NOTHING GOOD EVER CAME OF IT.

I have mentioned Marat Akchurin’s wonderful Red Odyssey: A Journey Through the Soviet Republics before, and I thought I’d quote this passage from his visit to Tajikistan in 1990, as the whole Soviet mess was in the process of falling apart; it resonates with the material I’ve been posting from Terry Martin’s book:

We tried to pay the counterman for the green tea that we had drunk, but he refused to take money, saying that he considered us to be his guests.
“If you had an opportunity to address Americans, what would you tell them?” I asked him.
“Americans?” he asked again in surprise. “Let them learn Tadzhik. It’s a very simple and beautiful language. Maybe they will make use of it one day!”
Safar and I went out and decided to go to the bookstore and then walk to my hotel.
“Is Tadzhik very different from Farsi?” I asked Safar. “Are they just dialects of one language?”
“Tadzhik is Persian-Farsi transliterated with Russian letters,” Safar replied. “But nothing good ever came of it. They took away the old alphabet and thus cut the Tadzhik people off from their ancient history and culture. This monstrously sly Bolshevik act did terrible damage to the national culture of the Tadzhik people. Why? Because letters are culture-producing for a Tadzhik. Can you imagine Pushkin writing in Russian but with Arabic ligatures? That would be crazy, wouldn’t it? But this nightmarish experiment was conducted in the U.S.S.R. on many peoples, Tadzhiks among them. I believe that it was a cunning policy.”

“What’s so cunning about this policy, tell me!” I snorted. Many Soviet and Western intellectuals are keen on ascribing refined cunning and slyness to the Bolsheviks, although they most often were led by nothing more than ordinary cruelty that resulted from their own lack of culture and purely proletarian hatred for the cultures of other peoples that are incomprehensible to them.
“Why? The formal reason they gave was that the Arabic alphabet is difficult to learn. But as a former teacher of Persian-Farsi in the Moscow Literary Institute, I am entitled to say that my students—Russians, Latvians, Georgians—learned the alphabet in just two weeks. And this language wasn’t native for any of them. Why then is it more difficult for Tadzhiks, whose ancestors were using this alphabet for ten generations? No, all this talk about the Arabic alphabet being too difficult for Tadzhiks is a blatant lie. So it turned out that in just seventy years Tadzhiks have lost their letters, their cultural legacy, and their cities.”
“What about Dushanbe? Or Leninabad? Are you going to give it back its ancient name of Khodzhend?”
“In Dushanbe Russians make up the majority of the population. In Leninabad Uzbeks are the most numerous ethnic group. As far as the restoration of its historical name is concerned, it’s true that the people are demanding the return of the former name. When it was renamed into yet another ‘Lenin’s city,’ for that’s what Leninabad means in Tadzhik, it was done on the pretext that it was ‘by request of the working people.’ In fact, as you know, no working people requested it. The Bolsheviks just impertinently renamed Khodzhend Leninabad and Dushanbe Stalinabad. Well, Stalin was dumped, but so far we can’t do anything with Lenin. The party functionaries stand firm on this point.”

Of course, Khojand was renamed the following year (and most of the Russians fled the new country).

Comments

  1. David Marjanović says:

    It’s interesting in this respect that Tajikistan still hasn’t returned to the Arabic alphabet and shows no signs of ever doing it.

  2. It still looks like a paradox to me that democratic and nationalist forces formed a close alliance in the last years of the USSR. In 1992 the Islamic Renaissance party of Tajikistan united with the Democratic party of Tajikistan and who could think that things as innocent as alphabet and national culture might cause the bloodiest (and one of the least studied) war in the post-Soviet USSR? Perhaps, most of those who were killed there of fled would rather write in Cyrillic (as they still do).

  3. I suddenly feel cut off from my Phoenician roots. Why aren’t the Turks complaining ?

  4. Joseph Dart says:

    Probably the bureaucrats who have been writing з all their lives are in no hurry to go struggle with ﺽ vs. ﻅ vs. ﺫ vs. ﺯ …
    The plan to go back to the use of classical Mongolian script in Mongolia has similarly been stalled for over a decade.

