How to Lose Your Accent in (American) English

they don't want to do 
it... it threatens their self-definition too much. And a third barrier is that 
they already have notions about how things are pronounced, and they 
rely on their notions rather than on what they actually hear. So at this 
point, slow down to one word at a time, and say it over and over again 
without an accent. Every time you say the word the wrong way, you are 
programming your neurons the wrong way... so say it as often as you 
can the right (all-American, apple pie) way. Don't speed up until it's easy 
to do so without losing your accent. You'll notice that whatever you can't 
do fast, you can't do slowly either. You'll also notice if you pay attention 
that you're often simply mistaken about which vowel is actually used... 
you're thinking it's rounded at the lips like 'o', when it's actually a nice 
wide 'ah', as in 'hot'. 
American Specific: 
All the above advice is equally applicable to learning any language. 
However, in addition to this, there are a few handy tricks they teach you 
in linguistics school about the differences between English 
pronunciation and those of other languages. 
1a. The vowels are usually the hardest for people. English has a lot of 
them compared to most languages, so the differences between them is 
likely to be more subtle than what you're used to. Like the other 
Germanic languages, English has a series of long vowels and another 
series of short vowels. Foreigners usually have the most trouble with the 
distinction between short 'e' (as in bed) and short 'a' (as in bad) and 
between long 'i' as in 'sheet' and short 'i' as in 'fit'. : - ) The long vowels 
are mostly diphthongized in English... this is not usually true of other 
Germanic languages (in case you're German or Swedish). This means 
that many of our long vowels are a combination of two vowels the last 
one is 'y' or 'w'. Our long 'a' as in 'fate' has a generous 'y' at the end. 
* You can draw the vowel in 'feeeeeeyyyyyyyt' out for a long time (2 
seconds) without gaining an *accent*... An accent happens when you 
pronounce the wrong vowels, not when you pronounce them for too 
long. When you pronounce them for too long, you just sound stupid, not 
foreign. And make sure that initial 'e' starts out low enough in the 
mouth... most languages let it get a little higher than we do, because they 
don't have to fit both short 'e' and short 'a' into such a small space. 
* Draw out also 'shuuuuuuuuuuut' (shoot) - that 'u' is not powerfully 
velarized in most dialects of American (unlike most languages), which 
means you don't lift the back of your tongue so high toward the velum... 
pronounce it much farther toward the front of your mouth - even 
between your lips and your front teeth. I don't perceive a shift in color 
for the duration of this vowel (unlike 'fate', it stays the same throughout) 
* 'looooowwwwwd' (load) - again the 'o' is not powerfully velarized. 
Pronounce it in a little space the size of a golf ball behind your front 
teeth. Most people have to work on the American 'w' as well. It's 
pronounced lightly... don't purse your lips. 
* 'laaaaaaawwwwwwd' (loud) - This 'a' isn't velarized either 
1b. Nearly 1/3 the vowels in my dialect of spoken American English are 
'uhhhhhh' as in 'duuuuhhhh' or 'ummmmm'. Practically all the unstressed 
vowels go to 'U'. Before an 'r' it become 'u' like in 'took'. Many dialects 
(New England) also let unstressed vowels go to 'i' in some positions. 
attentive -> UtentUv or utentiv, depending on the dialect. combination -
> kambUneyshUn, German -> JurmUn..., practically -> praektUkliy, 
color -> kUlur, unbelievable -> UnbUliyvUbUl. This is coincidentally 
also the stupidest, dullest and least vibrant of the vowels... speaking 
phonosemantically... in case you're wondering why we are as we are. 
1c. A following 'r' strongly colors our vowels. For me, the vowel in 'for', 
'fir' and 'fur' is all the same - basically the same vowel as in 'book'. 'Four' 
and 'far' each retain their vowels, except in the midwest where 'card' and 
'cord' are pronounced the same and 'horse' rhymes with 'farce'. In my 
(Colorado) dialect 'Mary', 'merry' and 'marry' are all pronounced the 
same, but in New England, they're all different 
1d. In New England and other regions, they have a more rounded 'a', 
which doesn't exist in the West. For me 'cot' and 'caught' are pronounced 
the same. Not so out here in the East. 
