Search Ace Linguist

May 27, 2019

The Colt-Cult merger?

I was once at a dinner party with some colleagues, and a co-worker of mine was explaining to another co-worker a story they had heard about the eclipse which had happened in 2016.

A: It turns out that they were actually in a cult.

B: A colt? Like a baby horse?

A: No, not a colt, a cult, with a leader.

B: Oh, you mean a cult!

What had happened in this exchange? Co-worker A pronounced 'cult' as [kolt], with a vowel similar to 'cold' [kold]. But B pronounced 'cult' with a low, unrounded vowel: [kʌlt], with the vowel of 'cut' [kʌt]. Moreover, co-worker B pronounced 'colt' and 'cult' differently: [kolt] and [kʌlt], while co-worker A pronounced the same: [kolt]. This is an example of a new merger I'm tentatively calling the 'colt-cult' merger (although I've also seen it called the hull-hole merger). This is when /ʌl/ and /ol/ merge to /ol/, so that words like 'colt', 'cult', 'hull', and 'hole' all have the same [o] vowel. The realization of the 'ul' vowel can be lower, so "culled" would be [kɔld] (and similar to old American 'called').

This seems to be a merger that happened on the phonological level, because before I started reading about English dialectology, I had no idea that "cult" and "colt" were supposed to have different vowel qualities. Perhaps "cult" had a slightly lower vowel, but to me they were similar enough that you could make a pun out of it. I recall watching an episode of Bones where the protagonist, Temperance Brennan, very clearly said "skull" [skʌl] with a low, back, unrounded vowel, and thinking that was odd. It turns out, historically, the odd one is me.

There are some particularities to this. All the people I know with this merger speak American English, but it's not limited to any region. Co-worker B was from Pennsylvania, whereas I am from Florida. The Americans I've met who have this pronunciation pronounce /ol/ without a diphthong. This is in contrast to /o/ anywhere else. This means "go" has a diphthong but "gold" has a monophthong.

While there are many historic vowel changes before /l/, this merger seems to be on-going because there isn't a lot of research or awareness on it. Information on it includes this uncited quote from Wikipedia, but about English English: "The hull–hole merger is a conditioned merger of /ʌ/ and /oʊ/ before /l/ occurring for some speakers of English English with l-vocalization. As a result, "hull" and "hole" are homophones as [hɔʊ]." The realization of this merger would certainly be quite different from the American one, because the American one retains /l/ in both situations. Labov, Ash, and Boberg's he Atlas of North American English (2006) also mentions that this is a merger in American English that might require more attention in the future, but don't discuss it in depth otherwise.

Wild Speculation on where it came from

Alright, so here comes some wild speculation completely off the top of my head on why this is happening. Don't quote me on this, because these aren't developed ideas. This is just some fun free association to think about what other sound changes might be related to this merger.

One interesting commonality between co-worker A and I is that we both have the cot-caught merger. This means that we do not use the [ɔ] vowel in words that used to have it, such as "bought", "caught," "caller", and "law." We instead use the [ɑ] vowel and put those words in the same category as "bot," "cot," "coller," and "la." If you are a North American and you also have the cot-caught merger, then you may not be familiar with which words have the /ɔ/ sound or what the /ɔ/ sound even is. [ɔ] is pronounced much lower and backer in the mouth than [ɑ] - I find it helps to imagine you're an aristocratic British man or some other character from Downton Abbey, and saying "law" with as low a vowel as you can muster. As a rule of thumb, words spelled with 'aw' (law, caw), 'au' (caught), and -all (ball, all) tend to have the /ɔ/ sound. Now, we are especially interested in the ones with an /l/, so words like "ball" and "caller." While there are many words with /ɔl/, there are relatively fewer words that are pronounced /ɑl/. This list from Wiktionary shows the dire state of words that rhyme with 'oll'. (Note that this excludes words that also have the 'oll' sound like 'acknowledge'):

  • boll (in some pronunciations; see also -əʊl)
  • coll
  • doll
  • gnoll
  • knoll (in some pronunciations; see also -əʊl)
  • lol
  • loll
  • moll
  • noll
  • pol
  • poll (in some pronunciations; see also -əʊl)
  • quoll
  • troll (in some pronunciations; see also -əʊl)
  • vol

This means that the cot-caught merger doesn't introduce as much ambiguity before /l/. Perhaps there might be some confusion between 'caller' and 'collar,' or 'mall' and 'moll,' but overall it's not the biggest deal. A curious result of this is that I've noticed that a lot of Americans who otherwise have the cot-caught merger will still keep [ɔ] before 'l'! So they'll pronounce 'caught' and 'cot' with the same vowel [kɑt], but 'all' will have a different vowel [ɔl]. And this vowel is normally a little higher, so it's [ol]. And then the kicker comes where words which used to have 'oll' start being pronounced with [ɔl]. This is anecdotal, but I've been told that there are people who pronounce 'acknowledge' as [əknɔlədʒ]. For some folks then, /ɔl/ is spreading and covering areas that used to be /ɑl/. (I do not have this pronunciation.)

