How can a speaker tell whether their idiolect has "dark l"?

Question

I believe my variety of English, General Australian, has "dark l", but I'm not sure.

I can't tell if I have it in my own idiolect or not.

It's pretty well accepted (I think) that it's hard to linguistically analyse your own speech.

In this case it could be tricky because in English dark l is an allophone of normal l (so it never contrasts directly as with a minimal pair).

Also I've largely lost my accent because of years of overseas travel and working with overseas travellers when I'm at home - so if I used to have it I'm still unsure whether I have it now.

So are there some methods or tricks a person can use to analyse their own various /l/ sounds, something like minimal pairs perhaps (I don't think there are exact minimal pairs).

How can a person analyse their own speech for presence of the "dark l" sound?

Answer 1

The main difference lies in what part of your tongue enters in contact with your velum (aka soft palate).

• If you try to pronounce "light" [try with a French accent if that doesn't work with your native Australian pronunciation ;-)] you will notice that the tip of your tongue enters in contact with your velum. It leans on the alveolar ridge and touches the upper front teeth. This is conspicuous when the "l" is at the beginning of a word.

• If instead you decompose the pronunciation of words like "feel" or "false" [as heard on the BBC may be], you will notice that your tongue makes itself more compact and touches an inner part of the velum (by which I mean more towards the back of your mouth).
  As you can see from the examples I cited, this actually happens when the "L" is at the end of the syllable (and often the word itself).
  Normally you should also notice, in a subsequent phase, that the tip of your tongue touches the front of your palate closer to the teeth, but since you already mentioned that you tend to finish your "dark Ls" with a "w" sound, it is possible that this part is absent in your pronunciation - which is actually quite common in America for instance. On the other hand Received Pronunciation does have both phases.

If you don't notice a tongue position difference between "light" and "dark" Ls in your way of pronouncing the words mentioned above, you might still notice a synchronisation difference.

• In a "light L", the tongue moves faster towards the front.
• In a "dark L" one clearly feels two different phases - try "false" or "tell" against "tale". Caution though: this is a different phenomenon from the length of the preceding vowel.

As for this more subtle difference, I have to quote "Phonology and Language Use" (p. 87).

Browman and Goldstein (1992, 1995) have discovered a generalization regarding the difference between certain sonorant consonants in syllable-initial and syllable-final position.

In syllable-initial position, the gestures involved in producing a consonant such as [n] or [l] are more synchronous than in syllable-final position, where the gestures are more spread out over time. Indeed, Browman and Goldstein (1992, 1995) report that multiple gestures following a vowel are phased such that the wider gesture – the one with lesser degree of constriction – comes earlier, yielding a structure in which the syllable nucleus has the widest opening and subsequent gestures are organized towards gradual closure. Studies of the relative phasing of syllable-final gestures leads to a unified description of the seemingly unrelated phonetic pattern of ‘velarization’ of [l] in syllable-final position and the nasalization of vowels before tautosyllabic nasal consonants.

American English [l], both syllable-initial and syllable-final (‘dark’) [l], is produced with two gestures: one involving the tongue tip and one involving the tongue dorsum. For [l] before a vowel, the two gestures occur almost simultaneously, while after a vowel, the tongue dorsum gesture begins much earlier than the tongue tip gesture. In fact, for the postvocalic [l], the end of the retraction of the back of the tongue occurs only a little after the beginning of the raising of the tongue tip. The acoustic impression of a velarized or ‘dark’ [l] results from this retraction of the back of the tongue occurring before the tongue tip raises (Sproat and Fujimura 1993).

A few more considerations:

• In Albanian, it should be easier to tell which is which because dark Ls are spelled with a double "l" whereas light ones are spelled with just a single one.
• An emphasised "dark L" is conspicuous in the Arabic pronunciation of "Allah" (الله‎) and since you are currently living in Albania, you should probably be able to feel the "depth" of the "l" in Allah (also noted /lˤ/ in IPA).

Answer 2

Perhaps you could start by studying how you pronounce English /w/, which consists of simultaneous labial and velar approximation. Ignore the labial approximation: you want to learn what velar approximation feels like. Of course it feels like the back of your tongue approximating your velum, but it's better to experience it yourself.

Now that you know what velar approximation feels like, you can try to assess how much velar approximation you use when pronouncing l. Light l has little or no velar approximation; dark l has strong velar approximation.

Answer 3

Try looking in a mirror into your mouth while producing some words like "fill feel hello ladle larry bottle etc". If you see a closure further back (by the dangling thing, your uvula), it is a dark l, while if a closure just behind your teeth, it is a light l. You could even do this with a small pocket mirror in high light -- or a webcam maybe for best results, so you don't have to look at the display while producing the sounds.

Answer 4

Most often when people don't make a dark L, they use their lips to make a substitute sound that is more like a W. This includes native speakers. The question is simple; can you say "simple" and smile while saying the L? Try to say any word that ends with L and pay attention to whether you're using your lips or your tongue to say the L.