  5. michael farris says:

    This sounds like the flip side of those that want to Romanize everything possible in the name of ‘progress’.
    I think in terms of writing systems, in practical terms, there are a few basic gradations.
    1. Not good enough
    2. Barely good enough
    3. Good enough
    Arabic script for Turkic was just barely good enough and something with vowels is far better – though the difference between Latin and Cyrillic is trivial changing from one to other is not worth the expense of the change absent very strong non-linguistic motivtions.
    I suspect the situation for Persian is similar though maybe the lack of most vowels doesn’t matter quite as much as for Turkic.
    Mongol script (and spelling) is barely good enough for modern varieties of Mongolian while Cyrillic is significantly better for Khalka(sp?) which is why the attempts to change back have mostly failed.
    My own idea is that while it’s possible (if difficult) to move a language’s orthography from 2 to 3 under some conditions, it’s harder to move from one 3 system to another and even harder to move back to 2 from 3.
    *full disclosure, I like to fiddle around with romanizations for various languages but that’s just for my own amusement

  6. the difference between Latin and Cyrillic is trivial changing from one to other is not worth the expense of the change absent very strong non-linguistic motivtions.
    For instance, for Russian it would be simpler to introduce modified, easier-on-the-eye Cyrillic fonts than to romanize.

  7. michael farris says:

    Yeah, I do think an internationally recognized romanization of Russian for academic purposes would be a useful thing to have, like pinyin for Mandarin. But a widescale switch from Cyrillic to Latin is not gonna happen because …. why? Russian speakers neither need nor want such a change.
    Bulgarian fontmakers have been working more with tweaking fonts and adding more descenders and ascenders, maybe Russian fontmakers should/could pay some attention.

  8. Cyrillic is a fine alphabet for Persian. I don’t anticipate Iran ever adopting it, though, which is unfortunate for the Tajiks—there were plans to start teaching Perso-Arabic script in school in Tajikistan, beside Cyrillic, but I don’t know if anything ever came of it.

  9. David Marjanović says:

    Yeah, I do think an internationally recognized romanization of Russian for academic purposes would be a useful thing to have, like pinyin for Mandarin.

    Absolutely, yes.
    Wikipedia article on the Tajik alphabets
    Wikipedia in Tajik – note the switch for automatic transcription to lotinī.

  10. Arabic script was barely acceptable for Ottoman Turkish, but only because the Perso-Arabic loanwords were so thick on the ground that Ottoman poets could go on for lines without using a single word of native origin. For modern Turkish it would be utterly unacceptable. Cyrillic might have been better technically, but certainly would have been the wrong alignment for Atatürk’s purposes.

  11. Just a note to say I loved this post, and the excerpt you quoted. Thanks.

  12. There is an internationally recognised romanisation of Cyrillic for academic purposes, with a standard and everything: ISO 9. It’s difficult to input it on the average Latin-alphabet keyboard, though.

  13. Yet another misguided attempt to use separate symbols for everything, creating monstrosities like “a diaeresis and dot below” and “z caron and cedilla.” Why do people consider it unscientific to use two-letter combinations?

  14. I believe the Uzbek Latin alphabet uses a two letter combination, either “sh” or “ch” (can’t remember which).
    I must say that as a speaker of both Tajik and Farsi I find the Cyrillic substantially easier than Arabic script. Part of that is to do with the fact that Cyrillic is nearer to Latin script, but it’s also to do with the fact that in Cyrillic: all the vowels are written; the ezafe is written; and that all the redundant letters have been combined into one.
    There’s no doubt that Tajiks lose a great deal culturally by not using the Arabic script, but in terms of ease of use, Cyrillic is the clear choice.
    Also, if you talk to Tajiks about their alphabet they will tell you there are plans to switch to Arabic script sometime in the nebulous “future”. Children take, I believe, 3 years of Farsi in school, so they are getting exposed to it on some level. However, I doubt a switch will ever happen as long as the current government is in power.