So now we have /ɔl ~ ol/ as a strange special category. Some Americans still distinguish between "caller" and "collar" while having the cot-caught merger, but what about 'call' vs 'coal'? I haven't had the opportunity to ask yet, but it would be interesting to wonder if these people who keep /ɔl ~ ol/ might merger 'call' and 'coal'. And if that's the case, might the absorption of 'cult' and 'hull' into /ɔl ~ ol/ be related? The 'uh' /ʌ/ vowel is otherwise safe in American English, but remember that strange things happen before /l/. If 'call' might be re-analyzed to be part of the same lexical set as 'coal', perhaps 'cull' might be being moved in that direction as well. Why, I'm not entirely sure from a theoretical basis. And this probably doesn't even describe the majority of people with the 'colt-cult' merger. For the record, I pronounce all /ɔl/ words with /ɑl/, so this doesn't describe me. But it seems that something is happening where words are moving between the COAL, CALL, and COLLER lexical sets, and now the CULL lexical set is joining the game.

Now that this wild and unfounded speculation on my part is done, I would love to hear if you all have any ideas on what is phonologically motivating this change. Is it related to the cot-caught merger? Are there people without the merger who still merge 'colt' and 'cult'? Are you aware of any research relating to changes in back vowels before /l/ and how it could apply to this situation? I would love to hear!

May 14, 2019

Dialect Dissection: Marina and the Diamonds' Aspirations

Marina, formerly known as Marina and the Diamonds, is a Greek-Welsh indie pop artist. You may have seen her in any number of 2012 Tumblr and Pinterest posts with a heart drawn on her cheek and wearing vintage-inspired styles, much like another alternative starlet. Although she's always been more popular in her native United Kingdom than in the United States, she is still a favorite player in the indie pop sphere. She is known for her witty lyrics, diverse production, strong aesthetic element, and most relevant to us, her immediately distinguishable style of singing.

Marina doesn’t have a well-known, persistent feature associated with her like Ariana Grande’s pronunciation or Taylor Swift’s adopted Southern accent, but she does have some interesting features that should interest fans and linguists.


Marina Diamandis was born and raised in Wales. This may be why so many people think her accent is Welsh (1, 2, 3, 4, 5). However, listening to her interviews, her accent sounds more London than Welsh. Indeed, others have noticed no trace at all of a Welsh accent in her speech. Per her own words, Marina says she used to have a Welsh accent that she lost:

Marina has never spoken at length about her accent, although she seems to feel some affinity with East London:

Interviewer: You’ve always been so vulnerable with the fans, like with the "FAQin Hell" sessions.
Marina: Does that translate in America?
Interviewer: Well, you guys have the accent, and we don’t!
Marina: Because in the U.K., it’s very, like, East London and kind of cockney to be like “Oh fackin’ hell!” But over here I didn’t know! I never even thought about how that might not actually work!

Unlike other English singers who purposefully hide their English accents in their music, Marina freely uses English English pronunciations in her music. Let's take a look at where these features appear.

  • Marina uses a short “o” [ɒ] sound in words like "orange," "forgot", and "gone." In other words, she does not have the FATHER-BOTHER merger like Americans, who instead use an 'ah' vowel for "forgot" and "gone" and the 'aw' vowel for "orange."
    • “That you are a horror [hɒrə], you're just as horrible [hɒrɨbl] as me” - Horror Pop, Mermaid vs Sailor (2007)
    • “Tall, tan, hot [hɒt], blonde [blɒnd] called Anya” - Hollywood, The Family Jewels (2010)
    • Orange, orange [ɒrɨndʒ]” - Orange Trees, Love + Fear (2019)
  • Non-rhotic "-ER" vowel. Marina often uses the [ɜ] vowel in words with an 'er' sound like 'girls', 'dirty', and 'birds.'
    • Girls [gɜlz] are not meant to fight dirty [dɜɾi]” - Girls, The Family Jewels (2010)
    • “I envy the birds [bɜdz] ... so purposefully [pɜpəsfʊli]” - Handmade Heaven, Love + Fear (2019)
  • Marry-vowel. Marina uses the 'aa' [æ] vowel of TRAP for words like 'marriage', 'guarantee', and 'paradise'. In other words, she does not have the MARRY-MERRY merger. This is in contrast to most Americans, who pronounce all these words with 'eh' [ɛ] vowel of SQUARE.
    • “Will that guarantee [gærənti] you a win?” - Mowgli’s Road, The Family Jewels (2010)
    • “Underneath it all, we're just savages hidden behind shirts, ties and marriages [mærədʒəz]” - Savages, Froot (2015)
    • "It's paradise [pærədaɪs]" - Handmade Heaven, Love + Fear (2019)
  • COT-CAUGHT distinction. Marina uses a o-like vowel [ɔ] similar to the one in the word CORE for words like 'thought', 'walk', and 'bought'. This is in contrast to most young Americans, who use the 'ah' [ɑ] vowel like in SPA for all these words, resulting in these words sounding like 'thot', 'wok', and 'bot'.
    • "Caught [kɔt] me cold so they could cut" - Rootless, The Family Jewels (2010)
    • “Baby you know what I'm talkin' [tɔkɪn] about [...] That chain of thought [θɔt] that followed me” - Forget, Froot (2015)
    • “[...] can't be bought [bɔt] or sold” - Gold, Froot (2015)
  • TRAP-lowering: In TRAP words like 'happy' and 'branch', Marina uses the [a] vowel instead of the [æ] vowel. This is used in Modern Received Pronunciation as well as in Welsh English.
    • "I never knew you had [had] such a dirty mind" - Hermit the Frog, The Family Jewels (2010)
    • "Are you satisfied, are you satisfied? [satɨsfaɪd]" - Are You Satisfied, The Family Jewels (2010)
    • Happy [hapi]” - Happy, Froot (2015)
    • "But in this handmade [hanmeɪd] heaven, I come alive. Bluebirds forever color the sky" - Handmade Heaven, Love + Fear (2019)
  • PRICE-backing. Marina pronounces the 'ay' vowel of words like PRICE further back in the mouth, so that the first part of the sequence sounds like the vowel in SPA: 'oi' [ɑɪ] or [ɒ̟ɪ]. She uses this vowel in her speech ("the creating it was quite noice"). This is a feature of Estuary English.
    • “I hoped you were a gemini [dʒɛmɪnɑɪ]“ - Horror Pop, Mermaid vs Sailor (2007)
    • “Easy to be sleazy when you've got a filthy mind [mɑɪnd], you stick to your yoghurts I'll stick to my apple pie [pɑɪ]“ - Girls, The Family Jewels (2010)
    • “Meant for a wedding toast to the bride [brɑɪd]“ - Shampain, The Family Jewels (2010)
    • “Baby I am plump and ripe, I'm pinker than shepherds delight [dilɑɪt]” - Froot, Froot (2015)
    • "But in this handmade heaven, I come alive. Bluebirds forever colour the sky [skɑɪ]" - Handmade Heaven, Love + Fear (2019)
  • Variable TRAP-BATH split. Words like "glass" are pronounced with the 'ah' [ɑ] vowel of SPA, and not with the 'aa' vowel of TRAP. However, Marina exhibits variability with this feature. In the same song, she'll pronounce "glass" with the SPA vowel as well as the TRAP vowel!
    • “When my heart just burst like a glahss [glɑs] balloon [...] We broke our glaass [glæs] balloon.” - Hermit the Frog
    • "I feel I've been riding up the wrong path [pɛæθ] but I'm gonna make sure I get the last [lɛæst] laugh [lɛæf]" - Girls, The Family Jewels (2010)

