Dem and Dose: “Th” in City Accents

The Strand, London

The Strand, London, in the 19th Century

I’ve lived in New York City for thirteen years.  In that time, I’ve learned that living in such an intense urban area has a palpable effect on how one communicates.  You have to talk faster, talk louder, talk more frequently, and communicate with a incredible variety of humankind.

It would be natural to assume, therefore, that certain features are common to all urban accents and dialects.  Now I personally think that’s a bit of a leap (I don’t see a lot of overlap between the Big Apple and Toronto).  However, there is one sound in the English language that, in city after city, seems to undergo some kind of noticeable change:  the th sound in words like mouth and this.

In Standard English, th is pronounced as a voiceless or voiced dental fricative (IPA θ or ð), meaning it is made with the tip of the tongue touching the top row of teeth.  But in many cities, typically very large ones, there is a noticeable tendency for this sound to go a bit off the rails:

–In the accents of New York City, Chicago, and Philadelphia, among many other American cities, this becomes a dentalized ‘d‘ or ‘t‘ sound.  Hence the famous (albeit inaccurate) caricature of New Yorker’s pronouncing 33rd Street as ‘toydy toyd shtreet.’

–In London, voiced th often becomes ‘d‘ at the beginning of a word: this becomes ‘dis.’  Meanwhile voiceless th becomes ‘f;’ mouth therefore is pronounced ‘mouf.’

–In Dublin, th simply becomes plain old ‘t‘ and ‘d:’ ‘ting,’ ‘dis,’ etc.*

Obviously, not all cities have this funny business with th.  But there are enough instances of this happening that I see a slight correlation between urban areas and accents with non-standard th pronunciations.

What accounts for this?  The most obvious explanation is that all of these cities have been subject to quite a bit of immigration.  Since standard English th (θ or ð) exists in few languages, many people who speak English as a second language use alternate pronunciations.  And this may have filtered down into the speech of native English speakers.

An even simpler explanation may be at play, though: th is frankly a cumbersome sound.  I’ve been speaking English for my entire life, and I still occasionally stumble over this consonant.  There is a reason young children have no problem pronouncing ‘m‘ in mama, but tend to say ‘mouf’ when they mean mouth. Linguists will send me hate mail for saying this, but I find something inherently unnatural and awkward about English th.

Now introduce this phoneme to a fast-paced urban environment, where you need to get a lot of information across quickly, and it’s not surprising that many ‘city accents’ skew this fricative a bit.

Since this sound can be easily replaced in many accents of English, then, I wonder what the future is for English th.  As we can see with the urban dialects above, other pronunciations can easily compete with (and beat) little old th. Will this sound survive another hundred years?  Or will non-standard pronunciations make it one more victim of English’s millenia-long process of simplification?

*This is different from the dentalized t and d that you hear for ‘th’ in other Irish accents.  In local Dublin speech, ‘math’ and ‘mat’ are pronounced almost identically.

Share

About Ben

Ben T. Smith launched his dialect fascination while working in theatre. He has worked as an actor, playwright, director, critic and dialect coach. Other passions include linguistics, urban development, philosophy and film.
This entry was posted in English Phonetics and tagged , , , . Bookmark the permalink.

52 Responses to Dem and Dose: “Th” in City Accents

  1. MT says:

    See also th-debuccalisation, found across the Central Belt in Scotland according to Wiki, although I’d associate pretty much solely wi Glaswegians. By this phonetic rule, we get word-initial /θ/ -> /h/, and /ð/ -> nothing at all, such that ‘thing’, /θɨŋ/ -> ‘hing’, /hɨŋ/ and ‘this’, /ðɨs/ -> /ɨs/, although you do sometimes hear /ɾɨs/, which I suppose is closer to ‘dis’ as found in lots of other English lects.

    • trawicks says:

      Thanks for mentioning that, MT! I don’t know much about how Glaswegian differs from other Scottish accents, but that’s another excellent example.

    • 'enry 'iggins says:

      That feature is also found in Northern Ireland, which isn’t too surprising, considering its history.

      • trawicks says:

        Not to mention (I can’t recall if it was you or another commenter who brought this up) the ‘l’ realization you find in Derry, NI. Hence “father” becomes, basically, “faller.”

