Cvrk
« previous post | next post »
If you're looking for words with lots of consonants and few or seemingly no vowels, try Eastern Europe, especially Czechia.
I have a friend named Stu Cvrk, and I asked him the story of his surname and how to pronounce it. Here's what he told me:
It is Czech. The Czech pronunciation is "tsverk". My grandparents Americanized it a bit to make it easier to say, as we now pronounce it "swerk."
The story of its derivation according to family lore is this:
In the Middle Ages, people were generally identified by a single first name (Joe, John, Mary, etc.). As the population grew, the need for second names became apparent, and the local lords and landowners assigned second names based on a person's employment (Butcher, Baker, Farmer, etc.). When they got around to my ancestor, they assigned the name "Cvrk," which ostensibly is a Czech word for "the sound a cricket makes," and thus you can see where my ancestors ranked in the grand scheme of things.
And then there's the wondrous Hungarian-American surname, Csikszentmihalyi (5 vowels out of a total of 16 letters), for which see this post (11/5/15).
Selected readings
- "Words without vowels" (/2/20)
- "Dzwil" (11/3/15)
- "Interslavic" (9/24/19)
- "Caucasian words for tea" (1/26/17)
- "The languages of the Caucasus" (8/25/08)
- "Obituary: The Ubykh Language" (1/25/12)
- "Ubykh: requiem and revival" (1/26/19)
- "Syllables" (2/24/20)
- "English syllable detection" (2/26/20)
.mau. said,
June 5, 2020 @ 5:11 am
Slovenian (and other languages from former Yugoslavia) also has a number of words without a vowel (the r is usually pronounced as a schwa): for example the city of Trieste is called Trst, and there is an island called Krk. When I had a brief vacation in Slovenia I was amused that the word for death is smrt; it kinda makes sense…
Athel Cornish-Bowden said,
June 5, 2020 @ 5:22 am
Hungarian names are nowhere near as difficult to pronounce (in a way a Hungarian would understand, though not exactly as a Hungarian would pronounce it) as they look.The rule is simple: you say each syllable in turn until you get to the end, pronouncing each one clearly (no mumbling) and putting the stress on the first. Once in a Hungarian restaurant in London I pronounced "paprikas" in a way that the waiters found so convincing that the called to owner to talk to me in Hungarian. Alas, speaking Hungarian demands more than knowing how to say "paprikas" , so he was disappointed.
Jim M. said,
June 5, 2020 @ 5:43 am
Mike Royko once wrote a column about an imaginary Bulgarian boxer named Brza Smrt—"Quick Death."
Phillip Minden said,
June 5, 2020 @ 6:24 am
A linguist will tell you that these are all kinds of pretty independent issues:
– consonant clusters, especially in phonotactic environment naive monolingual speakers of English aren't used to
– vocalic liquids (cvrk is pronounced cvrk, not cverk)
– schwa as a full-value, stressable vowel, but with the spelling suggesting it isn't similar or identical to stressed schwa in American English accents
– purely written multigraphs naive monolingual speakers of English aren't used to
David Marjanović said,
June 5, 2020 @ 6:32 am
For Americans, that's generally the same: they don't put any vowel at all in squirrel, just a syllabic /r/ and a syllabic /l/.
Slovene generally inserts a schwa next to every /r/ that finds itself between two consonant phonemes; FYLOSC* and Czech generally don't, and neither do rhotic American accents of English.
* the Former Yugoslav Language of Serbo-Croatian (© J. W. Brewer)
David Marjanović said,
June 5, 2020 @ 6:35 am
A vowelless Czech story. The translation follows a few comments later.
Chris Button said,
June 5, 2020 @ 6:52 am
I've always wondered about the last name of Belgian racecar driver Jackie Ickx. Does anyone know the origin of that one?
Chris Button said,
June 5, 2020 @ 6:53 am
Sorry Jacky (not Jackie) Ickx
Bloix said,
June 5, 2020 @ 7:03 am
It seems to me this is just a matter of orthographic convention. For example, the ordinary English word smirked could be spelled smrkd.
Ralph Hickok said,
June 5, 2020 @ 7:19 am
In the early 1950s, Temple University had a basketball player named Bill Mlkvy, of Slovakian descent. Sportswriters called "the Owl without a vowel" (the Temple sports mascot is an owl). Of course, that wasn't quite accurate because the final "y" functions as a vowel.
Dominik Lukes said,
June 5, 2020 @ 7:28 am
I think I commented a previous vowelless post, there is a much longer word without vowels in Czech: scvrnkls (transl. you flicked something off a surface). As somebody else commented, there is no schwa, the r and l have the sonorant role of vowels. Of course, there's also the famous sentence "Strč prst skrz krk."
Coby Lubliner said,
June 5, 2020 @ 7:30 am
For some reason there is a tradition of calling "vowels" the five letters that represent vowels in Italian (though "i" has a diacritic function as well, as in the digraphs "ci" and "gi").
The fact that "owl" is made to rhyme with "vowel" shows how arbitrary the placement of "vowels" before liquids is.
Chris Button said,
June 5, 2020 @ 7:40 am
I've said it many times on LLog. An obstruent/sonorant distinction is phonologically far more useful than a consonant/vowel one.
David Marjanović said,
June 5, 2020 @ 8:09 am
A lot of Belgian names are spelled like that. Hendrik's descendants have ended up as Hendrickx.
