16

Yes, I admit, as an AmE speaker, that all non-North American accents sound the same: BrE, Irish, Scottish, Australian and South African. Or rather, I can tell they are different if placed side by side as in the excellent 21 accents but I can't name them off in isolation, unless I pick out one very specific feature (e.g. he said 'bairn' for 'baby', must be Scottish, she said 'bruvvah' for 'brother', must be London).

Is there some way to say objectively, comparing them all to each other, that, say, dialects A and B sound more similar than A and C or B and C? And really, with specific dialects...To Australians, for example, does AmE and, say, BrE actually sound alike? I see a distance matrix:

     BrE  AmE  IrE ...
BrE   0    7    2  ...
AmE        0    4  ...
IrE             0  ...
...                ...

(of course the off diagonal numbers are picked out of thin air). Is there some less subjective comparison of distances (like how many people of one kind misunderstand those of another)?

Which ones are the most 'out there'? Do the accents in the British Isles share lots of similarities but the former colonies have little with each other or the isles?

tchrist
  • 134,759
Mitch
  • 71,423
  • 5
    Random note: A friend of mine and I were visiting Scotland and met a guy who grew up in Edinburg but whose parents were British. He insisted he had never spent any time around Americans but had a dialect that was nearly identical to ours (a fairly neutral AmE), save a word that would float in every 2 to 3 minutes that sounded completely different. We concluded that this was evidence that (if such a hypothesis can be made) the American dialect is an exact mix of the English and Scottish. – snumpy Apr 25 '11 at 21:30
  • 1
    @snumpy: There is also the influence of television and films... – Cerberus - Reinstate Monica Apr 25 '11 at 23:23
  • 2
    @snumpy: except that apparently Americans think Edinburgh has three syllables. – Tim Lymington May 11 '11 at 22:30
  • @TimLymington You are quite correct. I pronounced it Ed-in-burg (Which, consequently, is how you pronounce the name of a city in Texas which is spelled without the "h") until I learned that it's a lot closer to Ed-in-borough (without fully pronouncing the last syllable). – snumpy May 12 '11 at 18:21
  • 1
    The IPA chart for English phonetics gives the pronunciation for the lexical sets each major dialects of English. That may be usable to create a distance function. – Mitch May 16 '11 at 14:13
  • 2
    @TimLymington Frequently Scottish pronunciation does give Edinburgh 3 syllables - Ed-in-bruh – neil May 16 '11 at 15:17
  • I also thought it was pronounced ED-in-bruh. A brief search corroborated this. Can someone correct me if I'm wrong? – Andy May 16 '11 at 16:02
  • 1
    Maybe that American was telling you how to pronounce Edinburgh Indiana... or one of those other Edinburghs. – GEdgar Jul 20 '12 at 13:17

2 Answers2

13

Have a look at this highly relevant paper:

Towards an automated classification of Englishes
by Søren Wichmann and Matthias Urba
from the Max Planck Institute for Evolutionary Anthropology (Leipzig)

Have e.g. a look at the tree-like structure on page 4, Fig. 3.1. (unfortunately the referenced paper therein doesn't seem to be available publicly).

Laurel
  • 66,382
vonjd
  • 3,669
  • 3
    Interesting! Yes, they’re doing precisely what Mitch imagines. They start with IPA transcripts of how people in different dialects pronounce the same word. They calculate Levenshtein distances, which measure the amount of “editing” it would take to change dialect A’s pronunciation into dialect B’s. They aggregate these to produce a matrix of distances just as Mitch imagined. However: I wonder why this didn't produce a more compelling graph than the one on page 4. It should be more tree-like. – Jason Orendorff May 16 '11 at 15:33
  • To show what I mean by “It should be more tree-like”: if you take DNA sequences from different species and do the exact same thing, you get a tree that corresponds to what we think we know about those species’ evolutionary history from the fossil record. Or so I’m told. – Jason Orendorff May 16 '11 at 15:36
  • @Jason: yeah...I don't get it, where's the tree? But their description of the method is nice. – Mitch May 16 '11 at 16:17
  • Ah. See some beautiful pictures at the Wikipedia article for phylogenetic tree. It seems this quantitative approach revolutionized evolutionary biology. – Jason Orendorff May 16 '11 at 18:38
  • Why would you expect a tree? People are always moving back and forth between various English-speaking countries, and accents are influenced by this (Do you think New Yorkers and Bostonians started dropping their r's independently of England?) It's only guaranteed to be a tree if the accents develop in isolation. – Peter Shor May 17 '11 at 10:45
  • @Peter: there's reality and there's the model which is a simplification; we hope the model will capture reality well. A distance metric is one model which can then be further manipulated into the model of a tree (by well established methods (which is not to say they are the best)). Which is a long-winded way of saying that if you simplify out things like interactions, like ignoring friction in physics, one way to do it is by a tree. I'd really like a more complete set of dialects (all the ones from the British Isles) and then try to develop a model for them that shows mutual influence. – Mitch May 17 '11 at 14:16
  • 2
    Excellent...thanks for the link to the literature...that paper and its bibliography will start me off pretty well. – Mitch May 19 '11 at 00:18
  • The link has been broken or obsolete for a few years now. The next best thing I found was here: Chapter 54 Towards an automated classification of Englishes – Mari-Lou A Aug 26 '23 at 09:11
5

As an Australian, AmE, CaEn, NzEn, EnEn, ScEn, IrEn, ZaEn all sound quite "different".

  • Tv personalities from America pretty much sound the same regardless of whether they are east or west coast. The same is not true however of the average american in the street.
  • I'm not sure how you can't spot a Scot, especially one with a very heavy accent, as they are quite different.
  • Cockney is also easy to spot, just look out for the rhyming and "nonsense words" they insert.
  • New Zealand English is also quite different from Australian English, theres no confusion in spotting a genuine kiwi. As an example "six" becomes "sex", while "Fish and chips" becomes "Fesh and chups".
James Waldby - jwpat7
  • 66,660
mP01
  • 840
  • 3
    NzE is quite different? Are they easier of harder to understand than, say, a Scot to you? – Mitch Apr 26 '11 at 02:28
  • 2
    Cockneys don't actually use rhyming words, but slang derived from rhyming words. They might say something like "Ruby", which is short for "Ruby Murray" meaning "a curry". However, many rhyming slang words like this are now in wide use across England, so you can't always guarantee that someone using this language is actually a genuine Cockney. – Lunivore May 20 '11 at 16:00