Americans like myself ‘tap’ the ‘t’ in between vowels. This means that the ‘t’ in ‘butter’ is pronounced not with a /t/ sound, but rather with the ‘r’ consonant in Spanish ‘pero.’ The common impression, though, is that ‘t’ becomes ‘d’ in American accents: ‘bitter’ sounds like ‘bidder,’ ‘bit off’ like ‘bid off,’ and ‘cutter’ like ‘cudder.’
Yet Americans have no exclusive claim to this feature. London is known for its glottal ‘t’ (to put that very crudely, the ‘t’ in ‘butter’ becomes something of a ‘grunt’). Yet it’s clear that t-tapping is also common in the city, as J.C. Wells makes clear in his Accents of English:
Indeed, there is another variant which also has a strong claim to be considered ‘typically Cockney,’ namely the voiced tap … The use of [ɾ] appears to be connected to the rate at which the person is speaking, since [ɾ] does not occur in slow speech, in hesitation, or before pause.
This strikes me as true of several British accents*. To cite one example: in a transcription of an Estuary English (London) speaker in the International Dialects of English Archive, the transcriptionist notes the use of a tapped or voiced ‘t’ in the phrase ‘put on.’
And in Received Pronunciation, I’ve observed that it’s possible for the consonant to be tapped at the end of a word in rapid speech if the next word starts with a vowel (for example, in the phrase ‘A bit of luck’). Wells’ take on this (at least circa 1982) is that it is a matter of frequency: ‘occasional’ t-voicing is acceptable in RP; ‘frequent’ t-voicing would place an accent in the ‘Near-RP’ category.
So what makes American accents unique? I’ll answer that by making a comparison. In London English, the tapped t is only one of a number of possibilities. The phrase ‘bit off’ might be pronounced (for the non-phonetically inclined, forgive the jargon):
[bɪʔ ɒf] (t is a glottal stop)
[bɪʔt ɒf] (t is a glottally reinforced /t/)
[bɪts ɒf] (t is affricated — it becomes a /ts/ sound)
[bɪɾ ɒf] (t is tapped)
[bɪt ɒf] (t is simply a /t/ sound)
… and probably a number of others. Yet of all the times I, as an American, will utter the words ‘bit off’ in my lifetime, I will most likely have used a tapped /t/ every single time. It doesn’t matter if I’m saying the words slowly, quickly, with food in my mouth, or drunk, the tapped /t/ is almost compulsory in my accent.
So in the varieties of British English we’ve discussed, t-tapping is one of several options. For many Americans, however, the feature has near-exclusivity in many contexts, which is why it is so salient to people with other accents. The question is not so much why Americans do this, but rather, why we’re so consistent about it.
*There are many others, obviously, which I’m not mentioning here. I’m focusing on accents that aren’t associated with t-tapping/voicing.
It’s true that generally in american accents, word final prevocalic /t/ is not usually a place that conditions glottal stops like in ‘bit of”, unlike preconsonantal environments like “forget many”
But then what of the studies that show young-western-women to be leading the states in increase use of glottal stops in such environments? ie. [bɪʔ ɒf] rather than [bɪɾ ɒf]
If the increase in use continues to rise and spread, perhaps we might take a [baɪʔ ɒf] the perception that all americans turn their ‘t’s to d’s’ xD
Made a slight edit to that last paragraph, because I realize it suggests that all Americans use the alveolar tap this frequently. In fact, there are some accents where this is less the case: Western New England increasingly seems to use the glottal stop, and I would be unsurprised if several other American accents do as well. Generally speaking, I’ve noticed that various types of ‘t’ lenition, rather than tapping/flapping/voicing, are arguably a bit more common.
this t-tapping feature is prevalent in Canadian English as well, but I think Americans use it more, for example i’ve heard phrases like ‘what are we doing d’night’ spoken by Americans, but Canadians would pronounce the t in tonight, but would not pronounce the t’s in butter, is it because ‘tonight’ is at the end of a sentence that this occurs or part of a sentence? Would it be pronounced with a t if spoken as a single word? I heard Brad Pitt who I assume is Gen-Am, pronounce romantic as ‘romanic’ in the movie Moneyball, again in Canada this would not be normal, the t would be pronounced.
Any ideas why this feature would apply in Canadian English on butter, later, water but not tonight or tommorow (for example)? Or why does it in American accents?