Depends on the language. Your average Polynesian language has a clear-cut consonant/vowel distinction where the liquids behave like the other consonants. English, on the other hand…
Vincent Post said,
June 5, 2020 @ 8:12 am
I was excited to see a post about Czech because I'm fascinated by that language and posts about Slavic languages tend to be thin on the ground here. Too bad it boils down to 'look at all the consonants' as commented on by an anglophone named Stu who does not know the language.
Philip Taylor said,
June 5, 2020 @ 9:14 am
Is the "<vowel>ckx" cluster not primarily a Basque phenomenon ?
john burke said,
June 5, 2020 @ 9:23 am
@Chris Button: Could Ickx be Flemish? But if it is, I don't know how a Flemish speaker would pronounce it.
Robert Coren said,
June 5, 2020 @ 9:45 am
I'm pretty sure that "в" ("v") is a common word in Russian, although I forget what it means.
Peter B, Golden said,
June 5, 2020 @ 9:49 am
Georgian has the lovely ბრწყინვალე brcqinvale (brdsqinvale) (brilliant, glittering), ზღვადქცეული (zġvadk'c'euli) (like a sea) and many others
Victor Mair said,
June 5, 2020 @ 10:40 am
"…it boils down to 'look at all the consonants' as commented on by an anglophone named Stu who does not know the language."
Hardly.
Please read the comments.
This is one of the ways Language Log works and, as with this post, it has worked very well. Nothing "Too bad" about it.
Max Wheeler said,
June 5, 2020 @ 10:56 am
And if you're looking (outside Polynesia) for words with few or no consonants, how about Manx leoaie 'lead (Pb)', oaie, 'grave', ee-ee ee 'she will eat'?
Daniel said,
June 5, 2020 @ 11:02 am
Indeed. Also, the number of syllables is not clear-cut. When a schwa is elided, it can lengthen nearby consonants, which may be perceived to syllabify them, or not. I perceive the "el" of the words "vowel" and "towel" to be a lengthened [l] compared to "owl" or "howl", but they still might be considered one-syllable words.
I noticed a similar subtle distinction with Donald Trump's tweet that rhymed "violence" with "silence". This is not an exact rhyme because there is the remnant of the schwa in "violence", but I perceive it as not constituting a syllable but instead being incorporated into the [l], so these two words both have two syllables that rhyme close enough.
Chris Button said,
June 5, 2020 @ 11:02 am
@ David Marjanović
Vowels are sonorants. Nothing language dependent about it.
George Sprague said,
June 5, 2020 @ 11:52 am
@Robert Coren
В is in/at. Sometimes it will be Во, depending on what sound follows it.
Takes prepositional case or accusative. Pronounced "Ve."
Christian Weisgerber said,
June 5, 2020 @ 12:43 pm
@Robert Coren
Russian "в" ("v") is a preposition, meaning "in" (as far as prepositions can be translated).
The kicker is that it is actually cognate with English "in". Etymologically, they're the same word, both descended from PIE *én. For English that's not surprising. Slavic at some point didn't like initial vowels, so it added a prosthetic v- for *vъ(n). It also didn't like final consonants (law of open syllables), so the -n was dropped. And eventually the very short vowel was elided, resulting in, well, "v".
TR said,
June 5, 2020 @ 12:53 pm
they don't put any vowel at all in squirrel, just a syllabic /r/ and a syllabic /l/
No, a consonantal /l/: it's a monosyllable (for me anyway), rhyming with girl.
Nathan said,
June 5, 2020 @ 5:52 pm
For me, squirrel and girl have two syllables each.
Dara Connolly said,
June 5, 2020 @ 6:19 pm
Max Wheeler:
words with few or no consonants, how about Manx leoaie 'lead (Pb)', oaie, 'grave', ee-ee ee 'she will eat'?
In Irish these same words would be written "luaidh", "uaigh" and "ithfidh (s)í" but the pronunciation is similarly lacking in consonants.
For an Irish person, written Manx looks incomprehensible but can be deciphered merely by sounding out the words or reading them aloud.
ohwilleke said,
June 5, 2020 @ 8:44 pm
Welsh is going to start feeling jealous.
Truly though, there are letter combinations which have specific understood pronunciations in other languages or even other dialects of English that are just not in the American English codebook.
In New Zealand, for example, "wh" codes roughly as an "f" in American Englsh.
In Thai language words, "ng" codes rough as "w" or "wh" in American English.
You need to know whether a name or word has English or Spanish origins to know how to pronounce the letter "j" in a word, and the "ll" letter in Spanish doesn't code as the same letter combination does in English.
The letter "X" codes for all sorts of different sounds in different languages that are seemingly at odds with the American way to convert it into sounds, so you really need to know if you have a word with Chinese, Aztec or Greek origins to know how to pronounce it.
Consonant soup as in this case or Welsh or Hebrew or Arabic words with just the written characters directly transliterated all require knowledge of how to turn them into sounds.
Honestly, it would be nice to have a quick small single volume guide for the most commonly occurring such letter to sound coding rules for all Romanized written languages one has any chance of encountering for people who need to pronounce words that don't have obvious American English pronunciations.
