Talk:Ukrainian phonology

Ukrainian phonology 2019

Please keep in mind distinction between reliable and unreliable sources. Phonetics (and phonology resulting from it) is an experimentally-based science. Please don't use ill-based conjectures on phonetic characteristics made in Ponomariv 2001, Buk et al. 2008, Danylenko and Vakulenko 1995. The phonetic experimental results are lacking, but they are still present: Bilidid 1969 (this work was used in Ponomariv 2001 and Danylenko and Vakulenko 1995), Pompino-Marschall et al. 2016 (however, some ungrounded conjectures are taken from Buk et al.); Vakulenko 2018; Vakulenko 2019. Mova2016 (talk) 04:44, 2 September 2019 (UTC)

The reliable sources which are based on phonetic experiments, suggest appearance of [ʢ],[ɣ], and, possibly, [ɦ] (Bilodid 1969; Pompino-Marschall et al. 2016, Vakulenko 2019). Mova2016 (talk) 04:32, 2 September 2019 (UTC)

I support the [ʋ] for the Ukrainian /в/. Please see my explanations in the "w" section.

Mova2016 (talk) 08:39, 6 September 2019 (UTC)

I would like to comment on some sources I classify as unreliable.

• Antonina Bilous's thoughts presented in the doc file. If there were significant scholarly value, why not publish it in the scholarly journal? Anyway, some results of Bilous were duly cited in Vakulenko 2018.

• Danylenko & Vakulenko 1995. This is a compilation of phonetic results described in Bilodid 1969, with minor (and not experimentally proved) additions. And they use incorrect IPA notations (for example, h for Cyrillic "г"). So the value of this work is only as an English version of Bilodid 1969 for the English-reading audience. However, the link is not working.

• Press & Pugh 2015. They are Scottish linguists specializing on Russian linguistics. As we understand, this is not the same as the Ukrainian phonetics. The European phoneticians do not consider them as phoneticians at all. No experimental (instrumentally obtained) results are presented, just observations. So the value of this work is just as an additional side observation.

• Buk et al. 2008. Having read this article, I was surprised where the authors got the material and why they do not refer to the experiment at all. I contacted Solomija Buk and asked her about it. She confessed that there were no experiment and they described their own feeling.

• Ponomariv 2001. This is a journalist, not a phonetician. The authors do not carry out phonetic research and do not publish their results in the phonetic journals. This handbook uses the results of Bilodid 1969, but their arbitrary interpretation is given therein. Not a valuable source in the phonetic sense.

In addition to this. Much misunderstanding arises when the unknown authors of wiki-articles try to interpret the results of the sources. For Ukrainian, many of them use some more or less arbitrary Latin symbols fancying that they correctly represent the Ukrainian sounds (rendered in Cyrillic script). This is not the case, because correct IPA notation is a separate problem requiring profound investigation (see Pompino-Marschall et al. 2016; Vakulenko 2018, 2019).

Mova2016 (talk) 07:05, 9 September 2019 (UTC) The table provided in the article, does not include the rare (peripheric) phonemes mʲ, pʲ, ʒʲ, etc., regarding them as variants of their plain cognates. OK, this is acceptable for phonemic description. At the same time, the phoneme /g/ in Ukrainian is much more rare than those excluded from the table; its frequency, according to the encyclopedy "Українська мова", is less then 0.1%. In view of this, we must do one of the following:

• either keep /g/ and include all rare and peripheric phonemes; thus the table will be rather phonetic (this is not advisable for an article named "Ukrainian phonology");

• or exclude the /g/, too, regarding corresponding sounds as allophones of the /г/.

I suggest the second option.

Thus, the phoneme /г/ is realized in the following allophones: [g], [gʲ], [ɣ], [ɣʲ], [ʕ] or/and [ɦ] and, possibly, [ʕʲ] or/and [ɦʲ]. The appearance of [g], [ɣ], [ɣʲ], [ʕ] has been supported instrumentally. Please keep in mind that though [ɦ] was repoprted in Pompino-Marschall et al. 2016 and in Vakulenko 2019, this is only a visual or sensual perception. It may be both true or false. No reliable instrumental evidence of [ɦ] as [г] (opposed to [ʕ]) is known to me. If I am wrong, please provide due references.

