Assimilation is a phonological process by which one sound is pronounced like its neighboring sound. Assimilation, in Phonology, can occur between sounds within the same word or different words.
I mentioned in a previous article about the machine that produces mechanical speech. This machine can be compared with an audio dictionary since it contains all the English words recorded individually; or we can say, recorded in isolation. When we put together few of these words to make a sentence, even if the words were spoken by an English native speaker, the resulted sentence will not sound naturally.
That is because in natural connected speech, sounds belonging to one word can cause changes in the sounds belonging to a neighboring word. This is called assimilation. Assimilation varies according to speaking rate and style. It is more likely to be found in rapid speech and less likely in slow, careful speech. Sometimes the difference caused by assimilation is very noticeable but sometimes it is very slight.
The most common cases of assimilation are assimilations that affect consonants. Consider two words that are combined, where the first word ends in a single final consonant (FC – final consonant) and the second word starts with a single initial consonant (IC – initial consonant).
If FC (the single final consonant) changes to become like the IC (the single initial consonant) then, the assimilation is called regressive. If the IC changes to become like the FC, the assimilation is called progressive.
There are three main differences between consonants: differences in place of articulation, differences in manner of articulation, and differences in voicing. So, in parallel with these three kinds of differences we can identify three kinds of assimilation: assimilation of place, assimilation of manner, and assimilation of voicing.
Assimilation of place is very clearly observable in the cases where the final consonant with alveolar place of articulation is followed by an initial consonant with a place of articulation that is not alveolar. The best example that comes to my mind is the word "that" /ðæt/.
The final consonant in the word "that" is the alveolar "t". In rapid speech the consonant "t" will become "p" if the initial consonant is bilabial as in "that person" /ðæp pɜrsən/.
Before a dental consonant, final consonant "t" will change to a dental plosive "t" as in "that thing" or "get those".
The final consonant "t" will change to a consonant "k" if the initial consonant that follows it is a velar consonant as in "that case" /ðæk keɪs/ or "quite good" /kwaɪk gʊd/.
In similar context, "d" will become "b", dental plosive "b", or "g"; and "n" will become "m", "n", and "ŋ". The other alveolar consonants, "s" and "z" behave differently. There is only one noticeable change in their case and that is "s" become "ʃ" and "z" become "ʒ" when they are followed by "ʃ" or "j" as in "this shoe" /ðɪʃ ʃu/ / and "those years" /ðoʊʒ yɪərz/.
It is important to note that the consonant that undergone assimilation has not disappeared. Assimilation of place is noticeable only in this regressive assimilation of alveolar consonants. I don`t think this would be absolutely necessary for foreign learners to learn to do.
Assimilation of manner, compared with assimilation of place, is much less noticeable. It can be found in the most rapid and casual speech and the change in manner is towards a consonant that makes less obstruction to the air flow, called "easier consonant".
There are cases where a final plosive becomes a fricative or a nasal. For example, "that side" will be pronounced /ðæs saɪd/, and "good night" will be pronounced /gʊn naɪt/. However, there are very few chances for a final fricative or nasal to become a plosive.
There is a particular case where a word that begins with /ð/ sound follows a word that ends with a plosive or nasal sound. In this case, it is very common to find that the initial consonant /ð/ becomes identical in manner to the final consonant but with dental place of articulation. For example, "in the" will be pronounced /ɪnnə/ instead of /ɪn ðə/; "get them" will be pronounced /gɛttəm/ instead of /gɛt ðəm/; and "read these" will be pronounced /riddiz/ instead of /rid ðiz/.
As you can see, the /ð/ phoneme occurs with no discernible friction noise.
Assimilation of voice, unlike the other two kinds of assimilation, is found only in limited ways. Only regressive assimilation of voice is found across word boundaries and since it is important for learners of English, we are going to take a brief look at it.
If the final consonant Cf is a lenis consonant (which means voiced consonant) and the initial consonant Ci is a fortis consonant (which means voiceless consonant) then there are 99% chances that the lenis consonant has no voicing at all. I don`t consider this a noticeable case of assimilation anyway, since initial and final lenis consonants have little or no voicing.
When the final consonant Cf is fortis and the initial consonant Ci is lenis, a context in which in many languages the final consonant would become voiced, in English, assimilation of voice never takes place.
I like that black dog. /aɪ laɪk ðæt blæk dɒg/
Many learners of English allow regressive assimilation of voicing to change the final "k" of "like" to "g", the final "t" of "that" to "d", and the final "k" of "black" to "g". This creates a very strong impression of strong accent and obviously it should be avoided.
Effects of assimilation are observable across morpheme boundaries and within the morpheme. Sometimes, in the case of assimilation within the morpheme, it seems that the assimilation is quite different from the word-boundary examples.
The most obvious examples, can be seen in a syllable-final consonant cluster when a nasal consonant precedes a fricative or a plosive in the same morpheme. In this case the place of articulation of the nasal is always determined of the place of articulation of the other consonant.
Looking at these examples, or any other examples we can conclude that this assimilation is "standard", as part of the phonological structure of English language.
A particularly case, similar to the one presented above, is the progressive assimilation of voice with suffixes -s and -z. When a verbs ends in third person singular -s, or a plural noun ends in -s, or a noun carries an -s possessive suffix, that suffix will be pronounced /s/ if the preceding consonant is fortis (voiceless), and /z/ if the preceding consonant is lenis (voiced).
There are many more details to be said about assimilation but from the point of view of learning English (or teaching English as a second language) to continue detailing this topic will not be useful. Assimilation is a natural phenomena that can be observed and the only important matter is to remember restrictions on voicing assimilation already mentioned.