Michael Watts said,
June 5, 2020 @ 11:50 pm
This is interesting. My instincts:
1. "Vowel", "towel", and "owl" all rhyme with each other; they are two syllables. "Howl" doesn't rhyme with the other three and isn't two syllables.
2. "Girl" and "squirrel" rhyme with each other and are not two syllables. For example, in a song, they would occur in a place that fit a single syllable.
3. That said, "girl", "squirrel", and "howl" feel like a little more than one syllable. Maybe 1.1 syllables.
Well, speaking as an American, I have absolutely no idea what Phillip Minden meant by distinguishing "cverk" from "cvrk". But I also don't know what David Marjanović means by saying Americans don't have a vowel in "squirrel". My training is to represent the NURSE vowel as /ɚ/ (in all contexts), which is pretty transparently intended to be a schwa with rhoticity added. I'm not sure why this wouldn't be considered a vowel.
https://english.stackexchange.com/questions/192979/difference-between-ipa-%C9%9A-%C9%B9-and-%C9%9D seems relevant. The answer there specifies that the highlighted characters in "murder" represent a single "r-colored vowel"… but that in "murdering", they do not represent an r-colored vowel, since, as the following syllable begins with a vowel, the "r" must be interpreted as a consonant.
This point of view is then severely undermined by the analysis of "murderer" as /mɝdɚɚ/. (Where, according to the answer, ɝ is exactly the same sound as ɚ, but is the written form appropriate to stressed syllables.) I think it's also undermined by the fact that the syllables sound absolutely identical in "murder" and "murdering".
The only conclusion I can really draw from that answer is that there are no real principles involved in choice of transcription or in whether to call the NURSE vowel a "vowel" or not. I would be interested in seeing examples of the type of thing Phillip Minden / David Marjanović have in mind.
Philip Taylor said,
June 6, 2020 @ 3:14 am
A native British English speaker's response :
1) "bowel", "dowel", "foul", "fowl", "howl", "jowl", "owl","rowel", "towel", "vowel", "yowl" all monosyllabic in all but the most careful speech, and all perfect rhymes. Those ending "el" might be pronounced with "one and a half" syllables if the speaker were enunciating very carefully.
2) "girl" v. "squirrel" : almost as different as two words could be. /ɡɜːl/ v. /ˈskwɪ rəl/ (and the rhotic equivalents thereof). One syllable v. two, different vowels.
Bruce Foster said,
June 6, 2020 @ 5:57 am
I had a friend in college who noted that Czech fashion models were really beautiful. He created Rinap's Rule. The fewer the vowels the hotter the girl. Have no idea if it has any empirical validity, but it was a good joke!
Tom Dawkes said,
June 6, 2020 @ 6:22 am
Two thoughts on this.
1. For each language the orthography has its own conventions, which will in many cases run counter to this of other languages. For instance, the digraph 'ch' has in English at least THREE different pronunciations: 'cheese', 'champagne', 'character'. In Italian 'ch' is always as in 'character', but in Spanish it's always as in 'cheese' — which is why we should say 'machismo' and 'macho' alike with the 'cheese' sound.
2. On the syllabic 'r', Americans 'surely' know how that's pronounced? Have they not seen 'Airplane' and 'Don't call me Shirley!'?
Tom Dawkes said,
June 6, 2020 @ 6:27 am
Another thought on 'ou' or 'ow'. The two forms 'flower' and 'flour' are historically the same word. OED has the note:- Flour: Etymology. A specific use of flower n.; compare French fleur de farine the ‘flower’ or finest part of the meal. (Johnson 1755 does not separate the words, nor does he recognize the spelling flour.) a. Originally, the ‘flower’ or finest quality of meal; hence, the finer portion of meal (whether from wheat or other grain) which is separated by bolting.
anhweol said,
June 6, 2020 @ 8:36 am
@Bruce Foster: Sadly for Rinap's rule, pretty much all Czech female names in citation form are going to have at least 3 vowels (final -a or -e on given name, surname with feminine ending -ová). Admittedly adjectival surnames only impose a final -á and there are a few indeclinables but they tend to have vowels in the final syllable anyway. There are some surnames with a syllabic liquid in the root syllable to help keep the count down (e.g. Vrbová as against masc Vrba). But I am having trouble generating a plausible full name with under 4 vowels (e.g. Eva Vrbová or Blanka Bílá – can anyone improve? As against that, male names with just one vowel in the given name and none in the surname are easy, e.g. Petr Vlk. What that says about hotness, modulo your particular gender and orientation, is left as an exercise for the reader…
Andrew Usher said,
June 6, 2020 @ 10:39 am
OK, here's my summary:
Yes, owl, howl, vowel, towel, etc. are rhymes. Micheal Watts's assertion that 'howl' is different is most likely a misperception or idiosyncrasy. This is not limited to 'owl' but occurs for any of the 3 historic diphthongs followed by either liquid; although English has no 'oir' examples, we can be sure they'd follow the rule. The spellings flour/flower are a good example of this.
In most American accents, this additionally is possible for the sequences in mail, meal, and mule (but not 'mole' – that would be substandard); one and a half syllables. The final liquid can vary between a non-syllabic liquid, a syllabic liquid, and for /l/ a sequence of schwa + liquid. Brits apparently don't usually do this but may in specific words like 'jewel'.