I have a request to those who has access to articulographs (they are costly!) and Ukrainian speakers, too. Please make a short phonetic experiment aimed to establish articulation of the Ukrainian /г/ in different positions. I have a list of utterances needed for this. Please contact me for details. This may result in an interesting and important research article and, possibly, in a promising project. I am serious. My email is: science@dntb.gov.ua.

For the time being, based on current phonetic data, we must agree that ɦ cannot be regarded as the chief allophone of the /г/.

I suggest a ɣ.

Mova2016 (talk) 08:09, 6 September 2019 (UTC)

Please keep in mind that Ponomariv 2001 is NOT a reliable source. The authors try to use political arguments instead of phonetic facts. Actually, they did not carry out their own investigation and use Bilodid 1969 and some unknown Russian literature. Mova2016 (talk) 04:36, 2 September 2019 (UTC)

Please examine Vakulenko 2018 for more scholarly details. Mova2016 (talk) 16:22, 1 September 2019 (UTC)

Concerning the Ukrainian [и].

Thanks to W.M.Drossel for the formant frequencies table. In particular: F2([i]) = 3000-3500 Hz, F2 ([и]) = 1800 Hz (or 1650 Hz, if y is transliteration). This is pretty near the central vertical line. Please compare relevant vowel frequencies presented in Stevens 1998. Conventionally, the F2 value indicates horizontal position at the vowel (Jones) diagram. Please see Stevens 1998 for reference (or equivalent literature). According to this data, the Ukrainian [и] is a high lowered central vowel [ɨ̞].

It is remarkable that the experimental data of Vakulenko 2018 correlate well with these.

The references to other experimental data including perceptional experiments, are given in the text.

In view of this, we must denote the Ukrainian phoneme /и/ by the IPA symbol ɨ̞.

Objections? Please provide reliable and duly published experimental data, not assumptions or fantasies. For example, results of perceptional experiments showing equivalence of perception of [ɪ] as in "sit" and in [и] "сидіти", for example. And phonetic equivalence of an allophone of [i] as appearing in "безіменний", "передісторія", etc., and an [и].

Mova2016 (talk) 06:49, 6 September 2019 (UTC)

The Ukrainian [o] is mid high, not mid low. It causes the flattening effect. See details in Vakulenko 2018. Mova2016 (talk) 16:13, 1 September 2019 (UTC)

I suggest all [ɔ]'s to be changed into [o]'s. See above.

For those who still disagree.

1. Please provide reliable and duly published experimental data on vowel perception supporting an assumption of [ɔ]. The data supporting [o] are published in Govor (Vakulenko 2018) which is indexed in Scopus.

2. Please provide reliable and duly published experimental data on flattening effect of [ɔ]. The data supporting [o] are published in Govor (Vakulenko 2018) which is indexed in Scopus.

Otherwise, we must agree that the Ukrainian [o] is indeed a mid high back vowel, not mid low. Mova2016 (talk) 06:48, 6 September 2019 (UTC)

The experimental data published in Vakulenko 2019, clearly indicate that the Ukrainian /в/ is not a bilabial approximant [w].

I suggest to change the notation to ʋ, as in Pompino-Marschall et al. 2016, Vakulenko 2019.

Objections? Please provide the following:

1. The reliable and duly published experimental data on characteristic frequencies of the F1 and F2 (they should be below 700 Hz, if [w]).

2. The reliable and duly published experimental data on labiovelar articulation of a [в] (e. g. palatograms indicating tongue root rising during [в]). Please keep in mind that the palatograms of Bilodid 1969 indicate that this is not a [w].

3. The reliable and duly published data overcoming common agreement that diphtongs are not characteristic for Ukrainian.

4. The reliable and duly published experimental data overcoming Kawasaki's conclusion about unfavoured combinations wu, wo, uw, ow, etc.

Otherwise, we must agree on ʋ being realized through β̞ and v, vj. Mova2016 (talk) 06:48, 6 September 2019 (UTC)

The following statement is questionable: "/w/ is most commonly bilabial [β̞] before vowels but can alternate with labiodental [ʋ] (most commonly before /i/),[32] and can be a true labiovelar [w] before /ɔ/ or /u/.[16] It is also vocalized to [u̯] before a consonant at the beginning of a word, after a vowel before a consonant or after a vowel at the end of a word.[32][33] If /w/ occurs before a voiceless consonant and not after a vowel, the voiceless articulation [ʍ] is also possible.[16]"

No experimental evidence of rise of [w] in Ukrainian was given (please see above). On the contrary, there are facts indicating that appearance of [w] (/w/) in Ukrainian is not possible. — Preceding unsigned comment added by Mova2016 (talk • contribs) 09:56, 9 September 2019 (UTC)

A questionable, dubious, and unjustified conjecture: "/ɪ/ may be classified as a retracted high-mid front vowel,[14] transcribed in narrow IPA as [e̠], [ë], [ɪ̞] or [ɘ̟]".

