In general, consonants exhibit a far greater stability across different varieties. This is why here we’ll only discuss some of the options for variability that do exist, rather than listing all the defaults. Further details about actual usage in different varieties will crop up again on the relevant pages.


Initial plosives – that are generally assumed to be relatively moderately aspirated in English – may either be very strongly aspirated in certain accents (e.g. in Scouse) or completely unaspirated in others (e.g. in Yorkshire English). To distinguish between unaspirated and voiced plosives may often be very difficult, especially for the untrained ear, in which case it is probably best to consult a spectrogram in order to see when the voicing begins. Final plosives are often unreleased or at least have a weak release in many accents. Some – especially ‘non-native’ – varieties also ‘swap’ voiced and voiceless plosives and vice versa. Medial and final plosives may also be realised as glottal stops in some accents or be pre-glottalised. Medial /t/ may be flapped/tapped ([ɾ]) or rolled/trilled ([r]) in some varieties, so that it assumes /d/ or /ɹ/-like qualities.


It is primarily among the weak fricatives that we find a fair degree of variability. Here, it is especially /θ/ which may be realised as [f] (e.g. in Cockney) or [tʰ] (e.g. in Irish English), yielding [fɪn] and [tʰɪn] for thin respectively. This change in the place of articulation may also be observed in the voiced counterpart – but maybe to a lesser extent (?) –, possibly yielding realisations like [mʌvɐ] or [mɒdə˞] for the word mother.

Another very common feature of many accents is h-dropping, i.e. the non-realisation of initial <h>. This feature is widely stigmatised, but even in RP, at least in unstressed versions of the pronouns him, her, his, etc., is a normal feature of connected speech.


For nasal consonants, there are essentially only two issues. The first is whether the grapheme sequence <ng> is realised as /ŋ/ or /ŋg/ in words like sing or singer, where RP only has the first variant. There is only a small area left in the UK these days where this is the case, but it may also occur as a feature of some non-native accents. The second issue concerns the realisation of the suffix -ing, more specifically whether it realised as ending in a velar (i.e. [ɪŋ]) or an alveolar nasal (i.e. [ɪn]). The interesting thing about this feature is that, although it tends to be stigmatised, it not only tends to occur in lower-class accents, but also in the speech of the aristocracy.

Semi-Vowels & Continuants

(Non-)Realisations of /ɹ/

As we have already heard above, /ɹ/-like realisations of medial /t/ may occur in certain accents. The labio-dental approximant [ʋ] tends to occur in the speech of mainly younger (Southern) British speakers and is supposed to be one of the features of Estuary English. Rhoticity in general, i.e. the presence of postvocalic /ɹ/ ([˞ ]) in pre-consonantal or pre-pausal position, is a distinguishing feature for many different accents. For example, most accents of American English tend to be rhotic, whereas in England, rhoticity has been receding for a relatively long time, so that there are now relatively few rhotic accents left in England, most notably in the South-West and in very small parts of Lancashire. Within Britain, but outside of England, we have Scotland and Ireland as fairly large rhotic areas. To get a better idea, have a look at this map of rhotic accents in Britain.

(Non-)Realisations of /l/

As far as /l/-realisations are concerned, we have already discussed the difference between clear [l] & dark l [ɫ] previously, and stated that in general the clear variant occurs at/towards the beginning, while the dark variant occurs at the end of words/stem morphemes. However, some accents – most notably Irish English – tend to have a clear l in all positions. But l does not necessarily always get realised as a lateral at all, but is often vocalised to a back vowel. This back vowel is most likely to be an unrounded one, either [ɯ] or possibly also [ɤ].

(Non-)Realisations of /j/

There is at least some variability in the realisation of ‘plain’ /j/ in different accents of English, with those that exhibit an American influence tending more towards what is commonly referred to as yod-dropping. However, yod-dropping is not uncommon in British accents of English, only usually to a lesser extent and only in certain environments. So, whereas it is increasingly becoming less common to pronounce the semi-vowel between an alveolar fricative and /uː/, such as in words like super or suit, British English in general still tends to retain /j/ after nasal vowels, such as in new, neutral or music. There is one area in England, though, where the degree of yod-dropping is extreme, so that even following an initial bilabial plosive or labio-dental fricative, such as in the words beautiful or few, the /j/ is elided. The area where this occurs includes all of East Anglia and part of the South(-East) Midlands, as can be seen from this map, and the feature is so notorious that it is even exploited in ads that promote a certain East-Anglian brand of food.

In combination with an alveolar plosive, /j/ is increasingly subject to coalescence, i.e. a ‘blending’ together of the plosive and the /j/ in order to form an affricate. This results in pronunciations like [tʃuːn] & [dʒuːn] instead of [tjuːn] & [djuːn] for words like tune & dune, where effectively the semi-vowel is ‘converted’ into a fricative. A similar thing is supposed to have happened to words like sure & sugar in the 17th century, where a process of coalescence is assumed to have resulted in the change of the alveolar fricative to a palato-alveolar one with concurrent ‘loss’ of the /j/.

Furthermore, as we will see later, /j/ may also serve as a linking element in connected speech.

Cruttenden, A. (1994). Gimson’s Pronunciation of English (5th ed.). London: Edward Arnold.

Wells, J. (1982). Accents of English (Vols. 1-3). Cambridge: CUP.