Your Mouth Has an Accent – Even in Silence

Published November 3, 2026

Every language has a default mouth position that speakers return to again and again. What does the American English one look like, how much of it has science actually measured – and how can you train it?

An old idea, measured only recently

The idea that each language has its own mouth "gear" goes back to the mid-1600s (Laver 1978). German phonetics has used it as Artikulationsbasis almost continuously since Eduard Sievers in 1876, while British phonetics largely ignored it between about 1909 and 1964 (Jenner 2001).

In 1964 the phonetician Beatrice Honikman brought the term articulatory setting into English phonetics. Her advice to learners: shift your whole mouth into the new language's position first, and only then drill individual sounds (Wikipedia). Her description of English – a loosely closed jaw, neutral lips, a tapered tongue tip at the alveolar ridge – was based on observation of mainly British speech, not on measurement. She herself noted that Americans tend to open the jaw more (Esling).

Because a setting doesn't affect every sound, it is hard to hear directly. So researchers eventually stopped studying speech – and started studying silence.

What the lab has measured

English vs French: the tongue tip is the clearest difference

Ian Wilson repeated the X-ray comparison with modern tools: ultrasound for the tongue and optical sensors on the lips and jaw. He tested 7 Canadian English and 8 Québécois French speakers and analyzed 770 pauses (Wilson 2013):

Values are normalized distances (in mm) from Wilson's study – useful for comparison, not exact anatomical measurements.

The raised English tongue tip is the result that holds up across both studies. The lips are less clear: the earlier X-ray study found the English upper lip more forward but the French lower lip more forward, and Wilson's lip results even contradict Honikman's predictions (Wilson 2013). Neither study found a jaw difference.

American speakers: a pause is not a rest

Real-time MRI at the University of Southern California filmed 5 American English speakers. When the mouth was truly at rest, it was more closed and much more variable than during grammatical pauses. The measured jaw opening was 0.33 at rest, but 0.48–0.50 in pauses and 0.58 when "ready to speak." In other words, the pause posture is actively held, not a collapse into relaxation (Ramanarayanan et al. 2013). A follow-up found that native speakers of German, Hindi and Tamil held significantly different settings when speaking English (Ramanarayanan et al. 2011).

Where the "throaty" American sound may come from

An X-ray study of 4 American speakers found that schwa – often called a "neutral" vowel – actually involves a real narrowing in the pharynx, much like American /r/ (Gick 2002). Since schwa is the most frequent vowel in English (Styles), this small pull toward the throat repeats constantly in running speech.

Caveat: all of these studies are small (3–15 speakers), and the cross-language comparisons are mostly Canadian English vs Québécois French. No one has yet compared US English with German, Russian or Spanish using the same method. Treat the results as strong hints, not settled facts.

What accent coaches describe – and where they disagree

Coaches who train actors and learners describe General American by feel rather than in millimeters:

• Jaw – free and dropping. Let gravity, not muscle, open the jaw, especially on stressed syllables (Rachel's English). Since the lab found no jaw difference at rest, the open jaw is probably a feature of speaking, not of pausing.
• Lips – relaxed, not spread. Rachel's English warns against "speaking horizontally" (pulling the lips wide) and lets the tongue shape the vowels instead. But one academic ESL framework lists spread lips as American (Esling), while the lab found English lips less spread than French ones. The safest advice: don't push the lips wide or round – let them hang.
• Resonance – low and open. StageMilk describes the throat as the "engine" of the American sound and suggests imagining talking around a baseball at the back of the mouth (StageMilk). Rachel's English treats "uh" as the core American sound (Rachel's English), and the standard textbook notes that slight nasality is common among many Americans (Wikipedia, citing Cruttenden).

Tongue forward or back? Here coaches openly contradict each other. Some describe a retracted American tongue root (Talking with Teresa); others say the tongue rests forward and responsive (Language Academia). The research suggests both are partly right: the tip is high and forward at the alveolar ridge, while the body and root are pulled back by schwa, /r/ and the "dark" L in words like full and milk (Wikipedia: Velarization).

Native-sounding bilinguals switch settings

If the setting is part of an accent, people who sound native in two languages should have two settings. A study of 8 French–English bilinguals found exactly that: only those perceived as native in both languages held a separate pause posture for each one. When they didn't know which language would come next, they defaulted to the setting of the language they used most. The most balanced bilingual combined French lips with an English tongue tip (Wilson & Gick 2014).

The setting can also be learned: native German speakers reading English held their tongue lower and further back than in German (Benítez et al. 2014).