  15. Oh, and I’ll also add that Tajiki is not “Persian-Farsi transliterated with Russian letters”. It’s Persian-Farsi transliterated with Russian letters, plus a million Russian and Uzbek loan words, plus a lots of strange Persian words that are rarely (if ever) used in Iran, plus some (relatively small) grammatical differences.

  16. michael farris says:

    Problems with ISO9 for academic purposes.
    I can’t say I’ve ever seen it used.
    It’s a transliteration rather than a transcription and when you put it together you get some unfortunate combintions. I like â and û after consonants okay, menâ looks as good as menja (and much better than menya) but ja and ju would look better word initially or afeter a vowel. Compare
    moskovskaâ and moskovskaja
    There are other issues too but combinations like aâ are a deal breaker for me.

  17. I don’t know. One argument against digraphs is that with them you then end up having to use diaereses to write stuff like spermatozoön (or coöperation, to use a word more people have to write on a regular basis.)
    The Cyrillic approach to Morse code comes close to avoiding your monstrosities, and almost avoiding digraphs, at the expense of a very ugly transliteration: Hruqew for the leader of the 1960s, anyone?

  18. you then end up having to use diaereses
    No I don’t.
    Cooperation. There.

  19. Cooperation

    The abstract state of becoming a barrel-maker?
    Back on topic, perhaps all of Persian would be served by a return to a resurrected Avestan script, modified and modernized as needed. It couldn’t hurt the sense of cultural heritage.

  20. marie-lucie says:

    Digraphs: there is nothing wrong with digraphs for a practical spelling within the community, but they are inconvenient for international purposes (including phonetic transcription) since they vary from one language to another, and even within a language: words which had ch in Old French, pronounced as in English, gradually switched to the pronunciation sh in French, though not in English.
    Also, in setting up a spelling system one has to be careful that common digraphs are not ambiguous (eg if the sound [f] is written “ph”, that hiphop will not be pronounded as ‘hiffop’ or ‘heyefop’, for instance).
    Cooperation, etc: I always write it this way, and it seems to be much more common than coöperation or co-operation. Nobody confuses it wish “cooperage”. I find “noöne” for “no one” really weird, since the second o is pronounced [wa].

  21. they are inconvenient for international purposes (including phonetic transcription) since they vary from one language to another
    But single letters also vary from one language to another. Just as c, say, is defined as /ts/ for international purposes, so (say) sx (which does not occur normally) could be defined as /š/. There is no scientific/rational reason not to do it; it just doesn’t “look neat” or whatever. People would rather use weird symbols virtually impossible to reproduce.

  22. marie-lucie says:

    LH, It is true that single letters can also be misleading, but it is easier to deal with a single letter and convert it to one other letter than with a bunch of them which have to be taken as a unit. Consider trying to figure out how to read the name of Nietzsche where the five letters are needed where Italian would need only one (c)!
    The main problem with digraphs or multigraphs for practical international use (as opposed to professional use by phoneticians) is that sequences are unpredictable from one language to the next. For instance, you think that “sx” (is this “sks” or “sX”, where the X is as in Greek?) does not occur normally, but the first is perfectly normal in English (risks, desks) and “sX” is quite normal in some other languages; so is a sequence [s.h] (which is not English “sh”, but a sequence of [s] followed by h).
    I don’t think phoneticians are trying to mislead the public, or hide behind forbidding symbols – most linguists trying to set up or modify an alphabet for practical use by hitherto illiterate people are doing the best they can to make spelling as straightforward as possible for the particular languages they are dealing with, not trying to complicate matters unduly for the upcoming readers and writers.

