Dialectal variation

In BE, the /ɹ/ has no retroflexion; the tip of the tongue approaches the alveolar area in a way similar to that of alveolar stops, but does not make any contact with the roof of the mouth. This is commonly described as a post-alveolar approximant. Besides the difference in production, the distribution of this sound varies greatly. Whereas in AE and other so-called ‘rhotic’ (or ‘r-full’) dialects, such as IrE, ScE, Canadian English (CnE), /r/ can occur without much restriction, in BE (except the southwest of England) and in other ‘non-rhotic’ (or ‘r-less’) dialects, such as in New England and the Southern USA, Australia, New Zealand, and Wales, it can occur only before vowels. Thus, we observe differences such as car ([kɑ] / [kɑɹ]), farm ([fa:m] / [fɑɹm]). When a word ending in r is followed by a word beginning with a vowel, we see a ‘linking r’ in non-rhotic dialects (e.g. player of the game [pleəɹ əv...]). Another aspect of the ‘r-less’, non-rhotic dialects is the ‘intrusive r’, whereby an /ɹ/ is inserted between a word ending in /ə/ and a following word that starts with a vowel, as in India and Pakistan [Indiəɹ ænd...], the idea is [...aɪdiɚɪz]. In AAVE, /ɹ̣/ may be deleted intervocalically (e.g. during [dʊɹ̣ɪŋ] → [dʊɪŋ], Carol [kæɹ̣əl] → [kæʊ]), as well as in clusters (e.g. professor [pɹ̣əfεsɚ] → [pəfεsɚ]).

The other liquid, alveolar lateral approximant /l/, also presents appreciable differences among different varieties. While it is customary to see the groupings of ‘dark l’ and ‘clear l’ (‘light’ or ‘bright’ in some publications) in referring to both AE and BE, the reality is rather different. In BE, we find the ‘clear l’, which is articulated with the tongue tip in contact with the alveolar ridge (resembling an /i/ vowel, with no raising of the back of the tongue) in prevocalic (onset) position, as in like, law; in postvocalic (coda) position (e.g. fall, belt) the realization is the velarized ‘dark l’, which has a quality similar to /u/ with raising of the back of the tongue toward the velum. In AE, as well as ScE and IrE, however, we may hardly find the ‘clear l’; most commonly, the realizations differ in terms of shades of the ‘dark l’. Thus, we find a ‘dark l’ before front vowels (e.g. left), a more velarized darker variety before back vowels (e.g. loose, low), and the darkest one in postvocalic position (e.g. bolt, full). The syllabic [l̩] is invariably ‘dark’ in AE. In Welsh English (WeE), the /l/ is always ‘clear’. In AAVE, postvocalic /l/ may vocalize to [u] or [ʊ], as in bell [bεl] or [bεʊ], and /l/ may be deleted before a labial consonant (e.g. help [hεlp] or [hεp], wolf [wʊlf] or [wʊf]).

Words such as music [mjuzɪk], museum [mjuziəm], pure [pjuɹ̣], cure [kjuɹ̣], cute [kjut] have to have the same two-sounds sequence in their onsets in both AE and BE. However, the two dialects vary when we examine words such as tune, nude, dune, news, lute. While in BE, and to a lesser degree in New England, we see a /j/ after the first consonant in these words ([tjun], [njud], [djun], [njuz], [ljut]), the expected AE pronunciations are without a /j/ ([tun], [nud], etc.). The same difference is observed in words such as assume, resume ([əsjum], [ɹəzjum] in BE, and [əsum], [ɹ̣əzum] in AE). These examples may suggest that /j/ may not follow an alveolar in the same morpheme in AE (across morphemes this is possible, as in would you, bet you). This generalization, however, has to be amended, because words such as onion [ɑnjən], tenure [tεnjɚ], annual [ænjuəl], value [vælju], failure [feljɚ], million [mɪljən] have alveolars /n/ or /l/ followed by a /j/ in AE as well as in BE. Thus, the correct characterization of the AE restriction on alveolars should read as “/j/ cannot follow an alveolar obstruent; it can follow an alveolar sonorant when in an unstressed syllable”. In AAVE, /j/ can be deleted in a [CjV] sequence (e.g. computer [kəmpjutɚ] or [kəmputɚ]).

For several speakers of AE, as well as ScE, IrE, and NZE, /w/ has a voiceless version (phonetically shown as [hw] or [ʍ]) in words spelled with wh. Thus, these speakers make the following distinctions in pairs such as Wales - whales [welz] – [hwelz], witch– which [wɪtʃ] – [hwɪtʃ].