1. Why /ɪ/? No proofs are drawn for this choice. Only intuitive guesses.

2. "Narrow IPA" is designed, perhaps, to denote allophones? This assumption is not clearly formulated.

3. Determination of allophones is not an easy intuitive task. Where are the experimental proofs for this choice? Publications?

4. When the unknown authors use the formulation "transcribed as", they allegedly assume that this is the well-known fact. This is not so. Experimental proofs are required.

5. The use of IPA symbols for allophones is not a straightforward and intuitive deal. First of all, the general rules of vowel shifts on the Jones diagram should be met (see Stevens 1998 for details). In particular, in the absence of assimilation, the unstressed allophones move to the center, not to periphery. As I see from the notations, these principles are not observed. It seems that no theory and no experimental facts were used to write this assumption, rather someone's intuition. So please provide reliable proofs. These should be pretty revolutionary...

In view of this, I mark again this sentence as "dubious". Please keep it. If the proofs will not be cited, we must delete it.

Mova2016 (talk) 13:03, 9 September 2019 (UTC)

"Ponomariv 2001. This is a journalist, not a phonetician. The authors do not carry out phonetic research and do not publish their results in the phonetic journals."

What Ukrainian phonetic journals do you know? Костянтин Конь (talk) 05:28, 31 December 2020 (UTC)

[w] vs [u̯]

What is the difference between a true labiovelar [w] and vocalized [u̯]? Nardog (talk) 05:35, 2 March 2022 (UTC)

Even if there's a measurable phonetic difference between the two (with the coda variant being undershot), both should be written with ⟨w⟩. "If /w/ occurs before a voiceless consonant and not after a vowel, the voiceless articulation [ʍ] is also possible." proves that it's a consonant, just like Slovak and Slovene [w] (so ⟨u̯⟩/⟨ʊ̯⟩ are also wrong in those cases). Sol505000 (talk) 14:20, 2 March 2022 (UTC)

"stop taking out the detailed information on the allophones please"

@Fdom5997 What are you talking about? Literally all the information was kept in the edit. Nicodene (talk) 01:39, 12 March 2022 (UTC)

• 'Stressed vowels tend to be lengthened (1) while unstressed vowels tend to centralize (2)'
• 'Unstressed vowels are somewhat reduced in duration (3) and quality (4)'

1 and 3 are describing the same difference from opposite perspectives. 2 and 4 are describing the same thing, except that 2 uses much more specific terminology. There is no need for two sentences- it's repetition. Nicodene (talk) 01:48, 12 March 2022 (UTC)

@Nicodene I just re-added the info that you provided, as well as the info that you reduced. The first two things that you put are pretty necessary, but please stop taking away the other details from the other points. That is not repetition, it is also detail on the allophones, and it is much more organized as well. Fdom5997 (talk) 08:31, 12 March 2022 (UTC)

@Fdom5997 I'm sorry to ask this, but do you really understand what you are reading? Because my edit did not remove any information. It removed a few words, but all of the information contained in the original remained. If you disagree, please actually cite an example of information that my edit removed. 17:42, 12 March 2022 (UTC) Nicodene (talk) 17:42, 12 March 2022 (UTC)

really no ŋ in Ukrainian?

The article doesn't mention ŋ. Is nk as in Taras Shevchenko really not pronounced with a voiced velar nasal? --Espoo (talk) 20:52, 10 October 2022 (UTC)

Looking for clarity of IPA symbol for "и". Need letter-to-symbol transcription

I'm researching IPA symbols for Ukrainian vowels and found this file: https://commons.wikimedia.org/wiki/File:Ukrainian_vowels.svg It is very different from the image in this image. I don't know if it's right, but I do know that when I hear "и", the sound is not in this article's image. It's closer to /ɤ/ (close mid back unrounded vowel).