It is interesting that Marina clearly keeps a substantial amount of English English characteristics compared to major British acts, who Americanize their singing. In this way, she is more similar to The Beatles than to One Direction. Where the Beatles caught flak for continuing to use their Liverpool accents in their music, the members of One Direction were instructed to sound more American. One of the members of One Direction, Zayn Malik, made an interesting comment: "[Our song] 'What Makes You Beautiful' would sound more indie with a British accent."

Although Marina has always had ambitions of fame, she operates in the indie pop sphere, more comparable to Lana Del Rey or Florence Welch than Lady Gaga or Rihanna. The acceptable register of pronunciation for indie pop is much less strict than a genre like rock or mainstream bubblegum pop. Marina's decision to continue using her English accent is barely commented upon, and herself has not really discussed it. Perhaps, as with the 'FAQin Hell' joke, she may simply have never considered that it would be a problem. The world of pop music has become a little more linguistically accommodating.


Keeping her English accent isn't even the most linguistically notable thing about Marina, however. There is definitely something different about her pronunciation that many people have picked up on. Most people see that she's Welsh and therefore assume that any unusual pronunciations from her are because she's Welsh:

“Primadonna” had every girl wanting to put ribbons in her hair and heart on her cheek just to dance in front of a mirror with her friends, putting extra emphasis on the facial expressions and really enunciating the words with the most Marina-esque Welsh accent. Trust me, I know from experience. - (source)
She uses her Welsh accent and mezzo soprano range to good effect, switching between a powerful, heavy sound to a breathier, more feminine one effortlessly within the same song, giving the dramatic impression of different ‘characters’ or voices. - (source)
“I Am Not A Robot” is this weeks single and puts the vocals of lead singer Marina front and center. Her ability to own a song is displayed in the song as she gives us a taste of her unique Welsh accent. - (source)
She has a husky welsh accent and a very dirty laugh. - (source)

But Marina does not have a Welsh accent. A listen to her interviews is enough to show you that she has never had a Welsh accent while she's been producing music. By her own admission, she does not have a Welsh accent. (If you still think she has a Welsh accent, I encourage you to find a list of Welsh English features and point out which ones she has! There may be some lesser known features that are missing from most descriptions of Welsh English.) And although Marina's choice to use her South England accent on her music does distinguish her from American and Americanized peers, she is far from the only singer to keep her English accent - compare Ed Sheeran, who uses his local English accent, sounds nothing like Marina, and who nobody has accused of being Welsh. What's going on?

The reality is that Marina has a lot of idiosyncratic pronunciation choices that do not belong to any one accent in particular. Some of the features may be found in one accent, some in another, but there's no accent that collects all the following features together. Although I have always heard some of these in her first two major albums, The Family Jewels and Electra Heart, the first time I really noticed them was during her Froot era, where she seemed to have ratcheted their usage up. Now in the Love + Fear era, she has gone down again.

The following is a list of notable pronunciation choices that do not neatly belong to South England English, General American English, or Welsh English. Some of these features can be observed even in her independent 2007 EP Mermaid vs. Sailor, showing that these are features she has long used when recording. Don't be intimidated by the length - it's mostly a lot of examples!

