Pronouncing Kiev / Kyiv
« previous post | next post »
The Wikipedia article on Kiev or Kyiv gives this as the pronunciation of the Ukrainian form Київ, transliterated as Kyiv:
And here's a lesson from Twitter:
An explanation of how to pronounce the Ukrainian capital, for all those who are confused today 🇺🇦 pic.twitter.com/f9Geu0h5gN
— Nina Jankowicz (@wiczipedia) November 13, 2019
Putting the Wikipedia pronunciation together with Nina Jankowicz's two versions, we get
The main vowel of this word is hard for outsiders to say — or at least for English speakers. But Wikipedia gives the IPA transcription [ˈkɪjiu̯], which will puzzle those who know that [ɪ] is the vowel in English words like kit and give. That should mean that the pronunciation of Kyiv is basically the start of kit followed by Eve (or maybe ye've) — which it clearly is not.
A 2017 article in the Journal of the International Phonetic Association (Bernd Pompino-Marschall et al., "Ukrainian") underlines but does not entirely explain the mystery:
The Ukrainian vowel system comprises the six phonemes /i ɪ ɛ a ɔ u/ (corresponding to <i и e a o y> in Ukrainian writing). Vowel length is not contrastive. The vowel <и> (as phonetically different from Russian [ɨ], see e.g. Shevelov 2002: 949; Carlton 1991: 280) would be more correctly transcribed as [ɪ̠], since in contrast to [i] the tongue is quite retracted and lowered in the production of this vowel. (In the acoustical vowel space it may seem shifted to [ɪ] due to its articulation with strongly spread lips, see the F1/F2 plot of the vowel realizations of our informant in Figure 1.) Zilyns'kyj (1979: 45) points to a great variation in the realization of this vowel in Western dialects.
Here is the F1/F2 plot that they reference:
In that plot, the position of [ɪ] (or [ɪ̠] with the underbar indicating retraction) is more or less what we'd expect for American English [ɪ], though the articulation and sound seem quite different, at least in the pronunciations of Kyiv included above.
This strikes me as yet another example of why the International Phonetic Alphabet, useful as it is, is fatally limited as a way of describing and analyzing speech sounds. Dr. Seuss, we need you now!
Update — in the comments, Anatoly Vorobey suggests
[C]ould the difference be in the fronting of the initial [k]? To me, the [k] in English "kit" sounds more fronted than the [k] in the Ukrainian Kyiv, and when I try to combine the [k] of "come" with the vowel of "kit", that's when it sounds much like Kyiv to me (not a native English speaker).
As I note in response, the non-fronted /k/ hypothesis seems promising — and it would be easy to test it with modern ultrasound apparatus.
But we almost never bother to notate the fronted /k/ before front vowels in English, e.g. "kit" as [k̟ɪt] — and the IPA as usual gives little support for notation of less-common distinctions (such as front vs. back velars), as noted in Daniel Siddiqi et al., The Routledge Handbook of North American Languages, 2019:
So there's still no good way to communicate via IPA a possible Ukrainian lack of velar fronting before /ɪ/.
Misha Schutt said,
November 16, 2019 @ 4:08 pm
My understanding (seldom having actually heard Ukrainian spoken) is that Ukrainian и is similar to Russian ы, that is, IPA [ɨ] or [ɯ], but lower. (I’ve never been clear on the distinction between [ɨ] and [ɯ] as high back unrounded vowels.) Also, in literary Ukrainian, syllable-final в is [w].
So the Ukrainian pronunciation of Киïв should be roughly [kɨˡjiw] or [kɯˡjiw] (unaspirated, unpalatalized k). The Russian phonology is more different than the spelling; the Russian pronunciation of Киев is [ˡkʲiıf] (the k is palatalized before i).
Peter B. Golden said,
November 16, 2019 @ 4:38 pm
The pronunciation varies. Western Ukrainians (at least those who are old friends of mine, all born in Галичина/Halychyna (Galicia) before WW II, tend to pronounce Київ more like Keyiw. И in Ukrainian pronunciation is not quite like ы in Russian. I have heard Ukrainian all of my life. My grandfather, a native of SE Belarus', often lived and worked in Ukraine, in the Dnipro region. His native mixed Russian-Belarusian speech (Трасянка) was filled with Ukrainianisms (in addition to the considerable vocabulary that both languages share). Ukrainian friends, from other regions of Ukraine (not Halychyna), who have retained fluency in Ukrainian in the face of the Russification of the Soviet era all say that it was the grandmothers who saved the language.