I know ‘romantic’ -> ‘romanic’ has to do with intervocalic alveolar flapping [winner-winter merger]
Ive got nothing on ‘tonight’ tho
but do you know what i mean?…its hard to explain
I do, though in my accent, the /t/ in ‘tonight’ and ‘tomorrow’ both start with an aspirated /t/, [tʰ], and what I would probably hear as “d’night” or “d’morrow”sounds fine if i say it with an unaspirated [t] [especially if said rapidly], but using a flap sounds really odd.
Really? I’m American and I think can use a flap in tonight in that sentence. Maybe it’s an unaspirated [t] though.
My perception is that Canadian accents feature /t/ in some contexts where Americans would have a flap (although not consistently). Total speculation, of course, but I’d hazard to guess that if you compared a collection of American pronunciations of ‘bit off’ and a collection of Canadian pronunciations of ‘bit off,’ you might see a slightly higher rate of /t/ in Canadian English. It would probably still be a very small percentage.
Why it seems a tad more widespread here in the States, I’m not sure. We certainly understand that the underlying phoneme is /t/!
Could it be “what are we doing d’night” has voicing, but not flapping?
as do we…honestly its something ive noticed more because i’m listening for it, you are right its only a few words that I notice a difference, otherwise its identical
But do any accents outside North America have a merger of /t/ and /d/ in the relevant places? I know that not all t-flappers do have the merger, but many must do. Google “suddle” as a misspelling of “subtle”, for example.
Wells mentions that Cockneys, when emphasizing words/phrases where the tap might occur, may use [d] (he cites the example of ‘hospital’ as [ɔspɪdoʊ]). That would lead me to believe that neutralization between the two is at least possible. (Complicated, of course, by the fact that [ɾ ] is a less predictable allophone of /t/ in Cockney than it is in American English.)
Are intervocalic /t/ and /d/ really merged for most Americans after a stressed vowel? We use the same flap for both of them, but in American English stressed vowels are lengthened before a voiced consonant, and unless I’m fooling myself, I use this feature to distinguish /t/ and /d/ (not that it always works that well, but they remain distinct phonemes).
Definitely not merged for me. In addition to possible help from a length difference, in many cases, from hearing the same morpheme in other contexts we learn if it’s a d or a t. I think that applies in a large percentage of cases. Certainly a large enough number of cases that I can’t imagine anyone having a true full merger. For example, the phonemic difference between bit and bid doesn’t disappear just because a vowel follows, even if we don’t hear that difference.
American Betty does not sound like Scottish berry at all! It sounds like beddie (a little bed)
Take a look at this OED entry, it’s a [d], not an alveolar r:
Protozoa, n.
http://www.oed.com/view/Entry/153334?redirectedFrom=protozoa#eid
Pronunciation: Brit. /ˌprəʊtə(ʊ)ˈzəʊə/ , U.S. /ˌproʊdəˈzoʊə/
That’s just the convention used (misguidedly in my view) by the OED. It doesn’t say anything about the phonetic realizations used by Americans.
It’s OT, but there’s a great sighting of the “Mary, marry merry” merger, in this headline:
http://fivethirtyeight.blogs.nytimes.com/2012/01/05/perry-perry-quite-contrary/
1. T-tapping
a flap or tap sound produced by flicking (tapping) the tip of the tongue against the roof of the mouth – thus making only very brief and rapid contact
when occurs between vowels in words such as getting and better and across word boundaries in phrases such as get up, sort of and it is
got about ten people lined up like sardines on the floor of this little studio flat
Used in:
Birmingham
Banbury
SED Hackney
London RP
Harrow School
Cardiff
Bangor
2. T-voicing
a pronounced almost identically to a sound
when occurs before in words such as bottle or between vowels in words such as getting and better and across word boundaries in phrases such as get up, sort of and it is
got about ten people lined up like sardines on the floor of this little studio flat
Used in:
Melksham
Portesham
London RP
East Harting
Milland
Ballymoney
Londonderry
Belfast
Downpatrick
http://www.bl.uk/learning/langlit/sounds/case-studies/received-pronunciation/connected-speech/
”Because not everyone follows Ladefoged in distinguishing between flap (ballistic movement in one direction, striking a passive articulator on the way past) and tap (back-and-forth gesture bouncing back off a passive articulator). For those who distinguish the two terms, as I do, the AmE voiced t is a tap in most positions, not a flap. But the important thing is surely its voicing, not its precise manner of articulation. ”
by: John Wells
So, we could use [d] for a tap/flap T in words like bitter, thirty, writing…
since ”the important thing is surely its voicing, not its precise manner of articulation”.