  • De-aspiration.
    The English consonants “p”, “t”, and “k” are voiceless, meaning your vocal cords don’t vibrate when you make them. But they are also aspirated, meaning that there is a little puff of air that comes out after the consonant. You can test this by holding your hand in front of your mouth and saying “Poe,” “tee,” and “kay”. Whenever these consonants are at the start of a stressed syllable, they will be aspirated.

    Marina and the Diamonds has an interesting vocal tic when it comes to these consonants. She likes to lessen the amount of aspiration, or even remove the aspiration entirely. When you remove the aspiration from these consonants, they can sound “b”, “d,” and “g” respectively. Notice how the way she sings “my life is play” sounds like “my life is a blay.” She doesn’t do this all the time, but she does it consistently enough to make it a Marina-ism. This pronunciation pattern is noticeable in Indian English and varieties of English spoken by people whose native language doesn't have aspirated consonants(e.g. Greek, Spanish, Russian, Arabic).
    • “Musical c-ac-ophony [k˭ak˭ɒfəni] let insy-winsy spider free” - Seventeen, Mermaid vs Sailor (2007)
    • “Can’t let your c-old [k˭old] heart be free” - Obsessions, The Family Jewels (2010)
    • “My life is a p-lay [p˭leɪ]” - The State of Dreaming, Electra Heart (2012)
    • “And I'm sad to the core, core, core [k˭ɔ]... take a p-icture [p˭ɪtʃə] I'm with the boys ... Anything for the c-rown, c-rown, c-rown [k˭rɑʊn]” - Primadonna, Electra Heart (2012)
    • T-ough [t˭ʌf] to t-alk [t˭ɔk] to” - Starring Roles, Electra Heart (2012)
    • “I’ve seen seasons c-ome [k˭ʌm] and go from winter [wɪnt˭ə] sun to [t˭u] summer snow. This ain't my first t-ime [t˭am] at the rodeo.” - Froot, Froot (2015)
    • “It's a p-ower, it's a p-ower, it's a p-ower [p˭aʊwə] move” - Better Than That, Froot (2015)
    • “I even c’ried [k˭raɪd] but I never meant it.” - Blue, Froot (2015)
    • “I love that he pret-ends [prɪt˭ɛndz] to care […] I was born to be the tort-oise [tɔt˭ɔɪs]” - Forget, Froot (2015)
    • "But in this handmade heaven, I c-ome [kʌm] alive. Bluebirds forever c-olor [kʌlə] the sky" - Handmade Heaven, Love + Fear (2019)
    • "All of the money in this t-own [t˭aʊn]" - Karma, Love + Fear (2019)
    • "Other p-eople [p˭ipl] wanted to hurt me" - Soft to be Strong, Love + Fear (2019)
  • Very Back U. While modern London English has a fronted 'u' vowel that sounds like 'ew'[ʉ], Marina likes using an exaggeratedly backed 'u' [u] sound. This is a feature that might be inspired by Welsh English, which uses a back 'u' [u] (Wells, 1982). Greek also uses a back [u].
    • "I'm rootless [rutlɛs]" Rootless, The Family Jewels (2010)
    • "Who [hu] are you [ju] to tell me" - Hypocrates, Electra Heart (2012)
    • “Like froo-oo-oot [frut]” Froot, Froot (2015)
    • “Oh yes she did yes she did what she wanted to do [du] ... But you [ju], you [ju] can do better than, you can do better than that” - Better Than That, Froot (2015)
    • “I'll ruin, yes I'll ruin [ruɪn] you, I'll ruin... [ru]” - I’m a ruin, Froot (2015)
    • "You don't own me, but I can cut you loose [lus] ... cuz it's all about you ... you [ju]" - You, Love + Fear (2019)
  • MOUTH-backing. Marina uses a vowel in words like "down" [daʊn] with the tongue pulled further down and back, [ɑʊ]. This pronunciation can be found in Multicultural London English.
    • "Going up, going down down down [dɑʊn] ... When the lights, they went down down down [dɑʊn]" - Primadonna, Electra Heart (2012)
    • "You're only sorry when you're coming down, down, down [dɑʊn]... you carried on without a doubt [dɑʊt]" - Karma, Love + Fear (2019)
    • "I wanna change, but I don't know how [hɑʊ] ...people look so lonely with eyes turned down [dɑʊn]" - Just Too Afraid, Love + Fear (2019)
  • Voiced /d/. In English, the consonants 'b', 'd', and 'g' are often unvoiced and unaspirated. This is in contrast to 'p', 't', and 'k', which are unvoiced but aspirated. Marina likes to use a /d/ that is clearly fully voiced. She sometimes keeps the full value of /d/ even between vowels, instead of using a flap. While not flapping is normal in British varieties of English, using a strongly voiced [d] at the start of a syllable is not.
    • "Going up, going down down down [dɑʊn] ... When the lights, they went down down down [dɑʊn]" - Primadonna, Electra Heart (2012)
    • "Murder lives forever and so does war" - Savages, Froot(2015)
    • "It's paradise [pærədaɪs]" - Handmade Heaven, Love + Fear (2019)
    • "I wanna change but I don't know how, people look so lonely with eyes turned down" - Too Afraid, Love + Fear (2019)
  • Marina pronounces "-or" words with a diphthong "o-a" [ɔə] instead of a pure vowel "aw" [ɔ]. This is reminiscent of the way that New York English speakers pronounce words with an 'aw' sound. It is also found in some varieties of Welsh English (Wells, 1982).
    • "Waking up too early in the morning [mɔənɪn] ... Auntie Emy was a medium, oh she worked the Ouija board [bɔəd]" - Daddy was a Sailore, Mermaid vs Sailore (2007)
    • "Every day is a chore chore chore [tʃɔə]" - Primadonna, Electra Heart (2012)
    • "Gotta be lookin' pure [pjɔə], kiss him goodbye at the door, and leave him wanting more [mɔə]" - How to be a Heartbreaker, Electra Heart (2012)
    • "Give me good and pure [pjɔə]" - Blue, Froot (2015)
    • "Survival of the fittest, rich against the poor [pɔə]" - Savages, Froot (2015)
  • Most varieties of Welsh English like using a light [l] made with the tip of the tongue. London English uses a light l at the start of syllables, but a dark l [ɫ] made with back of the tongue raised up at the end of syllables. Marina often uses a dark l even at the beginning of syllables, which is a common feature in General American English and Northern Welsh English (Wells, 1982).
    • "You're not horrible like [ɫɑɪk] me" - Horror Pop, Daddy vs Sailor(2007)
    • "But the melody [mɛɫədi] went stale" - Seventeen, The Family Jewels (2010)
    • "Who you never really liked [rɪɫi lɑɪɫkt] and you never trusted" - I Am Not a Robot, The Family Jewels (2010)
    • "I wanna be a bottle blonde [bɫɑnd]" - Teen Idle, Electra Heart (2012)
    • “Baby I am plump and ripe, I'm pinker than shepherds delight [diɫɑɪt]” - Froot, Froot (2015)