Suzanne Valkemirer said,
November 16, 2019 @ 5:10 pm
It is one thing to describe the pronunciations of the Ukrainian name of the capital city of Ukraine; another, to prescribe what the standard pronunciation should be in Ukrainian; still another, to describe the pronunciations of its names in other languages; and yet another, to describe what its names should be in other languages.
Nina Jankowicz is going on the — unstated — assumption that the pronunciation in Standard Ukrainian — even its pronunciation at the phonetic level — should also be the pronunciation of the English name of the city. Her assumption is defenseless.
I am reminded of a true story told by David L. Gold in one of his courses at the University of Haifa about a native speaker of Polish of his acquaintance who, though expressing admiration for Yidish language, literature, music, and so on, once remarked to him that the one thing she could not stomach was how yidishophones speaking Yidish mangle "the names" of places in Poland (as if their Yidish names had to be identical to their Polish ones). His reaction was as follows:
D. L. G.: The Latin name of the capital of the Roman Empire is Roma. The Italian name of the capital of the Italian Republic is Roma. What is the Polish name of the city?
X. Y. Z. Rzym [I hope the diacritics appear here: [žɨm].
D. L. G. Why is t he Polish name not [roma']?
X. Y. Z. Thank you for enlightening me.
Bob Ladd said,
November 16, 2019 @ 5:29 pm
Suzanne Valkemirer: I agree with your take on this issue, but a lot of people (and especially, I think, a lot of anglophones) don't. We've been here on Language Log before (e.g. here.
martin schwartz said,
November 16, 2019 @ 5:51 pm
Re the etym. of Kiev, N. Golb and O. Pritsak, in their book on the
Khazars, traced the name to kuyava (Kujawa) from
Khwarezmian *kuyâw, an ethnicon 'pertaining to [M. bin] Kuy,
the Khwarezmian leader of the Khazar garrison. The Khwar. name would derive from *kuy = Sogdian kaw-i 'hero, warrior', on which I wrote in an article on Avestan kauui- in Gifts to a Magus, Fs. Kotwal.
Martin Schwartz
martin schwartz said,
November 16, 2019 @ 5:52 pm
Re the etymology of Kiev, N. Golb and O. Pritsak traced it to
Kujawa (Kuyavsa) from Khwarezmian kuyâw, an ethnicon, from
M(ahmûd?) bin Kuy, head of the Khwarezmian garrison of the
Khazars. Kuy would be the cognate of Sogdian kawi 'warrior, hero'.
For the further history of that word, see my article on Avestan kauui-
in Gifts to a Magus, ed. Choks(e)y and Dubeansky
Neil Dolinger said,
November 16, 2019 @ 7:07 pm
@Suzanne Valkemirer, Bob Ladd is correct, as shown in the link he provided as well as in many others that can be found on LL. A lot of people here, myself included, do not consider *defenseless* the desire that anglophones pronounce – or at least attempt to pronounce – local place names in their local pronunciation.
I won't take up anyone's time here to rehash the argument. It is clearly one of those political arguments that never changes anyone's mind. I'd be happy to take the discussion off line if you are interested.
Rebecca Stanton said,
November 16, 2019 @ 8:27 pm
Ukr. и and Russ. ы are not, in fact, alike (although each is the "nonpalatal" equivalent of the "i" sound in their respective languages). Ukr. и sounds more like Fr. é (which is also technically the sound of the "i" in "kit"). It's a pure vowel whereas the Russian ы often sounds somewhat diphthongized depending on what consonants surround it (this is particularly audible when it's preceded by a labial consonant). Also, the combination KЫ is orthographically "illegal" in Russian so one never hears that vowel pronounced after а /k/ sound.
