Meeting of the Working Group on Romanization Systems
United Nations, Vienna

Thursday 5th May 2011

1. Discussion of various romanization systems:

a.) Arabic It was reported that a modified system for the romanization of Arabic had been agreed within the Arabic Division in 2007. This system incorporated a number of changes to the 1972 Beirut system and was likely to be presented to the 10th UNCSGN in 2012. Special characteristics of the various countries of the Arabic Division had been taken into account in the compilation of this system. Details would be forwarded to WG members for perusal. The Convenor reminded members that evidence of a system’s application on national cartographic products was necessary before the system could be adopted by the United Nations. The benefit of the resulting Roman characters being easy to reproduce in a digital format was also restated.

b.) Georgian The experts from Georgia reported that a new system for the romanization of Georgian had been officially adopted by Presidential decree earlier in 2011. The new system had been developed to ensure a standardised form of geographical names in Roman script and was seen as a practical and technically accurate and reversible solution which generated internet-friendly spellings. The success of the system’s implementation was to be monitored, with a potential view to adopting the system for personal names. Geographical names were currently being converted from their existing Russian forms into Georgian. The Convenor encouraged the experts from Georgia to provide examples of the system’s implementation.

c.) Persian The experts from Iran described a national system of transcription for Persian which had originally been presented to the 23rd UNGEGN in 2006. This system had been applied to names in Iran’s national geographical names database, although not yet to names on printed maps. The system would be submitted to the WG for consideration. The current UN system was based on transliteration which the WG considered to be technically preferable. Whilst the merits of transcription were in its simplicity, transliteration was more easily reversible and preserved better the integrity of the name and its original script form. It was possible that experts from Iran would be considering a second system based on transliteration in the future.

d.) Ukrainian The WG had already agreed in principle to adopt Ukraine’s official system for the romanization of Ukrainian. A slightly amended system had been approved by Ukraine in January 2010 which had incorporated several changes to that approved in 1996, and had also brought in line all systems used for the romanization of proper names in Ukraine. It was anticipated that this amended system would be presented to the 10th UNCSGN in 2012. Charts and online maps containing names in the new system were already available. The Working Group discussed the difficulties of producing a system yielding spellings which could be regarded as “language-neutral”. The absence of diacritical marks on the resulting romanizations was considered to make application of the system more straightforward, although a perceived orientation towards the English language was not favoured by all Working Group members.

e.) other languages/scripts The Working Group noted that a new system for Bulgarian had been implemented nationally. New or amended systems for Byelorussian, Korean and Maldivian were also in work and were potential candidates for future Working Group consideration.


The launch of internationalised and extended domain names had reinforced the need for close cooperation between ICANN and the Working Group. ICANN had recently launched a project to study variant and dialectal spellings of names for which linguistic assistance was being sought. Teams had thus far been set up to examine inter alia Arabic, Chinese, Cyrillic, Devanagari and Greek. A case study looking at the implications of including diacritical marks on Roman-script names was also in place.

3. WGRS and the Unicode Transliteration Guidelines

As part of its Common Locale Data Repository (CLDR) project, Unicode was creating automatic script conversion models. Transliteration rules were currently available for most of the UN romanization systems and the WG was encouraged to let the Unicode Technical Committee know of any newly-approved systems or changes to existing systems. Problems of text encoding were being encountered by some WG members, although this could be due to font deficiencies. The representative of Unicode noted that newer operating systems incorporated greatly improved font packages and several specialist linguistic fonts were also available.

4. Preparation for the 10th UNCSGN

Preparation for the next Conference had been covered under previous agenda items.

5. Membership

New members were encouraged to join the WG. Dr Gerd Quinting, a former long-serving Convenor of the Working Group had announced his retirement from WG activities. The WG recorded its grateful thanks to Dr Quinting for his contributions and support over the years.

6. Any other business

The Convenor would continue to circulate WG-related items by email.