Filler words for non-native English speakers: what to work on
Updated 13 October 2026 · Speaking Coach guides · 10 sources
A practice tool, not therapy. These guides and Speaking Coach help with everyday speaking habits. App numbers are estimates from automatic transcription of English speech and can be wrong. If you stammer or stutter, or have voice problems, a speech-language professional is the right person to see.
- Native speakers use fillers and pauses too; listeners judge native and non-native fluency on the same cues. 1
- Pauses and speed predict how fluent you sound more than self-corrections do. 2
- A strong accent does not necessarily make you harder to understand. 3
- Your first language can carry its own filler sounds into English; notice them, then practise English ones or silence.
- Automatic transcription is less accurate for some speakers, so app counts are estimates. 4
Speaking in a second language means planning harder: finding words, choosing grammar and building sentences all take more effort. That extra planning shows up as pauses, fillers and restarts. If you are a non-native English speaker worried about saying "um" or "eh" too often, the good news is that research points to a few specific things that matter most to listeners, and accent is not the main one.
Everyone hesitates
Disfluencies such as pauses, fillers and repairs occur in both native and non-native speech. When researchers manipulated pauses and speed in recordings, listeners' fluency ratings changed in the same way for native and non-native speakers, which suggests they apply the same standards to both 1. Native English speakers produce plenty of fillers; their disfluency rates rise with harder topics and roles 5.
What listeners actually notice
In a study of second-language speech, untrained listeners' fluency ratings were best predicted by measures of pausing and speed. How often speakers corrected themselves contributed very little 2. In practice that means:
- Long pauses in the middle of a phrase stand out more than pauses between phrases.
- A very slow overall pace lowers perceived fluency, but rushing to sound fluent usually adds errors and fillers.
- Fixing a word ("on Tuesday, sorry, Thursday") matters less than people fear.
Accent is not the same as clarity
Many learners focus on accent. A classic study found that listeners could rate speech as heavily accented yet still understand it well: a strong foreign accent did not necessarily reduce comprehensibility or intelligibility 3. Spending your practice time on phrasing, pausing and pace is often more useful than trying to erase an accent.
Fillers from your first language
Most languages have their own hesitation sounds, and they can follow you into English. Researchers have tracked forms like uh and um across English, Dutch, German, Norwegian, Danish and Faroese 6. Listeners generally understand a hesitation sound, but one from another language can stand out. Two options work:
- Replace it with the English "um" or "uh", which listeners process as normal hesitation 7.
- Better, replace it with a short silent pause. See how to pause.
Practice that helps second-language speakers
- Prepare chunks, not just words. Learn useful phrases whole: "The main reason is…", "What I'd do differently is…", "Let me give you an example." Ready-made chunks cut planning time at the points where fillers appear.
- Repeat the same one-minute topic. Record yourself on the same topic three days running 8. Fillers usually fall as the content becomes familiar, and you can hear what changed.
- Read aloud, then retell. Read a short paragraph aloud, then tell it again without looking. Reading builds rhythm; retelling practises planning.
- Use the transcript as a mirror. If your iPhone supports Voice Memos transcription in English, read the transcript to spot repeated words and missing words 9.
- Pace yourself. Conversational English varies widely in speed, and people speak more slowly to strangers than to friends 10. You do not need to match the fastest native speakers. See speaking rate.
A note on apps and automatic transcription
Speech-recognition systems do not work equally well for everyone. A study of five commercial systems found much higher error rates for some groups of speakers than others 4. Any tool that counts your fillers from automatic transcription may miss or mishear words, especially with accents it was trained on less. Speaking Coach analyses English practice takes only; its counts are estimates, and listening back to your own recording remains the most reliable check. If you hear a filler the app did not count, or the other way round, trust your ears.
Common questions
Do non-native speakers use more filler words?
Often, because planning in a second language takes more effort. But native speakers use fillers too, and listeners judge pauses and speed the same way for both groups. 1
Should I fix my accent or my fillers first?
For being understood, pausing, pace and clear phrasing usually matter more. A strong accent does not necessarily reduce how understandable you are. 3
Is it bad to use my native language’s filler sounds in English?
It is not wrong, but it can stand out. Swapping it for a short silent pause, or for the English um, is usually smoother for listeners.
Sources and further reading
- Bosker HR, Quené H, Sanders T, de Jong NH (2014). The perception of fluency in native and nonnative speech. Language Learning 64(3), 579–614. doi:10.1111/lang.12067 doi.org
- Bosker HR, Pinget A-F, Quené H, Sanders T, de Jong NH (2013). What makes speech sound fluent? The contributions of pauses, speed and repairs. Language Testing 30(2), 159–175. doi:10.1177/0265532212455394 doi.org
- Munro MJ, Derwing TM (1995). Foreign accent, comprehensibility, and intelligibility in the speech of second language learners. Language Learning 45(1), 73–97. doi:10.1111/j.1467-1770.1995.tb00963.x doi.org
- Koenecke A, Nam A, Lake E, et al. (2020). Racial disparities in automated speech recognition. Proceedings of the National Academy of Sciences 117(14), 7684–7689. doi:10.1073/pnas.1915768117 doi.org
- Bortfeld H, Leon SD, Bloom JE, Schober MF, Brennan SE (2001). Disfluency rates in conversation: effects of age, relationship, topic, role, and gender. Language and Speech 44(2), 123–147. doi:10.1177/00238309010440020101 doi.org
- Wieling M, Grieve J, Bouma G, Fruehwald J, Coleman J, Liberman M (2016). Variation and change in the use of hesitation markers in Germanic languages. Language Dynamics and Change 6(2), 199–234. doi:10.1163/22105832-00602001 doi.org
- Fox Tree JE (2001). Listeners’ uses of um and uh in speech comprehension. Memory & Cognition 29(2), 320–326. doi:10.3758/BF03194926 doi.org
- Apple Support. Make a recording in Voice Memos on iPhone (iPhone User Guide). support.apple.com
- Apple Support. View a Voice Memos transcription on iPhone (iPhone User Guide). support.apple.com
- Yuan J, Liberman M, Cieri C (2006). Towards an integrated understanding of speaking rate in conversation. Proceedings of Interspeech 2006, paper 1795. doi:10.21437/Interspeech.2006-204 isca-archive.org