More to the point, isn't the confusion here caused less by the main vowel than by the final consonant? It's not hard for English speakers to pronounce [kɪ], and that will in fact give a pretty correct first syllable. But what seems to trip everyone up (including the diplomats testifying before Congress) is the fact that the final "v" is actually pronounced "w."
[(myl) That's true in the Wikipedia audio, but not in Nina Jankowicz's renditions. Is that a case of dialect variation? Or is she doing it wrong? ]
Ben Zimmer said,
November 16, 2019 @ 10:23 pm
See also: "Wait, How Do You Pronounce Kiev?" (New York Times, Nov. 13), and the bit at the end of this NPR "All Things Considered" segment from Nov. 14:
Gali said,
November 17, 2019 @ 1:46 am
As an aside, using i/y for и/і/ї seems like a sensible strategy for transliteration, but it's very jarring for someone not acclimated to it to read e.g. "Zelenskiy/Zelenskyi" (per his twitter account it seems even he's not sure!).
Anatoly Vorobey said,
November 17, 2019 @ 2:46 am
Rebecca: the combination [кы] is not found in words of Slavic origin, but is not forbidden phonetically, and is realized by Russians in place names (like Kyrgyzstan) and in borrowings, of which I think the most notable might be акын (aqyn). [гы] is a common lowbrow interjection (expressing either surprise, like "gee", or derision).
On why Kyiv doesn't seem to sound like "the beginning of kit + Eve": could the difference be in the fronting of the initial [k]? To me, the [k] in English "kit" sounds more fronted than the [k] in the Ukrainian Kyiv, and when I try to combine the [k] of "come" with the vowel of "kit", that's when it sounds much like Kyiv to me (not a native English speaker).
[(myl) The non-fronted /k/ hypothesis seems promising to me — and it would be easy to test it with modern ultrasound apparatus.]
Maksym Vakulenko said,
November 17, 2019 @ 4:57 am
This is a good idea – to find out correct pronunciation. It should be understood that Ukrainian phonetics lags well behind the world level that resulted in numerous myths.
I did research in this field since 1997. I got the Fulbright scholarship in 2003, participated in the phonetic project of Lionbridge Technologies in 2016-2018. Some results were published in the articles:
Vakulenko, Maksym O. 2015. “Practical transcription and transliteration: Eastern-Slavonic view.” – Govor 32, 1: 35-56. http://www.hfiloloskod.hr/images/HFD/Casopisi/Govor/web-GOVOR_32_1.pdfVakulenko, Maksym O. 2018. “Ukrainian vowel phones in the IPA context.” In: Govor 35 (2): 189-214. Available at: https://www.hfiloloskod.hr/images/HFD/Govor/Govor-2-2018-web.pdf. DOI: 10.22210/govor.2018.35.11.
Vakulenko, Maksym O. 2019. “Ukrainian Consonant Phones in the IPA Context with Special Reference to /v/ and /gh/.” In: Linguistica online 22: 36-61. Published online August 22, 2019. Available at: http://www.phil.muni.cz/linguistica/art/vakulenko/vak-001.pdf.
Vakulenko, Maksym. 2019. “Calculation of Phonetic Distances between Speech Sounds.” In: Journal of Quantitative Linguistics [WoS, Scopus. 2018 impact factor 0.824]. Published online: October 23, 2019. DOI: 10.1080/09296174.2019.1678709.
So, the name of our capital Kyjiv should be pronounced as [̍kɨ̞+-jiβ̞].
I am open to propositions to collaborate in the field of Ukrainian phonetics.
Kyїvan said,
November 17, 2019 @ 9:49 am
Consider using Kyїv, Ukraїne :)
Kyїv is better than Kiev/Kyiv.
Sniffnoy said,
November 17, 2019 @ 11:51 am
To me, as an English speaker, the vowel used there sounds like the [əɪ] of the coil-curl merger…
Tetiana Katchanovska said,
November 17, 2019 @ 12:32 pm
There's something here about how to pronounce "ы" in Russian: https://burupo.com/ru/%D1%8B-sound-ru/
Andrew Usher said,
November 17, 2019 @ 3:17 pm
So all that can be agreed is that the monosyllabic 'Keev' is wrong? Though, who knows, it might become standard in future English.