Several of the pronunciations Marina uses (de-aspiration, back /u/, voiced /d/) can also be found in Greek phonology. Marina is half Greek - her father is from Greece and met Marina's mother while studying in the United Kingdom. Marina also lived with her father in Greece from 16 to 18 to go to an international school in Athens. She was interested in living there "to connect with [her] heritage and learn to speak the language." She started making music when she was 18 and had moved back to London, so the time frame would be right for her to be influenced by Greek music and language. She has mentioned her Greek heritage in several songs - "you better make way for a Greek gold rush" in "Gold" from Froot, a clip of her grandmother singing in Greek on "Fear and Loathing" from Electra Heart, and "Fly to Athens, pass the Parthenon, see the village where my father's from" in "To Be Human" from Fear + Loathing.

These features found in Greek are also rather common cross-linguistically, being found in languages like Spanish, Portuguese, and Russian, to name a few just from Europe. This might not be coincidental. Marina is clearly interested in topics like human nature and finding connections between different cultures - the Love + Fear album and specifically the song "To Be Human" deals with these topics. In the linked video, she says "I moved around a lot as a kid. I spent summers in Russia, in the Ukraine, my dad worked in Japan for a long time [...] I just wanted to create this feeling of bringing people together."

She adds, "The message that we keep on seeing in the media, which is that, if you're from a different country, or a different continent, then you are so obviously different from us, that you're a threat. Whereas, I believe that people aren't that different, actually." Although the Love + Fear album is her first to explicitly ponder the shared experience among human cultures, it makes sense that Marina's childhood spent in different countries would make her more receptive to art from different cultures and that she would seek to imitate that. The timeline isn't right to say that she purposefully does it to sound 'international' - but it could be a result of her international life.

Wales Millennium Center

While I mentioned in the first section that Marina does not have a Welsh accent and does not use Welsh features consistently, it is worth pointing out that many of these features (back /u/, dark /l/, diphthongized /or/) could arguably be from Welsh English. They are not part of her spoken speech, which uses the London English equivalents instead (front /u/, clear-dark distinction of /l/, monophthongal /o/), but it is worth considering that she was borrowing features from Welsh English. At the same time, there are alternate explanations for some of these - dark /l/ is used frequently in American English, and both The Family Jewels and Electra Heart have a fascination with American culture.

If there is an overlap between some features (back /u/ and voiced /d/ in Welsh English and Greek; dark /l/ in North Welsh English and American English), how can we tell which one is the 'real' origin? It might not be possible to tease apart which English variety it is coming from. Indeed, she may not even be purposefully choosing to invoke any variety of English. But we could say that these different language varieties reinforce each other. Marina has been exposed to Welsh English and to Greek, both of which have a back /u/. Marina was interested in American culture, where Americans frequently use a dark /l/, and North Welsh English also uses a dark /l/. That these features are repeated across different dialects might not be accidental. There are other features of Welsh English and American English and Greek that Marina does not use at all.

The MOUTH-backing is from London Multicultural English. She doesn't have this feature in her speech, but per her FAQin Hell interview above, she hangs out a lot in East London, and it's likely she would have been exposed to this feature.

