Are local accents doomed?
« previous post | next post »
Annie Joy Williams, "The Last Days of the Southern Drawl", The Atlantic 1/4/2026:
By the end of my life, there may be no one left who speaks like my father outside the hollers and the one-horse towns.
On Sundays after church, my family would pile into our crank-window GMC truck and head to Kentucky Fried Chicken. “Can I get me some of them tater wedges?” my father would say into the speaker, while my sisters and I giggled in the back seat. My dad has always had a southern accent: His words fall out of his mouth the way molasses would sound if it could speak, thick and slow. But his “KFC voice,” as my sisters and I call it, is country. It’s watered-down on work calls and during debates with his West Coast relatives. But it comes out around fellow cattle farmers and old friends from Kentucky, where he grew up.
My mother’s accent isn’t quite as strong. She’s a therapist, and she can hide it when she speaks with her patients and calls in prescriptions. But you can always hear it in her church-pew greetings, and when she says goodnight: “See you in the a.m., Lawd willin’.”
I was always clear on one fact: I wasn’t going to have a southern accent when I grew up. I was raised in Murfreesboro, Tennessee, near Nashville, where the accents grow stronger with each mile you travel from the city. I watched people snicker at the redneck characters on television who always seemed to play the town idiot. I knew what the accent was supposed to convey: sweet but simpleminded. When I was 15 and my family went to New York for the first time, the bellhop at our hotel laughed when my mom and I spoke; he said he’d never met cowgirls before. That was when I decided: No one was going to know I was from the South from my voice alone.
The article sketches a conversation with Margaret Renwick, links to two of her studies ("Boomer Peak or Gen X Cliff? From SVS to LBMS in Georgia English" and "Demographic Change, Migration, and the African American Vowel System in Georgia"), and lays out some of the reasons for homogenization of local varieties, including migration and ethnocentric prejudice.
And then there's a series of (positively-evaluated) discussions about code-switching, offering hope that the future of American speech may be less homogeneous than the title suggests.
The whole article is well worth reading.
It doesn't discuss the process by which new varieties emerge and spread, but that would be a distraction from its nostalgic tone. Still, it's worth noting that a similar set of issues form the background of George Bernard Shaw's 1916 play Pygmalion, and in fact have been around, in one form or another, since the origins of spoken language. It's true that the internet and social media are a new source of change, just as in the past there were effects of agriculture, writing, empires, universal education, and broadcasting. But it's been hundreds of years since (for example) the Romance dialect continuum coalesced into a few national languages, with the associated gradual loss of tens of thousands of local varieties.
And there's plenty of evidence that American regional varieties are diverging rather than converging — see Bill Labov's 2012 book Dialect Diversity in America, whose blurb says
The sociolinguist William Labov has worked for decades on change in progress in American dialects and on African American Vernacular English (AAVE). In Dialect Diversity in America, Labov examines the diversity among American dialects and presents the counterintuitive finding that geographically localized dialects of North American English are increasingly diverging from one another over time.
Contrary to the general expectation that mass culture would diminish regional differences, the dialects of Los Angeles, Dallas, Chicago, Birmingham, Buffalo, Philadelphia, and New York are now more different from each other than they were a hundred years ago. Equally significant is Labov's finding that AAVE does not map with the geography and timing of changes in other dialects. The home dialect of most African American speakers has developed a grammar that is more and more different from that of the white mainstream dialects in the major cities studied and yet highly homogeneous throughout the United States.
Labov describes the political forces that drive these ongoing changes, as well as the political consequences in public debate. The author also considers the recent geographical reversal of political parties in the Blue States and the Red States and the parallels between dialect differences and the results of recent presidential elections. Finally, in attempting to account for the history and geography of linguistic change among whites, Labov highlights fascinating correlations between patterns of linguistic divergence and the politics of race and slavery, going back to the antebellum United States. Complemented by an online collection of audio files that illustrate key dialectical nuances, Dialect Diversity in America offers an unparalleled sociolinguistic study from a preeminent scholar in the field.