There's nothing surprising about us using the Russian name for a Ukrainian city, given the history. It's also apparently easier to adapt to English phonology!
k_over_hbarc at yahoo.com
bianca steele said,
November 17, 2019 @ 4:15 pm
I have a vague memory that my grandmother, who was born in Kyiv/Kiev and moved to Philadelphia when she was small, may have pronounced it Kyiv, but we would probably have chalked it up to her generally idiosyncratic pronunciation of Yiddish words, which I assume was also regional, and often involved similar fowl shifts.
Rebecca Stanton said,
November 17, 2019 @ 6:02 pm
> But what seems to trip everyone up (including the diplomats
> testifying before Congress) is the fact that the final "v" is
> actually pronounced "w."
>
> [(myl) That's true in the Wikipedia audio, but not in Nina Jankowicz's
> renditions. Is that a case of dialect variation? Or is she doing it wrong? ]
Her version sounds non-native to me, to be honest, but yes there is some accent variation. What I was taught in Ukrainian language class is that typically the letter в [v] is pronounced [w] in word-final position. It would be closer to [v] if a case ending were added (e.g. the locative "In Kyiv" – у Києві – [u Kɪɛvi]. In other words it oscillates between "v" and "w" (and even "u", which sometimes is written interchangeably with it) depending on the environment.
One can hear lots of different speakers pronouncing it at https://uk.forvo.com/word/%D0%BA%D0%B8%D1%97%D0%B2/
–some with [v], some with [w], but to my (non-native, critical) ear, the ones that are closer to [v] sound more Russian-influenced, especially when the [v] is devoiced (which is normal in Russian but not in Ukrainian).
The [k] at the beginning is indeed not fronted/palatalized, because Ukr. и is a "hard" (non-palatalizing) vowel. So consonants get palatalized before я, є, і, ю, ь
but are firmly non-palatalized before а, е, и, о, у.
Anatoly, of course you're right that кы appears in some non-Russian words, I was oversimplifying. Technically this combination can of course be pronounced, but it isn't a "Russian" sound, is what I meant. Whereas ки is perfectly at home in Ukrainian.
Finally, about the transliteration of the Ukrainian President's name: in Ukrainian, it's Зеленський, so according to the Library of Congress transliteration system*, it should be written in English as Zelens'kyĭ (that last letter is equivalent to English y-as-consonant, which is where the media spelling "Zelenskyy" comes from — the first "y" is a vowel [ɪ], the second is a consonant [j]). Both "e"s are hard as well as the и, so the only consonant in his name that gets palatalized is the "s" (which is followed by a "soft sign" ь, represented in LOC transliteration by the apostrophe).
*https://www.loc.gov/catdir/cpso/romanization/ukrainia.pdf
Rodger C said,
November 18, 2019 @ 7:40 am
Then there are the NPR newscasters who pronounce it [kjɛv], being presumably more familiar with French poultry cookery than with Eastern Europe.
J.W. Brewer said,
November 18, 2019 @ 10:24 am
I am impressed by the Polish approach, which is to introduce a third vowel into the mix rather than take sides, at least according to Polish wikipedia, which gives the Polish name for the city as Kijów while noting that the Ukrainian cyrillic spelling transliterates as Kyjiw and the Russian spelling as Kijew. The Finnish name (Kiova) seems to reflect a similar plague-on-both-your-houses approach.
J.W. Brewer said,
November 18, 2019 @ 11:23 am
Let me note another problem that arises when spelling changes (especially those motivated by political/historical/military changes in who is in control of the location of the toponym in question) are expected to encourage pronunciation changes on the part of Anglophones. Sometimes-to-often a proposed change in English spelling simply reflects a change in spelling in some non-Latin orthography and then the mechanical application of some transliteration system to the change. But transliteration systems, especially those devised and controlled by speakers of the foreign language in question rather than by Anglophones, are often confusing or unhelpful if the goal is to tell non-specialist Anglophones how they should pronounce things.
