Author:
Alexei Savchenko
Search for other papers by Alexei Savchenko in
Current site
Google Scholar
PubMed
Close
Free access

A book discussing, among other matters, the original sounding of Sogdian place names communicated by a speaker of Persian to an Arab traveller in the Turkic corner of Mawarannahr inevitably has a large number of transliterated words in different languages. Arabic, Persian and Sogdian have enjoyed a long tradition of academic study, so text in those languages is rendered according to the accepted conventions. For Syriac, I follow the Eastern a-pronunciation; for Russian, the transliteration used follows the PCGN conventions; the few Chinese words are written with Pinyin.

Compiling the index of geographical names, I encountered a problem with standardising Uzbek place names in the Romanised alphabet. That particularly applies to the spelling principles regarding the common Turkic ö and γ, formerly expressed by ў and ғ in Extended Cyrillic. In the first version of the new alphabet adopted in 1993, the letters for those sounds became ö and ğ. In 1995, they were replaced by o‘ and g‘, supposed to be handwritten as õ and . In 2018, it was proposed to change o‘ to ŏ and g‘ to ğ. One year later, another change was announced: ŏ to ó and ğ to ǵ, and a new reform is currently being planned. In reality, none of those standards has ever been in effect since the additional symbols are missing from the standard computer keyboard. In practice, a surrogate of apostrophe has been used, producing o’ and g’ absent from European orthographies and thus hard to identify with any familiar sounds.

To dispense with deceptive diacritics, I decided to use u for both the mid ö and close u. A few homographs, inevitable in any alphabet-based writing system, seem to be a lesser evil than a changed spelling of Uzbek and its derivatives, which would be too radical an innovation. As for the γ, there is hardly a better solution than gh since Arabists have long used this digraph for the same sound.

Uzbek orthography does not differentiate between the front and back allophones of i, prescribing i in all cases. I will distinguish between these audibly distinct sounds by rendering them as i and y, respectively.

Also, I will ignore the rounding of a to o in modern literary Uzbek, recorded in the official spelling. That Persian-influenced feature, peculiar to the ‘upmarket’ dialects of Tashkent and Ferghana, was codified as standard in 1937; earlier, both the Latin and Cyrillic spellings followed the ‘inurbane’ a-pronunciation. Most of the archival materials I quote go back to before 1937, so the older spelling standard occurs more frequently. I decided to stay with it in all cases to avoid an excess of cross-referenced index entries (‘Andijan: v. Andijon’). I keep the o in Tajik place names (Sar-i Hisor), where it is a matter of course.

With the help of this set of conventions, I was able to put in order the toponymic material covered by this short book, except for Kazakhstan and Kyrgyzstan formalised by the United Nations. Under the rules above, they would be Qazaqstan and Qyrghyzstan.

To make the text readable while remaining within the academic genre, I have chosen to use diacritical marks for direct quotations and keep them to a minimum elsewhere. Original spelling is always retained in citations.

  • Collapse
  • Expand