        • 'enry 'iggins says:

          Not to mention the elision of intervocalic /ð/ in Northern Ireland (“faer” for “father”, “norern” for “northern”). The “th” sounds are very interesting there. Anecdotally, I’ve also heard th-fronting in the speech of more than one young (from about 16-20 years old) Northern Irish male believe it or not. I don’t know if this is an ongoing change or something a few of them do when they’re young and “grow out of” later in life.

        • trawicks says:

          I also think I’ve heard a dentalized ‘th’ or two in Northern Irish speech, despite the general thinking that this feature is entirely a Southern Irish thing.

        • TT says:

          “Anecdotally, I’ve also heard th-fronting in the speech of more than one young (from about 16-20 years old) Northern Irish male believe it or not.”

          I know this discussion is very old now, but apparently you’re not the only one who has heard this. The linguist Karen P. Corrigan wrote about it in her 2010 book Irish English: Northern Ireland.

        • TT says:

          I forgot to mention that NIE = Northern Irish English and US = Ulster Scots (not the United States) in that book.

  2. Sarang says:

    Are Dublin and Glasgow really high-immigration cities?

    • trawicks says:

      I can’t speak for Glasgow, but Dublin has been influenced by more immigration than I think people give it credit for. Let’s not forget that the city’s most famous literary character, Leopold Bloom, is of Hungarian-Jewish descent!

      Not to mention many of the internal “immigrants” who flooded into Dublin in the 18th-Century WERE foreign language speakers–many, if not most, of them would have spoken Irish Gaelic. I think this had a pronounced impact on the local accent (for example, very strong Dublin accents use an the Irish “velarized r” that is more common in Western Ireland).

    • MT says:

      Glasgow’s high-immigration by Scottish standards (with the biggest immigrant community being Irish folk, funnily enough). Compared to other places in Britain (e.g., London, Bristol, Brum), though, not particularly – and compared to the US or Canada or Australia, Glasgow is very homogeneous in terms of ethnic group demographics.

      • MT says:

        Those aren’t definite numbers, by the way, just vague recollections of figures that I can’t bring to hand at the minute. I would be surprised to learn, though, that Glasgow was immigrantier than any of the English cities I mentioned. Which reminds me, I also forgot a bigger immigrant group in Scotland than Irish folk – English folk!

      • trawicks says:

        Both Irish and Scottish cities, of course, have seen this trend reversed in the past few decades. Irish cities like Dublin, Limerick and Belfast are surprisingly multi-ethnic these days, even if they still lag cities like London or Birmingham.

  3. Hannah Lynes says:

    This is also found in Liverpool with the infamous quote “dee do dough don’t dee dough” (they do though don’t they though). Although it’s also heard throughout Lancashire with people in my hometown saying “shut de dowa” (shut the door). So it’s interesting to know whether Liverpool picked that up from the Irish migrating into the city along with other nationalities or if it’s something that’s stayed around from it’s days as a Lancashire accent?

    • Hannah Lynes says:

      *but Iesu my grammar is poor this evening. Apologies, it’s been a long day!

    • trawicks says:

      I’m guessing the Liverpool dentalized ‘th’ is related to Irish immigration. It notably has a few other consonant features that tie it back to the Emerald Isle as well, such as what we call ‘t lenition,’ meaning the letter ‘t’ tends to soften and become almost like an ‘s’ sound. Although the Lancashire explanation is also a possibility.

  4. 'enry 'iggins says:

    I’ve heard from native speakers of NYC English and others (Wells of course) that the ‘d‘ and ‘t‘ sounds that they use there for th are articulated in a weaker and less deliberate way than a regular [d] or [t]. This guy mentions that in his video. I’m not sure how one would narrowly transcribe that in the IPA.

    Oh and I forgot to mention that you’re not the only one who has heard a dentalized ‘th’ from some Northern Irish speakers. Some linguists whose work I’ve read say that some Nationalist politicians in Northern Ireland have been found to use some “Southern Irish features”, like dentalized ‘th’ sounds, in their political speeches. Also Northern Irish people from relatively close to the border with the Republic of Ireland sometimes have that feature. This young lady from a place called Lissummon in a District Council called Newry and Mourne uses a dentalized ‘t‘ in three (Sorry, I get a bit “link-happy” sometimes). This was an interesting topic today as usual.

