A readily observable example of an allophone is /l/, which may be pronounced as a
clear [l], as in million (it has some of the quality of the vowel // as in fit associated with
it). This pronunciation typically occurs when /l/ precedes a vowel in RP. However, the
/l/ may be dark [
], as in pull, which means it has some of the sound quality of the vowel
/
υ/ as in foot. This is the way an /l/ is pronounced in RP when it is not followed by a
vowel. The difference between the two is easy to hear; however, if they are exchanged
one for the other, the words in which they occur do not become different words or uniden-
tifiable sound sequences. (For more on /l/ see 4.3.1, p. 72.)
This is not always unproblematic, for in some accents of English (e.g. Cockney, various
areas in the United States, New Zealand), /l/ is completely vocalized, that is, realized
more or less like the vowel /
υ/. In this case there is the possibility that new homophones
(words which sound alike, but carry different meanings) may be created. The following
words may, for example, be pronounced similarly in Cockney: Paul [p
ɔυ] or [pɔə], paw,
pore, poor [p
ɔ] or [pɔə] (e.g. Wells 1982: 316). The theoretical question is whether the
[
υ] of [pɔυ] (Paul) is an allophone of /l/ or whether it has merged with the phoneme /υ/.
4.3 THE CONSONANTS
The inventory of English consonants has remained stable to a remarkable degree over
several hundred years. As a result it is the consonants which contribute most to the phono-
logical unity of the English language in its many and often quite different sounding accents
throughout the world. The form of any English word is most easily characterized by the
position and type of combination of its consonants.
Since the first Germanic sound shift (also known as Grimm’s Law) in the third or
second century
BC
, there have been no major changes. However, in the Middle English
period (roughly between 1050 and 1500) the three sounds [
ð], [] and [ŋ], which until
then had been allophones of /
θ/, /ʃ/ and /n/, became independent phonemes. In the same
period the phoneme /x/ (the consonant sound of German ach or ich, which once regu-
larly appeared in words still written with <gh> such as right or thought) disappeared
in all but a few regions, most particularly the regional dialects of Scotland. In addition,
the phoneme /hw/ (as in which) is presently losing its status as an independent phoneme
for more and more speakers as it converges with /w/ (as in witch) – something that has
already happened in RP and for most GenAm speakers.
The consonants may be divided up into the following types as far as the degree of
their consonant-like nature is concerned.
Semi-vowels Semi-vowels or approximants or frictionless continuants are conson-
ants which are usually produced without audible friction in, or stoppage of, the air coming
from the lungs; phonetically, therefore, they are vowel-like. However, they do not form
the centre of a syllable, but are peripheral; that is, they are found initially or finally. In
this phonological sense, therefore, they are consonants. The semi-vowels of English
include /w, r, j/, though each also has variants (allophones) which involve friction and/or
stoppage. /h/ may also be said to belong here; for, although it is not sonorous (that is, it
is not produced with vibration of the vocal cords), it is voiceless and it has as many
variants as there are vowels which may follow it. For this reason it will be called a voice-
less vowel (i.e. it is whispered). However, it is also often termed voiceless glottal
fricative, which would put it in the group of obstruents below.
1111
2
3
4
5
6
7
8
9
1011
1
2
3
4111
5
6
7
8
9
20111
1
2
3
4
5111
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
44
45
46
47111
THE PRONUNCIATION AND SPELLING OF ENGLISH 69