Thursday, June 08, 2006

Intrusive intrusive /r/

In the summertime we'd drive my grandfather from Hallandale, Florida to Chicago, and as we'd pass through Tennessee he'd invariably remark, "Chattanooger. That's how the Bostonians say it." They don't, but he's certainly not the only one who thinks they do. The actual phenomenon underlying this belief is known technically as 'intrusive /r/.' When words ending in certain vowels (non-high ones) are followed by words beginning with vowels, people with this feature insert an /r/. So a Bostonian wouldn't just call Chattanooga 'Chattanooger,' but he or she certainly might say "Chattanoog[er] is in Tennessee." Or, to draw an example from my favorite corpus - Beatles lyrics - John Lennon sings "I saw[r] a film today, oh boy."

Two questions arise: who does this, and why? In answering one we will answer the other.

First off, as you undoubtedly know, some dialects of English are non-rhotic - that is, they drop the /r/ sound when it occurs after vowels - but not before them. A consequence of this is that the dropped /r/s come back when they are immediately before a word starting with a vowel. Thus in "Let It Be" Paul McCartney sings "In my hour of da(r)kness" and "There is still a light that shines on me." So intrusive /r/ happens when a word sounds like it has a dropped /r/ at the end but doesn't really, and the phantom /r/ appears precisely where a dropped /r/ would reappear, before a vowel.

Do all English speakers who drop /r/s have intrusive /r/? In a word, no. The key phrase in the paragraph above is "when a word sounds like it has a dropped /r/ at the end but doesn't. See, in some /r/ dropping dialects (such as African American, or the dwindling /r/ dropping white Southern dialects) the dropped /r/ alters the preceding vowel. In these dialects, then, 'manna' and 'manner' don't sound the same, so the confusion that gives rise to intrusive /r/ isn't present, and as a result there is no intrusive /r/.

All this is very complicated, and it's no wonder, then, that my grandfather had trouble mimicking it. Though Peter Trudgill's examples of people misusing intrusive /r/ in British pop songs may be problematic, the phenomenon, which I will dub 'intrusive intrusive /r/' is a very real one, and the point Trudgill is trying to make - namely, that people aren't as good as they think they are at miimcking other dialects - is entirely valid.

In the comments to my last post the question arose why a British professer would pronounce the name Echa as 'Eker.' After all, don't non-rhotic Brits have intrusive /r/? Shouldn't he therefore know how to use it. I would suggest two possible explanations:

1. He was mocking rhotic American dialects, but misanalyzing them and overgeneralizing. More likely, though, is the explanation suggested by the fact that

2. In British English intrusive /r/ is stigmatized, and this professor spoke a fairly posh dialect. I suspect, then, that intrusive /r/ is as foreign to him as was to my grandfather, and he too misanalyzed the phenomenon, and, attempting to employ it mockingly, misused it.

On an unrelated note, I want to acknowledge the sad fact that Billy Preston died on Monday. Honor his memory by reading my two posts about him.

14 comments:

~J said...

Interesting comments. Even though I was born and raised in Chicago, I've got a lot of Bostonians in my family and blame them for the times I have found myself saying something like 'so I sawr a thing on the television the other day', my tendancy to add an 'r' on the end of 'saw' before a vowel was brought to my attention when I moved to southern Indiana for school and for some reason that particular affectation was more noticable there than where I grew up (otherwise I might have quit doing it, now that I'm in my 30's it takes effort not to, and I can't be bothered).

Ben said...

Hi, Tilde Jay -

As my brother-in-law in New Zealand would say, Good on you! Keep talking the way you talk, even if people laugh. I myself had a similar experience of being dialectologicaly radicalized by college. It's interesting, but not entirely surprising that your Bostonian feature played better in Chicago than Bloomington. In my experience, a Chicago accent gets a lot more attention in Oregon than in New York. This is probably due mostly to the greater cosmopolitanism of New York and Chicago, compared to Oregon and Southern Indiana, respectively, but I think another factor is that there is a deep kinship among the urban dialects of the northeastern quadrant of the US. This is why Dennis Franz can play a New Yorker and no one objects.

Ben Zimmer said...

What you're calling "intrusive intrusive /r/" is common among speakers of a non-rhotic background who are shifting to a new rhotic prestige model. (This is related to the explanation given for the South Midland pronunciation of "wash" as "warsh", which supposedly developed as a hypercorrection resulting from the non-rhotic Southern pattern's loss of prestige. When the rhotic pattern became prestigious in the region, "war" changed from [wO:] to [wOr], and some speakers added r's to other forms like "wash".)

