Meeting of the UNGEGN Working Group on Romanization Systems

Gdańsk, Poland, 18 May 2012


  1. Information on the status of various romanization systems:
  2. Preparations for the 10th UNCSGN (including any recommendations if necessary)
  3. Report of the Working Group to the 10th UNCSGN
  4. Cooperation with other bodies (ICANN, ISO TC 46/WG 3)
  5. Membership and convenorship
  6. Any other business

Present: Maria del Mar Batlle Gutiérrez (Spain, Catalonia), Herman Bell (United Kingdom), Pavel Boháč (Czech Republic), Ojārs Bušs (Latvia), Sungjae Choo (Republic of Korea), András Dutkó (Hungary), Khalil Jamshidi (Iran), Peter Jordan (Austria), Gholamreza Karimzadeh (Iran), Helen Kerfoot (Canada, UNGEGN Chair), Drago Kladnik (Slovenia), Mária Kováčová (Slovakia), Małgorzata Mandola (Poland), Phil Matthews (New Zealand), Peeter Päll (Estonia, WG’s convenor), Béla Pokoly (Hungary), Maryia Shcharbakova (Belarus), Paul J. Woodman (United Kingdom), Bogusław R. Zagórski (Poland), Maciej Zych (Poland).


The meeting of the working group took place in Gdańsk, Poland, in conjunction with the meeting of the UNGEGN Working Group on Exonyms and was hosted by the Polish Head Office of Geodesy and Cartography.

The convenor opened the meeting and drew the attention of the participants to three important UN resolutions: I/9 (A single romanization system), IV/15 (Guidelines for the consideration of romanization systems) and IX/8 (Implementation of romanization systems by sponsoring countries). He also reminded the audience on the importance of reversibility but acknowledged that only a few UN-approved systems were fully reversible (in fact, only Serbian qualified as 100 % reversible, all other systems had at least some ambiguous romanization equivalents).


A new proposal for a Standard Arabic System for Transliteration of Geographical Names from Arabic Alphabet to Latin Alphabet was presented at the 3rd Arab Conference on Geographical Names, Beirut, 30 May 2007, and adopted by the Arab experts who attended the meeting. There were some technical problems with some of the romanization equivalents, as explained in Technical Notes on Arabic Romanization (2007), noted by the WG. Currently there was no information if that proposal would be submitted to the 10th UN Conference on the Standardization of Geographical Names (UNCSGN) for approval.

Mr. Zagórski noted that the 2007 system did not fulfil scientific requirements and it was not applied by any country. Short vowels were not marked in Arabic script, thus one word could be transliterated in various ways. One of the main problems was that the official language, Modern Standard Arabic (MSA) was not used by the local population. There were large differences between MSA and the local Arab dialects (as large as between the Slavic languages). Mr. Zagórski considered preferable to join definite articles to a word with a hyphen (rather than a space) and not to capitalize articles within a name. He suggested not to accept the 2007 proposal before more information was supplied.


The convenor noted that discussions with the Armenian experts regarding their proposed romanization system were still ongoing and recently a new reply was received from Armenia. This helped to clarify many aspects of the romanization system but the issue of using apostrophes differently in the case of plosives and affricates still remained open.


The convenor noted that the current UN-approved system (1977) had diacritics. In 2006 a new system was proposed and submitted at the UN conference in 2007. Unfortunately nobody presented it and thus it could not be discussed. In the new system digraphs were applied, there were no diacritical marks. Mr. Zagórski noted that the new system was irreversible. Mr. Jordan proposed that an irreversible system should not replace a reversible one. The new system seemed to favour a particular language and its Roman script. Mr. Woodman stated that reversibility was a good criterion but not the only one, other aspects must also be considered, like implementation and stability (since 2006 when it was officially implemented), also user-friendliness. In Bulgarian Internet sites the names were already in this form. Stability was also an important aspect in emergencies. Mr. Zagórski asked if reversibility was a basic criterion in approving romanization systems for international use. The convenor replied that reversibility was recommended but it could not be strictly mandatory.

It was agreed that as Bulgarian and Ukrainian new systems shared many characteristics, this question would further be discussed later at the meeting.


Belarus had already presented a system of romanization at the 9th UNCSGN in New York in 2007. Ms. Shcharbakova informed the participants that the system was in use in Belarus in numerous books, road and tourist maps. Mr. Zagórski noted that the system was reversible and thus should be approved if presented at the 10th UNCSGN. There were no objections.


The convenor informed the participants that as the 2002 national system was not yet approved and there seemed to be objections to the use of the apostrophe, a new provisional system was being devised where letters with the apostrophe were replaced with digraphs. However, some of the romanization equivalents were not systematic in the new system. It was highly questionable if the new system would be better than the previous one. The convenor offered to send a letter to Georgian experts with these considerations. Mr. Zagórski added that the former system was more practical with media.