Let me take my favorite example of this from the post-Soviet transition. The portion of the USSR formerly known in English as the Kirghiz SSR ended up as independent Kyrghyzstan. The what now? Well, as I understand it, the new regime changed the standard Cyrillic spelling of the ethnonym from Киргиз to Кыргыз, to replace the Russian pronunciation with the indigenous one, and that then flowed through via mechanical Romanization conventions as a change from "Kirghiz" to "Kyrghyz." But while the first-step change in Cyrillic spelling is perfectly transparent to e.g. Russophones — i.e. it accurately communicates what desired change in pronunciation is being suggested by the change in spelling, the meaning of the second-step change in the romanization is totally opaque for Anglophone readers and does nothing for them but sow confusion. "Kyrghyz" does not represent, per standard English orthographic conventions, any predictable pronunciation that is predictably different from how you might have guessed (or been taught) "Kirghiz" should be pronounced. All the Y's in "Kyrghyz" do is signal "THIS IS A WEIRD FOREIGN WORD SO ALL BETS ARE OFF REGARDING PRONUNCIATION."
For another, perhaps more immediately relevant, example, take the given name of the current Ukrainian president. It's the Ukrainian name conventionally romanized these days as Volodymyr, which contrasts with the usual English spelling (which matches up with the standard Russian version of the same name) Vladimir. The difference between "Vla-" and "Volo-" is a reasonably useful cue for Anglophones as to how to change the pronunciation, but changing "-dimir" to "-dymyr" does nothing whatsoever to help non-specialist Anglophones pronounce the name other than, again, send the "THIS IS A WEIRD FOREIGN WORD" signal.
BZ said,
November 18, 2019 @ 2:14 pm
@J.W. Brewer, what would you suggest the English spelling to be though? To tell the truth, English doesn't have a sound that matches either the "i" or the "y" in your examples, and the English short "i" is somewhere between the two. But at the same time, the pronunciation *has* changed. Does it need to change in English? Maybe not, since it's already "close enough".
J.W. Brewer said,
November 18, 2019 @ 3:26 pm
I don't think English spelling or pronunciation needs to change whenever the relative political power of rival pronunciations and/or non-roman-alphabet spellings (especially when neither/none of the contending pronunciations can be perfectly approximated by an ordinary Anglophone anyway) has shifted in foreign parts. A politically-motivated change of toponym that reflects something other than a pronunciation fight (e.g. Leopoldville -> Kinshasa, or St. Petersburg -> Petrograd -> Leningrad -> St. Petersburg) is a different situation presenting different issues.
Admittedly, one factor that can be hard to judge is what percentage of Anglophones are familiar with the traditional English spelling for a given toponym and how frequently it's in fact used. If a toponym is obscure enough, then the transition costs associated with adopting a new spelling may be minimal – except for the potentially significant factor I already noted of whether as compared to the old one the new spelling is more or less helpful versus opaque to the Anglophone trying to match tongue to eye. And come to think of it, my own perspective is probably in part selfishly motivated by being an Anglophone of a certain age who perhaps idiosyncratically looked at atlases and world maps quite a lot as a boy — so any change from what was the standard English spelling in atlases in wide distribution in the 1970's risks imposing a personal transition cost on me while transitions that had already been largely accomplished by then have no similar effect on me, so (to go o some shifts that were caused by the Soviet regime but continued by the post-Soviet successor regimes) I do not feel personally irked by the demise of e.g. "Tiflis" for "Tblisi" or "Wilno" for "Vilnius."
Plenty of other latin-scripted languages have no problem respelling our toponyms to fit their orthographic and/or phonotactic conventions rather than require their readers to understand ours. Latvian, for example, is happy to respell New York as Ņujorka and Washington as Vašingtona, and it would be silly for us to complain about that. Perhaps English ought to be more cosmopolitan than Latvian in this regard, but the notion that we are required by some universal norm to adopt whatever romanized spelling is preferred by the current regime in control of the place (often reflecting orthographic and/or transliteration conventions that were not developed for the convenience of Anglophones) should not be treated as obligatory and exceptionless.