I heard one truly amazing example of a speaker with "intrusive intrusive /r/" - Norman Siegel, a civil rights attorney and housing advocate from Brooklyn. Listen to this interview with Siegel on WNYC. I discussed Siegel's hyper-rhoticity in more detail on alt.usage.english last year.

Ben said...

I must admit that I'm a bit dubious about that explanation of the origin of "warsh" - first of all, to my ear non-rhotic Southern accents are way more prestigious than rhotic Southern and South Midland ones; the former sounds like plantation owners, and the latter sounds like toothless hicks. Also, the geography seems a bit screwy - how far east does "warsh" go? I don't know and I don't have any maps or data about it, but I'd be surprised if it went too far east of the mountains - still far from any non-rhotic area I know of. To me it seems simpler to explain it as an odd development of an odd vowel in an odd environment.

But I guess I don't know a lot about the history of /r/-lessness in the south, and it could be possible that it was once much more widespread and not prestigious.

The link to the Siegel interview didn't work - a shame; I'm eager to hear it. Siegel is a combatant in the fight over the future of my Manhattanville neighborhood. Me, I'm neutral.

Ben Zimmer said...

The WNYC link works fine for me. Here's a direct link to the streaming audio:

http://www.wnyc.org/stream/ram?file=/bl/bl022305a.ra

The explanation I remember hearing about "warsh" is that the Southern non-rhotic pattern became stigmatized in the South Midland region after the Civil War. But I'd have to check Labov to see if my recollection has any validity.

Anonymous said...

The south did go through a sudden reversal of prestige of rhoticity, yes—but the South Midland was always rhotic. Indeed, Labov's account of it is that the model for the prestige Southern accent shifted from the coast (nonrhotic Richmond and Savannah) to rhotic inland areas. So I don't see how you'd get hypercorrection in the South Midland from that.

Ben Zimmer said...

OK, I must have misremembered the explanation I heard in a lecture on phonological change. So what is the Labovian explanation for South Midland "warsh"?

Anonymous said...

I don't know if he has an explanation for it, actually.

Ben said...

Wow, that turned into an interesting exchange. A thread, if you will. I'm even more partial now to the theory I made up a few days ago to explain this phenomenon, that is, that it's a weird vowel development in a weird environment. The only other words I can think of where there's intrusive /r/ in southern Midland speech are "squarsh", "gorsh", and "Warshington." How many common words even have this vowel before that consonant? Not many, right?

Anonymous said...

Yeah, I was just talking about this with Keelan. The fact that it happens in all of "squash", "wash", and "Washington" indicates that it's a regular sound change, not a lexical diffusion or analogy of any kind—or at least that it started out that way. [S] is sort of articulatorily similar to [r] anyway, nu?

Ben said...

[S] is articulatorily similar to [r], emes. Speaking of [r], [S] and Yiddish, a development I like in Yiddish is the shift from [rs] to [rS], so that NHG 'erst' corresponds to 'ersht', 'anders' to 'andersh', etc.

Tony Dinosaur said...

What do you mean when you say the intrusive r is stigmatized in British english?, do you mean it belong to the posh side of society or the other way around?. I'm just getting into these new concepts, I'm just inlove with phonetics and languages. Your blog is terrific

Anonymous said...

Intrusive "R"s between words such as 'saw it' are common (virtually standard) in English, Australian, MZ, and Sth African brands of English. Saying 'saWit' would round ridiculous in those accents, so it's either 'sawRit' (Beatles style) if spoken quickly/normally, or, if the words are being emphasised, they are almost said individually (so no R required).

I think it's accents like the Boston accent, which used to be rhotic but lost the R for prestige reasons (in 19th Century), who at that time almost hypercorrected their non-R's to the point that 'saw it' has no R, and even a word like 'mannerism' is pronounced without the R! ('mannah-ism') When the r would always be pronounced in England and Australia.

Anonymous said...

Just now listening FOYLE'S WAR, Season 1, Episode 3: young "high-born" woman says, "Of course, it was a bad idea(r), it was my idea." I'm a midwestern American, fairly void of an accent, despite my being dropped into the Texas, where only God-knows-what is fairly frequently spoken. But i have to say that the intrusive r annoys me. It's just not there.