Philadelphia Street (Photo: Adam Jones)
I heard rumors in college of a speech teacher with an exceptional knack for guessing dialects. He could supposedly pinpoint, within ten miles, where a student was from. “Ohio,” he would deduce. “About seven miles west from Akron.” “Bangor, Maine.” “St. Louis. The western suburbs.”
Such legends have floated around drama schools and linguistics departments for years. All are variations of “Higgins’ Boast,” the claim Shaw’s Pygmalion protagonist makes in Act I:
You can spot an Irishman or a Yorkshireman by his brogue. I can place any man within six miles. I can place him within two miles in London. Sometimes within two streets.
Alas, I find that most such boasts are exaggerations, although more plausible when Shaw was writing. Higgins was “living” in the 1910’s, a time when streets, boroughs and villages may very well have had distinguishable accents. While recently reading Douglas W. Rae’s The City, a book about American urbanism, it struck me how dense and insular communities were during the industrial era. So perhaps Epsom and Lisson Grove were as linguistically unique as Higgins suggests.
That being said, Henry Higgins was “speaking” before modern tape recorders existed, so I don’t entirely buy his claim. He implies, for instance, that he can make such deductions as far away as Ireland. But unless Higgins had bottomless financial resources and travelled constantly, it’s unlikely he could spot the difference between, say, two nearby villages in County Kerry.
If someone were to make a similar boast today, we would find the opposite problem. One can imagine the endless hours Higgins would spend on YouTube; given the “corpus” of English speech available online, a man of his talent could come close to achieving such mastery.
But conversely, the kind of fine lines Higgins draws in Pygmalion are rarer now than they were in the late Edwardian. Given our economic and physical mobility, I have a hard time believing that there is a “Cambridge Heath Road accent” distinguishable from a “Vallance Road twang.”
Which is not to say that such guesses are impossible. For instance, while recently watching a documentary about the making of Hitchcock’s Psycho, I accurately guessed the neighborhood of the exact city where screenwriter Joseph Stefano grew up (answer after the clip):
Stefano is from South Philadelphia. My shot-in-the-dark might sound impressive, but like a cheap psychic, I relied on luck and probably some unconscious cheating. That Stefano is from the Philadelphia area is unquestionable. That I guessed he’s from South Philly, however, may have been influenced by his Italian last name.
So while I think such geographically precise guesses are possible in some cases, I find the idea that someone can pull this trick off for the entire English-speaking world (or even the entirety of America or the UK) fairly implausible. People differ too greatly in the degree to which their accents are “local.” And regions differ too greatly in the degree to which they differ!
Have you ever met someone with an ability as impressive as Higgins? Or are such feats the stuff of legend?
People think they can recognize a Brooklyn accent when in reality it’s a working class accent that is much more widespread. So you’d have to require your expert dialect detector to be specific about social as well as regional dialects. And don’t forget occupational groups, gender and age cohorts.
Insider and outsider status counts for a lot, too, as your County Kerry example suggests. New Englanders sound alike to most General North American speakers, but to insiders Massachusetts sounds different from Rhode Island or Connecticut, or New Hampshire. In Massachusetts, Boston sounds different from Worcester, etc.
I think William Labov (and associates) could place speakers somewhat reliably with the help of their extensive database. Otherwise, I think that people think they are much better at dialect detection than they really are.
Nothing like some good cognitive bias
I know exactly what to say on this one.
There were reports that Stanley Ellis, one of the fieldworkers in the Survey of English Dialects, had this ability. One case that received media attention was when (apparently) he located Wearside Jack (a hoaxer who impersonated the Yorkshire Ripper) to the village of Castletown in Sunderland. The police diverted their attention to this area. When Jack was eventually caught, he turned out to be from the Castletown area.
When Ellis died, this was mentioned in some of the biographies. For example, this in the Guardian or this in the Telegraph.
You can hear the recording here: it’s the first audio file on this page by Jack Windsor Lewis.
However, I’m not convinced. I think that either Ellis made a lucky guess or that the reports got exaggerated (I can’t find any direct comments from Ellis on the subject). He worked in County Durham during the SED, but the speech that he was investigating was on the brink of extinction at the time and Wearside Jack did not have a particularly old-fashioned accent. In addition, the recording was only short and of poor-quality, and Jack spoke slowly. To locate someone to a village, you must need more material than that.