Rebecca Stanton said,
November 18, 2019 @ 7:52 pm
J.W. Brewer but this isn't a fight over transliteration, it's a fight over the name of the city. Which is "Kiev" in Russian, "Kyiv" in Ukrainian. Different spelling. Different Cyrillic letters. Same with "Volodymyr." That's not a matter of "i"s arbitrarily being changed to "y"s. The Ukrainian name has a different vowel there than the Russian name does. The difference between the Russian "i" and the Ukrainian "y" is no less than the difference between the Russian "a" and the Ukrainian "o" in the same name (which you have no trouble conceding are different).
J.W. Brewer said,
November 18, 2019 @ 10:03 pm
It's a fight over transliteration because of the unexamined assumption that Cyrillic words should be romanized in ways that mechanically preserve orthographic differences that are meaningful in Cyrillic, not least because they are show different pronunciations in various Cyrillic-scripted languages, without thinking about whether the output of the romanization process will actually make much sense to a native speaker of any given latin-scripted language, such as for example English. There is no one-to-one correspondence of Cyrillic glyphs to Latin-alphabet glyphs, and there is no one-to-one correspondence of East Slavic vowels to English vowels. Acting as if there ought to be is a recipe for trouble. I and Y-as-a-vowel in English do not contrast with each other in pronunciation; whatever orthographic distinction they draw is largely historical/etymological/ornamental. Cyndi might be a variant spelling of Cindy but it is pronounced identically; Cindi and Cyndy also work as variant spellings of the same underlying pronunciation despite the fact that the name in question actually has two different vowels in it. Y, like I, can variously represent in English the KIT vowel, the FLEECE vowel (or at least an unstressed word-final variant of it), or the PRICE vowel, but using Y instead of I is not, by itself, going to tell you which of those three it is. Thus, using a transliteration scheme that proceeds as if systematic alternation between I and Y is phonemically meaningful is not a good approach if you're trying to be user-friendly for Anglophones trying to pronounce loanwords in an English context. Now maybe that sort of user-friendliness is not the primary goal of your transliteration scheme, which is instead meant to be useful to people who already know the language of origin but sometimes need to write about it without access to a Cyrillic keyboard. Fine. But that's why Anglophones trying to figure out how to spell and pronounce words that come from your language that they may have occasion to use when speaking their own language may not feel obligated to follow your system in figuring out what spelling or pronunciation to use for foreign toponyms in English.
Digging into the google books corpus, I see that a number of 19th-century English texts used "Kieff" for the city in question before fashions shifted, and elsewhere on the internet I see the claim that if you go back a bit further than that the good old Anglo-Saxon toponym was Cænugeard. I'm not sure if that's actually attested in contemporaneous texts or just a modern adaptation of the better-attested Old Norse Kænugarðr (also sometimes seen as Kønugarðr).
Gali said,
November 19, 2019 @ 7:49 am
Using transliterated names that do not aid or even hinder an uninitiated reader (see: Mandarin) is something I can only see as an extension of the general English language consensus not to respell loanwords and just let hapless native and foreign learners alike contend with yet another word that defies any orthographic conventions.
Andrew Usher said,
November 19, 2019 @ 7:01 pm
We are hardly the only Latin-alphabet language to do that, though. And in many cases there's no agreement on pronoucing a foreign word in English so spelling it phonetically would not be easy; following here, how would we spell the authentic pronunciation of 'Kyiv'?
J.W.Brewer:
The familiar 'old name' was kee-ev or kee-eff, stress on either syllable. So 'Kieff' made just as much sense as 'Kiev'; it was following the French transliteration convention to explicitly mark the devoicing of 'v'.
It may be of interest that that Anglo-Saxon name you cited would come into modern English as 'Caneyard', presumably pronounced in the manner of 'vineyard'.
Ksenia Lena M said,
November 20, 2019 @ 1:34 am
As they sow, so shall they reap :(
Listen for 90 sec starting 10:13
https://www.cbc.ca/amp/1.5360367
J.W. Brewer said,
November 20, 2019 @ 12:07 pm
Of course, in dealing with other Latin-scripted languages, with certain historical examples for important toponyms like Cologne and Munich, we often keep the original spelling at the expense of the original pronunciation — you wouldn't get echt-Deutsch pronunciations from Americans if you respelled Frankfurt and Berlin as e.g. Fronkfoort and Bearleen, but you'd probably get somewhat closer approximations. But we don't seem to care about that, and most of the time even those AmEng speakers who can do the echt-Deutsch pronunciations credibly because they actually possess some fluency in German will not do so when mentioning the toponyms in the middle of a discourse in English, because doing so will seem weird and/or pretentious unless there is some context-specific reason to emphasize the un-Anglicized pronunciation. And I daresay German phonology is less of a stretch for the typical AmEng speaker than East Slavic phonology.