  23. For instance, you think that “sx” (is this “sks” or “sX”, where the X is as in Greek?) does not occur normally, but the first is perfectly normal in English (risks, desks) and “sX” is quite normal in some other languages; so is a sequence [s.h] (which is not English “sh”, but a sequence of [s] followed by h).
    You’re misunderstanding me. I’m talking about “sx” as a purely graphic symbol. I chose it because as a graphic symbol it does not normally occur in written languages, so it is a good candidate for a conventional digraph. Obviously “sh” would not work for /š/, because it does occur in written languages. Similarly, “cx” could be used for /č/, etc. The particular symbols are not important; what is important is that you would end up with an international system that could be written by a normal person on a normal keyboard.
    I don’t think phoneticians are trying to mislead the public, or hide behind forbidding symbols
    I don’t think they’re trying to, but intent is irrelevant here: the systems they create are in fact forbidding, misleading, and virtually impossible to use unless you’re an initiate, because they simply don’t care about practical matters. I say that with confidence, because I have seen the results. Nobody with an ounce of practicality or good sense would create the “phonetic alphabets” that exist.

  24. David Marjanović says:

    Yet another misguided attempt to use separate symbols for everything, creating monstrosities like “a diaeresis and dot below” and “z caron and cedilla.” Why do people consider it unscientific to use two-letter combinations?

    It’s not just any transcription, it’s a transliteration. The idea behind a transliteration is that you can represent the original letters 1 : 1 and therefore reconstruct the original 1 : 1 from the transliteration – without having any idea about the language. That’s among the explicit reasons Acta Palaeontologica Polonica (pdf!) gives in its instructions to authors for using ISO 9:1995.
    (I don’t know of another science journal that has an explicit rule on this, however.)

    I believe the Uzbek Latin alphabet uses a two letter combination, either “sh” or “ch” (can’t remember which).

    Those times are over. Don’t bother remembering, look it up on Wikipedia! :-)

  25. The idea behind a transliteration is that you can represent the original letters 1 : 1 and therefore reconstruct the original 1 : 1 from the transliteration – without having any idea about the language. (emphasis added by GS)
    Confirming my worst, previously withheld suspicions !

  26. I remember being puzzled by reading the word straphangers as a child &mdash not the meaning, I knew well enough that it meant ‘subway riders’ — but by whether the pronunciation was /strəˈfeɪn(d)ʒərz / or /ˈstræfən(d)ʒərz/. Finally I asked my mother, and she told me that it was simply strap-hangers, people who hang from a strap.

  27. marie-lucie says:

    The idea behind a transliteration is that you can represent the original letters 1 : 1 and therefore reconstruct the original 1 : 1 from the transliteration – without having any idea about the language.
    If you are a Western librarian having to classify books in Russian or Armenian, for instance, it helps to have a system of transliteration which will allow you to write the title and author with letters you and the readers can recognize. If there is an equal number of symbols in the two versions, it is easier to match the words and to spot errors if any. With a syllabic script such as Japanese or Ethiopian, the transliteration similarly would have two letters per character. Of course, it would be desirable to have multilingual librarians or clerks, but in the absence of enough qualified persons, transliteration is a handy tool for those limited uses.

  28. Yes, marie-lucie, but David wrote “reconstruct the original 1:1 from the transliteration”. I was assuming that he too meant phonetic quality, but perhaps I’m wrong about that, because he also wrote

    represent the original letters 1 : 1 and therefore reconstruct the original

    as if he were talking only about changing fonts, so to speak: replacing marks by marks. But in languages with an alphabetic writing system, “letter” and its equivalents are ambiguous: the word can refer to “just a mark”, or to “a mark representing pronunciation”.
    Above, you yourself write as if you were thinking of a notation for sounds:

    in setting up a spelling system one has to be careful that common digraphs are not ambiguous (eg if the sound [f] is written “ph”, that hiphop will not be pronounded as ‘hiffop’ or ‘heyefop’, for instance).
    Consider trying to figure out how to read the name of Nietzsche where the five letters are needed where Italian would need only one (c)!

    The assumption that phonetic quality is being referred to here seems natural, because my layman’s understanding of “alphabetic writing system” is that it is a “phonetic writing system”. Even Mandarin Chinese was mentioned above by michael farris, thus strengthening my belief that people here are talking about a notation system for sounds, not only one for changing fonts:

    an internationally recognized romanization of Russian for academic purposes would be a useful thing to have, like pinyin for Mandarin.

