Two directions

Here are three members of the boy band One Direction singing a phrase to the repeated words tearing me apart/tearing up my heart:

The historic r in apart and heart survives in ‘rhotic’ accents of English like General American and Irish. But in ‘non-rhotic’ accents like most accents in England and Wales, r was lost except before a vowel – becoming ə or a lengthening of the preceding vowel – and so is not pronounced in such words. Of the three clips here, the one with the least r-colouring to my ear is the first, which features Ireland’s Niall Horan. The others feature Zayn Malik and Louis Tomlinson, both from Yorkshire. So they’re modifying their spoken rhoticity in two opposite directions: the rhotic Irishman reduces his rhoticity when he sings, while the non-rhotic Yorkshiremen introduce rhoticity which their normal speech lacks.

The same phenomenon was illustrated in my post on last year’s Eurovision song contest, with Bonnie Tyler from non-rhotic Wales using r-coloured curse the rain and worth the pain, and Ryan Dolan from rhotic Ireland de-rhoticizing but in our darkest hour:

The explanation for this apparently contradictory behaviour lies in the fact that all these singers are Americanizing; this is indicated by other features in their singing, such as the voicing/flapping of t and the quality of the LOT and THOUGHT vowels. The American accents heard in pop/rock singing are often less than fully rhotic (hypo-rhotic). We tend to notice what differs from our own speech more than what is the same: so non-rhotic speakers notice rhoticity, and often imitate it, while rhotic speakers notice its absence, and often imitate that.

Over the past half century or so, rhotic General American (GenAm) has become the dominant spoken accent in American popular culture; see my previous post. But in popular singing, GenAm is not so dominant. African Americans have played a seminal role throughout the history of popular music — from ragtime, jazz, blues, gospel, R&B, through rock and roll, soul and disco to house, rap and hip hop — and their accents commonly reflect the traditional hypo-rhoticity of the South. (Another Southern influence in singing is the frequent use of a front, relatively monophthongal PRICE vowel, so that eg I, my are often sung a, ma.) In the previous post I gave an example of the blues singer Bessie Smith’s non-rhoticity in millionaire, bootleg liquor and round their door:

White artists from Al Jolson to Adele have adopted African American musical styles. Jolson even painted his face black; that’s no longer acceptable, but Adele has said, “I taught myself how to sing by listening to Ella Fitzgerald for acrobatics and scales, Etta James for passion and Roberta Flack for control.” Flack, James and Fitzgerald, of course, are/were all African American. Here is Flack singing the first time ever I saw your face; her first has the r-coloured NURSE vowel often heard in otherwise non-rhotic American accents, but her ever and your are r-less:

Similarly, Etta James sings my heart was wrapped up in clover with a non-rhotic heart but an r-coloured final vowel in clover (probably a neutralization of NURSE and lettER):

Ella Fitzgerald was an essentially rhotic singer, eg music I hear and I can not forget:

which is probably related to the fact that she was influenced by the white, rhotic singers of the 1930s Bing Crosby and Connee Boswell. “I tried so hard to sound just like her,” Ella Fitzgerald said of Boswell. (Note also that Fitzgerald’s PRICE vowel tends to be more diphthongal than Flack’s or James’s.)

Given the influential nature of the Southern/Black accent, it’s not surprising that many GenAm speakers, like the Irishmen above, reduce their rhoticity when singing. Here Jon Bon Jovi explains in his rhotic spoken accent why he respelled his Italian name Bongiovi (for phonetics, in our culture the G I proper spelling of the word reads as though it’s a J):

But when he sings we get decidedly non-rhotic we’ll make it I swear, living on a prayer:

Not all rhotic speakers reduce their rhoticity when singing; it depends somewhat on the musical genre. The old ‘crooners’ like Bing Crosby tended to sing in their own accents (Crosby was from Washington state, and rhotic). A genre notable for its rhoticity is country music, which has roots not in African American culture but in traditional white folk music. Here country singer Jim Reeves is rhotic in she’s waiting now for me in Heaven’s open door:

Elvis Presley may have been the King of Rock and Roll, but he also crooned and sang country; his speech was basically rhotic but he was heavily influenced by African American artists. (The man who launched him famously said “If I could find a white man who had the Negro sound and the Negro feel, I could make a billion dollars.”) Unsurprisingly, Elvis’s singing rhoticity is variable, eg r-less it’s one for the money, two for the show and r-full you can burn my house, steal my car:

Crooning and country aren’t the only rhotic genres. We also get rhoticity from American artists that I lump under the crude heading of Arty White Bands – some might prefer terms like ‘alternative’ or ‘indie’ – who are as far removed from country music as they are from African American styles like blues and soul. These would include David Byrne of Talking Heads (I’m a real live wire… psychokiller):

Michael Stipe of REM (sure you’ve had enough):

They Might Be Giants (I’d be fired if that were my job):

and Billie Joe Armstrong of Green Day (boulevard of broken dreams):

Rhoticity can even be varied systematically within a song. My favourite example of rhoticity semiotics is the 1998 house hit Horny ’98. The song is introduced (and interspersed) with the phrase All night long I’m horny, repeated obsessively on two notes like a neurotic inner monologue. Here horny is rhotic:

The song proper is sung by an apparently different voice, overt rather than internalized, relatively unbuttoned – and non-rhotic:

Note also that I’m is more diphthongal in the neurotic version, less so in the unbuttoned one. The former definitely sounds more ‘white’, the latter more ‘black’.

