Starting to learn the IPA and having a few questions. One is about the L sound. Here are the L sounds I've seen:

  • l
  • ɬ (No language uses this).
  • ɫ: both velarization and pharyngealization
    • (velarized lateral).
    • lˤ (pharyngealized lateral).
    • ɫ̪

Sidenote: Wondering how the l̥ is pronounced in English. If it is just a whisper with your mouth in the L position.

I don't understand any of them except the first one. I can sort of hear the difference of the dark ɫ, but it doesn't seem like enough to mean a different letter sound, just a different way of saying L.

But the shape of the L means putting the tip of your tongue at the front roof of your mouth, not touching the teeth. Then you voice it, and press air around the the tongue. But if you just left it at that it sounds barely different than a hmm... or nggg.... So you finish it off by whipping your tongue off the roof to say "lah" or "luh". That gives it (to me at least) the "L sound".

My questions are, if there is a way to pronounce the L without that last "finishing it off" part (giving the aspirated "lah" or "luh"). It seems that by itself you can't not aspirate it, but when it is part of a word like hello it doesn't get aspirated and instead is just basically a transition between two vowels. But the "clicking" or "snapping" of the L seems to play a part in the L sound. So wanted to see if there is a name for that snappiness, and the terminology around the part that: you aspirate when saying the letter by itself, but don't aspirate it when it's part of a word like hello.

The second question is, if it is related to the S or T or D sounds. I've seen T and D related as the same mouth shape yet just voiceless vs. voiced. It seems S is the same mouth shape but your tongue never touches the roof of the mouth. Then it seems L and D are almost exactly the same, but it just seems that with the L you are rolling the tongue off the roof, while in the D it is more of an air-pressured snap (an ejective, voiced T perhaps, I'm wondering). So I'm trying to name the L and the S relative to the T. I understand their IPA terminology (voiceless avioloar). But it's like: L = approximant, S = sibilant, T = stop. Also L has the word "lateral" in the IPA description. I guess the difference between approximant and sibilant is that the L doesn't produce turbulence, while the S does. Also, the L is a "lateral consonant", so the air goes to the sides, while the S it goes down the middle. Let me know if there are any other major differences.

The final part of the question is if there are any "equivalent" sounds to the L in different parts of the mouth. Similar to how there is the p and t sounds in different parts that are similar (voiceless, popping sounds). Maybe the ng is one of them. Would be interested to know the others.

  • The place that the tongue tip anchors on is not important -- it can be on the teeth (even the lower teeth!); it just needs to be fixed so that the sound is not R. The 'breakaway' is not part of the L -- you can keep the tongue tip anchored for words ending in L like 'bell'.
    – amI
    Commented Aug 30, 2018 at 22:21
  • 1
    If you're teaching yourself phonetics you need to get J.C. Catford's A Practical Introduction to Phonetics, which is full of little experiments for the autodidact to get the feel of a sound from the inside.
    – jlawler
    Commented Aug 30, 2018 at 22:47
  • 3
    It's wrong that "no language uses ɬ": see phoible.org/parameters/1F0C89A7E99CD5113AA994AD2CC86CBC#1/19/…
    – drammock
    Commented Aug 30, 2018 at 22:54
  • 4
    Why do you say that no language uses /ɬ/? Welsh, Classical Nahuatl, and Biblical Hebrew immediately spring to mind, and I'm sure there are many more.
    – Draconis
    Commented Aug 31, 2018 at 3:38
  • 3
    Also, the reason you say you can't hear the difference between alveolar L and velar L is because they're allophonic in English. They aren't allophonic in every language.
    – Draconis
    Commented Aug 31, 2018 at 3:40

3 Answers 3


It is possible to pronounce various kinds of lateral without any vocalic release (the "aspiration" thing -- aspiration actually means something else in phonetics, though it's similar), though without practice, it may seem more natural to say [lə] rather that just [l̩]. Perhaps if you say "bottle" very slowly and just prolongate the l part, and then leave off everything before the l, you can get an l without a following vowel, but then what you have is closer to [l̩]. The thing that's pretty much impossible to do is just say [l], not [l̩], with no preceding or following vocalic sound. The same does for [m,n]. In general, we can't just pronounce single consonants as they would occur in real speech for example "low" without the "ow". You can listen to recordings of [l, m, g] with the vowel part edited out from "low, mow, go" – this isn't something that can be done in normal speaking. These consonants have a duration of 40-50 msc, which is really short, and it's impractical (maybe impossible) to articulate an isolated consonant that is so short. Thus the vocalic support that you get with "luh" i.e. isolated l, as well as "guh" and so on, is there because you want to make the consonant sound "snappy" like it would be in "lift" etc. when it is in the syllable onset. You could just make a really long 500 msc [l], but then it sounds like syllabic [l̩] (or [m̩, n̩]). The same goes for the stops [p,b] and so on, only more so because being stops, you don't get any sound without that vocalic support.

Unfortunately, I don't understand the second question. In words like "lip, dip, tip", [l, d, t] are very similar in having tongue-tip contact with the alveolar ridge, but they differ in whether they are voiced ([l,d]) or not ([t]); or whether there is complete closure involving the tongue ([t,d]) versus a lowering of the tongue at the sides ([l]).

  1. Pronouncing l without the "finishing off" part: I'm not sure what your native language is, but aspirated approximants are very rare in general, so I doubt your "finishing off" is aspiration. If you are referring to English, the phoneme /l/ is realized as velarized [ɫ] in coda position: so that would be the difference between the /l/ in "hell" [hɛɫ] and the /l/ in "hello" [hɛloʊ]. It's possible to pronounce an unvelarized /l/ in the coda of a syllable. You could try listening to /l/ in other languages if you want to practice it.

  2. The difference between [l] and [s] and [t] is manner of articulation. [t] is a stop, meaning it stops the airflow completely for a moment. [s] is a (sibilant) fricative, so it pushes the air through a narrow space. [l] is an approximant, which is similar to a fricative but leaves more space open.

  3. You seem to be looking for non-alveolar lateral approximants. Wikipedia has a list here. The IPA has symbols for the palatal lateral approximant [ʎ] and the velar lateral approximant [ʟ].

  • I have "dark l" in "hello": some American English accents use dark l in non-foot-initial contexts rather than just in non-syllable-initial contexts. Commented Aug 31, 2018 at 7:30
  • @sumelic The OP says that his /l/ is "aspirated" at the end of a word but not in the word "hello," so I would guess that what I wrote is the case for his accent
    – b a
    Commented Aug 31, 2018 at 11:22
  • I understood Lance to be saying that L is "aspirated" specifically when he tries to pronounce it in isolation, not in all word-final positions. I think it's fairly common for people to produce an epenthetic schwa sound after a consonant when they try to pronounce it "by itself". E.g. someone might describe /d/ as "a duh sound". Commented Aug 31, 2018 at 16:05

This is a partial answer to the second question whether there is a relation between l and the coronal stops (t, d):

Yes there is a connection, and historically sound shifts like d -> l (lambdacism) are attested. They occurred sporadically in Latin, e.g., lingua < dingua "tongue" or lacrima < dacrima "tear" and on a more regular basis in Finno-Ugric languages.

In the Turkic languages, there are lir and shaz languages with mutual sound correspondences l <-> sh and r <-> z.

The l sound also relates to n, in Ancient Egypt all Proto-Afroasiatic l's became n's.

  • The exact explanation of the d/l alternations in Latin seems to be disputed. I've read about a hypothesis that the words that show variation originally had dl- clusters, rather than a sound change of d -> l. Commented Aug 31, 2018 at 16:08
  • And what happened to the "l" in the Germanic cognates (English tongue, German Zunge for lingua, and English tear, German Zähre for lacrima)? Evidence from other branches of Indo-Germanic also speaks against dl- clusters in the beginning of these two words. Commented Aug 31, 2018 at 16:52
  • Pulju ("Indo-European *d, *l and *dl", 1995), citing Hamp (1972), says that the tear words come from original *draḱru, with dissimilation causing the loss of the *r from the initial cluster in Germanic and Celtic, and the change of it to *l in Italic. The resulting *dl- cluster in Italic would then have been simplified to Latin l-, like the simplification of tl- to l-. Pulju says Italic spellings with "d" might either represent an alternative simplification of dl- to [d], or be an imprecise way of spelling a sound like [dˡ]. Commented Aug 31, 2018 at 17:10
  • 1
    The additional r after the d is attested in Old Armenian only, neither Greek nor the Celtic languages show a reflex of it, and I doubt it was still present in Old Latin. But it is a possible hypothesis, indeed. Commented Aug 31, 2018 at 17:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.