P.S. Ian Catford’s Wikipedia article says that he was able to “identify where people were from exclusively through their speech.” Presumably, this applied within his native Scotland. There’s no reference for this claim.
I looked him up after finding this In Memorium for him and Stanley Ellis in IPA News. Perhaps someone else knows more about his work and can add a comment.
Re “In Memorium” — am I wrong to continue spelling it “In Memoriam”? Will I one day lose a job, or lose the respect of others, because of knowing this?
I’d have said Sunderland without the heads-up. It wuold have to be from either there or very close by. I don’t see any way you could pinpoint to Castletown, though. It’s very much a part of Sunderland rather than a village in itself, with regard to dialect as with everything else.
That’s interesting. Ellis worked at Washington (the one near Sunderland) during the SED, and you can hear his talking with two retired miners here. Castletown is fairly close to Washington.
The Telegraph obituary above mentions that Humble lived on the opposite side of the river from Castletown (I believe that there’s a large council estate around there). I interpreted this to mean that he lived there when he was arrested, but it might mean that he’d always lived that side.
Given that Ellis had done fieldwork across the north of England, it’s plausible that he could locate northerners within 10 miles but he couldn’t have known the speech of every little village.
According to Wikipedia, he’d been to school in Castletown so, if true, that would give some credibility to the claim. I can’t think what it would be though. There’s nothing about his speech that obviously distinguishes it from any other part of Sunderland.
The phonetician Jack Windsor Lewis, who worked on the Wearside Jack tape alongside Stanley Ellis, has posted this response to our discussion. He says that Stanley Ellis noticed similarities between Wearside Jack and an informant on the dialect of of Washington (England) who he recorded. In addition, he agrees with Amy Stoller’s comment below that the Higgins Boast was probably meant to be fictional.
Cities are way harder, as accents varying within those are more class-based. If a person from Los Angeles is non-rhotic, you can reasonably assume they’re from the northeastern part of the city, but they could be from any neighbourhood. The same with the allegedly “saath lawndawn” accent.
But neighbouring cities are easier: there’s no confusing an Angeleno for a Pasadenan, or a San Franciscan from a San Josean.
“But neighbouring cities are easier: there’s no confusing an Angeleno for a Pasadenan, or a San Franciscan from a San Josean.”
I highly doubt that (and most of the other things you’ve said), but whatever.
Meh. The research out west is spotty as hell, but people who live in those little pockets are well-aware of the difference. SF, Oakland, and the older suburbs around them are much more like Philly/Baltimore than the rest of California. Pasadena has a weird backing trend.
Northern New England is like that too. New Hampshire sounds like a mix of Boston and Maine that’s somehow more extreme than either of the two.
I know that for elder speakers what you say is somewhat true but younger speakers are generally absorbed into the dialect leveling (SF still! taking its time cot-caught merging xD)
Who says “meh”? You’re the least hip hipster I’ve ever met.
you DONT say meh?
Actually, when you are fluent, it is quite easy to distinguish San Diego, LA, San Francisco, and Sacramento. It is even easy to distinguish Latino accented English from San Diego, LA, San Francisco and Sacramento, and Asian accented English from these cities. When I was more fluent, I could imitate each one to demonstrate. One trouble spot I have is with the Huntington Beach to Long Beach area: I can’t detect distinguishing accents from those areas. So if I know someone is from California, but I can’t place the accent, I can be pretty sure they are from that area.
It definitely depends on my personal familiarity with the regional accent. I can peg an older (maybe ~60+?) speaker from the area of the South Downs (UK) where I grew up at 200 paces. Not so much with younger speakers, as I suppose tv etc has smoothed out the accents a fair bit. And if it’s an area I’m not so familiar with I can’t be nearly as specific.
I agree that when you are intimately familiar with a region, a number of minute distinctions come to the forefront. That’s another factors in my deduction about Stefano: he has a slightly opener-than-usual vowel in functional words like ‘can’ (the verb) which I strongly associate with South Philly. It’s a connection I probably wouldn’t make if I didn’t spend as much time in the region as I do.
Though I was born in Canada, my family is from the Netherlands. I’ve heard stories about people in the past who could guess where someone was from in the Netherlands (and maybe the Dutch speaking part of Belgium) within 10km just based on their accent/dialect, though I’ve never met one. Since the Netherlands is much smaller, and does appear to show a great deal of accent/dialect variation, it’s more plausible, though in light of similar claims for English, perhaps it was just a myth. Although, there is an astounding degree of linguistic diversity present in my mom’s home province of Friesland, which you could drive across in about an hour. There are many dialects of Frisian alone, not to mention different dialects of Dutch and Dutch-Frisian hybrids.