Increasing divergence doesn't imply stasis — on the contrary, obviously. But still…
Update — Williams' description of her father's speech ("His words fall out of his mouth the way molasses would sound if it could speak, thick and slow") is evocative, but may not be empirically accurate. See "Regional speech rates", 10/13/2007.
And for a striking example of inter-ethnic phonetic prejudice, see Michael Lewis (who's from New Orleans) ridiculing the pronunciation of a lawyer from southern Indiana, discussed in "Lazy mouths vs. lazy minds", 11/26/2003.
Update #2 — For those who aren't familiar with the way people from rural Tennessee speak, here's a clip of Trae Crowder:
Jerry Packard said,
January 6, 2026 @ 8:15 am
Are they serious? The very idea that it is local dialects that are doomed is laughable to say the least. It seems to me that if anything, it is standard dialects that are most subject to change while local forms go on their merry way. My guess is that standard dialects would have the most difficulty maintaining their unchanging status quo. The non-standard dialects exist in the pockets, canyons and hollers, and may well persist and outlive their more visible cousins.
Jenny Chu said,
January 6, 2026 @ 8:56 am
Here's the most interesting part of the excerpt, for me:
> I watched people snicker at the redneck characters on television who always seemed to play the town idiot. I knew what the accent was supposed to convey: sweet but simpleminded.
This is clearly someone who associates the "redneck" quality with something negative / non-prestigious. And given that she's now writing for _The Atlantic_, it's easy to imagine her point of view.
But given the increasing sociological divide / culture war between rural and urban populations, it wouldn't surprise me to see some speakers of rural dialects embracing – deliberately or unconsciously – their local variety of speech as a way to differentiate themselves from the (perceived) effete/woke urbanites. It's not materially different from any other in-group signifier.
Philip Taylor said,
January 6, 2026 @ 11:39 am
As I know next to nothing about American dialects/topolects/whatever, I have nothing to contribute to the discussion, but there is one point about which I am intrigued : how does one pronounce "Murfreesboro" (a) if one comes from Murfreesboro, and does not seek to conceal the fact, and (b) if one speaks "Gen.Am" ?
Fritz Newmeyer said,
January 6, 2026 @ 11:48 am
As a linguist, I would love to believe that local dialects in the United States are not dying out, or, as some claim, even getting stronger. But sadly that claim seems totally refuted by my personal experience. I was a child in the New York City area in the 1950s. At that time, one could often tell what neighborhood one was from by their accent. I meet young New Yorkers now and in general cannot even tell that they are New Yorkers from their speech patterns.
Or to take a very different example, consider Newfoundland. Traditionally Newfoundlanders have had the most distinctive accent (actually, group of accents) in North America. My colleagues there now complain that young Newfoundlanders have tended to adopt standard Canadian English, with only the slightest trace of their historical dialect/accent.
Now, maybe this claim applies only to people below a certain level of income or education and who have pride in their local area. Even so, I am skeptical. That certainly does not seem to apply in New York, though I admit that I have not made a neighborhood by neighborhood study.
I am wondering if other have made the same observations that I have.
Roscoe said,
January 6, 2026 @ 12:02 pm
"His words fall out of his mouth the way molasses would sound if it could speak, thick and slow"
Interesting – that’s almost exactly how onecharacter on the U.S. version of “The Office” instructed another character to speak in a convincing Georgia accent (after diagnosing her attempt as “more of a Florida Panhandle twang”).
Jarek Weckwerth said,
January 6, 2026 @ 12:26 pm
@Fritz Newmeyer: In England at least the leading analysis these days seems to be one of "accent levelling" where differences are still there between larger areas but are increasingly levelled within them. So, instead of dozens easily distinguishable accents (down to the level of boroughs in London), you have an increasingly homogenous group in the South-East, another larger grouping in the Midlands, etc. And also, new accents appear that do not follow traditional accent geography, e.g. Multiethnic London.
Fritz Newmeyer said,
January 6, 2026 @ 1:31 pm
@Jarek Weckwerth: The last part of what you write is certainly true. Vancouver, where I live, has an enormous population of Asian-descended people. It has been noted that a new accent is developing, used even by second-generation speakers, with features from (in general) Chinese or Punjabi. The same sort of thing is true in Montréal, where younger people's French shows influences from Haitian Creole, English, Arabic, and other languages spoken in Africa.
Vince said,
January 6, 2026 @ 1:53 pm
@Philip Taylor: I don't know how to pronounce "Murfreesboro" in either the local dialect or more general American, but you might find the pronunciation in Tom Waits' song "Pony" of interest. Maybe it matches one of the two? And the song (I think) is great.
Stephen Goranson said,
January 6, 2026 @ 4:18 pm
How to pronounce Murfreesboro? Depends whether you all are from Murfreesboro, Tennessee or Murfreesboro, North Carolina.
JPL said,
January 6, 2026 @ 4:31 pm
@OP:
"Williams' description of her father's speech ("His words fall out of his mouth the way molasses would sound if it could speak, thick and slow") is evocative, but may not be empirically accurate."
Yes, I too found that paragraph puzzling, because I couldn't figure out just what was happening in the situation being described. Could the following possibly be a more empirically accurate account of what was going on there?
"… giggled in the back seat. This was because, as usual, his words poured out of his mouth the way molasses pours out of a cooking pot, but we could tell that, in an attempt to appear less local, he was putting on his "KFC voice" for the benefit of people who he thought might be more cosmopolitan. Unfortunately, the leisurely crawling of his delivery was not erasable, and gave him away."
J.W. Brewer said,
January 6, 2026 @ 6:22 pm
Murfreesboro is maybe not a place you would have expected dialect stability, because its population has quintupled since 1980 – presumably some of the newcomers are from not very far away but others are. Some are from outside the U.S. – there's a school there that provides supplemental Japanese-language instruction on the weekends to children whose parents desire that. It's also a college town and even if Middle Tennessee State doesn't draw too many out-of-state students you would expect the faculty to be of primarily out-of-state and often out-of-region backgrounds. (Indeed, MTSU would like you to know that Bangladeshi Nobel Peace Prize winner Muhammad Yunus' first post-Ph.D. job was as an assistant professor of economics there.)
This doesn't mean there hasn't been substantial shift in accent over the last generation or two in other parts of rural Tennessee where the population has been relatively stable and there have been relatively few incomers, of course, but you need to look at data (or at least anecdotes …) from those places to assess that.
Richard Hershberger said,
January 7, 2026 @ 5:19 am
I can personally attest that AAVE is alive and well. I commute on public transit in Baltimore and hear AAVE spoken daily.
Jerry Packard said,
January 7, 2026 @ 6:25 am
I traversed Murfreesboro twice a year for many years, often stopping for the night. The folks in contact with we travelers had more of a minor drawl than southern Illinois, and definitely had the expected [I/i] merger. But the pronunciation of Murfreesboro was phonetically the same as its spelling, as far as I could tell. This site doesn’t do IPA well, so I’ll just say it was as you’d expect, with the [s] pronounced as a voiced [z], and boro pronounced as [ber-oh].
Philip Taylor said,
January 7, 2026 @ 7:39 am
Thank you Jerry. In the U.K., final "boro" is normally rendered as /ˈbʌr ə/ or even /ˈbrə/, whence my question. So now a subsidiary question — from whom does Murfreesboro take its name ?
Chris Button said,
January 7, 2026 @ 10:16 am
It's weird how the whole perceived speed of language is often put down to stereotypes.
While actual differences in speed are usually negligible to none across languages, the reason for the mistaken perception of differences seems to come down to reasons of stress and vowel reduction.
So, for example, Spanish is perceived as fast to non-speakers relative to Portuguese because most varieties of Spanish don't reduce vowel quality in unstressed syllables.
J.W. Brewer said,
January 7, 2026 @ 10:22 am
@Philip T.: I know no more than what your own internet searching could have suggested, but that would have suggested https://en.wikipedia.org/wiki/Hardy_Murfree. Murphree looks to be a variant of the Irish surname much more commonly spelled these days as Murphy.
Rodger C said,
January 7, 2026 @ 10:24 am
Murfreesboro takes its name from a prominent Tennessee family whose best known (to me) member (by marriage) was the writer Mary Noailles Murfree.
It is pronounced MUR-frees-burra (/ˈbʌr ə/ indeed) or, rapidly, something like "Murphy's burr."
Philip Taylor said,
January 7, 2026 @ 11:06 am
Thank you Rodger. So it would seem the final "boro" can be rendered both as [ber-oh] (Jerry's transcription, which I take as something close to /bɜːr oʊ/, tho' I am uncertain of the vowel implied by [ber]) and /ˈbʌr ə/ in <Gen.Am>. Interesting. Incidentally, Jerry, if you are still following this thread, what exactly did you mean by "[t]his site doesn’t do IPA well" ?
Terry K. said,
January 7, 2026 @ 12:13 pm
Specific local accents, yes, they will disappear eventually. Change happens. But others will take their place. Maybe even with the same names to describe them.
Jerry Packard said,
January 7, 2026 @ 2:07 pm
When I say this site doesn’t do IPA well it’s based on my memory of having tried to copy and paste certain vowels, such as the ‘retroflex’ and ‘apical’ vowels in shi 是 and si 四 respectively. In such cases the site produced them as null square boxes.
Mark Liberman said,
January 7, 2026 @ 6:39 pm
@Jerry Packard: "When I say this site doesn’t do IPA well it’s based on my memory of having tried to copy and paste certain vowels, such as the ‘retroflex’ and ‘apical’ vowels in shi 是 and si 四 respectively. In such cases the site produced them as null square boxes."
That's almost certainly not the site, but rather the font used by your browser, which presumably lacks coverage for certain regions of unicode.
And note that font choice can be modified in the browser's settings (though html can also specify font-family for specific regions).
Philip Taylor said,
January 8, 2026 @ 4:50 am
As I can't be certain as to exactly which "‘retroflex’ and ‘apical’ vowels in shi 是 and si 四" you refer, Jerry, let me try a simple copy-and-paste and we can then compare how it renders in our browsers :
Philip Taylor said,
January 8, 2026 @ 6:39 am
Which looks like this when viewed in Firefox under Windows 11. The DOM inspector reveals that the font used to render ɹ̩ [U+0279 : LATIN SMALL LETTER TURNED R
U+0329 : COMBINING VERTICAL LINE BELOW] is Arial Unicode MS.
Philip Taylor said,
January 8, 2026 @ 6:40 am
Which, bearing in mind an earlier comment, may be more easily seen if I include the URL as plain text — https://www.dropbox.com/scl/fi/hubkjx0f1s1bazgcho4zc/Screenshot-2026-01-08-11.26.16.png?rlkey=w2amhfvwbkoasmevym94622by&dl=0
David Marjanović said,
January 8, 2026 @ 7:53 am
Usually, yes, but it does happen. Most people all over France have the exact same accent, with the same level of vowel reduction (t'veux conduire, ou t'veux qu'j'conduise ?), and still the Parisians pack, I'd say, twice as many syllables into the same time.
This can change particularly fast. My two sisters (5 years apart) speak at very noticeably different speeds, and so do their friends of the same ages.
David Marjanović said,
January 8, 2026 @ 7:54 am
Most young people. Trump jumps to the next sentence while he's still saying the previous one, I jump to the next word while I'm still writing the previous one…
Jerry Packard said,
January 8, 2026 @ 9:16 am
@Philip
Yes, good work, thank you. I have always, since the work of my colleague and mentor C-C. Cheng (Chin-Chuan Cheng, 郑锦全）, thought of the two vowels as the non-standard, non-IPA [ɿ] and [ʅ] for the apical and retroflex vowels respectively. Until now, I had not been successful in pasting them onto this site. Those two phones are accurately and validly represented as simple vocalic continuations of their preceding consonants, as many phonemic analyses have posited and as you have accurately stated, with all kinds of repercussions for the Mandarin phonemic system, most notably that the Mandarin syllable minimally requires a V (main vowel) element.
@Chris, David
The differing speeds of syllable articulation across languages is not a matter of perception but of fact, as acoustically measured by Pellegrino et al. (2011) and Coupé et al. (2019).
Philip Taylor said,
January 8, 2026 @ 9:58 am
"Most people all over France have the exact same accent" — well, I'm in no position to argue against "most", but I can think of two occasions (one recent, one distant) where the accents differed so much that an English ear could easily discern the difference. The first (the earliest) was in Lille, where I was taking lunch with a group of French delegates to a conference at which I was presenting a paper, and during the course of the lunch I heard one speaker say (sorry, my IPA skills are not really up to transcribing French accurately) EX-ˈOM-PULL while the second said EX-ˈOM-PLUH. I asked about this, and was told that they came from geographically distinct regions of France. The second, very recent, is when a native French speaker pronounced "Armagnac" in such a way that the second syllable appeared to have been completely swallowed. I asked our Head Chef (also a native French speaker, born and raised in the banlieues of Paris) about this, he pronounced "Armagnac" almost exactly as I do, and said I needed to find out from which part of France she came (which I have not yet done).
Rodger C said,
January 8, 2026 @ 10:44 am
Philip, I suspect that Jerry with his "oh" is rom the North. I'd be surprised if I heard it pronounced that way by anyone in Tennessee.
Jerry Packard said,
January 8, 2026 @ 2:01 pm
I am indeed from western Massachusetts, considered part of the northeastern NY state dialect area – like Albany, Syracuse and Rochester.
Chris Button said,
January 9, 2026 @ 8:45 am
@ Jerry Packard
Perhaps it's both?
For example, "diferente" in Brazilian Portuguese "sounds" much slower and relaxed than in Spanish. Admittedly, the Brazilian Portuguese one does end up as three syllables (dif.e.rente) rather than four (di.fe.ren.te) though.
Jerry Packard said,
January 9, 2026 @ 12:13 pm
Well it may well be both. But the acoustics is measured and presumably replicable.
Barbara Phillips Long said,
January 9, 2026 @ 3:42 pm
My family spent most of the 1990s in Maysville, Kentucky, along the Ohio River about an hour’s drive east of Cincinnati. That area of Kentucky had a wide range of accents, even among locally raised folks.
My recollection is that one local woman was particularly admired for her “pretty” Southern accent. My neighbor, from southern Alabama, had a distinct accent that I thought was similar, but I don’t remember women gushing over it the same way (I don’t recall men discussing accents). Women seemed to like an acquaintance’s Georgia accent as much or more. The “pretty” accent seemed to be a status marker of some sort, but the speaker said she’d just always spoken that way.
Whatever accent the Atlantic writer wanted to shed, it probably did not convey the higher status of the “pretty” accent. It’s also interesting that an encounter with a derisive New York City male precipitated the writer’s desire to speak differently. In my experience, living in rural upstate New York where “summer people” came to stay in the northern Catskills, derisivenness was a common trait in visiting males from New York City. Locals talked for years about a memorable tantrum thrown by a visitor who could not believe that there was not a single one-hour shirt laundry in the entire county.
Mark Liberman said,
January 10, 2026 @ 9:06 am
@Jerry Packard: "The differing speeds of syllable articulation across languages is not a matter of perception but of fact, as acoustically measured by Pellegrino et al. (2011) and Coupé et al. (2019)."
Those papers (Pellegrino et al. (2011) and Coupé et al. (2019)) don't measure subjective perceptions of speaking rate, which I suspect are empirically complex at best. They're focused on showing that "information rate" is similar across languages, which of course requires that languages with lower syllabic entropy must produce syllables at a more rapid average rate.
And they don't consider, much less test, the possibility (indeed certainty) of cultural, individual, and contextual differences in information rate (causing or caused by differences in syllabic rate, lexical complexity, etc.). Even independent of pure ethnocentrism, distributions of such differences might give rise to folk-ethnological beliefs/prejudices.
The point in the cited blog post was that the standard prejudices about "Southern drawl" in American English appear to be false to (average) fact.