2. Now some words about consonants... Linguists distinguish different 
kinds: 
labial (pronounced at lips): b,p,v,f,m,r(initially),w 
dental (pronounced just behind the teeth): th (thing) and th (that) 
alveolar (pronounced in English behind the alveolar ridge): 
d,t,j,ch,z,s,n,r 
palatal (flatten the tongue in the middle of the mouth): sh,l,y 
velar (pronounced at the velum): g,k,h,ng 
stop (fully stops the airflow in the mouth): b,d,g,p,t,k 
fricative (air flows through a narrow opening in the mouth): 
v,th,z,f,th,s,sh,h 
affricate (stop and fricative at the same time): j (d+zh), ch (t+sh) 
nasal (air flows through the nose): m,n,ng 
liquid (historically always mixed up): l,r 
glides (vowels used as consonants): w,y 
voiced (pronounced while using the voice): 
b,d,g,v,th(this),z,j,m,n,ng,r,l,w,y 
unvoiced (pronounced while voice is still): p,t,k,f,th(thing),s,sh,h,ch 
3. What I have called alveolars here are dentals in Slavic and Romance 
and many other language families. That is, they are pronounced in Slavic 
by touching the tongue to the back of the teeth. When you pronounce 
them in English, the tongue does not touch the teeth, but rather hits just 
behind the alveolar ridge (that bump in the roof of your mouth just 
behind the teeth). 
4. The unvoiced stops in American are aspirated at the beginning of a 
syllable. This is not true in Slavic or Romance. So there's a slight puff of 
air after the 'p' in 'pot' and 'prize', the 't' in 'top' and 'train', and the 'c' in 
'cat' and 'climb'. This is not true of the unvoiced stops in 'stop', 'spill', 
'ship', 'skip',... After 's' and at the end of the syllable, the sound remains 
unaspirated. The aspiration in English is much lighter than in the Indian 
languges, at it's limited to unvoiced stops. 
5. That brings us to the 't' at the end of 'fate'. The 't' doesn't even have 
to be fully pronounced. If you want, you can just let the tongue hit the 
top of your mouth... or you don't even have to do that. 't' at the end of a 
word like that can in many dialects even be pronounced with a glottal 
stop way at the back... like the first consonant in Hebrew 'Aleph. The 
glottal stop precedes all words in English that start with a vowel too... 
we just don't perceive it as an actual consonant like the Semites do 
because its distribution in our speech is completely predictable. 
6. When you have two stops in a row in American English, you don't let 
go of the first one before you pronounce the second. In French or 
Norwegian, for example, you pronounce a full 'c/k', then let it go and 
then start the 't' in 'ct'. But in the American word 'act', you don't let go of 
the closure of the 'c' before you start the closure for the 't'. Your tongue 
is actually touching both the velum and the alveolar ridge at the same 
time. 
7. Dentals, labials and palatals in most languages is 
usually velarized more than in American English. Don't lift your tongue 
up in back so much. This makes the sounds 'lighter', higher pitched. 'l' is 
an exception. 'l' for me is quite heavily velarized. You Slavs have a 
distinction between 'hard', velarized consonants and 'soft' palatalized 
consonants. We don't have that distinction. All our consonants are sort 
of in the middle between these two. 
8. In a lot of languages, particularly Slavic, voiced sounds are 
pronounced with stronger voicing than in American. In Slavic, you start 
moving your vocal chords before you even start articulating the 'z' or 'v'. 
In American, you start articulating the 'v' or 'z' before the vocal chords 
start moving. You Russians... have your American friends pronounce the 
Russian word 'Zdravstvyujte' for you. See if their lame 'z' in the 
beginning doesn't sound almost like an 's' to you. 
9. In a great many languages, there can be no voiced consonants at the 
end of the syllable. This is not true in English. 'Log' is different from 
'lock', 'bid' is different from 'bit', 'raze' is different from 'race'. Be 
especially careful of those plurals and 3rd person singular verbal 
endings. Most of them are spelled 's' but actually pronounced 'z'... 
anything after a vowel or another voiced consonant is pronounced 'z', not 
's' - 'has', 'watches', 'bugs', 'news', 'pans', 'initiations'. 
10. The American 'r'. When you guys make fun of our pronunciation, 
you overdo the 'r'. It is pronounced with rounded lips in initial position, 
but not final position. And don't velarize it so much. Our 'r's and our 
vowels vary the most dialectally, so you have to decide what version of 
American you want to speak. 
11. 'th' is pronounced behind the teeth, where most languages 
pronounce 'd' and 't'. It's not achived by sticking your tongue between 
your teeth. 
12. 'd' and 't' between vowels in American are pronounced the same. 
It's what we call a 'flap', in some ways similar to one strike against the 
roof of the mouth of a rolled 'r'. It's a voiced sound. 'Bitter' and 'bidder' 
are pronounced the same. It also happens in speeach between words 
"What did you do?" becomes 'wUDijUdu?' The final 'd' there is not 
pronounced as a flap, because it comes at the beginning of a stressed 
syllable. 
13. Once you've got individual words down cold, start listening to the 
intonation. American rises and falls less dramatically in normal speech 
than most languages... except when we get emotional. 
That's all I can think of for now.