I guessed South Philly too. It was the fact that he had a Mid-Atlantic accent and an Italian last name. I figured there must have been some “clue” that would allow us to correctly guess where he was from, otherwise you wouldn’t have brought that up.
His accent could easily pass for Baltimore, or South Jersey (Burlington County, but not Ocean County or further south). His accent has been softened from South Philly, by years in Southern California. He has also taken on a bit of a Californian lilt (he hangs on to some syllables that would be rough and clipped in South Philly). Spend a day on Market Street, or buy a cheesesteak, and you will hear an much rougher accent, more like Rocky Balboa. It is fun to listen to for a while. Stefano is definitely speaking with a Philly/South Jersey/Baltimore accent with an overlay of something smoother.
I think Mark Twain used a lot of dialect/accent-variations that have now disappeared or morphed into larger regional dialect/accent-variations.
I find I’m more able to identify accents (or accent differences) when they are socially meaningful. I can travel to a pretty wide area before I notice that people have accents different from mine. And the only time I notice the area where I live having an accent different than mind at all (250 miles different between the two, in the U.S.) is occasionally with the pin/pen merger, which they have here but I don’t have.
But, yet, hick accents (there’s a better term for that, I’m sure, but I don’t know it; an accent that identifies someone as a hick) from this area (including both where I grew up and where I live now in “this area”) I notice.
In Ireland, Dublin area accents sound distinctly different to me than Galway accents. People from Dublin nearby areas sound like people I know (American residents I know from the area). People from Galway sound like movie characters. I could not, though, tell the difference between the different accents in and around Dublin.
I have never met anyone who could live up to Higgins’ boast. That some people within their own communities can accurately identify others from nearby communities based on their speech I do not doubt. But that is a very far cry from what Shaw decided Higgins could do.
That some people are both especially knowledgeable and/or especially gifted is also true. But this? “Brought up in Hounslow. Mother Welsh, I should think.” Let us remember that Pygmalion is fiction, and comedic fiction at that. Does it contain large truths? Yes. Is every detail a verifiable fact? Absurd.
Moreover, I’d like to see the actor, then or now, who could produce such an accent accurately. There is a difference between stagecraft and science, as Shaw well knew.
Ben, do you get tired of people asking you to place them on the basis of their speech during casual conversation? I do. Even if I could do so, my work is not a party trick; it is a profession. I don’t ask doctors to diagnose me (for free!) when I meet them at cocktail parties – and how many of us have encountered a faulty medical diagnosis at some point in our lives? How many people have been sent to prison, or their deaths, though innocent of any crime? Henry Higgins, like Sherlock Holmes, is not real, no matter how real Henry Sweet and Joseph Bell may have been.
In other words: Grrrrr.
Amy,
I tend to be a bit–maybe “irked” is too strong a word —wary of such requests. Such games are fun in certain controlled circumstances, like this quiz where you match up broad British regional accents to their correct locations. But it’s a different story when your “test” is someone from Manchester who spent a few years in London and the past decade in New York. People differ tremendously in how recognizable their regional accent is, a fact that many people don’t grasp.
Were those real accents? They all sounded put on, especially the Southern Irish one. And there were mistakes too. It said the northern Irish one was Southern Irish. I should probably complain to them, not here.
When you click on the answers, the numbers for each accent don’t match the numbers that you used to answer the quiz.
Nope, it’s simply wrong. That’s okay though.
I doubt that anyone can do it. I am reasonably good at identifying fake Southern accents, having lived in the South since birth, but even I have trouble identifying the locale of specific Southern accents. I think I can identify the particulars of my own region (NW Georgia), and maybe that of the eastern, near-coastal region of Georgia, tucked up against the Savannah River between Augusta and Savannah. But beyond that, and maybe even within that, it’s hit or miss. I always thought Jodie Foster did a good job with her accent in the Silence of the Lambs. It sounded genuine, with essentially none of the fake features so many actors end up with. But I could not have told you where that accent was supposed to have originated.
Sorry to jump on an older post, but wanted to second this. I’m NE GA raised, and can generally peg “South Georgia” vs North, and Augusta or Savannah comes through every time (they both sound like mint juleps, but Savannah seems a bit more refined to me, while Augusta runs high).