Not all non-American pop/rock singers Americanize when they sing. Thirty years ago the sociolinguist Peter Trudgill published a paper on British pop singing (‘Acts of Conflicting Identity: The Sociolinguistics of British Pop-Song Pronunciation’, in On Dialect, Blackwell, 1983), in which he gave evidence for a decrease in phonetic Americanization during the 1960s. For example, the Beatles’ non-prevocalic r dropped from around 50% to very low levels as they developed their own style(s) and became more imitated than imitating. (Indeed, the link between popular singing and hypo-rhoticity, though due mainly to Southern/Black influence, has probably been reinforced by the major influence of the Beatles and other non-rhotic Brits.)

I don’t question Trudgill’s data or interpretation, but it would be a mistake to conclude from his work that there was a once-and-for-all shift away from Americanization in the 60s and 70s. For one thing, even before the Beatles came to fame there was Anthony Newley singing hit pop songs in his London accent (and in so doing influencing David Bowie). Here in 1960 Newley shows BATH-broadening, t-glottaling and L-vocalization in would you like to pick from my basket, come and get an earful:

Since then there’s no doubt been an increase in the number of Brits who’ve used their own accents to sing: Bowie when he feels like it, assorted punk rockers, Morrissey from Manchester, the Proclaimers from Scotland, rappers like the MLE-accented Dizzee Rascal, the Arctic Monkeys from Sheffield, etc, etc. A link can probably be made between these and the ‘Arty White Bands’ in the US, all of them resisting the standard tropes of pop/rock singing. But this resistance has, I think, remained the marked case. Audiences are very aware of the Sheffield accent of the Arctic Monkeys’ Alex Turner; I think they notice much less the hypo-rhotic American singing accent of Elton John or Adele (or Bon Jovi) because it is so much the default. And Americanization shows no sign of going out of style: England’s latest success Jake Bugg (b. 1994) sings about his Nottingham council estate of Clifton as if it were somewhere in the Appalachians:

(Bugg sings rhotically, like his cited influence, the folky American Don McLean.)

There are of course hazards when non-rhotic speakers go rhotic. Non-rhotic Brits merge the vowels of lettER, START and NORTH/FORCE with those of commA, PALM and THOUGHT; and they generally remain unaware that rhotic speakers have r in the former but not in the latter, despite the evidence of the spellings. As a result, they may create rhotic forms that never existed in America or anywhere else. I’ll end with a few examples of such hyper-rhotic singing. The first is somewhat notorious among Americans, coming from the early days of the Beatles’ stardom. Like Elvis, the Beatles were not just youthful rockers but also performed ‘standards’ that appealed to a wider and older audience. One such was the Broadway number “Till There Was You”, in which Paul sang about birds in the air, I never sore (saw) them winging, no I never sore them at all:

Then there’s the Kinks’ Ray Davies in 1967 lamenting that My girlfriend’s run off with my car, and gone back to her mar and par:

John Harris has pointed out to me a more recent and doubly funny example from London folk rockers Mumford & Sons, performing the Canadian Neil Young’s “Dance, Dance, Dance”. They sing the song with a tendency towards rhotic North American, but pronounce dance in its south-of-England BATH-broadened form. The result is the otherwise non-existent darnce, darnce (feel it all around you). It’s phonetically subtle but definitely there:

Hyper-rhoticity can even be perpetrated by Americans, if their native accent is sufficiently non-rhotic. The BBC recently broadcast a wonderful documentary about the veteran singer-songwriter Neil Sedaka, born Brooklyn, New York in 1939. His song “The Queen of 1964”, written with fellow Brooklynite Howard Greenfield, contains references to Mick and Bianca Jagger. The documentary allowed us to compare his non-rhotic spoken Jagger with his rhotic sung version:

He’s sufficiently non-rhotic to sing well I saw her last night with unwritten linking r (sɔɹə):

And the song goes on to pull off the feat of rhyming conquer with Bianca. The stressed vowel of both words is the ɑ of American LOT/PALM. In the original record, both words end with non-rhotic schwa; but for the documentary, Sedaka opted to rhoticize conquer and to hyper-rhoticize Bianca to Biancar: