Thursday, 20 August 2009


Kyle Steed writes
I'm a college student from Florida and native speaker of American English, and I've got a problem: I can't make an alveolar trill. I've heard tons of methods for acquiring it, like saying "tee-dee-va" or "ladder" over and over again, but I find myself eventually saying "tee-thee-va" and "lather," as if I've got a lazy tongue. I understand that the point of these exercises is to develop that flick of the tongue that usually happens with the d-sound, but I don't recall ever having been able to do that with my tongue. As of now, I'm making sounds that almost sound like I'm on the right track to trilling, but my tongue remains flat and I'm only using one side to vibrate. So what would you
say is wrong here?

It is notoriously difficult to learn to make an alveolar trill. It took me nearly a year to acquire this skill myself, even though I was a highly motivated postgraduate ambitious to become a proper phonetician, professionally required to be able to make all the sounds on the IPA chart. The teacher who finally helped me conquer this hurdle was Marguerite Chapallaz, and she did it by getting me to relax. I was lying supine in the bath at the moment I first succeeded, so it might be worth experimenting with different body orientations.
What it comes down to is that you have to hold the organs of speech in the right place, relax, and then produce an airstream. You don’t actively move anything to make the separate vibrations: it’s all done by aerodynamics. (Physicists will tell you it’s the Bernouilli effect.)
As Ladefoged and Maddieson put it (The sounds of the world’s languages, p. 217),
This is very similar to the vibration of the vocal folds during voicing; in both cases there is no muscular action that controls each single vibration, but a sufficiently narrow aperture must be created and an adequate airflow through the aperture must occur.

One useful hint is to start with trills you can make. Most people can make a voiced bilabial trill, brrr [ʙ], or its voiceless counterpart. Some can make a uvular trill (think Edith Piaf, nɔ̃ , ʀjɛ̃ də ʀjɛ̃…). Use these to get a feel of how trills work.
(If you can do a voiceless uvular trill [ʀ̥], it’s fun to do it while whistling. It makes you sound like a referee’s whistle.)
If you still can’t manage an alveolar trill, it may be a consolation to know that there are tens of thousands of native speakers of Spanish and Italian who can’t make one either, but replace it by something or other that is easier. Yet they manage to function in those languages.
If you can make [r], and want a further challenge, try the retroflex trill of Toda. Only the onset is retroflex; the actual trill is alveolar.
Like spɛkjələtɪv ɡrəmeriən, you could also attempt a nasal-ingressive voiceless velar trill. Or not.
Sorry, Kyle: I haven't got a magic bullet.


  1. I've just realised that I can produce all of the trills, namely /ʙ/ (sort of easy), /r/ (in my native language) and /ʀ/ (I wish I could use it even when pronouncing French).

    Btw, I use Vista and Google Chrome but now a few of your IPA symbols are missing for me...

  2. I can only assume that what I make is not a proper /r/ because I don't even know how I picked it up. It seems easy enough to me (of course, I was the one without /ɬ/).

    I can't whistle while doing doing an /R/, though.

    Speaking which: What is it Roger Whittaker does?

  3. I think Roger is whistling while saying [dldldldldldl] inside his mouth (however one would transcribe it—diddle I guess is the English orthographic version).

    I can do all the trills too. Some of them I learnt in childhood games.

  4. Just a note of encouragement. I grew up in the midlands of England, and in the speech of my youth my only R-like speech sound was a labiodental approximant (sounding to many people more like a W than a R). I found it incredibly hard to learn the alveolar trill: but I eventually succeeded. The two keys are:

    * Keep trying, and

    * Don't listen to anyone who suggests that you may be genetically unable to produce the sound. Unless you have been clinically diagnosed with a condition such as tongue-tiedness, this is complete nonsense.

    If you're interested, the method that finally worked for me is at under the heading '"Vision dream" Method'.

    Good luck!

  5. If you're in for at real challenge try out the glottal trills presented by John Esling here:

    Also it is possible (and not that hard) to make a bilabial, alveolar and uvular trill all at once.