If there’s a bit of a hill twang, I can tell Tennessee, Kentucky, or Alabama/South Carolina uplands, but North Carolina Mountain and North Georgia Mountain sound exactly alike to me.
Do you suppose it has to do with geological pocketing? North Georgia has a lot of different landforms in a fairly tight space.
My mom said she met someone at a party who correctly guessed she was from Western Montana, which I think is pretty impressive.
Also, on the wildly improbable assumption that everybody reading doesn’t already know about this website: http://accent.gmu.edu/browse_atlas.php
I agree with Amy; identifying micro-accents is a nice party game for language nerds.
I wonder, is it possible to make predictions about accents you’ve never actually heard? “You live north of X and east of Y, therefore you should say Z”
Ben, could you tell us what features of Stefano’s speech alerted you to the facted that he’s from Philadelphia?
Sure! Mostly it had to do with the relatively open vowel in words like ‘began,’ his consistently raised /ai/ diphthong before voiceless consonants (as in “Psycho”), and the inconsistent way he produced the vowel in words like “goat” (which wavers between being quite fronted, as is often the case in the Mid-Atlantic, and somewhat more back).
Incidentally, another non-linguistic factor that helped me out was that I could deduce Stefano had been born in the early 1920’s or so. It might have been a bit more tricky if he were a bit younger, as the accent started to “move to the suburbs” after WWII. (As is the case with many American urban accents).
Henry Lee Smith of Trager-Smith used to tell that on his radio show. Check here:
http://www-personal.umich.edu/~jlawler/aue/smith.html
His radio program was “where are you from”
http://www.buffalo.edu/ubreporter/2011_10_27/flashback
Another phonetics/phonology Bob Ladd says the same about Smith: “Expert humans used to be able to make such judgements with detail comparable to that achieved by the computer in the cartoon. When my father was an undergraduate at Brown in the late 1930s, he was on a radio program called “Where are you from?”, hosted by Brown linguist Henry Lee Smith (the Smith of Trager and Smith, for linguists of a certain age). As my father used to recall it, Smith asked him to say “marry merry Mary” and promptly judged that he was from “within 15 miles of the Cape Cod Canal”, which in fact he was. Whether the story was embellished in the retelling I have no way of knowing, but I have no reason to doubt that there was such a radio program or that Smith could actually do this sort of thing. Since (according to Mark’s first graph) human listeners are still substantially outperforming machines, it seems unlikely that we will be seeing scenes like the cartoon any time soon, and an electronic Smith is probably still several decades off.” From http://languagelog.ldc.upenn.edu/nll/?p=904
I don’t know about other areas, but in New York, it seems to me the lines aren’t geographical (for all the talk of a “Brooklyn accent” or a “Bronx accent”) but ethnic. Jews in Brooklyn sound less like Italian-Americans in Brooklyn than Italian-Americans in the Bronx do.
More relevantly: my girlfriend has gotten over the phone at work “I can tell you’re in New York because of your accent.” She’s from Atlantic City.
I think Labov and other sociolinguists have concluded that the Brooklyn accent is a working class accent (not geographic), which would correlate somewhat with ethnicity.
I wonder, can one tell where in the former USSR a native Russophone comes from by his or her accent in Russian? In English?
I don’t know anything about Russian accents, but if the variation in Russian accents is similar to that of Dutch accents, then it might be possible for a native Russophone to guess where another one comes from based on their accents in English. I once witnessed two Dutch people speaking to each other in English doing the same. To the extent that it’s possible in the native languages, though, I would suspect the precision is lower when guessing based on their accents when speaking English.
Pingback: This Week’s Language Blog Roundup: Election, WOTY, and terrifying origins | Wordnik
Based on my experience, I would expect experts to be quite accurate for uncorrupted accents. I spent a decade in the Navy, working with people from all over the country. It was a game to imitate all the accents and local idiom you could hear, and from that it became quite easy to differentiate many accents and place them. It was easy to distinguish, for example, accents from major Southern cities, or cities and there distinct surroundings (Indianapolis from southern Indiana, Chicago from the North Shore, etc.) . If amateurs can get that accurate, experts who actually study dialects should be able to do even better. But when accents become corrupted, as when someone spends decades in a new setting and adopts that accent on top of their native accent, I bet it is pretty hard.
Pingback: Henry Higgins, Language Snob | Dialect Blog