Cause or effect? These studies show that a native-like setting and a native-like accent go together. They don't prove the setting causes the accent – people who sound native also produce native-like individual sounds.

Coming from German, Russian, Spanish or French?

Most of these contrasts come from classic phonetic descriptions, not lab measurements – treat them as hypotheses to test in the mirror.

Your hesitation sound is another giveaway. French "euh" is front and rounded, German "äh" is central and unrounded, and American "uh" is schwa-like or a slightly backer vowel, as in but (Lo; Muhlack, Trouvain & Jessen 2023). Researchers have suggested a link between hesitation sounds and articulatory setting, but the evidence is still circumstantial (Lo).

Exercise: find your American home position

Do this in front of a mirror for 2–3 minutes before practice. It combines coaching techniques with the research above – a practice routine, not a lab-tested method.

1. Release the jaw. Massage the jaw joint just in front of your ears, then let the jaw fall slightly open by gravity. Back teeth apart, not clenched (Rachel's English).
2. Find a low "uh". Sigh out a relaxed schwa and feel it low in the chest. Picture the space at the back of your mouth opening up.
3. Mirror check: lips. While saying "uh", your lips should neither stretch sideways nor round forward. German and French speakers: watch for automatic rounding. Russian speakers: watch for a smile-like spread.
4. Park the tongue tip high. Rest it lightly near the alveolar ridge, and touch the ridge for t, d, n (Language Academia) – matching the raised tip measured in the lab.
5. Slide "uh" into "er". Either curl the tip up slightly or bunch the middle of the tongue, keeping the front of the tongue from touching anything (Rachel's English). Choose the shape that gives the right sound – both work.
6. The freeze test. Read a sentence aloud and freeze at the period. Jaw free? Lips quiet? Tongue tip up? That is the inter-speech posture the researchers measured. Train it until it becomes your default.
7. Compare your thinking sound. Say your native hesitation sound ("euh", "äh", "eh", "э"), then American "uh". Feel jaw, lips and tongue change – a quick check of which setting you're in.

The takeaway

The research changes where learners should practice: in the silences. An accent is not only a set of sounds but also a posture the mouth returns to between them – and people who sound native in two languages keep a separate posture for each. The American version is best summed up as a raised tongue tip, quiet lips and a free jaw, with the tongue body pulled back toward the throat by "uh" and /r/. Treat it as a hypothesis to test in your own mirror, and judge it by the sound it produces.

Sources

• Acoustics Today (2018), on Flege 1984 – link
• Benítez, Ramanarayanan, Goldstein & Narayanan (2014), Interspeech – link
• Duval, "Phonétique – base articulatoire" (compiles Honikman, Sweet, Hall and others) – link
• Esling, "Pronunciation Considerations in ESL: Voice Quality Settings" – link
• Estudios de Fonética Experimental, articulation of [t–d] in Spanish – link
• Gick (2002), Phonetica 59 – link
• Gick, Wilson, Koch & Cook (2004), Phonetica 61 – link
• I'm All Ears, "The Spanish Vowels" (University of Washington Pressbooks) – link
• Jenner (2001), Historiographia Linguistica 28 – link
• Language Academia, "Oral Posture, Strength and Control" – link
• Laver (1978), Historiographia Linguistica 5 – link
• Lo, hesitation sounds in German and French (accepted manuscript) – link
• Marsh, Elfenbein & Ambady (2003), Psychological Science – link
• McAllister Byun, Hitchcock & Swartz (2014), JSLHR – link
• Mielke, Baker & Archangeli (2016), Language 92(1) – link
• MIT Press Reader, "Mysteries of Foreign Accent Syndrome" – link
• Muhlack, Trouvain & Jessen (2023), Languages 8 – link
• Ohio State Working Papers in Linguistics (1994), German vs English vowel rounding – link
• Rachel's English – vertical vs horizontal speaking, vocal warm-up, R consonant
• Ramanarayanan, Goldstein, Byrd & Narayanan (2011), ISSP – link
• Ramanarayanan, Goldstein, Byrd & Narayanan (2013), JASA 134(1) – link
• ScienceDaily (2007), on Weikum et al., Science – link
• Scientific American, "Unspoken Accents" – link
• StageMilk, American Accent Guide – link
• Styles, "Which vowels are used most in English?" (Figshare) – link
• Talking with Teresa (Towson University) – link
• WebMD, Foreign Accent Syndrome – link
• Wikipedia, Basis of articulation – link
• Wikipedia, Velarization – link
• Wilson (2013), ultrasound and Optotrak study (working paper) – link
• Wilson & Gick (2014), JSLHR 57(2) – link