Further, pace Philip Taylor, for Americans in even the most careful speech 'squirrel' does rhyme with 'girl' in having NURSE in the first syllable. If I make it two syllables in slow speech it could be /skwɝ əl/ – unpacking the second syllable, but leaving the invariable NURSE.
All these words can be analysed as one or two syllables and could be set as either in singing or poetry; however, always choosing two could sound humorous or childish.
Describing the American NURSE sound as a 'vowel' or not is a matter of convention – there is no real disagreement over its quality. I say it's best seen as a syllabic 'r' simply because it has essentially the same sound as syllabic 'r' in unstressed syllables, as is shown with the examples of 'murder' etc. So it would be better to write, even though it's against convention, /ɹ̩/ or /r̩/ for both /ɚ/ and /ɝ/ in GA or similar rhotic dialects.
k_over_hbarc at yahoo.com
Philip Taylor said,
June 6, 2020 @ 11:14 am
"English has no 'oir' examples" — we have 'coir' (made from coconut), and arguably "choir" (does not rhyme with 'coir' in <Br.E&gr;). The first is monosyllabic (I don't think anyone will disgree with this), the latter to my mind has "one and a half" syllables.
Thomas Rees said,
June 6, 2020 @ 2:05 pm
OED: The spoken word is still quire, though since the close of the 17th cent. this has been fictitiously spelt choir, apparently as a partial assimilation to Greek-Latin chorus, or French chœur.
The spelling quire has never been altered in the English Prayer-book. Some people affect to pronounce choir /kɔɪə(r)/.
Philip Taylor said,
June 6, 2020 @ 2:32 pm
Well, I think it is at least arguable that modern "choir" comes not by "partial assimilation to Greek-Latin chorus, or French chœur" but rather from Old English "chor", which translate into modern English "choir"
Michael Watts said,
June 6, 2020 @ 3:11 pm
To me, "choir" rhymes with "fire" — or to use an even closer match, it is the last syllable of "require". This is the prototype of the awkward syllable-and-a-half.
"oir" examples that come to mind are "foyer" (in a pronunciation that was used on Downton Abbey; the normal American pronunciation has no R) and "lawyer". But these are possibly not examples of words that underwent the full course of historical development that I believe Andrew Usher has in mind.
Philip Taylor said,
June 6, 2020 @ 3:21 pm
I have never listened to (?watched?) Downton Abbey, but I believe that it was set (and therefore probably made) in the UK. But as a lifetime UK resident I have never heard "foyer" pronounced other than /ˈfɔɪ eɪ/, rather different to our /ˈlɔː jə / — how do you remember it from Downton Abbey, Michael ?
Andrew Usher said,
June 6, 2020 @ 7:07 pm
Correct; in my speaking of 'oir' words, I meant the (empty) set that was originally monosyllabic. 'Choir' doesn't count because that's not how it is or ever was pronounced, and 'coir' – a good find – doesn't either because it comes from a foreign word that was 2 syllables, showing the merger had already happened!
I propose to call this phonemic developement, if it's not already named, the roil/royal merger, reflecting one homophone produced by it, following the 'cot/caught merger' and similar.
For 'foyer', the standard American pronunciation is, and has been for a century or so, as spelled; what Michael presumably heard (or thought he heard) on 'Downton Abbey' is the non-rhotic version of that i.e. /fɔɪ ə/.
Daniel said,
June 6, 2020 @ 9:40 pm
Andrew Usher, I appreciate your summary, especially the notion of one-and-a-half syllables.
Michael Watt's assertion that "howl" is different may be an idiosyncrasy, but I share it. To me, "howl" ends in a light L, but the ones that have an extra (half) syllable end in a dark L. This may be because "howl" as a verb often takes additional endings, so that the L ends up beginning the next syllable (e.g. "howling", "howler"). In contrast, I perceive "howled" and "toweled" to be rhymes, both with dark, half-syllable Ls.
Michael Watts said,
June 7, 2020 @ 2:34 am
I find this claim surprising; to me the standard is /fɔɪ eɪ/. (As you can see, I noted above that the American pronunciation "has no R".)
Michael Watts said,
June 7, 2020 @ 2:41 am
What I remember from Downton Abbey is a use of the word "foyer" in what I would describe as an "as spelled" Britishized pronunciation. It struck me as a Britishism. I particularly remember a strong-weak stress pattern on the word (where /fɔɪ eɪ/, in my mind, is more of a strong-strong).
I should note that the transcripts available online at http://transcripts.foreverdreaming.org/viewtopic.php?f=151&t=9393 do not include the word "foyer" at all. Maybe I'm remembering something wrong! But in my defense, they also note that they need help improving the quality of the transcripts. I don't want to rewatch the show trying to find the mention, or pay $50 for the official transcripts. :/
Chas Belov said,
June 7, 2020 @ 3:52 am
For me, squirrel and girl have 1.25 syllables each.
Anthony said,
June 7, 2020 @ 7:43 am
I pronounce "violence" like "silence," and also "violet" like "eyelet." But "inviolate" has four syllables for me, and it has been deemed worthy of note that in Yeats's "Far-off, most secret, and inviolate Rose" it must be elided to three syllables for the line to scan. Does anyone normally pronounce it that way?
Steve Bacher said,
June 7, 2020 @ 8:56 am
(Ralph Hickok: In the early 1950s, Temple University had a basketball player named Bill Mlkvy, of Slovakian descent. Sportswriters called "the Owl without a vowel" (the Temple sports mascot is an owl). Of course, that wasn't quite accurate because the final "y" functions as a vowel.)