    When my daughter was around a year old she often made sound that sounded like a lateral trill. It might be impossible for older children or adults because their tongue is too massive.

  6. My thanks to Ruben for mentioning that complicated trill--a "bilabial, alveolar, uvular trill all at once". I often made that triple articulation just for fun when I was a small kid. I've tried the sound just now and succeeded in making both voiced and voiceless versions.

    Fortunately, I've found it quite easy to make a glottal trill (made by raising the larynx) but I tend to cough immediately after making it.

    As far as a lateral trill is concerned, I can only make an "ingressive" version of it.

    Another challenge would be a dental trill.

    1. Hello ! Any news on the dental trill? I have looked for a while and didn't find anything on the internet

  7. Boris Blagojević21 August 2009 at 05:39

    I wouldn't have expected the Americans to find the alveolar trill so difficult, since most of them have the flap. But perhaps I'm biased because of my native Croatian, in which intervocalic /r/ with more than one contact would sound unnatural, so the flap and the trill are basically allophones, and are supposed to sound like the same thing to me. It's still probably the best starting point, though.

    As for the uvular trill, It always sounds far to gargle-y when I try to make it. Definitively not the same thing Edith Piaf's singing. To bad speaking a language that has long syllabic alveolar trills doesn't make you able to do a convincing uvular one too.

  8. In the traditional Japanese dialect/accent of downtown Tokyo, the alveolar trill used to be very fashionable mostly among men as an allophone of /r/, used to show toughness and manliness.
    Nowadays, it is in most cases used in aggressive speech (again, as an allophone of /r/) to threaten, intimidate or bulldoze someone. Naturally, therefore, gangsters and violent youngsters are characterised by their frequent use of this sound.
    Since Japanese women rarely produce the alveolar trill (some women never make it in their lives), it seems an extremely difficult sound for them.

  9. The retroflex trill doesn't seem to difficult, but of course I have no way to check whether I'm actually making one. As for the nasal-ingressive voiceless velar trill, unfortunately I have a very bad cold, so I'm producing it all the time :).

    As for pronouncing /r/, although it is mostly about relaxing the tongue, I can pronounce something close to it (i.e. it's more than just a flap) without any airstream at all, it seems.

  10. Oops! The phrase "downtown Tokyo" is not precise. It should be "the traditional Shitamachi area of Tokyo".

  11. We Japanese make a "nasal-ingressive voiceless velar trill" whenever we mimic the grunting of pigs.

    Sorry for my consecutive posting..

  12. Many people can produce a nasal-ingressive voiceless velar trill. It's even so easy they can do it while they're sleeping.

    The sounds their bedfellows make then are much more interesting.

  13. I could only do the Uvular trill until recently. It seems I was the only one in my phonetics class to not be able to produce the alveolar. I can do it now but I have build up to it. I start with a uvular and sort of work forward. It's not so useful for actually using in speech but thankfully, I don't have to.

  14. My party piece is a v- alveolar lateral click trill, but I have the greatest difficulty with bilabial trills, unless I take myself by surprise.

  15. I seem to be developing some slight sense for phonetics.

    One of the most recent podcast from "365 Days of Astronomy" had a speaker with an r that sounded awfully close to /w/.

  16. Boris: The flap is a single contact, whereas the trill is many contacts that must be made by the airstream alone without any conscious movement of the tongue. It's a very different matter.

    Furthermore, Americans are not at all aware (unless they have had phonetic training) that they use lateral flaps. They will tell you in all sincerity that the consonants in "rider" and "writer" are [t] and [d] respectively, since they are indeed /t/ and /d/. Furthermore, many Americans make distinctions in the preceding vowel, which means that the words are still a minimal pair.

  17. I remember hearing an [R] construction spoken by old old men when I was a young young lad in the North of England, or maybe it was just Hovis commercials, e.g. ... see "cRedit" (0.05) "bRass" (0.07) a glottal stop, an uvular trill, something else?

  18. David Marjanović3 October 2009 at 20:38

    In the unlikely event anyone reads this anymore...

    The flap is a single contact, whereas the trill is many contacts

    No, a trill can easily have a single contact. In fact, most trills the world over do. But in spite of the strong tradition of using the same IPA symbol for both, a flap is a deliberate, conscious, muscle-controlled retroflexion of the tongue, a movement. It doesn't even sound terribly similar to any trill – unless, I suppose, the flap is familiar to you but no trill is.

  19. I can do a sub-apical/-laminal click-trill (if there is such a thing). Try curling your tongue so much that it forms a loop, then make a click. The trick is that if you release the click little by little, the vacuum is going to suck your tongue tip back for another click, and so on. The more apical the trill is, the crisper it sounds. I can make it last at most about 2 seconds, and in so doing I kinda sound like a dying dolphin. So have fun!

  20. D. Vancouver. Canada

    I agree with Ruth, above, about moving an uvular trill forward to (eventually) get an alveolar (or retroflexed) trill, and I read a similar description elsewhere last night, and wrote an expansion on the matter. I regard this as the 'last resort' method, for people who have spent years of frustration with doing 'dr...' type practice.



    I'm taking an acoustic linguistics class, and this perennial question arose. I personally could always do an alveolar, though monolingual English, but I've been scouring the internet for tips for those who can't. The uvular-to-palatal-to-alveolar sounds like the best bet for those who can do an uvular trill, or even any kind of raspy, trillish, rollish, German ach-ish, somewhere along the way. I think what happens is, once you can work forward to some kind of 'trill'/vibry-rasp thing at the palate, and can bend back the tongue-tip a bit at the same time, it will pick up vibrations from behind and then the practicer will be able to FEEL, get feed-back, as to when, where, and how, the tip is at its best RESONANCE configuration.

    I've been fiddling with this method for an hour now, and without doing my own trill, just pushing a back-trill forward, it really does seem to transfer quite naturally to a forward alveopalatal trill. (The dental-trill is the sine qua non of trills. I cannot do it; it is found in Marshallese, it's said, and our Hungarian-born TA does it quite easily.)

    If you can't do any back trill, don't worry about it. Just force air out with your tongue-placement point moving forward with a rasping, scrape-a-gob-of-spit-up-and-out sort of racket.

    ['with your tongue-placement point moving forward...'
    By 'point', I don't mean the tongue-tip, of course. I mean the point of contact of the tongue with the mouth: tongue-back touching uvula, tongue-middle to palate, etc. and all places in between. Think of those rounded rubber-stamp things you roll onto the document, rather than bang down once. Only upside down.]

    Do this a few minutes day (so as not to hurt yourself) until you can do a 'trill', or raspy buzzing, at the soft-palate/hard palate junction, and keep it buzzing there continuously (until you run out of breath).

    Doing that, start the tongue-front bent-back-maneouver while palate-buzzing/trilling. Keep adjusting position, buzzing, air-flow/pressure all over the place, till you feel the vibrating getting picked-up at the front of the tongue. The tip doesn't even have to be bent back for starters. If it's starting to vibrate in any configuration, that's good. Keep it going until you can get it bent upward toward the gum ridge while keeping it vibrating. After a few days of practice, the tip will hopefully not only vibrate in some sort of sympathetic partial resonance, but oscillate on its own, retroflexed or bent up somewhere in the alveolar or post-alveolar region; where and how exactly will depend on personal alveolar or postalveolar shape, tongue flexibility.

    Good luck.

  21. I believe his problem comes from a lack of work from the kindergarten teacher he had that didn't employed emphasis in the pronunciation of his students that prolonged all the way through middle and high school. It's not a lazy tongue it was a lazy set of teachers.

  22. > One useful hint is to start with trills you can make. Most people can make a voiced bilabial trill

    I have the opposite problem. Alveolar trills are totally easy, but I can't bilabial for my life. The only way I seem to produce a bilabial is when I coarticulate an alveolar.