I teach French to people from various background and first languages, but one thing that most English speakers do (even very proficient ones sometimes) is adding R sounds in words. Saying 'droi' for "doigt", 'par' for 'pas', 'mitroyen' for "mitoyen". Other speakers don't do this at all. Any idea why this happens?

  • When you say an ‘r’ sound, do you mean an English [ɻʷ] or a French [ʀ]? In two of your examples, note that you have a dental followed by [wa], of which [w] is a velarised consonant. If you’re concentrating on correctly producing the foreign (and to many people quite difficult) velar/uvular trill [ʀ], the velar narrowing in [w] could perhaps end up getting so narrow that it becomes actual friction – but of course only if you’re talking about the French r sound. As for pas, English doesn’t have short [ɑ] and wouldn’t allow short [a] in an open monosyllable, so [ɑː] would be closest. Oct 21, 2020 at 20:55
  • 2
    Many English dialects have a rounded /r/, which is sometimes reduced to /w/. Since initial /dr/ is common in English, while initial /dw/ is rare, droi for doigt may be a case of mishearing the French. To an English speaker, French /r/'s are very strange animals.
    – jlawler
    Oct 22, 2020 at 1:50
  • 1
    What sort of English do your English speakers speak? British? American? Irish?
    – fdb
    Oct 22, 2020 at 12:44

1 Answer 1


This is conjecture, but it's conjecture that's too long to put into the comments.


French /dʀw/ and /tʀw/ are likely to be analysed as /dw/ or /tw/ by English speakers. Because the realisations of /ʀ/ and /w/ in such words are co-articulated and therefore overlap, they mimic what would be analysed as a single /w/ in English. Some English speakers will therefore replicate this pronunciation for all /dw/ and /tw/ sequences.

Full answer

Doigt and mitoyen

Firstly in a word like droit, there is nothing resembling what an English speaker would recognise as an (English) /r/, the sound [ɹ]. The French /ʀ/, often realised as [ʁ], is unusual enough for English speakers to recognise it in those environments where we would have no comparable sound. However, the [ʁ] in a word like droit is very similar to a sound that many English speakers sometimes encounter when they hear a word-initial /w/ in English, especially if this word is particularly stressed.

English /w/ is a co-articulated consonant. In other words, it involves two strictures (narrowing obstructions of the airflow through the vocal tract). The first is labial, caused by the narrowing and rounding of the lips. The second is velar, cause by the back of the tongue raising towards the velum. In English, it is the labialisation of /w/ which is most prominent, because this stricture is the narrowest. In a language like Spanish, the velar stricture for Spanish /w/ is narrower. It would be fair to say that English speakers find their /w/ more like the vowel /u:/ and Spanish speakers their /w/ more like the velar consonant /g/. However, occasionally, when an English speaker hears an accented or stressed word beginning with /w/ they will hear frication caused by the back of the tongue getting very close to the velum and causing there to be frication, turbulence in the air stream. This is caused by the air from the lungs becoming disturbed as it is forced through the narrow space between the raised back of the tongue and the velum, the soft skin covering the back of the roof of the mouth. In other words the velar aspect of the /w/ will become much more prominent.

Why does this matter? Well, when an English speaker hears a word like droit, they don't hear the beginning of the word as /dʀwɑ/ but just as /dwɑ/. When we speak, our articulators, our mouths, are always preparing for the forthcoming sounds and the articulations necessary to produce them. A French speaker saying droigt will already be rounding and narrowing their lips during the production of the [d] in anticipation of the [w]. In other words the [d] will be labialised. This lip rounding will continue through the /ʀ/, or rather [ʁ], segment. At this point the co-articulation of the French /ʀ/ and the French /w/, will mimic a hyper-articulated English /w/. An English speaker will not recognise the frication as a separate phoneme.

My semi-educated guess (and that's what this is) is that an English speaker's phonological and phonetic brain, when they hear droit, interprets it as /dwɑ/ but detects what they feel to be a specially French articulation of the /w/. They then carry this pronunciation through to all /dwɑ/ sequences they encounter.

One may wonder why an English speaker would interpret a uvular fricative [ʁ] with that component of a hyper-articulated English /w/, the velar fricative [ɣ]. The reason is simply that English has no uvular sounds made at the very back of the mouth. So any fricative made at the back of the mouth counts as the same single type of noise for an English listener (there is no differentiation between velar and uvular sounds).


The case with pas is completely different. There are two possibilities that come to mind. The first is that French /ɑ/, [ɑ], is a sound which, in English is associated with words with an orthographic R. For rhotic speakers [those that pronounce English /r/, even when there is no following vowel], it may be that they include an /r/ there because it would be present in English. For non-rhotic speakers [those that only pronounce English /r/ when there is a following vowel], they may be inserting an /ʀ/ in the French word when the following word begins with a non-high vowel. This mimics what non-rhotic English speakers do in English with English /r/.

  • 1
    English doesn’t permit /vw/ and /vr/? Perhaps not in inherited, lexical words, but loan words like voyeur /vwɑˈjɜːr/ and onomatopoetic words like vroom /vruːm/ definitely do have those clusters initially. And of course English definitely permits /dw/ and /tw/, so if speakers perceive French /Tʀw/ clusters as a sort of ‘special’ /Tw/ cluster, it’s because they recognise them as different from their own /Tw/ clusters – but they should then still find it easy enough to distinguish these from French /Tw/ clusters, which are like the English ones. Oct 28, 2020 at 19:46
  • @JanusBahsJacquet Point taken re vroom and voyeur, but I do think such cases are exceedingly rare, and as you point out are both exceptional in other ways. However, I don't agree with your second point there. These two pronunciations would be allophonic in English for English speakers [and hence the words droit and doigt would appear to be homophones]. The ones that are the same as the English ones won't stick out or sound particularly French for them and they'll pass unnoticed. However, the "French-sounding" ones will be more noticeable and prominent. ... Oct 28, 2020 at 20:32
  • @JanusBahsJacquet You conjecture English speakers should find it easy to tell French /tʀw/ clusters from /tw/ ones (and I believe that if you gave them minimal pairs, they'd be able to consistently pick up on which one had that particularly French sound). However, problem is that as evidenced by the question, (which I assume to be in good faith), they don't seem to be able to on their own. And then we'd need a better explanation of why. There almost definitely is a better or fuller explanation, including why they find droit and doight homophonesbut I'm stuffed if I can think of it ;-) Oct 28, 2020 at 20:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.