Eagles producer Bill Szymczyk was called "the vowelless producer." Of course, the same observation about "y" applies.
Philip Taylor said,
June 7, 2020 @ 9:30 am
"inviolate Rose". Pronounced by itself, "inviolate" has (for me) four distinct syllables. But in the context of "inviolate Rose", at most 3½, perhaps because I cluster the final /t/ of "inviolate" and the initial /r/ of "Rose".
Robert Coren said,
June 7, 2020 @ 10:28 am
George Eliot's Middlemarch, written in the 1860s, uses the spelling "quire".
Andrew Usher said,
June 7, 2020 @ 11:15 am
Michael Watts:
All American dictionaries prefer my pronunciation of 'foyer', and certainly I've never gotten any sense of its being wrong or questionable. I'm astonished that anyone could get the opposite impression.
Anthony:
No, I wouldn't think anyone categorically reduces 'inviolate', and doing so is, Ii think, surely more common for 'violet' than for 'violence' – I think my usual pronunciation for all three would have a syllabic 'l'. Again that works acceptably in poetry without having to make an effort to say 'invi'late'.
Andrew Usher said,
June 7, 2020 @ 11:21 am
I also wanted to reply to David Marjanovic's observation. The distinction between the schwa and non-schwa pronunciations of syllabic 'r' is really a continuum, at least auditorily. Only if you could prove the schwa to be _non-r-colored_ could you say that it was phonemically there.
To demonstrate, I recorded myself saying a NURSE word with a long trilled r, then cut it short after one contact. The result was hardly distinguishable from my rendition of the same with schwa (like British NURSE) + tapped r.
David Marjanović said,
June 7, 2020 @ 1:34 pm
Chris Button:
Of course, but remember what I was replying to:
That's what I disagree with. If we want to distinguish just two basic types of sounds, the usefulness of where to put that distinction will depend on the language.
=========================
Not at all. Basque doesn't use the letter C.
However, it uses x for /ʃ/, and tx for /t͡ʃ/; maybe that's what you're thinking of?
[ʔɪks] of course.
Welsh does not have unusual consonant clusters or syllabic consonants. W is used for [u], because u is used for [ɨ] (or even [i], or, when unstressed, for [ə]); y is used for [ə] (even when stressed).
Distinguishing [t͡svɛrk] from [t͡svr̩k]. Very much outside any American English sound system (but [ɛr] is common in Scottish accents, let alone Spanish).
Phonetically, it's an approximant, specifically [ɻ] – the same thing as non-syllabic /r/. It's not really more a vowel than the syllabic [l] in the same word is, or than the [n] in button, or than [r] and [l] are in Czech.
Phonologically, the NURSE "vowel" differs from the other syllabic consonants of rhotic American English in being stressable. But that's a fact of how the language treats it, not a phonetic fact. In Czech, [r] and [l] are stressable, but they're always short, while the vowels all come in short and long versions; in the closely related Slovak, however, the length distinction of the vowels is extended to the syllabic consonants.
Yeah. [mɻ̩dɻ̩], [mɻ̩dɻ̩ɪŋ].
Compare generic non-rhotic British: [mɜːdə], [mɜːd(ə)ɹɪŋ].
Recently I came across Weird Al's song about aluminum foil and was very surprised to find the foil (first occurrence at 0:48) to consist of two syllables – for a non-rhotic audience I would not hesitate to spell that foierl.
Meanwhile, disyllabic milk has been creeping into the Queen's English (there was a scientific paper on this 20 years ago), and anyone soccer-adjacent will have encountered trisyllabic England.
See also: diaper without the a.
Does anyone normally pronounce inviolate at all? :-) I don't think it's in the active vocabularies of many people; and if there are any who do use the word on a regular basis, they've probably learned the word from reading and may just have given it a spelling-pronunciation.
Andrew Usher said,
June 7, 2020 @ 2:48 pm
I don't know if that's what Phillip Minden meant by "cvrk, not cverk" because he was addressing an English-speaking audience for whom the distinction doesn't make a lot of sense. I am aware of and can easily pronounce foreign 'er' as in 'Verdi' – but did he believe the same of everyone?
> Recently I came across Weird Al's song about aluminum foil and was very surprised to find the foil (first occurrence at 0:48) to consist of two syllables – for a non-rhotic audience I would not hesitate to spell that foierl.
This follows the rule as first stated. Read more carefully – I stated that in words that rhyme with 'mole', it would be substandard. 'Foil' does not rhyme with 'mole'! Not to mention that that pronunciation of 'foil' is normal almost everywhere (among native speakers) even if the rhythm of the song exaggerates it.
> Meanwhile, disyllabic milk has been creeping into the Queen's English (there was a scientific paper on this 20 years ago), and anyone soccer-adjacent will have encountered trisyllabic England.
This is schwa-epenthesis between consonants, a phonetically different phenomenon – and 'Engeland' is not likely to expand beyond soccer chants.
Philip Taylor said,
June 7, 2020 @ 3:37 pm
Singing is, of course, not speech, but listen to Maddy Prior between 00:36 and 00:38 singing Hard Times of Old England — a very clear tri-syllabic "Eng-ge-land". And (perhaps because of exposure to Czech for over 25 years) "cvrk, not cverk" makes perfect sense to me — I was once congratulated by two Czech hitch-hikers in Egham (Surrey) on my pronunciation of "Brno".
