Is the phenomenon of liaison developed by dark L in British English in some areas?

Question

There has been discussion about the dark L being heard as a vowel by L2 learners, though this view is often denied and corrected by L1 speakers, who point out that the dark L is indeed a consonant rather than a vowel. However, a study shows that the vocalization process is indeed completed in some areas in England (see page 13 on https://core.ac.uk/download/pdf/4187965.pdf).

Any why, in the circumstance of a vowel following the dark L, is the dark L vocalized as well (in the u-areas on the map)? For example, is "all over" pronounced as "o over" or "o-l'over". If the former, is there any other mechanism that indicates the existence of two o's rather than one, for example by a glottal stop? If the latter, can I say that English in these areas has developed liaison as in French?

In French, liaison occurs as well after a pause caused by hesitation or thinking while speaking, e.g. "les... [1 sec. pause] s'enfants" where they keep the s after the pause. Does the after-pause-liaison also happen in a sequence like "a personal... [1 sec. pause] and..." in some British English accent?

In British English, the r in the end of "worker" is not pronounced, while in the sequence "worker of", it is pronounced. So how about if there is a short pause between the two words? Is it become "worker... [pause] r'of"?

Answer 1

Liaison phenomena vary a lot in English, even from context to context. A 2003 study which named this particular phenomenon /l/-sandhi said that:

Coda (re)syllabification of /l/ is not subtle or flexible enough to condition the distribution of vocalisation. Prosodic, segmental and phrasal factors are all required.

Gimson's classic tome on pronunciation says under "Liaison" that:

It is unusual for a word-final consonant to be carried over as initial in a word beginning with an accented vowel.

... and under one example for "Juncture":

ill eagle /ɪl iːgl/ : dark [ɫ] in word-final position; possibility of glottal stop before [iː]

In British accents with fuller L-vocalisation, the glottal stop would be even more likely.

In most pre-pausal positions in British English, liaison is rare. That goes for the well-established, standard phenomenon of linking-R (hence "worker... of" is more likely to be [ˈwɜːkə # ɒv] than [ˈwɜːkə # ɹɒv]), and is even more marked for L-vocalisation. This stands in contrast to most varieties of French, where the liaison can happen with enchaînement through certain pauses, particularly through hesitations (although this is not necessarily considered very 'good' French).

Is there thus a intrusive-L in English, similar to intrusive-R? It has been attested, but is more dialectally marked than L-vocalisation in general. Intrusive-R on the other hand is widespread and pretty much normal in the colloquial English of most of England, even crossing syntactic boundaries in certain cases.