GH said,
November 21, 2019 @ 3:38 am
@J.W. Brewer
Where I fail to follow you is in the implication that spelling these names with a Y is less "user friendly" to English speakers than using I. If both signal the same English pronunciation and both are orthographically permissible, why not do so in accordance with transliteration conventions, even if the reasons are opaque and seem arbitrary to people unfamiliar with the original languages involved?
Maksym Vakulenko said,
November 22, 2019 @ 4:00 am
A catastrophic drawback of the form “Kyiv” has been revealed in this discussion.
When doing phonetic analysis (which is one of the language processing procedures), we build the chain “phone – phoneme – grapheme” in the direct or reverse order where each chain member contains, in general, a number of variants. All of these should be properly transliterated into Latin script in the way that keeps necessary distinctions between initial Cyrillic characters. We also should make sure that no false identity between any transliterated character and established IPA symbols arises. This procedure was thoroughly carried out in the articles:
Vakulenko, Maksym O. 2018. “Ukrainian vowel phones in the IPA context.” In: Govor 35 (2): 189-214. Available at: https://www.hfiloloskod.hr/images/HFD/Govor/Govor-2-2018-web.pdf. DOI: 10.22210/govor.2018.35.11.
Vakulenko, Maksym O. 2019. “Ukrainian Consonant Phones in the IPA Context with Special Reference to /v/ and /gh/.” In: Linguistica online 22: 36-61. Published online August 22, 2019. Available at: http://www.phil.muni.cz/linguistica/art/vakulenko/vak-001.pdf.
Please pay attention to this.
I have to remind that the only technique to do so is the so-called scientific transliteration providing isomorphic (simple-correspondent) correspondence between initial and latinized characters giving rise to full text reversibility (equivalence). The form “Kyiv” (and the corresponding pseudo transliteration system) fails to satisfy this requirement, so it cannot be seriously taken as material for phonetic analysis and other language processing procedures.
The translilteration system of the Ukrainian Latinics meeting necessary conditions was grounded and described in the paper:
Vakulenko, Maksym O. 2015. “Practical transcription and transliteration: Eastern-Slavonic view.” Govor 32, 1: 35-56. http://www.hfiloloskod.hr/images/HFD/Casopisi/Govor/web-GOVOR_32_1.pdf.
For example, analyzing (segmenting) Ukrainian texts or words, we should understand that the Ukrainian Cyrillic letters «і», «й», «ї» have distinctly different readout, and their false identification by the single Latin letter “i” following from “Kyiv” provokes false phonetic conclusions.
The starting point to analyze the needed pronunciation is the Cyrillic form «Київ» which has the equivalent latinized form Kyjiv (or Kyïv), and its uk-UA pronunciation is [̍kɨ̞+-jiβ̞].
Much attention is paid here to the article “Ukrainian” by B. Pompino-Marschall et al in JIPA. I have to say that when starting the phonetic project with Lionbridge Technologies in 2016, it was me who recommended to take this publication as a starting point of further research. However, we should be aware of its shortcomings, too, including the following:
(1) The informant involved speaks the Pokuttja subdialect of the South-Western dialect which is not the base of the literary Ukrainian language;
(2) The authors analyzed the speech of just one informant that is not sufficient to make reliable concludions;
(3) The results obtained are intuition-driven;
(4) No frequency measurements were carried out.
My recent publications cited above account for these.
And please be informed that Nina Jankowicz does not speak uk-UA. And her first readout of “Kyjiv” is not clear.
Philip Taylor said,
November 23, 2019 @ 11:58 am
In your opinion, Maksym, would "Kiyiv" be a reasonable transliteration, given that for English speakers the letter "y" would represent IPA /j/ in such a context ?