Barbara Phillips Long said,
June 7, 2020 @ 4:22 pm
“The owl without a vowel” example seems to me to be more about the way the English language is taught in the U.S. and the strict categorization of the alphabet into vowels and consonants. In my experience, the “and sometimes y” business about vowels gets brushed over, and there is little or no discussion of liquids. The lack of linguistics input into U.S. K-12 language arts textbooks has often been discussed here.
For native English readers, the “no vowel” issues in words and names from other languages seems to be more about orthography than pronunciation, as Phillip Minden has already noted.
Michael Watts said,
June 7, 2020 @ 10:39 pm
I tend to agree with the later point that under normal circumstances, the word "inviolate" would never be used at all, making it difficult to talk about what it would sound like in normal use.
But what struck me more about this is the idea of clustering the final /t/ of a preceding word with the initial /r/ of "rose". I would definitely not do this, because a /tr/ cluster obligatorily becomes [tʃr], and making this reduction across a word boundary is unusual outside of common collocations. I would instead reduce the final /t/ to a glottal stop.
Vanya said,
June 8, 2020 @ 12:01 am
czikszentmihalj
Funnily enough in Czech orthography this would become more vowel heavy — čiksentmihaj
In English Chicksaintmeehigh.
Vanya said,
June 8, 2020 @ 2:00 am
Or better Chick Saint-Meyhigh
That might be perceived as „consonant heavy“ by a non-English speaker, but looks pretty normal to a native speaker.
Philip Taylor said,
June 8, 2020 @ 2:51 am
Michael, I can't detect any [ʃ] in my /tr/ cluster when I say "inviolate Rose". Might that be because I trill my 'r's, and why do you say that 'a /tr/ cluster obligatorily becomes [tʃr]' ?
Andrew Usher said,
June 8, 2020 @ 7:58 am
The /tr/ affricate automatically 'becomes' (sounds like) [tʃr] if you use the 'bunched r', as Michael presumably does but I don't in this environment, and the same for the analogous /dr/.
I agree, though, that I don't think I'd make this assimilation in reading poetry, even though it's common in speech – the final /t/ of 'inviolate' would likely get pronounced as my usual unreleased stop.
Finally I find that making /r/ part of a /dr/ or /tr/ cluster absolutely prevents trilling it; it must remain a fricative, which is my most common form of initial /r/ anyway. So I'm not sure what Philip Taylor is really doing.
Chris Button said,
June 8, 2020 @ 8:09 am
@ David Marjanović
Respectfully, I'm afraid you're missing the point. A sonorant may become a vocalic nucleus at the phonetic level, but that does not require a phonological distinction. It also affords a certain flexibility when accounting for surface phonetic realizations. Forget "squirrel", let's talk about how many syllables "real" has for example.
If you wanted to challenge the supposition, you could have pointed out that "sonorants" aren't always voiced, despite their name. But there is a good reason why voiceless nasals and the like are often unstable in the same way that English "voiced" obstruents usually aren't actually voiced.
Philip Taylor said,
June 8, 2020 @ 8:12 am
I adopted the trilled-r at a very young age, when I discovered that my untrilled 'r's sounded like 'w's. And perhaps because of this, my /tr/ retains that trilled 'r' and it doesn't become a fricative as far as I can tell.
David Marjanović said,
June 8, 2020 @ 10:07 am
Not in the case of [mɪɫ̩k], which you'd have to analyze as /mɪəlk/ if you wanted to assume an underlying /ə/, definitely not as **/mɪlək/.
How did it get into soccer chants in the first place?
I've listened to the whole thing. The refrain contains England twice. Every time the refrain comes up, I hear England with syllabic [l] the first time and with [əl] the second time… I find that quite striking! But yes, three syllables every time.
I certainly am, because I don't understand what any of this has to do with the original blanket statement that "[a]n obstruent/sonorant distinction is phonologically far more useful than a consonant/vowel one."
Unrelatedly as far as I understand:
That, too, is language-dependent. In southeastern Standard German Karl "Charles" and kahl "bald" are a minimal pair for syllabicity: /kaːl̩/, /kaːl/. Assuming an underlying /ə/ would require assuming a vowel phoneme that never surfaces as a phonetic vowel.
Chris Button said,
June 8, 2020 @ 10:40 am
@ David Marjanović
It's to with what's called the sonority hierarchy.
Certain languages will only allow the most sonorous to occur as the syllable nucleus; other languages will allow things far lower down on the scale as the nucleus. That distinction is language dependent. However, what is not language dependent is that all of them are sonorants.
Now, here too, if you would like a counterpoint (again one that we have discussed on LLog many times), some syllables in some languages do not rely on a sonorant for the vocalic nucleus but can also use a fricative (think of something like "pssst" in English)
Chris Button said,
June 8, 2020 @ 10:41 am
@ David Marjanović
It's to with what's called the sonority hierarchy.
Certain languages will only allow the most sonorous to occur as the syllable nucleus; other languages will allow things far lower down on the scale as the nucleus. That distinction is language dependent. However, what is not language dependent is that all of them are sonorants.