It's the Brand, Baby

Marina's vocal profile doesn't just involve her accent. Her singing style is immediately recognizable as well. She likes using harsh breaks between chest and head voice, resulting in a yodeling sound. She likes slipping into falsetto for more delicate portions of songs. In her early career, she liked using a stuttering vibrato at the end of phrases. Her lower register is so dark and smoky sounding compared to her airier upper register than an audio engineer mistook it for formant manipulation:

I do wonder whether there's something a bit strange going on with the vocal formants in this song. Formants are resonance peaks in every singer's frequency response that we humans are very sensitive to, and which provide us with strong clues as to a performer's age, size, and gender. Shifting human vocal formants unnaturally low gives the impression of some kind of mythical giant or monster (an effect used to death on film and TV), but as the formants shift upwards, you move through the natural vocal ranges for males, then females, and, finally, children, before heading off into Tombliboo territory. Of course, even the briefest comparison between Michael Jackson and Tracy Chapman illustrates that there's bountiful overlap between these notional formant ranges in practice, and it's not the specific formant range of Marina's voice in this production that bugs me — it's that the formants appear to be moving about! So "someone else's fault” at 0:19 comes across like a fresh-faced learner driver, whereas "pop that pretty question” packs more testosterone than the entire cast of Glee.

Now, it's conceivable that this is simply a virtuoso performance, because there is a significant degree of physical control you can exert over your own formants. Indeed, it's the stock-in-trade of voice-over artists the world over. However, I'd hazard a guess that there's some artificial manipulation going on here. It's not something that's difficult to do in something like Celemony Melodyne (as Will.I.Am has demonstrated with depressing regularity), so it wouldn't surprise me at all to discover that some producers are now deliberately modulating formants more subliminally during post-production, to add synthetic expression to their lead vocals.

- Mike Senior, Sound on Sound

Although her lower register is definitely used as a jarring break, there's no reason to believe her voice was electronically manipulated. Similarly, although Marina has a noticeable 'accent profile,' there's no evidence to suggest that it's the influence of Welsh English. Both her vocal style and accent profile seem to have been developed by her.

The end result of this is that Marina has an immediately recognizable vocal profile. When Marina is on a track, her voice is unmistakeable. Even on the rare occasions when she has appeared on other musicians' tracks, such as in "Baby," she carries her style with her. She is never at risk of being mistaken for a generic 'hook girl'; indeed, you could say that her peculiar pronunciation choices constitute a sort of branding.

Although Marina isn't the only musician to use her English accent in her music - the Beatles were groundbreakers in that sense - and she isn't the only singer to use features that don't belong to any particular accent - Lady Gaga has also done that - she manages to combine the two consistently such that she has a recognizable 'linguistic brand.' She does not switch through accents like Taylor Swift or Lana Del Rey; she has displayed all these traits throughout her career. Her singing style on her early, 'quirky' and 'camp' album The Family Jewels isn't very different from her more 'mature' and 'psychological' album Love + Fear. The instrumentals change - they have probably been the least consistent aspect of Marina's music. But her linguistic profile has not, and neither has her singing style. Combined with her lyrical content, which tends towards introspection, workaholism, human relationships, and musings on the nature of humanity, it is clear that she has a strong lyrical-vocal identity.

It's very common for people commenting on interesting linguistic aspects of singing to dismiss them as "people trying to be different and stand out." It's a thought-terminating cliche that limits any notable linguistic characteristics to individual innovations for no purpose other than to "stand out." For instance, the notion that "indie voice" is just an attempt to "stand out" has limited the scope of research on indie voice and ignored how it has developed over time and the particular linguistic contexts in which "indie voice" occurs.

However, we also cannot deny that having a strong musical or personal brand is an important part of the music business, and it will undoubtedly influence how musicians approach their craft. I cannot say that I know what Marina's thought process is when she pronounces "my life is a play" as "my life is a blay." Perhaps she thinks it's more melodious, or she's imitating some artist she heard in her teenage years that's influencing her. Without any interviews or writings, we cannot know her internal thought process, and any proclamation on "what she wanted to do" is only speculation. But we can say that when she consistently says "my life is a blay" and "going dawun, dawun, dawun" and "like froo-oo-oot," throughout all her albums, the effect is that she stands out and is memorable, at least in some part for her curious linguistic choices. Her fans seem to enjoy the way she plays with language, so it is successful in that sense:

Having said that, he [my voice teacher] did say one of my problem is because I probably try to mimic Marina's singing style and accent (which is true). [...][He said] "I just want to listen to her song so I can decide whether it's suitable for you (me) to sing her song. The answer is yes, but unfortunately you can't do it in her style," because he said something about Marina's way of vocalising her words (could be her pronounced accent) is unhealthy. [...] The reason that I fell in love with Marina's songs is because the way she sings a song is very unique. It's like she's telling a story and just having fun with singing without the restrain of "Okay, i think i need to sing it this way that so the notes comes out beautifully". Marina's singing style is not flat and bland and I love her unique accent so much. - (source)

Is Marina aware of it? Might she disagree that she's doing it to "stand out"? It's certainly possible! But every time a fan says she has a "Welsh accent" (she does not) or talks about her "recognizable style" or wants to imitate the way she sings, it's proof that people are noticing that there's something different about the way that Marina sings, and they're remembering and talking about it. Like it or not, it's part of her brand.

Works Cited

Przedlacka, Joanna (2001), "Estuary English and RP: Some Recent Findings" (PDF), Studia Anglica Posnaniensia, 36: 35–50

May 8, 2019

Mobile Formatting

It's been a while since I did one of these updates, but at long last, the update to mobile formatting is done. It's very simple: on the mobile version of Ace Linguist, the font will be bigger and there will be more space between the lines. Previously, the font on the Ace Linguist mobile version was terribly cramped - it was basically the desktop version of the site scaled down. This change should make it easier for mobile readers!