    • trawicks says:

      “Some linguists whose work I’ve read say that some Nationalist politicians in Northern Ireland have been found to use some “Southern Irish features”, like dentalized ‘th’ sounds, in their political speeches.”

      Good example of this is Gerry Adams. You can hear quite a few of these pronunciations in this video.

    • dw says:

      the ‘d‘ and ‘t‘ sounds that they use there for th are articulated in a weaker and less deliberate way than a regular [d] or [t].

      Affricates? [tθ] and [dð]?

  5. th says:

    It’s probably noteworthy that most other Germanic languages (German, Danish, Swedish, Norwegian) have replaced the old Germanic th sound by d a long time ago. English, Icelandic and Faroese are the only remaining Germanic languages that still feature this sound. For example: (English) that – (Dutch) dat, (German) das, (Swedish) det … (English) thistle – (German) Distel, (English) thorn – (German) Dorn. This change has happend in German about 1000 years ago.
    Thinking about the fact, that all remaining Germanic languages that still have the th sound are in fact those spoken on islands – British Isles, Iceland, Faroe Islands – contact with constantly changing languages might be an explanation why German, spoken in central Europe, abandoned the th sound first, followed by the other continental Germanic languages and now finally by major English speaking cities. If that’s really a trend, then the future of the th sound in English seems to be obvious.

    • dw says:

      @th:

      English is (along with some dialects of Dutch) the only major Germanic language to retain /w/. In German and the Scandinavian languages it’s lost its velar component to become /v/ ~ /ʋ/. By your logic it seems that English /w/ is doomed to follow the same route, yet this doesn’t seem likely to me.

  6. Erica Walch says:

    It’s interesting that the /s/ and /z/ for voiceless and voiced “th” haven’t taken hold, though, in areas with high immigration rates of speakers who make that (or zat) substitution.

    The French who migrated to Canada and French and French-Canadians who migrated to the Northeastern US in the past and the Russians and former-Soviet country people who are continuing to migrate to the Northeast in huge numbers are all /s/ and /z/ substituters. One might think that in Montreal and Quebec /s/ and /z/ ought to be doing what /t/ and /d/ are doing in NYC.

    Any thoughts on why the /t/ and /d/ have dominated, if the substitution or shift is indeed influenced by migration?

  7. trawicks says:

    @Erica,

    That’s a good question, one which I wish I had a clear answer to! I’m not an expert in the specifics of how sounds shifts work, but my layman’s intuition about ‘s’ and ‘z’ is that they would cause far too much trouble to be adopted as an alternate ‘th’ pronunciation by a native English accent. ‘S’ and ‘z’ have much more morphological importance, since we use them constantly to indicate plurality and conjugate verbs. I think you would end up with countless confusions: e.g. did he say “runs away” or “runs the way?”

    I also think ‘t’ and ‘d’ have dominated because most people used dentalized ‘t’ and ‘d’ at least occasionally as allophones.

    @th,

    That’s very true. English ‘th’ may have been reinforced by the Brithonic languages, even as most of its Germanic cousins dropped the sound. Now that the world is a bit more tolerant of regionalisms, this sound may go the way of the dodo.

    • Erica Walch says:

      Well we use final /d/ and /t/ as past tense and adjective markers, too. “I bade the cat” for “I bathed the cat” would be equally confusing. I think it’s easier for my lazy tongue to make /s/ and /z/ than /t/ and /d/, though /t/ and /d/ are certainly easier than either of the “th” sounds.

      Something else that interests me about this sound is that my ESL students who pronounce “th” as /s/ or /z/ will often (regardless of voicing) write the “th” as the letter “t” without the “h” (but will never write a word with a “d” for the voiced version). For example, a Russian speaker might say “I sought zis was zere” and write “I tought this was there”. Funky, eh?

      • trawicks says:

        Haha of course I just had one of those “duh” moment: reading your comment, I realized that the more simpler explanation for why ‘t’ and ‘d’ take hold in native accents rather than ‘s’ or ‘z’ is perhaps just orthography. Since these words are written as ‘th,’ this spelling would probably reinforce the ‘t’ sound in literate speakers.