Now, here too, if you would like a counterpoint (again one that we have discussed on LLog many times), some syllables in some languages do not rely on a sonorant for the vocalic nucleus but can also use a fricative (think of something like "pssst" in English)
Chris Button said,
June 8, 2020 @ 10:43 am
The sonorants at the top of the sonority hierarchy being the "vowels"
Andrew Usher said,
June 8, 2020 @ 6:13 pm
Philip Taylor:
The inability to post sound files here means that and more detailed discussion about how you or I pronounce /r/ would be unproductive. I suspect the difference is not as large is you seem to claim.
Also, I had a similar experience when younger, only it was not /r/, but /l/ that sounded too w-like. And indeed I did try to adopt a clearer /l/, and no longer get misunderstood.
David Marjanovic:
I guess I mistook you. I assumed /mɪlək/ was the disyllabic pronunciation referred to (on the model of e.g. /fɪləm/ 'film'). The other one I do know, and think it clearly a result of l-vocalisation, which can create new syllables as r-vocalisation did.
I think your Karl/kahl example overreaches in claiming the vowel _never_ surfaces, especially since it must have been there before the loss of /r/. Are you trying to postulate two different /l/ phonemes, one syllabic and one not? I think that's what was being denied.
Chris Button:
Vowels can also be devoiced (even in English) … that's not a sound basis for a distinction.
Chris Button said,
June 8, 2020 @ 6:30 pm
@ Andrew Usher
The classic example for voiceless vowels is Japanese. What distinction are you referring to?
Andrew Usher said,
June 8, 2020 @ 6:47 pm
The distinction you implied between vowels and 'sonorants'. Both vowels and non-vowel sonorants can be devoiced, but are normally voiced.
Chris Button said,
June 8, 2020 @ 7:18 pm
I didn't imply that. I think you are confusing what David M said with what I said. My point is that vowels are sonorants and that all sonorants can be voiceless but are called sonorants because they are sonorant by default as opposed to osbtruents, which are not. The discussion came from me saying that a sonorant/obstruent distinction is far more useful than a consonant/vowel distinction for phonological (not phonetic) analysis. I further pointed out that the only thing that we can put down to being dependent on a particular language, dialect or even idiolect in question is how far down the sonority hierarchy a sonorant can appear before it cannot occur as a syllabic nucleus.
Bloix said,
June 8, 2020 @ 10:56 pm
"think of something like "pssst" in English"
We could think of "pistachios," couldn't we, where the first "i" pretends to be a vowel.
But we had this argument over "rhythm" a couple of months ago.
https://languagelog.ldc.upenn.edu/nll/?p=46316&utm_source=rss&utm_medium=rss&utm_campaign=words-without-vowels
Bloix said,
June 8, 2020 @ 11:00 pm
PS – Chris Button, thank you for the word "sonorant." I now understand that, when I commented two months ago that to me there are no "vowels" in rhythm, I was saying that there are two sonorants – "th" and "m" – instead.
David Marjanović said,
June 9, 2020 @ 8:07 am
…I think I get it now: you used sonorant as a phonetic category (language-independent), but vowel to mean "syllable nucleus" (language- to idiolect-dependent)?
If so, we agree on everything! I just wouldn't use vowel that way. It isn't hard to come up with a purely phonetic definition for that term that has little to do with syllable nuclei.
It was never there in this particular word since its first attestation. Charlemagne seems to have been a monosyllabic [karl], and in partially and fully rhotic German accents today you get [karl̩] or [kaʀl̩] if I'm not gravely misremembering things – two syllables, but still just one vowel. At some point before Charlemagne, a vowel was lost at the other side, at the end of the word.
(Funnily enough, the name is Karel in… wait for it… Czech. But that nominative singular is backformed by analogy from the other cases that lack this e.)
You're right that the vast majority of syllabic consonants in modern German do result from the historic loss of preceding vowels. But I'm not trying to make a point about history, but about current sound systems: in southeastern Standard German, there is no [ə] in any word, there's nothing between [ɛ] and zero – [ɐ] exists, but it's strictly an allophone of /r/.
Yes, exactly. I'm aware this would be a very rare feature worldwide, but I can't see a way around it.
The two pronunciations of rhythm I know (some people use one, some the other, and some both) are [ɹɪðm̩], with one vowel, one non-syllabic sonorant ([ɹ]), one syllabic sonorant ([m]), one fricative ([ð]), and two syllables, and [ɹɪðəm], with two vowels, two non-syllabic sonorants, no syllabic sonorants, the same fricative, and again two syllables.
Philip Taylor said,
June 9, 2020 @ 9:11 am
Andrew — Here is a link to an audio file to enable us to compare our pronunciations of /tr/, and perhaps endeavour to establish whether or not there is an intrusive [ʃ] in my trilled version — I am reasonably confident that there is not, but that there is one in my untrilled version. Could you provide a similar link to your version(s), please ?
Chris Button said,
June 9, 2020 @ 10:15 am
@ David Marjanović
No that's not what I mean. A vowel (a useful category in phonetic analysis) is simply a kind of sonorant. All languages allow vowels to form the syllabic nucleus. Other sonorants may or may not form the syllabic nucleus depending on the language. Hence, obstruent/sonorant is more useful in phonological analysis than consonant/vowel.
@ Bloix
Your pronunciation of "pistachios" perhaps includes a voiceless vowel in the first syllable.
