Listening to Ira Glass the other day, I noticed his 'l', to my ears, sounds exclusively velar with little to no dental component. Here's a clip (he says the word "like" a couple times in rapid succession):


Is this an American regional variant? I hear it more on his pronunciation of light 'l' and though I also hear more nasality in his voice. The American Journalism Review in this article called his voice "adenoidal" and said it has a "slight stutter, not a speech defect, but a verbal tic, a device" but I am not hearing a stutter, nor a tic. I can't tell if this article acknowledges what I'm hearing or is commenting more on the cadence and style of his speech, rather than focusing on the way he pronounces the letter 'l'.

I realize it could be caused by an anatomical or physiological difference, but I couldn't find a technical description that mentioned specifically pronouncing 'l' in this way as an articulation disorder (unlike the heaps of descriptions of the more common "ar" -> "aw" rhoticism problems).

Another American English speaker I've noticed with this speech difference (who also has a public presence) is Ben Shapiro, though to a lesser degree. Here's a clip (I hear it in "tolerance", but less so in the words "mayflower" and "religious"):


  • 1
    @jlawler I've never heard velarized /l/ from native Hebrew speakers -- in fact velarizing /l/ is a common way to make fun of the accents of non-native speakers (e.g. Russians or Americans).
    – TKR
    Feb 17, 2021 at 22:11
  • 1
    @jlawler I think this question is more specifically about dark l’s which are purely velar, with no dental articulation, essentially a debuccalised version. In many southern English dialects, of course, l in syllable coda is commonly debuccalised but also has labial rounding, so it’s realised as [w]; in Polish this development even happened in onset position. In Armenian, dark l was debuccalised with no attending rounding, and it’s now become [ʀ]. I have heard English speakers pronounce l as [ʟ], but only as part of a speech defect, never as a dialectal feature. Feb 18, 2021 at 2:43
  • 2
    @JanusBahsJacquet, I actually don't think it's a dark /l/ phenomenon. The words "like" and "tolerance" are usages of English's light /l/ right? Feb 19, 2021 at 13:57
  • 1
    @jk-ReinstateMonica, I am asking about English, specifically, but more broadly speaking I'm asking about people who natively speak a fronted /l/ language either having trouble or varying their light /l/ to be verlarized. Feb 19, 2021 at 13:57
  • 1
    @speedfranklin In dialects that distinguish dark and light l, yes – but many dialects do not, including Standard American (which has generalised the dark allophone) as well as both Irish and Scottish English (which have mostly generalised the light allophone). Ira Glass is from Baltimore, so you wouldn’t expect him, a priori, to have the distinction. Feb 19, 2021 at 13:59

1 Answer 1


In the third volume of his Accents of English (pp.550-1), Wells notes that, in the southern United States, dark /l/ may be realised as velar [ʟ] rather than velarised alveolar [ɫ], especially in the sequence /əl/.

I've also heard anecdotal reports of dark /l/ being produced as [ʟ] by speakers from the UK but am not aware of any published work corroborating this.

There are though also varieties of English, such as Manchester English, where it has been reported in the literature that velarised [ɫ] is found in all positions.

  • @Mitzli, do you have any links for any of the UK stuff? Also, I'm not hearing it in the speakers dark /l/, only in words like "like" and "tolerance". Feb 19, 2021 at 13:58
  • @speedfranklin Have a look at this review chapter of Manchester English, for example (search for "/l/"). I don't have an explanation for why you might only be hearing it in the words you say though.
    – Miztli
    Feb 19, 2021 at 14:52

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.