names which is the area of interest for the United Nations. The second method, transliteration, could be applied for rendering the original script forms, e.g., in bibliography. Such systems fall into the sphere of activities of the International Organization for Standardization (ISO). For one such language there may exist two totally differing systems. For example, for Thai there is the UN system 1967/2002 and ISO 11940:1998. Or these two can be combined into one integral system, as, for example, for Greek for which exists the UN system 1987 (ELOT 743) containing two versions of romanization: one for general purposes and the other for documentation, i.e., a fully reversible one. Reversibility is achieved by adding diacritical marks in case there could be ambiguities.

A new romanization system is typically submitted by the proposing country in form of a resolution to the next full United Nations Conference on the Standardization of Geographical Names. When the resolution is adopted, the subject romanization system becomes the United Nations standard.

The present report consists of two main parts. The first part lists languages with non-Roman writing systems and the romanization systems used to render geographical names from these languages.

The second part of the report presents an overview of romanization systems for each of the languages. A language section usually consists of an introduction that gives the origin of the UN system, describes the implementation of the system both nationally and internationally and briefly characterizes the system itself regarding the ease of applicability, reversibility, possible ambiguities and other problems.

The romanization system is presented with notes on its application if appropriate. While trying to preserve accurately the content of the original tables published in the UN materials, an attempt has been made to harmonize and consolidate their presentation. Where possible, all non-Roman characters are listed in the order which is traditional for the given language. Examples of romanization and details on pronunciation have been left out of the tables in this version as it is difficult to find sources for all the presented languages. Examples may be found in supplementary notes.

Briefly listed and compared with the main system are also other systems of romanization.

Updates

Major changes between Version 1.3 (March 2000) and Version 2.1 (June 2002).

  1. The general layout of the report was improved, some of the fonts (esp. vowel marks for the Perso-Arabic script) were replaced with more legible ones.
  2. For reference purposes all characters in romanization tables were numbered.
  3. New romanization systems were provided for Byelorussian, Georgian, Korean and Tigrinya.
  4. Substantial new information was added in the sections for Armenian and Maldivian.
  5. Turkmen and Uzbek sections were moved to Annex: Languages that have recently adopted Roman alphabets.

Changes between Version 2.1 (June 2002) and Version 2.2 (January 2003), following the Eighth United Nations Conference on the Standardization of Geographical Names in Berlin (2002).

  1. Romanization system for Thai was updated according to Resolution VIII/13.
  2. The term Serbo-Croatian Cyrillic was changed into Serbian Cyrillic according to Resolution VIII/14.
  3. A new national provisional romanization system for Dzongkha of 1997 was introduced in the report, replacing the system of 1994.
  4. Information on the proposed modifications to the UN recommended system of romanization for Arabic was included.
  5. Details of the national provisional romanization system for Khmer (1995) were provided.
  6. Information on the romanization of Burmese was updated and notes added to the table of abbreviated vowel characters. Note 5 was rephrased, some other examples added.
  7. References to new documents submitted to the Eighth UN Conference were included in the sections for Arabic and Korean.