Assimilation of /t/, /d/ and /n/


Assimilation is a natural process which happens in every language. It is also carried out unconsciously, so speakers don’t normally realize what they are doing and even tend to be surprised when told that the actual sounds they produce don’t always match the spelling. The reason behind assimilation processes is quite simple: our articulators (tongue, lips, teeth, etc.) have to move from one position to another -from /n/ to /b/, for example-, but certain changes are difficult to make in the required time, so they take a shortcut.

Let’s see this case in detail, as an example:

English                                               Spanish

One boat                                           Un barco

Changing from alveolar /n/ to bilabial /b/ is not easy because it involves too much movement in the mouth. It can only be done accurately in slow, careful speech. If we are speaking rapidly we turn the alveolar nasal /n/ into a bilabial nasal /m/, so that we are already prepared for the bilabial /b/. In short, we skip one step (alveolar) and go directly to the next one (bilabial). That’s why we normally say:

ǀ wʌm ˈbəʊt ǀ    ǀ um ˈbarko ǀ

And that’s also why, in Spanish, the prefix en  becomes em  when it’s followed by bilabials p and  b  (empaquetar, embotellar). As you see, it’s a completely natural process.

Three points to remember

We’ll first address the assimilation phenomenon by explaining what happens to the phonemes /t/, /d/, /n/ (we’ll deal with other sounds in the future). The cases studied below have a number of common features. Keep the following three points in mind (it would be useful to have this table at hand):

  • Alveolar. All the sounds subject to variation here are alveolar. Alveolar consonants are especially unstable in English (my next article will be about assimilation of /s/ and /z/).
  • Place of articulation. The only change these sounds undergo is one of place of articulation. E.g., /t/ (alveolar) can become either /p/ (bilabial) or /k/ (velar). But the manner of articulation remains the same (plosive, in this case).
  • Voicing. Also the voicing remains unchanged. Voicelss /t/ can turn into voicelss /p/ and /k/. Voiced /d/ can become voiced /b/ and /g/.

/t/ → /p/


/t/ (followed by /p/, /b/ or /m/) becomes /p/

sound_loud_speaker That person.      ǀ ðæp ˈpɜːsən ǀ

sound_loud_speaker It boils.      ǀ ɪp ˈbɔɪlz ǀ

sound_loud_speaker That money.      ǀ ðæp ˈmʌni ǀ

sound_loud_speaker But, yeah, at that point I thought, “My God, I could be much more charming than Hugh!” (Rupert Everett, BBC4).

sound_loud_speaker There were sources that believed that they had a hand in Philip’s assassination (Paul Cartledge, BBC4).

sound_loud_speaker So, clearly, in a way, not much of a regular rythm there (Michael Rosen, OpenLearn)


/t/ → /k/

/t/ (followed by /k/ or /g/) becomes /k/

sound_loud_speaker Is that clear?      ǀ ðæk ˈklɪə ǀ

sound_loud_speaker Put down that gun.      ǀ ðæk ˈgʌn ǀ

sound_loud_speaker He greets ticket collectors and stationmasters and they return his salute (Julian Barnes, KUSP).

sound_loud_speaker As ideas emerge, get going, start writing and a form will start becoming aparent to you (Michael Rosen, OpenLearn)

/d/ → /b/


/d/ (followed by /p/, /b/ or /m/) becomes /b/

sound_loud_speaker It could be better.      ǀ kʊb bi ˈbetə ǀ

sound_loud_speaker You could publish it.      ǀ kʊb ˈpʌblɪʃ ɪt ǀ

sound_loud_speaker She could modify it.      ǀ kʊb ˈmɒdɪfaɪ ɪt ǀ

sound_loud_speaker But, yeah, at that point I thought, “My God, I could be much more charming than Hugh!” (Rupert Everett, BBC4).


/d/ → /g/


/d/ (followed by /k/ or /g/) becomes /g/

sound_loud_speaker You should come.       ǀ ʃʊg ˈkʌm ǀ

sound_loud_speaker He should go.       ǀ ʃʊg ˈgəʊ ǀ

sound_loud_speaker He had very low self-esteem, so he didn’t feel he could go and actually approach this person about what they’d said about him (Theresa Gannon, BBC4).

/n/ → /m/


/n/ (followed by /p/, /b/ or /m/) becomes /m/

sound_loud_speaker Ten percent.      ǀ tem pəˈsent ǀ

sound_loud_speaker Ten boys.        ǀ tem ˈbɔɪz ǀ

sound_loud_speaker Better than me.        ǀ ˈbetə ðəm ˈmi ǀ

sound_loud_speaker Though I have, certainly on one book, gone back to writing the first draft entirely by hand (Julian Barnes, OpenLean).

/n/ → /ŋ/


/n/ (followed by /k/ or /g/) becomes /ŋ/

sound_loud_speaker One king.       ǀ wʌŋ ˈkɪŋ ǀ

sound_loud_speaker Then go for it!      ǀ ˈðeŋ ˈgəʊ fər ɪt ǀ


Is assimilation obligatory? No it isn’t. But it’s natural and it makes the speaker sound more fluent. One important point is that it doesn’t have to be done all the time, that is, on every single occasion. But, by the same token, a speech totally deprived of assimilation sounds very artificial, even stilted.
Listen to this interesting recording:
sound_loud_speaker This is now going to be a struggle between the good guys and the bad guys. Now the army, he says, the army top brass are the bad guys, the small number of intellectuals who are campaigning to reopen the trial are the good guys (Robert Gildea, BBC4).
The speaker doesn’t assimilate the phoneme /d/ to a /g/ in the phrase good guys, but he does do it in bad guys. So he says
ǀ gʊd gaɪz ǀ      but
ǀ bæg gaɪz ǀ
Why does he choose to do it so? This is probably a matter of personal taste or habit, or it might depend on the moment, the pace of the speech or other factors. Actually, it doesn’t matter. Some speakers consistently assimilate certain sounds and not others, and the same can be said about some expressions or combinations of words. The fact that someone pronounces should go as ǀ ʃʊg ˈgəʊ ǀ doesn’t mean that he or she assimilates every /d/ to a /g/ when it’s followed by a velar. What is certainly true is that assimilation is extremely common in the speech of native speakers and that it makes non-native speakers sound more natural as well.