        Interesting about the way your students spell “th” words. The only explanation that I might posit is that cyrillic “t” is the same as English “t,” where as the “d” in each alphabet is different. But that doesn’t explain why they atrribute “th” exclusively to a sound which they pronounce ‘s’ and ‘z!’

    • dw says:

      @trawicks:
      English ‘th’ may have been reinforced by the Brithonic languages, even as most of its Germanic cousins dropped the sound. Now that the world is a bit more tolerant of regionalisms, this sound may go the way of the dodo.

      Greek has had [θ] and [ð] for at least 1500 years, and those sounds aren’t going anywhere, are they?

      Hindi and other Indo-Aryan languages have had the even more typologically marked voiced aspirates/breathy voiced stops for over 4000 years (assuming the predominant theories about Proto-Indo-European are correct), and they are still going strong.

      • trawicks says:

        Sorry, actually meant this sound may go the way of the dodo in English.

        • dw says:

          Sorry, actually meant this sound may go the way of the dodo in English.

          Yes, I understand that. But your argument seemed to rest on some kind of inevitable loss of the dental fricatives because they are weird/unusual sounds. I was trying to point out that other languages have retained weird/unusual sounds for millennia, and that there’s nothing inevitable about their loss in English. See also my reply to th above.

        • trawicks says:

          I think languages tend to become more simplified as they grow in number of speakers and become less heterogeneous. “Tend” is the operative word here, because there are any number of counterexamples. But in terms of this simplification process, I find that the more uncommon features tend to be the ones to go first: Spanish /θ/, Portuguese’s constant usage of /ʃ/, Latin’s endless inflections, pitch accent in many languages, etc. I wrote about English ‘th’ in the first place precisely because it is one of the notable exceptions to this “rule.”

        • dw says:

          @trawicks:

          I think languages tend to become more simplified as they grow in number of speakers

          Do you have a basis for saying this? I’m just asking.

          and become less heterogeneous.

          Why should a language “become less heterogeneous” as it grows in number of speakers?

        • dw says:

          I find that the more uncommon features tend to be the ones to go first

          The typologically most uncommon feature of English phonetics is probably its stressed central vowels (in NURSE words). Do you think that this is going to be lost soon?

        • trawicks says:

          I admit ‘simplification’ is probably not the right word there; but I think it’s reasonable to say that when you introduce a language to a large population of foreign language speakers who adopt this language either by choice or through force (as most European languages were spread), some of the “quirkier” aspects of that language are going to fall by the wayside. Not always, of course. The rather exotic phoneme /β/ is heard throughout Latin America, for example, so it seems the Amerindians in that part of the world had no trouble picking it up.

          But I don’t want to come off as generalizing, here: there are an infinite number of factors that go into language spread. If America decided to invade China tomorrow and force everybody there to speak English, you would have a very different English dialect within a few decades than if, say, we were to invade Denmark and do the same thing! But if you were to invade the entire WORLD, you would probably find local dialect after local dialect dropping the same uncommon features again and again.

          As to the NURSE vowel, I don’t know that I would classify that as an unusual feature. Stressed schwa (or something quite similar) can also be found in Albanian, Standard Mandarin, Vietnamese, and Korean. Although I don’t know much about any of those languages, so I’m probably missing some of the phonotactic nuances.

        • LanguageNerd321 says:

          I didn’t realize that /θ/ and /ʃ/ were disappearing from Spanish and Portuguese, respectively. I thought they were still going strong.

        • trawicks says:

          /θ/ is still going strong in Castillian Spanish, but it never really planted itself in the New World (partially because many of the earliest Spanish settlers in the Americas were probably from regions without this sound, and partially because it would be hard for many Amerindians to pick up). It’s almost a certainty that non-/θ/ Spanish speakers outnumber those with /θ/ these days if you’re looking at speakers of Castillian worldwide. /ʃ/ is still used in all Portuguese dialects, BUT the very unusual tendency for it to replace /s/ post-vocallically is much less common in Brazil.