    Take the case of Russian, for which there are various font-changing systems for “rendering” in English (I’m avoiding the word “transliteration” here). If these were merely font-changing systems, it would be hard to imagine why there are several of them. My explanation is that these systems are not intended “merely” to change fonts, but also to approximate pronunciation.
    In one of these Russian system for “rendering into” the Latin alphabet (call it a]), that lenis b-looking-thing becomes a single quote. In another one (call it b]), it has no equivalent mark. In both cases pronunceability is being taken into account. a] uses the single quote to indicate “here is something to pronounce, although you may not be familiar with the pronunciation of a single quote at the end of a word”. b] drops the lenis b-looking-thing, because it is concerned to render a text that is pronounceable by the phonetic-letter rules of the target language in question, even though the result is only an approximation to the Russian.
    In your example of Western librarians there can still be hitches. In older Spanish dictionaries the “ll” was treated as a letter in its own right that comes after “l” (doesn’t the RAE hold to that still ?). So you had to know that llave comes after lobo. An English librarian and a Spanish librarian, in a correspondence about classification and sorting systems, will not initially understand each other in the “ll” cases, until each learns that the sorting system of the other is based on slightly different interpretations of marks. The Spanish librarian takes “ll” to be a digraph, but the English one takes it to be two consecutive letters.
    The IPA is a whole-hog attempt to separate the “spelling mark” aspect from the “phoneme signifier” aspect, and is for that reason rather hard for people to learn who use a phonetic writing system that is part spelling convention, part pronunciation indicator. Maybe the Chinese would find the IPA easier to learn, in principle ?

  29. It’s interesting that alphabetic writing became conventionalized over the centuries (millenia ?). The notion of “what he means” gradually emerges and becomes ever more important, while the fiddly details of “what was the exact sound of what he said” become a special application of alphabetics, say when reproducing “dialect”. The exact-soundists ultimately split off and found the IPA, where “the exact sound of what he said” is all that counts, and supposedly “what he means” is irrelevant: thus David’s “without having any idea about the language”.

  30. founded the IPA

  31. David Marjanović says:

    With a syllabic script such as Japanese or Ethiopian, the transliteration similarly would have two letters per character.

    That wouldn’t be a transliteration. If a script isn’t (similar enough to) an alphabet, it can’t be transliterated, and some other kind of transcription is needed.

    In your example of Western librarians there can still be hitches.

    Yeah. The university library of Vienna treated J as I till 1972. <headdesk>

  32. Lots of linguists write the underlying phonological representation in IPA for a makeshift orthography of their field languages. So it’s not really sure if the exactists had won themselves a permanent bulwark.

  33. David Marjanović says:

    IPA is primarily meant for phonemic transcriptions. These are often very inexact from a phonetic point of view.

  34. A transliteration is a bidirectional mapping from one script to another. Japanese kana can certainly be transliterated into Latin, and there are several systems for doing so; etymology is not destiny. Chinese cannot be in practice either a source or a target of transliteration, simply because there are no comparably complex morpho-syllabic systems to pair it with; one could “transliterate” Chinese characters into their Unicode codepoints written down in European digits.
    If a single symbol X on one side of a transliteration is equated to multiple symbols YZ on the other side, it’s necessary to be careful that Y+Z can never occur as a result of transliterating two symbols, or the transliteration will not be bidirectional.
    A transcription is a unidirectional mapping from one written language to another. It generally uses the orthographic conventions of the target as much as possible, but may introduce conventions of its own, as in the use of “zh” in English-based Russian transcription.
    A romanization is a Latin-based orthography for a language, which may or may not correlate to any existing non-Latin orthography. There are corresponding terms cyrillicization and arabization, and more could be created as needed.