Prior to the meeting, Iran had proposed a new romanization system, based on broad transcription, used in the Iranian National Geographical Names Database. There had been fruitful discussions with the working group regarding some of the romanization equivalents, and as a result of this, Iran submitted a revised version of the system. Mr. Karimzadeh presented the romanization system at the meeting, stressing that as the Persian language was an Indo-European language, very different from Arabic, its script should also be romanized in a different way from Arabic. Some names were pronounced incorrectly based on the present romanization making them sound like Arabic names. The basic principle of the new system was that one character stands for one sound but knowledge of Persian would be important for the correct reading of names. Mr. Karimzadeh emphasized that the transliteration system had an artificial and complex character and it had not been used in modern Iranian maps and other cartographic products.

Mr. Zagórski stressed that reversibility was an essential requirement for a romanization system and therefore the current UN system was better than the proposed one. It was up to the Iranians to reform and simplify their own orthography, omitting duplicate characters for the same sound but a potential user of Iranian names would wish to know which characters were used in the original script. Mr. Karimzadeh noted that the original Persian spellings were available in the Iranian National Geographical Names Database. Before the 10th UNCSGN, a map of Iran would be published with the new romanized names. There was lengthy discussion on some of the aspects of transliteration, notably the rendering of the sound v in names where it is not pronounced. Mr. Woodman noted that, while the new system remained based on pronunciation, it had been significantly improved since its first submission to the working group. However, more implementation was crucial. Mr. Zych observed that it was not without precedent that two or three non-Roman characters were romanized by the same letter. Mr. Bell reminded that Ferdowsi had already 1000 years ago remarked that Persian was different from Arabic (full statement is given in Annex A).

It was agreed that comments on this proposed system would be sent via the convenor to the experts in Iran (for eventual comments and the reply see Annex B). The convenor suggested that Iran could provide additional information such as transliteration in its national geographical names database. Mr. Karimzadeh said that the old romanization could be kept in the database like other information, for example old names, other names, coordinates, etc. This seemed to be a rational option for other participants as well.


Ukraine had slightly revised its system of romanization in 2010, omitting the apostrophe, used for the representation of the soft sign. As a result, the system became less reversible. The convenor noted, however, that the system had been implemented well for many years. In particular, the Euro 2012 soccer championship had provided a good impetus for its wide implementation on street name signs and elsewhere. Mr. Jordan drew the attention of the participants to a statement agreed by the Permanent Committee on Geographical Names (StAGN) which stated that the German-speaking countries would favour a reversible system that would be consistent with other Slavic Roman orthographies. There was no need for an English-based system. Mr. Woodman stressed that implementation and practicality counted, as did proper liaison with the donor countries involved.

There followed a discussion on both Bulgarian and Ukrainian proposed systems. Mr. Zagórski added that to his knowledge the old system of Bulgarian romanization was still used by the local academy of sciences. The convenor cited the Bulgarian transliteration law of 2009 which prescribed the use of the new system to all institutions and contained even fines for failing to do this. As it was obvious that experts present could not agree on a joint position with regard to these two systems, it was proposed to continue the discussion at the informal meeting of the working group during the 10th UNCSGN.

Other languages

Mr. Choo informed the participants that the romanization system used in the Republic of Korea would be included in the toponymic guidelines for that country, submitted to the 10th UNCSGN. Participants took note of a document on the French transcription of African toponyms and proposed to discuss it with the Task Team for Africa. There were no further notices.


Regarding preparations for the 10th UNCSGN, the convenor noted that there were four likely romanization systems that could be put forward for approval: Belarusian, Bulgarian, Persian and Ukrainian. Summarizing the discussions, there were no objections to the proposed systems of romanization for Belarusian and Persian, in the latter case it was recommended to make the geographical names of Iran also available in transliteration, as the proposed system is not fully reversible. Also the importance of implementation was stressed. Concerning the romanization systems for Bulgarian and Ukrainian, there were differing opinions on the priority of criteria, required for the adoption of romanization systems, particularly on the importance of reversibility and of the system’s implementation in national cartographic products.


The reports of the working group to the 27th session of UNGEGN and the 10th UNCSGN were briefly discussed, and the convenor asked to send in corrections as soon as possible.


The convenor gave a brief account of contacts with ICANN, and the possible cooperation with ISO TC 46/WG 3. An annual meeting of the ISO TC 46 had just been held in Berlin and more information was to come.


The convenor said that new members to the working group would be welcome.


At the meeting it was decided to update the spelling of the language name, from Byelorussian to Belarusian, as used in the UN documents and in the document submitted by Belarus in 2007. However, the final form of the name will be determined in consultation with the other UNGEGN working groups and the geographical names authorities of Belarus (there were also other variants proposed, such as Belarussian or Belorussian).

The next meeting of the working group would be held during the 10th United Nations Conference on the Standardization of Geographical Names in July-August 2012.

The convenor thanked the hosts for excellent arrangements for the meeting.

- - -

Minutes based on notes taken by Mr. Béla Pokoly.

Annex A. A note sent by Mr. Bell.

Annex B. Comments by Mr. Woodman, 28 May 2012. Comments by Iranian experts, 30 May 2012.