May 7, 2019

Bay-zhing or bay-jing: Let's go to Beijing

This impromptu post was brought to you by my setting up an RSS feed with my favorite language blogs and deciding to look at Language Log's recent posts. One of them was called "Why we say "Beizhing" and not "Beijing"" by Victor Mair. He immediately clarifies that he does not use the pronunciation:

Well, I don't say "Beizhing", and I think it sounds ghastly, so much so that I cringe when I hear it and my flesh creeps. I never could figure out why English speakers would use this hideous pronunciation when it would be so much easier, transparent, and direct just to pronounce the name the way it looks: "bei-", like "bay", as in "Beirut" (we don't have any trouble with that, do we?), and "-jing" as in "jingle". BEI- -JING! Voilà! We don't have to say "bei- -zhing". I realize, though, that almost everybody, including many China specialists who surely know better, say "Beizhing", not "Beijing".

Victor Mair is an American sinologist, and I know people who study or specialize in a particular language often have a fondness for pronouncing words from that language in a way that resembles the origin. I can say myself as a minor in Russian language that I get a twinge when I hear people say "baBUshka" instead of "BAbushka", which is how it would be said in Russian. Of course, "baBUshka" is the common pronunciation in English, at least in American English, and so though I continue stubbornly saying "BAbushka," I can't fault anyone for using the average pronunciation. Similarly, I understand Dr. Mair's feelings at hearing pronouncing "bay-zhing" when he knows that "bay-jing" is closer to the Mandarin pronunciation.

However, there is always a certain amount of elitism and posturing that occurs with regards to pronouncing words from foreign languages in another language. The "curmudgeonly correspondent" that writes in to propose an origin for the "bay-zhing" pronunciation refers to it as "egregious mainstream pronunciation." Another commenter refers to "English speakers getting it wrong (as per usual)." An earlier Language Log post by Bill Poser has an identical complaint: "This word [Beijing] is routinely mispronounced by newscasters and other people who are supposed to know better." And another commenter: "the only reason to use "Beijing" at all is the desire to match the official Chinese pronunciation, so going on to mispronounce it seems like fair game for some criticism."

I am not sure why "fidelity to names of places in foreign languages" is something to strive for in the first place. What does it matter if "bei-jing" is closer to how it is pronounced in Mandarin than "bei-zhing"? It reminds me of all the monolingual US American newscasters who really want everyone to know that they took a Spanish class by pronouncing Latin American country and place names with an approximation of the Spanish pronunciation, as if Spanish speakers were incapable of understanding that the same place has a different pronunciation in different languages. As a native bilingual speaker of Spanish and English, I occasionally struggle with words that aren't fully incorporated into English and alter between a Spanish-y and English-y pronunciation (do I go with "croqueta" in Spanish, "croqueta" mangled into English, or the deceptively French "croquette"?). But a place that has an established English pronunciation, distant from the original as it may be, is perfectly fine. I would never demand anyone say "ar-hentina" (Argentina) or "cooba" (Cuba) or "abana" (Havana) despite these being much closer to the originals. I wouldn't even ask Russian speakers to stop pronouncing "Habana" as "Gavana," and that one also makes my "flesh creep" (the spelling pronunciation [h] turned into a [g]...).

The origins of the "zh" pronunciation may be nonsensical or hyperforeignisms or a mistake or just plain lost, but the notion that people continue to pronounce it with "zh" as a hyperforeignism is misguided. For as far as I can remember, the capital of China was "Beijing" with the middle consonant pronounced as a "zh" [ʒ]. As such, I have pronounced "Beijing" as "bay-zhing" for almost my entire life, with the exception of the brief period of time when I pronounced it as "Bay-jing" (and also pronounced "Moscow" as "Moss-ko" instead of the common American "moss-cow" because it also "made more sense"). I made that brief change because I wanted my pronunciation of everything to be "correct". (Not coincidentally, this also coincided with the time in my life when I was an enormous pedant about "correct language use.") I dropped both pronunciations because they never sounded right and likely sounded affected to my peers. Am I mispronouncing "Beijing"? If I were speaking Mandarin, yes, I would be mispronouncing it. But the pronouncing with 'zh' [ʒ] is common enough that it seems odd to refer to it as a "mispronunciation" from a descriptivist point of view.

Indeed, Ben Zimmer pointed out a research paper from 1994 (Joseph, B. Systematic Hyperforeignisms As Maximally External Evidence for Linguistic Rules) describing this pronunciation as an example of 'hyper-French/pseudo-Mandarin'. It is compared to the similarly hyperforeign American pronunciation of "parmesan" as "parma-zhan," when it is pronounced as "parmezan" in French and "parmejano" in Italian - no 'zh' anywhere. "Bay-zhing" has apparently been a common pronunciation for nigh-on a quarter of a century. At which point does it cease being a mispronunciation worthy of goosebumps?