  35. michael farris says:

    “IPA is primarily meant for phonemic transcriptions. These are often very inexact from a phonetic point of view.”
    I’d say it’s the other way around, for me, the IPA is more a hindrance than help if I’m doing an initial field-like analysis.
    With IPA I’d be too concerned with matching symbol to sound (not so easy when ‘catch on the fly’ is the buzzword). In initial fieldwork, figuring out the oppositions and distinctive features and how they fit into a coherent system (which might differ quite a bit from the surface phonology) is more important than phonetic exactness, which is best left till later.
    For my own purposes, I would follow an Americanist tradition with common diacritics, that is I’d rather use š (until I have a good reason not to) than the elongated s of IPA.

  36. I’d rather use š (until I have a good reason not to) than the elongated s of IPA.
    Same here.

  37. David Marjanović says:

    Again, IPA sucks at phonetic exactness. It can be done, but often it requires multiple diacritics on the same symbol. Several sounds that are common the world over lack IPA symbols: for instance, there’s no symbol for the open unrounded central vowel, because there is apparently no language that has a phonemic opposition between that and the open unrounded back vowel ([ɑ]) – even though I’m sure more languages have the central one than the front one ([a]).

    For my own purposes, I would follow an Americanist tradition with common diacritics, that is I’d rather use š (until I have a good reason not to) than the elongated s of IPA.

    IPA is specifically designed to look pretty when printed. This goes so far that Greek-derived symbols are modified to better fit in with Latin letters: ɛ isn’t ε, ɣ isn’t γ, ɸ isn’t φ. It’s also where the convention comes from to stick, in phonemic transcriptions, as closely to the nearest symbols derived from Latin letters as possible: if a language has the five vowels [a ɛ i ɔ u] or [a ɛ ɪ ɔ ʊ], they’ll be written /a e i o u/. It’s also one reason why symbol-inherent diacritics are avoided for “basic” symbols (…what “basic” means depends largely on Standard Average European, of course).
    In contrast, the Americanist notation is designed to be easy to print with widely available typesets. This is why š and č are used and why the Greek-derived symbols are not modified.
    In terms of handwriting suckage, I guess they’re probably rather similar, but I have no experience with this whatsoever…

  38. Does the Americanist notation have the “open unrounded central vowel”? I wouldn’t know how that’s different from the front or back vowels.
    Is anybody developing a system for transcribing nonhuman sounds, such as animal sounds, bird calls or machinery?
    In court somebody might be asked to testify about what sound he/she heard and it might be useful to transcribe it instead of trying to verbally describe it.

  39. David Marjanović says:

    Does the Americanist notation have the “open unrounded central vowel”?

    AFAIK it uses a for this and consequently lacks a symbol for the front one.

    I wouldn’t know how that’s different from the front or back vowels.

    The back one (pharyngealized, though) occurs in English, the front one in French, the central one in Russian and Mandarin.

  40. marie-lucie says:

    In court somebody might be asked to testify about what sound he/she heard and it might be useful to transcribe it instead of trying to verbally describe it.
    There is now a specialty of “forensic linguistics”, since linguists are frequently called as expert witnesses in court (eg in cases of recorded telephone messages). With all the diacritics available together with the main symbols, it is possible to write down a large variety of sounds, especially vowels, One can also represent those vowels on a graph showing the approximate position of the tongue, etc while uttering them. A purely verbal description would be meaningless for most people apart from experienced phoneticians.
    A famous example is the Prinzivalli case in which the pioneer sociolinguist William Labov (known particularly for his studies of the varieties of New York speech) testified. Prinzivalli, an Easterner, was accused of having telephoned bomb threats to the LA airport, where he worked. Labov heard the tapes of the calls and recognized that the caller could not have been Prinzivalli, as one of the men had a New York accent and the other one a Boston accent (I forget which was which). Labov spent a day in court playing and analyzing tapes of the two voices and showing the court the difference in their vowel diagrams, and the judge threw out the charge against Prinzivalli without bothering to complete the trial.
    The back one (pharyngealized, though) occurs in English, the front one in French, the central one in Russian and Mandarin.
    The back one also exists in traditional French (eg what I speak, or what is spoken in Canada), together with the front one.

  41. David Marjanović says:

    Yes, sorry for my laziness.

Speak Your Mind

*