Meeting of the UNGEGN Working Group on Romanization Systems
Gdańsk, Poland, 18 May 2012
Comments by Mr. Woodman, 28 May 2012
Transcription procedure for Iranian toponymic items
Paper presented to the UNGEGN Working Group on Romanization Systems, Gdańsk, 18 May 2012
My understanding of the Iranian position
Although Persian and Arabic share the same script, they are nevertheless very different languages from different language families. The Persian language wishes to express its distinctive individuality when seen in romanized form, and so it is necessary that toponyms in romanized Persian should differ visually in style from toponyms in romanized Arabic. As a result, the current UN system for Persian, based on the same visual style as the UN system for Arabic, is inappropriate.
Thoughts on the proposed transcription procedure
I was very sceptical about this proposal when it was first mooted in 2011. It was (and to some extent remains) very much orientated towards the spoken form of a toponym, whereas the UN conventionally regards the written form as paramount. But the work subsequently undertaken by the UNGEGN Romanization WG, culminating in the convenor’s letter to Iran dated 18 April 2012, and the results of the excellent meeting held in Tehran the following week (Minutes dated 25 April 2012), have combined to improve the proposed system significantly.
General observations
- The proposed system would need to be much more widely implemented than we have seen so far, if it is to be considered as a UN romanization system.
- The proposed system is still essentially based on pronunciation. But many users have a genuine and legitimate need to know the properly written form of a toponym. Hence Iran would need to ensure free access to the toponymic database where these proper forms are stored.
Technical observations
- Looking at the transcription table, I am unsure why the letter combination <iy> is required in romanizations such as <Hājiyābād> and <'Aliyābād>. It seems to me that, according to the table, the letter <i> should be sufficient in such circumstances: < Hājiābād > and <'Aliābād>.
- Looking at the transcription table, it might be useful under Consonant 32 to add a note saying that this consonant is ignored when not pronounced. Otherwise the romanization example <Xormiz>, shown later in the table, is not easily understandable. (I realise that this silent <vāv> is explained in the introduction to the table, but it might be useful to have a note within the table as well).
- I have a general question about the intention to represent pronunciation. In the preceding bullet point, we see that <vāv> is not to be represented when it is silent. But what about other silences, such as the <ah> in <čahār> (the numeral ‘four’), which is usually pronounced simply as <čār>? Should the <Čāhārmahāl> on the new country divisions map in fact be <Čārmahāl> – or is this spelling considered just too far removed from the written form?
- I am puzzled by the representation of the conjunction ‘and’. The transcription table (Note 3) suggests that this should only be <-o> following numerals greater than 20 (on other occasions being a separate word <va>). But my understanding is that the colloquial pronunciation of this conjunction is usually <o> rather than <va>, and perhaps this is borne out by the fact that two provinces are romanized on the new country divisions map by use of the <-o> option, even though in neither instance does the conjunction follow a numeral greater than 20: <Čāhārmahāl-o Baxtiyāri> and <Sistān-o Balučestān>. So I am wondering whether Note 3 of the table should be amended to say that <-o> is in fact the usual representation of the conjunction ‘and’?
Comments by Iranian experts, 30 May 2012
About your understanding of the Iranian position
We also agree that the current UN system for Persian, based on the same visual style as the UN system for Arabic, is inappropriate.
About thoughts on the proposed transcription procedure
It is all right.
About general observations
- Tendency to pronunciation is an unavoidable fact. Now, many disapproved romanization systems, based on transcription, are implemented in our country, e.g. in geographical encyclopedias, books, etc. We are very much expanding the use of a unique system in the whole country – this proposed system.
- The original Persian written forms for geographical names are available in the maps and databases for all users.
About technical observations
- Iranian linguists who work on pronunciation believe that two vowel sounds should not occur consecutively. Therefore, the letter <y> was inserted between the two vowel sounds to prevent the clash of them. However, this may be ignored, for the reason that the letter <y> whether written or not, will be pronounced. Therefore, it can be eliminated.
- This subject has been already clarified by Iranian experts. The letter <و> may have more than one sound, such as /v/, /o/, or /u/, and on the other hand the sound /o/ is rendered as <ُ> and <و>. Finally, in the Persian pronunciation, there is nothing like silent <vāv> (و) or vowel <vāv>. Silent <vāv> is an incorrect term that may be created due to consideration of only written form. There is no need to add a note saying that "this consonant is ignored when not pronounced".
- The pronunciation <Čār> is used in colloquial Persian, whereas in the official Persian it should be pronounced as <Čāhār>. Also for the name of Tehran city, we should use <Tehrān> instead of <Tehrun>. The official Persian should be used for maps, databases, etc.
- The Persian letter <vāv> when used to link components of a place name, is romanized as -o after consonants. When after vowels, there are many exceptions depending on the type of the vowel. For example, شهرکي و ناروئي is romanized /Šahraki-yo Nāru’i/, مانه و سملقان /Māne va Samalqān/.