Can anyone add the phonetic symbols for Ukrainian vowel letters? DBlomgren (talk) 04:03, 14 November 2022 (UTC)

On the reliability of the statement regarding realization of /w/ as a bilabial approximant or a labiovelar approximant

I would like to raise an issue based on my concerns regarding a quote from the article, specifically the "Phonetic details:" list of the "Consonants" part, which is as follows at the moment:

/w/ is most commonly bilabial [β̞] before vowels but can alternate with labiodental [ʋ] (most commonly before /i/), and it can be a true labiovelar [w] before /ɔ/ or /u/.

First, I would like to review the first part of it, namely, /w/ is most commonly bilabial [β̞] before vowels but can alternate with labiodental [ʋ] (most commonly before /i/). It is reinforced by a reference to Žovtobrjux, M.A.; Kulyk, B.M. (1965). Курс сучасної української літературної мови. Частина I. Kiev: Radjans’ka škola, which is available by following this link.

Below is the direct quote from the source:

Звук в в українській літературній мові найчастіше є губно-губним приголосним, або білабіальним (у транскрипції його його часто позначають буквою w: wода, wулиц'а), але в деяких позиціях він виступає і як губно-зубний (у транскрипції такий в передається буквою ʋ: ʋ'ін, ʋ'іс'т'). Як губно-губний, тобто утворюваний верхньою і нижньою губами, в виступає перед усіма голосними, крім і: wага, wухо, wеду, wир (орфогр. вага, вона, вухо, вир). Перед голосним і звук в може вимовлятися як губно-губний, і як губно-зубний, тобто утворюваний нижньою губою і зубами: w'іл, w'ітер і ʋ'іл, ʋ'ітер (орфогр. віл, вітер).

I am providing my own translation of it below:

The sound в in Ukrainian literary language is most often a lip-lip [sic] consonant, or bilabial (in transcription it is often described with the symbol w: wода, wулиц'а), however in some positions it is lip-dental [sic] (in transcription such a в is described with a ʋ: ʋ'ін, ʋ'іс'т'). As a lip-lip [sic] consonant, meaning that it is formed by the upper and lower lips, в is formed before all vowels but і: wага, wухо, wеду, wир (orthogr. вага, вона, вухо, вир). Before the vowel і the в sound can be pronounced either as a lip-lip [sic] or a lip-dental, meaning it is formed by the lower lip and the teeth: w'іл, w'ітер і ʋ'іл, ʋ'ітер (orthogr. віл, вітер).

I was originally confused by this solely on the grounds of my experience as a native speaker of Ukrainian living in Ukraine and being exposed to the manner of speaking from many regions, and having performed a rudimentary questionnaire on how people pronounce a sample word вода to validate. In my experience, it is most often pronounced with either a labiodental fricative (v) or a labiodental approximant (ʋ) in the beginning of a word before a vowel, with a larger degree of reduction towards a bilabial or a labiovelar approximant (β̞ and w respectively). When questioned further, participants specifically confirm that the sound they pronounce is done with their lip and their teeth. While the usage of labiodental fricatives can be attributed to a deviation induced by the influence of Russian language, which is also mentioned in the article, the prevalence of labiodental approximants can not. I have not yet conducted a comprehensive comparison with either older generation speakers or older media containing relevant Ukrainian speech, but am planning to do so.

Aside from that, there is a number of problems with this source and how it is interpreted into the content of the article.

• The source states that The sound в in Ukrainian literary language is most often a lip-lip consonant, or bilabial (in transcription it is often described with the symbol w: wода, wулиц'а), which is incorrect since w is used to describe a labiovelar approximant. Assuming this is a mistake on their part, there's little ground to either an assumption that what was meant was in fact a bilabial (I need to emphasize that they do not specify if it's a fricative β or an approximant β̞), or indeed a labiovelar w. I am unfortunately not educated about the transcription standards of the mid-soviet period, so if anyone has any input please do provide it.
• The source states that however in some positions it is lip-dental [sic] (in transcription such a в is described with a ʋ: ʋ'ін, ʋ'іс'т'). This is, while not incorrect, incomplete to the point where it is not exactly easy to draw a concrete conclusion when we have the above-mentioned inconsistency of transcription standards and the ambiguity of description. It is not specified whether it is a fricative or an approximant, but rather we assume that they mean an approximant from the transcription symbol, which the source uses incorrectly and ambiguously in the previous statement in the same sentence. While I do not find this particular part problematic, it highlights the ambiguity of the source.