The "th" in "rhythm" is not a sonorant. It's a voiced fricative which is an obstruent. It is reliably voiced as an obstruent (which are by default unvoiced) due to it being surrounded by voiced elements, but it is the coda of the first syllable (not the onset of the second as some might have, and certainly not the nucleus)
Narmitaj said,
June 9, 2020 @ 1:53 pm
Maybe trisyllabic Engerland originates in and has historic echoes of the original Englaland – "The name 'England" is derived from the Old English name Englaland, which means "land of the Angles"'
https://en.wikipedia.org/wiki/England#Toponymy
or in Old English:
https://ang.wikipedia.org/wiki/Englaland
Victor Mair said,
June 9, 2020 @ 7:59 pm
From Olga Lomová:
I have never met anybody with the Cvrk family name. More common is Cvrček, which means cricket. I have checked and there are currently 143 people of this surname, 19 of them in Prague. Given the other occurences it is possible they originated in South Bohemia (https://www.kdejsme.cz/prijmeni/Cvrk/). To my knowledge, surnames started to be given to ordinary people only in 18th century, during the rule of Joseph II. This Austrian emperor – among other things – admired the „enlightenment despotism“ of China as portrayed in Jesuit sources and emulated them in his reforms (and even in perfoming farm work: https://cs.m.wikipedia.org/wiki/Soubor:Joseph_II_is_plowing.jpg). Fascinating topic.
Chris Button said,
June 10, 2020 @ 7:08 am
We've probably been remiss not to bring up the notion of mora (syllable weight) and its association with sonorance/vocalism
Andrew Usher said,
June 10, 2020 @ 7:23 pm
Philip Taylor:
I did get around to replying in kind. I normally don't upload any files directly on the internet (rather than by e-mail or FTP), but try this: https://file.io/FxQghfwL for my similar sound file. I needed several takes to get the 'bunched r' examples straight, and I know they still won't sound like someone natively using it, but hopefully at least suggest it.
I do think my normal /tr/ sounds quite like yours in the untrilled 'trill', with perhaps a trace of sibilance if listened for but certainly not enough to justify transcribing [tʃr]; but as you admitted you couldn't really do it in 'inviolate rose', as I think I can.
Which – now that I've looked at the original poem it's from – actually seems to have been beside the point! It has many extra unstressed syllables in its meter, and they clearly can't all be elided. It gives me no reason to think he thought of 'inviolate' as three syllables.
Chris Button said,
June 10, 2020 @ 8:55 pm
@ Philip Taylor
Your comment on "tr" reminded me of this old blog post by the phonetician John Wells (author of the excellent Longman Pronunciation Dictionary, which I know from your other posts you appreciate as much as I do)
http://phonetic-blog.blogspot.com/2011/03/how-do-we-pronounce-train.html?m=1
Andrew Usher said,
June 10, 2020 @ 9:01 pm
Well, that link disappeared immediately, and I tried again and the same thing happened. I did make the file and can certainly provide it by e-mail any time.
Sorry, I'm not going to be messing around further right now.
Philip Taylor said,
June 11, 2020 @ 3:27 am
Thank you Andrew/Chris. Chris, you are quite correct, I have my LPD open every time I read an article here, and greatly miss John Wells' blog, but one thing leapt out at me when I read the section you cited :
It was the "Obviously, century is basically ˈsentʃəri" that hit me. Whilst I can say /ˈsentʃəri/, it sounds very 'casual' to my ear, and were I to use the word in any sort of a formal context I would pronounce it /ˈsentjuri/ — the /tj/ -> /tʃ/ shift always grates when I hear it on Radio 4. My LPD has /ˈsentʃ ər‿i/, so at least we know that JW is consistent !
Andrew, to e-mail me your sound file, just prepend "p.taylor@" to the domain linked from my name in the comments.
And totally OT, but given the mention of John Wells — does anyone have news of John Maidment ? I had very much hoped to meet up with him when we moved to Cornwall, but his "blogjam.name" disappeared shortly after we moved, and I have heard nothing of/from him since.
Andrew Usher said,
June 11, 2020 @ 6:03 pm
File sent, and no further comment needed on that.
Your pronunciation of 'century' is one more piece of evidence that your accent is in some ways out of step – as John Wells himself speaks fairly conservatively, and even he didn't consider that an option (in normal use).
Andrew Usher said,
June 12, 2020 @ 7:34 pm
I should have said that having 'out of step' pronunciation is not itself a problem; it's not realising they are and expecting dictionaries to reflect your preferences that seems strange. I have many such myself, actually.
I would like at least to know that you got and listened to the file (thus you weren't just trying to make me waste my time).
Philip Taylor said,
June 13, 2020 @ 4:15 am
Audio file well received, Andrew, and already listened to, but tied up with other matters at the moment so feedback on temporary hold.
As to "out of step", my 1933 OED has /se·ntiŭri/, which I regard as ample justification for allowing me to believe that my pronuncation is entirely in step with RP and therefore with accepted norms. And much as I admire John Wells, his work and in particular his LPD, I cannot help but wonder whether his much-publicised research into (and therefore familiarity with) Estuary English may in some very small part have influenced his choice of pronunciation / phonetic transcription for words such as "century", for which both 'evolved' and more conservative pronuciations are in widespread use.