Assimilation is a process of alternation of speech sounds as a result of which one of the sounds becomes fully or partially similar to the adjoining one.
The influence of the neighbouring sounds in English can act in progressive, regressive or reciprocal (double, mutual) direction.
1. When some articulatory features of the following sound are changed under the influence of the preceding sound, which remains unchanged, assimilation is called progressive.
e.g. The pronunciation of the plural suffix –s of nouns depends on the quality of a preceding consonant: it is pronounced as [z] after voiced consonants (pens [penz], calls [kO:lz]) and as [s] after voiceless consonants (desks [desks], books [bVks])
Within the words sandwich, grandmother, etc under the influence of [n] the consonant [d] changed into [n] and then disappeared
(sandwich ['sxnnwIdZ →'sxnwIdZ])
2. When the following sounds influences the articulation of the preceding one assimilation is called regressive.
e.g. Within the word width and in the word combination in them, the alveolar [d] and [n] become dental before the interdental [T] and [D].
3. Reciprocal (mutual, double) assimilation means complex mutual influence of the adjacent sounds.
e.g. Within the word tree [tri:] the sonorant [r] is partly devoiced under the influence of the voiceless [t] and the alveolar [t] becomes [post-alveolar before the post-alveolar [r].
Assimilation can affect the place of obstruction and the active organ of speech; the work of the vocal cords; the position of the lips; the position of the soft palate; the manner of the release of plosive consonants.
1. Modification of the place of obstruction and the active organ of speech.
Assimilation may take place within a word and also at word boundaries. The following three important cases should be noticed:
a) The alveolar allophones of [t, d, n, l, s, z] are replaced by the dental variants when immediately followed by the interdental [T] or [D].
e.g. within a word: breadth, tenth
at word boundaries: Put that down! Read this! on the desk.
b) The post-alveolar [t] and [d] are heard before the post-alveolar sonorant [r].
e.g. within a word: trip, true, trunk, dream, drink, dry.
At word-boundaries: at rest, would read
2. Changing in the work of vocal cords (voicing or devoicing).
a) The sonorants [m, n, l, w, r, j] are partially devoiced when preceded by voiceless consonants [s, p, t, k, f, T, S].
e.g. within words:
[ṃ] - small
[ṇ] - sneer
[j̣̣̣] - stupid, tune, pure, few
[ẉ] - sweep, square
[ṛ] - spread, try, cream
[ḷ] - slow, place climb
At word boundaries the sonorants [l, r, w] are slightly voiced if with the adjacent words they form a phrasal word or a rhythmic group
e.g. at last, at rest.
b) Contractive forms of the verbs “is” and “has” may retain voice or be devoiced depending on the preceding consonants.
e.g. That’s right [Dxts raIt]
Bob’s gone out [bPbz gPn aVt]
c) The assimilative voicing or devoicing of the possessive suffix –’s or –s’, the plural suffix –(e)s of nouns and of the third person singular present indefinite of verbs depends on the quality of the preceding consonant. These suffixes are pronounced as:
[z] after all voiced consonants except [z] and [Z] and after all vowel sounds
e.g. girls [gE:lz], rooms [ru(:)mz]
[s] after all voiceless consonants except [S] and [s],
e.g. books [bVks], writes [raIts]
[Iz] after [s, z] or [S, G]
e.g. dishes [dISIz], George’s [dZO:dZIz]
d) The assimilative voicing or devoicing of the suffix –ed of regular verbs also depends on the quality of the preceding consonant. The ending –ed is pronounced as:
[d] after all voiced consonants except [d] and after all vowel sounds
e.g. lived [lIvd], played [pleId]
[t] after all voiceless consonants except [t]
e.g. worked [wE:kt]
[Id] after [d] and [t]
e.g. intended [In'tendId], extended [Ik'stendId]
In English regressive voicing or devoicing is found only in few cases of historical assimilation within a compound word when the semantic independence of the first component is lost
e.g. fivepence ['faIfpqns], gooseberry ['gVzb(q)rI]
Regressive voicing or devoicing may also take place in closely connected pairs of words
e.g. I have to [aI hxftu], I used to [aI ju:sttu], does she [dAS SI].
3. Changes in the lip position.
Consonants followed by the sonorant [w] change their lip position. They become lip-rounded in anticipation of [w].
e.g. twinkle, quite, swan, language
4. Changes in the position of the soft palate.
Nasal consonants may influence the adjacent plosives. This type of assimilation is not typical of English. Sometimes [d] changes into [n] under the influence of the preceding [n].
e.g. handsome ['hxndsqm → 'hxnnsqm → 'hxnsm]
Nasalisation affects mainly the alveolar consonants, especially adjacent to the negative n’t, and is characteristic of very rapid speech.
e.g. She wouldn’t do it [SI wVnnt dH It]
5. Changes in the manner of the release of plosive consonants.
English plosives don’t always have a sudden oral release of air. The main variants are:
a) Loss of plosion
A plosive loses its plosion, if it is followed by another plosive or affricate,
e.g. within a word: accommodation, attraction, bookcase
at word boundaries: what time, went down, that child, that joke
b) Nasal plosion
When a plosive is followed by a the syllabic [n] or [m], it has no release of his own, the so-called nasal plosion is produced. In such sequences the closure for the plosive is made normally, but the release is produced not by a removal of the oral closure, which is retained, but by the lowering of the soft palate, which allows the compressed air to escape through the nasal cavity to form the nasal consonant.
e.g. within a word: happen ['hxpn], shipmate ['SIpmeIt]
at word boundaries: sob noisily, stop moaning
c) Lateral plosion
In the sequences of a plosiveimmediately followed by [l] the closure produced for the plosive is not released till after [l]. Before [l] the release is made by a sudden lowering of the sides of the tongue, and the air escapes along the sides of the tongue with the lateral plosion.