I am therefore skeptical of the quote referenced in the very top of the post, and would like external review and considerations from other participants.

Second issue I have is with the and it can be a true labiovelar [w] before /ɔ/ or /u/. part of the quote, which is reinforced by a reference to Buk, Solomija; Mačutek, Ján; Rovenchak, Andrij (2008), "Some properties of the Ukrainian writing system", Glottometrics, 16: 63–79, arXiv:0802.4198, Bibcode:2008arXiv0802.4198B, which is available by this link. The quote that it relies on is: voiced labialized velar approximant [w] before /ɔ, u/ and voiced consonants (not after a vowel): вниз, вона, вухо; on page 4. There is nothing wrong with the quote, but it is unclear where this comes from, since this is essentially a meta-study of other sources, and they do not cite the specific study it was taken from. I have not reviewed every study mentioned, but they do use the source which is used for the first part of the quote (Žovtobrjux, M.A.; Kulyk, B.M. (1965). Курс сучасної української літературної мови. Частина I. Kiev: Radjans’ka škola). I am skeptical about this part, again, based on my experience as a native Ukrainian speaker in Ukraine, and based on my rudimentary questionnaire I've performed recently, mentioned above.

I would be extremely grateful if anyone with more experience could review these concerns and provide input on if this is relevant and comprehensive enough to warrant a change to the article, or further deliberation on the subject matter is necessary. I am grateful to anyone willing to provide their opinion on the topic.

RollerKnobster (talk) 19:41, 28 September 2024 (UTC)

The JIPA article on Ukrainian has this to say:

• “The approximant /ʋ/ is realized in different ways due to co-articulatory effects: it surfaces as labiodental [ʋ ʋʲ] when preceding the front vowels [ɪ i] or labiovelar [w] when preceding the rounded back vowels [u ɔ]. In the syllable-final position this approximant is realized as [u̯]. Buk et al. (2008) also note that /ʋ/ can be realized as a devoiced labio-velar approximant [w̥] before voiceless consonants (but not after a vowel), e.g. vperše [w̥pɛrʃɛ] ‘for the first time’”.

The article also features a narrow transcription of a story as read by “a 38-year-old speaker born in the Bukovina area […] who received his academic education in Kyjiv”. I have found the following occurrences of /ʋ/:

• With a following vowel:
  • [i]: [ˈʋʲitɛr] (4x), [vʲin], [pɔˈʋʲitrʲa], [prɪˈvʲitnɪmɪ] // [ʋʲ] (5x) and [vʲ] (2x)
  • [ɪ]: [ˈʋɪznanɪi̯], [ˈʋɪmusɪtʲ], [xʋɪˈɫɪn], [ˈʋɪmuʃɛnɪi̯], [ˈʋɪznatɪ] // [ʋ]
  • [a]: [ɔˈbɪdʋa] // [ʋ]
  • [ɔ]: [prɪwɔdu], [ˈdwɔx], [ˈdvɔx], [wɔˈnɪ], [swɔˈjɛ] (3x), [swɔˈjimɪ] // [w] (7x) and [v] (1x)
• With a following consonant:
  • [n]: [mandrʲivnɪˈka]/[-k]/[-ɡ] (4x) // [v]
  • [nʲ]: [pʲiu̯ˈnʲit͡ʃnɪi̯] (2x), [pʲivˈnʲit͡ʃnɪi̯] (2x) // [u̯] (2x) and [v] (2x)
  • other consonants: [pɔu̯z], [ˈkutau̯sʲa], [u̯ˈrɛʃtʲi], [u̯ˈʒɛ] // [u̯]
• At the end of a word:
  • [prɔˈxɔdɪu̯], [duu̯] (2x), [pɛrɛˈstau̯], [zʲnʲau̯], [ˈbuu̯] // [u̯]

This seems consistent with the description quoted above, apart from the (variable) occurrence of a fricative. However, it should be noted that the sample size is just one, and the speaker comes from a western region. Nicodene (talk) 21:38, 28 September 2024 (UTC)

FWIW, in anglophone literature at least, the distinction between fricatives and approximants wasn't always made, and the term "approximant" didn't exist until the 1960s (they were both classified under fricatives, and whenever clarification was needed, approximants were called "frictionless continuants"). I don't know if that was the case in Ukrainian/Soviet literature too, but it might explain why old sources are unclear about that distinction. Nardog (talk) 22:00, 28 September 2024 (UTC)