Considering that Language Log normally uses a descriptivist approach to language - and with some nuance, I may add - I am surprised to see that both major contributors and commenters find it comfortable to refer to a mainstream pronunciation of a city as "incorrect," "egregious," and "fair game for some criticism." Contrast this with this article on the stigmatized pronunciation of "nuclear" as "nucular" (emphasis mine):

What about Pinker's second point, that Palin's pronunciation of the word is "not a sign of ignorance"? Well, not of her ignorance, anyway. It's fair to assume that "nucular" was the dominant pronunciation in the ambience she grew up in, as it was for Bill Clinton, and that she acquired it "naturally." But at its inception, the "nucular" pronunciation was the result of ignorance, or at least of unfamiliarity with the item, which is why it tends to be more frequent in the varieties used by less-well-educated speakers (or maybe I should say it's less frequent in the varieties used by literate ones).

That doesn't mean that speakers who pick up the "nucular" pronunciation from family, friends, or teachers can be accused of ignorance themselves — they weren't the ones who came up with the reanalysis that motivated the pronunciation. But it does explain why such speakers might want to correct their pronunciation once they're made aware of it — not just because the "nuclear" variant happens to be used by better educated speakers, but because it conforms more closely to the word's orthography, and because this is, in its nature, a word that belongs to literate discourse.

Palin has to be aware that many people consider her pronunciation nonstandard, and she (or her handlers) seems to have made some effort at correction, which is presumably why she pronounced the word as "new clear" when reading off the teleprompter in her convention speech. Since then, though, it's been "nucular" all the way, which may be part of the "let Palin be Palin" strategy.

The pronunciation of "nuclear" as "nucular", much like the pronunciation of "Beijing" as "Bay-zhing," is one that deviates from the etymology of the word. The difference is that "nucular" is strongly stigmatized as an "uneducated" pronunciation while "Bay-zhing" is used even by Dr. Mair's colleagues in sinology, meaning it is clearly present in multiple class strata. Both pronunciations can be said to be the result of not knowing the standard pronunciation (nuclear) or the intended pronunciation (Bay-jing). However, both forms are common enough now that it is clear they are no longer one-off errors - they are distinct variants. Being that all these variants are recognized and understood, there is no risk of miscommunication occurring. I therefore see no reason other than aesthetic preference or class signaling to choose a certain pronunciation. What do we have to gain by stigmatizing or mocking perfectly common and understandable variants? Well, we create an additional linguistic shibboleth by which to recognize and exclude people who have not had access to education. It creates additional class anxiety (which in turn leads to the creation of hypercorrective forms which in turn creates even more shibboleths). You get people feeling like the commenter Chris Johnson:

I consistently pronounce 'nucleus' as [nukliəs], and I imagine most, if not all, other 'nucular' speakers do too. I don't think I've ever heard "nuculus".

I use the two pronunciations for 'nuclear' in more or less free variation. I never notice which form a speaker uses unless I'm specifically listening for it. I think I am more likely to use 'nucular' when saying the word 'thermonuclear' than when saying 'nuclear' itself. I have no idea whether my usage varies by domain or not. It doesn't seem like I had any trouble treating 'nucular' as an adjectival form of 'nucleus'.

I think the 'nucular' pronunciation has hopped over social and class boundaries more than people realize. It's not a very common word. I think it may be spread more by mass media than by the family dinner table. I had an upper middle class upbringing that valued education, and I even wanted to be a physicist in 5th grade. Nevertheless, I think 'nucular' was my usual form growing up. (I've shifted towards [nukliɚ] over the years, but my girlfriend still catches me sometimes.) Bush could very well be 'slumming' linguistically, but I'm sure there are more than a few kids from elite families that picked up 'nucular' from the culture at large.

(I've known for a long time that 'nucular' was a less prestigious pronunciation, but in the past few weeks I've been shocked at just how stigmatized this feature of my speech is – especially by people 'on my side' politically. As a speaker of a pretty standard variety of American English, this is a new and unpleasant position for me.)

I understand the sentiment of wanting people to treat foreign languages as worthy of respect. For example, many US Americans (and likely people in other English speaking nations) do not try to learn how to pronounce names from other languages and instead shorten them or even give people "English nicknames" instead. It makes sense to want to counter this Anglo-centric sentiment by saying that we ought to consider the native pronunciations of loanwords and other places. But cities are not people, and no cultural harm will come to Beijing if English speakers use the form 'Bay-zhing.' This is of course assuming that the name is not offensive or otherwise troublesome to the people it is applied to. Sometimes the standard name is offensive and in such a case, it is worth making an effort to switch to a preferred name. But I have seen no reason so far to believe that there is any harm to the people of Beijing by using the 'zh' pronunciation.

If you hang out with a lot of sinologists, maybe you will want to start using the "j" pronunciation to fit in. And if you feel compelled to start saying "Bei-jing" after reading this, that is fine. Indeed, it is fine to even tell people "did you know that if you say 'Beijing' with a 'j' sound, it's actually closer to the Mandarin pronunciation? Nobody knows where that 'zh' sound came from!" But I question what there is to gain from calling this now entrenched pronunciation a 'mispronunciation' and judging those who use the 'zh' variant as (sarcasm on) ignorant class-climbers foolishly betraying their station by applying French pronunciations when the English pronunciation would be closer to the Mandarin! (sarcasm off) Considering the amount of insecurity, angst, and division this sort of language policing causes, I question any perceived multicultural benefit that could come of it - and I am not confident that cultural understanding is gauged by hewing to the 'original' pronunciation, anyway.

This is a more freeform discussion style article. I would love to hear any thoughts or comments you may have on the topic!