        • dw says:

          @trawicks:

          Sorry to be such a devil’s advocate on this thread. But your first assertion, that

          I find that the more uncommon features tend to be the ones to go first: Spanish /θ/

          seems to be contradicted by your second (which I think is closer to the truth), that

          /θ/ is still going strong in Castillian Spanish, but it never really planted itself in the New World

        • trawicks says:

          No problem! I’m always game for a bit of devil’s advocacy 😉

          My point about /θ/, though, is that it has found limited “success” outside of Castile and Northern Spain. Shouldn’t have used “Castillian Spanish” there, since that technically refers to the language spoken in Latin American (although I’m pretty sure I’d get into fisticuffs if I told a Mexican or Peruvian that they “speak Castillian!”)

        • LanguageNerd321 says:

          Yeah, but some Brazilian accents have /s/ as [ʃ] post-vocalically, perhaps most notably the Carioca (Rio de Janeiro) and some Nordestino accents. I’ve heard it elsewhere too though. That means there are probably millions of Brazilians who have this feature, considering that there are 190,732,694 people in Brazil. I know you never wrote that no Brazilian accents had that feature, but my point is, if it’s so exotic and/or difficult, then why did any Brazilians pick up that feature from the Portuguese a while ago and why do many still have it in their speech today?

        • trawicks says:

          Your’e right: I looked it up, and it actually looks like post-vocalic /sh/ developed later in Portugal, so that was a bad example. Anyway, seeing as I’m REALLY no expert in the phonology of Spanish and Portuguese outside of the most broad of differences in dialects, I’m going to go ahead and keep mum about them.

        • LanguageNerd321 says:

          Well, it’s possible that that guy is just wrong of course.

        • dw says:

          As to the NURSE vowel, I don’t know that I would classify that as an unusual feature. Stressed schwa (or something quite similar) can also be found in Albanian, Standard Mandarin, Vietnamese, and Korean.

          Not quite the same thing, but:

          “[R-colored] vowels are very unusual, and occur in less than one percent of the world’s languages”

          Ladefoged and Maddieson, The Sounds of the World’s Languages, p. 313 (Google Books link)

          PS: We really need more levels of comments 🙂

  8. Hannah Lynes says:

    @ Trawicks: lenition in Liverpool is something I’m currently researching along with lenition in North Wales although /d/ isn’t a focus. Lenited /t/ has its roots in Ireland but it did get me thinking last night (as you do) that maybe it has been left over from Lancashire. Maybe the Irish reinforced an already existing feature?

    It’s certainly a very interesting post though, has got my brain cells whirring 🙂

  9. Pingback: /th/ variation – noffing new | David Wright

  10. dan says:

    It’s interesting to see that the th – d shift is apparent in text messaging too among many urban teenagers. There’s a lot more dis, dat and dose in the texts of my south London students than those of other groups whose data I’ve seen. Might have something to do with ethnicity (they’re mostly African-Caribbean and African/British African) as you would have thought the older cockney alternative (wiv instead of with, vis instead of this) would appear in some London pronunciations and cross over into text speak but it really doesn’t seem to.

    Your initial point about similar features cropping up in different urban areas is interesting too, because the work done by Paul Kerswill and Sue Fox on Multicultural London English seems to suggest that many of these patterns are actually occurring in the speech of teenagers in urban areas like Birmingham, Nottingham and elsewhere too.

    • trawicks says:

      One thing I actually forgot to mention was that in the US, African American Vernacular English follows a similar pattern to Cockney. Similar to multi-cultural British English, there seems to be some influence from older creoles that accounts for this. Another example of alternate ‘th’ pronunciation filtering down into a native English accent.

  11. Pingback: The thorny thicket of "th-" sounds - G7Finance.com - Special Finance Resources

  12. Ellen K. says:

    While [θ] doesn’t exist (that I know of) in Spanish in the Americas, [ð] does, as an allophone of /d/.

  13. 'enry 'iggins says:

    I believe it’s actually [ð̞], i.e., the approximant version of [ð].

  14. To tell the truth this was a fantastic advanced post however as with all wonderful writers there are some items that may be labored on. However by no means the particular much less it absolutely was interesting.

  15. Kyle Thomas says:

    I live in Swansea, Wales. I’ve noticed I and a lot of other people around here tend to just completely drop ‘th’. Them become ’em there becomes ‘er ( pronounced ‘eh’ as the accent is non-rhotic) etc. However it’s not done all they time, some words the ‘th’ gets droped but with others it’s not, the stay the.