French consonant assimilation: why je pense can start with a ch sound
Follow the two changes in casual je pense: a vowel drops and neighbouring consonants interact. Compare absent and anecdote, then check the process.
In short
Assimilation is a sound taking on a feature of its neighbour. In casual je pense, the vowel of je can drop; the remaining /ʒ/ can then lose voicing before /p/, giving a “ch” beginning: [ʃpɑ̃s]. The words still mean “I think”.
Yellow the idea to rememberBlue French to learnRed watch out
Why does je pense sound like ch’pense?
| Stage | Sound sketch | What happens? |
|---|---|---|
| Fuller form | [ʒə pɑ̃s] | The vowel of je separates /ʒ/ and /p/. |
| Vowel drops | [ʒpɑ̃s] | The consonants come into contact. |
| Consonant adapts | [ʃpɑ̃s] | Voiced /ʒ/ becomes voiceless before /p/. |
Je pense que tu as raison.
I think you’re right.
The beginning can be fuller [ʒə pɑ̃s] or more compact [ʃpɑ̃s].
The middle row isolates the vowel loss so you can see the two processes separately. In speech, they can happen together. Removing /ə/ is schwa loss; changing the neighbouring consonant’s voicing is assimilation. The spelling stays je pense in standard writing.
What does voicing change?
/ʒ/ → [ʃ]losing voice
Focus: Vocal-fold vibration is reduced or absent under a voiceless neighbour’s influence.
Je peux répondre.
I can answer.
/k/ → [ɡ]gaining voice
Focus: A voiced neighbour can make a voiceless consonant voiced.
Elle raconte une anecdote.
She tells an anecdote.
In a reduced je peux, /p/ can influence the preceding consonant, just as in je pense. In anecdote, /k/ can become [ɡ] before voiced /d/: [anɛɡdɔt]. The second consonant influences the first in both examples, although the direction of the voicing change differs.
Does assimilation happen inside French words?
Il est absent aujourd’hui.
He is absent today.
In absent, the consonant written b often sounds [p] before voiceless /s/: [apsɑ̃].
The lips still make a closure for [p], as they do for /b/. What changes here is voicing. This gives you three useful checkpoints: which sounds are adjacent, which one influences the other, and which feature changes. The degree of voicing varies with the speaker and context; the bracketed sketches show the main pattern.