No we can’t. [d] is defined to be a voiced alveolar stop, not a flap or tap. The IPA symbol for an alveolar flap or tap is [ɾ].
John Christopher Wells in ”Accents of English: An introduction”
(3.3.4 Tapping and T Voicing)
says
” …the medial consonant in GenAm atom is a tap, and not a flap.
It remains true, however, that it is not identical with the [ɾ] allophone of Spanish /ɾ/, which has a somewhat different configuration of the front of the tongue (see X-ray tracings in Monnot & Freeman 1972).”
So, you are advocating a very narrow IPA phonetic transcription, but the one which is not correct at all. As a speaker of Italian and Portuguese, I can hear a clear difference between an American voiced/tap T (almost the same as the Italian or Brazilian Portuguese intervocalic [d]) and an (intervocalic) [ɾ]. Many Americans can speak Portuguese, they never hear (Brazilian) Portuguese CADA ”every” and CARA ”the face” or ”a dude” the same.
[d] – [ɾ] merger in American English is a merger in perception, but not the one in pronunciation/production (I can hear the difference). But at least 50% of Americans don’t have this merger in perception, they would never hear the Italian verb ADORARE as [ aɾoɾaɾe].
So, please find another symbol for the tapped/voiced T in American English and don’t spoil our [ɾ] which has nothing to do with your phone. (J C Wells does not use [ɾ] in his pronunciation dictionary).
[ɾ] is apical, with a stronger contact with the alveolar ridge, and there is one single vibration. [ɾ] in Italian, Spanish, Portuguese is an apical vibrant.
American tap/voiced T is a laminal consonant, with a larger portion of the
tongue involved in the production of the sound, although the contact be more incomplete. There is never any vibration at all.
“American tap/voiced T is a laminal consonant, with a larger portion of the
tongue involved in the production of the sound, although the contact be more incomplete.”
I’m American and my voiced is definitely not laminal; it’s apical. I find it nearly impossible to make a laminal voiced tap.
That should say “my voiced T is definitely not laminal” sorry.
@Sooryan:
[P]lease find another symbol for the tapped/voiced T in American English and don’t spoil our [ɾ] which has nothing to do with your phone.
Languages do not “own” IPA symbols. Just because [ɾ] is appropriate for sounds of Spanish or Portuguese, that does not mean that it may not be used for other, somewhat different, sounds in other languages. Each basic IPA symbol covers a range of phones, not a unique phone. Where further distinction within the sounds covered by a basic symbol is required, we use IPA diacritic symbols.
By the way, for what little it’s worth, your assumption that I speak American English is incorrect.
[d] – [ɾ] merger in American English is a merger in perception
I’m not sure exactly what you’re trying to say here, but a “merger” usually refers to phonemes, not phones. In this case, the merger that some American speakers have is of /t/ and /d/ (note slashes not square brackets), which may both be realized phonetically as [ɾ].
I’ve noticed this too. Spanish the r is Spanish “para” sounds different than the “tt” in English “butter”.
However, they are similar. I remember once students at my university were campaigning for immigrant rights. The chant “Sí, se puede” came out as “puere”. Nevermind that it should be pueðe.
The doubled “T” in American “butter”, perhaps, but generally not English.
I don’t think anyone can properly recite “Betty Botter bought some butter” all the way through without T-flapping, no matter the dialect(including RP) unless they(singular they) do so with great and fairly slow exaggeration. I used to read it to my daughter at least once a week, and the fun of it wa to recite it as fast as possible without making a mistake – a variant on “Peter Piper picked some peppers” and other tongue-twisters.
I’m sorry, but your claim is absolute nonsense, the like of which could only be made by a T-flapper.
For someone who maintains a voiceless /t/ in all positions (such as me), “Betty Botter” is no more difficult than “Peter Piper”. Or do you claim that “Peter Piper” is similarly impossible without voicing /p/s?
Whether or not I’m mistaken, your anonymous incivility is totally uncalled for.
Apologies: I think I must have got out of the wrong side of bed this morning, and the accent chauvinism that I perceived in the original post set me off.
Pingback: This Week’s Language Blog Roundup | Wordnik ~ all the words