How do you pronounce r?

I don't know how and why. I was taught r very very early at the elementary school, and ... I was only told to "curl".

To be exact, my question is: how is your r?

1) can you describe it? (if possible, do tell me which r you use, alveolar or retroflex approximant or any others...) I really find liquid consonants(?) so difficult to describe.

2) if you are native, where are you from?

I find it hard whether mine is alveolar or retroflex, but probably the latter, since... I treat it almost like sh (http://en.wikipedia.org/wiki/Voiceless_retroflex_fricative), since I learned the basics of Mandarin when I was around 8.

I can't search the site, so...

I'll give my pronunciation of /r/ here, but you probably really should not use it as an actual example for pronouncing English, as it is very highly nonstandard (even though it is phonetically not that strongly apparent to listeners).

For starters, I really have two main types of pronunciation of /r/, which are articulated completely differently. The more common type, which shows up in all positions other than after a coronal, is a uvular approximant or a weak voiced uvular fricative. This is articulated purely with the very back of the tongue and with, at times, changes in phonation (vocal cord vibration), and can be articulated with the rest of the tongue lying very close to the floor of the mouth. It tends to be the most closed prevocalically (especially when stressed or word-initially), tends to be somewhat more open but still may be very closed when stressed when postvocalic, tends to be rather open intervocalically, and tends to be very open syllabically (such that it is quite vowel-like then). When stressed and postvocalic, it may also be accompanied by creaky voicing of varying degrees.

The other type, which only shows up after coronals, is a laminal postalveolar approximant. This is very much like what is found in most English dialects than the former pronunciation of /r/. One key thing about it is that it does not involve any retroflexion at all, and rather is actually very similar to a voiced palato-alveolar fricative [ʒ] pronounced more openly so it is approximant rather than fricative. (Mind you though that when people refer to rhotics in English dialects as being "retroflex", they really are not - what people really mean by "retroflex" is postalveolar, which does not involve any actual curling back of the tip of the tongue.)

Oh, I forgot to mention, that I am a native speaker of English (even though some have confused me for a non-native speaker thereof), and am from Milwaukee, Wisconsin. Of course, many of you probably already know this.

I spent some time on Christmas Island, Kiribati. The local people there roll their "R"s to sound like a d or even a t. Also since the ti is used as S the name of their country sounds like "kid-a-bus". In areas where the r is not rolled it sounds like Ke-re-bus.

>>voiced uvular fricative

why? This sounds more German and French...

>>laminal postalveolar approximant

When you say it's like "zh", then I know... it may be more rounded than the average r (as in "red" in most varieties of English, where this r is often labialized).

Until I read my phonetics textbook carefully (no time for it now)... I can't imagine how _not_ to put it retroflex.

My testimony: here, a lot of folks confuse r with w (as in wh- words, among others), and many of us still often say the Japanese suck in English just like us (in particular, they can't produce r too, they say). A lot of us do know the r somehow, but more are the girls who .... try to speak with a very stilted pseudo-American accent. It almost sounds like they want to put everything as r...

>>>>voiced uvular fricative

why? This sounds more German and French...<<

The reason for such being found here is probably substratum influence here from High and Low German dialects that were brought here by immigrants (and which have since died out). But that said, it actually sounds not too different from the pronunciations of /r/ found in most other NAE dialects unless you listen carefully.

>>>>laminal postalveolar approximant

When you say it's like "zh", then I know... it may be more rounded than the average r (as in "red" in most varieties of English, where this r is often labialized).<<

Actually my [ʒ] is not labialized, and neither is any of my realizations of /r/. This is unlike many English dialects where non-affricate postalveolar consonants ([ʃ], [ʒ], [ɹ̠]) in general *do* tend to be labialized. One often cannot rely on such kinds of phonetic details to be consistent across English dialects.

I think that you are referring to the intrusive r. I think that it applies to every word that ends with the schwa vowel. Non-rhotic English introduces an r at the end of a word if the one after it starts with a vowel. Because in non-rhotic English they replace the "er" at the end of words with the "schwa" this is crept into words that don't end with "er" but with the "schwa" itself. They claim it's something to do with sounding "better" and I have to admit they have a point. I've listened how Americans pronounce "I SAW IT". Now, im my opinion when Americans say "SAW IT" it sound like one word sort of like SOIT. Non-rhotic forms reinforce that they are two separate words by introducing an r so it sounds like "SORE IT".

Travis, why don't you upload a recording to demonstrate your special uvular r? I never heard of such a thing before in English, and I have no idea how can it not sound completely different from normal GA r.

even know im 9 I find it hard to pronounce 'R'.evev trying to say 'RED'

ooh I forgot to say that I'm German Dunka means thanks.I will reply later.

And all this time I thought it was "Danke".

<<I've listened how Americans pronounce "I SAW IT". Now, im my opinion when Americans say "SAW IT" it sound like one word sort of like SOIT. Non-rhotic forms reinforce that they are two separate words by introducing an r so it sounds like "SORE IT". >>

Why? "Sorry" doesn't come off as two words, does it? And all words are run together in actual speech.

<Non-rhotic forms reinforce that they are two separate words by introducing an r so it sounds like "SORE IT". >

This analysis is baseless. The introduction of the intrusive /r/ has nothing to do with reinforcing that they are separate words. If that were the case, it would not occur inside words, such as <DRAWING> which is often pronounced /"drO:rIN/ in non-rhotic accents. The intrusive /r/ is a consequence of making the linking r rule universal regardless of historical <r>.

Travis,

I vaguely remember one of your speech samples that you posted a looooong time ago. A lot of people wrongfully accused you of having a speech impediment, but it was actually because of your accent.

Would you mind posting another speech sample, preferably one where you're reading and where you're speaking extemporaneously.

I am very curious as to how your uvular-fricative R sounds in American English.

thanks!