Filler words and silence: how to stop padding when you speak
Introduction
"Um… like… you know… I mean…" Everyone has filler words, and almost nobody knows which ones are theirs until they hear a recording of themselves. Then comes the uncomfortable surprise: that one word repeated fourteen times in three minutes, the "um" at the start of every sentence, the "right?" at the end of them.
It's worth starting by taking the drama out of it. Fillers aren't a moral failing or a sign of clumsiness: they're a natural mechanism of spontaneous speech. They show up when thought lags behind the mouth and you need to buy a fraction of a second. Everyone uses them, in every language, and in casual conversation they go completely unnoticed.
The problem appears when they pile up in situations where the listener is paying attention to how and not just what: a presentation, an interview, an important meeting. There, fillers stop being background noise and start reading as insecurity.
The solution isn't to speak perfectly. It's to replace the padding with the one thing that does the same job at no cost: silence.
Why they happen
Speaking in real time is a cognitively demanding task: you have to find the word, order the sentence, monitor the other person's reaction, and decide what comes next, all at once. When that chain jams for an instant, the brain needs to signal "I'm still here, I haven't finished" so nobody interrupts. And to do that, it emits a sound.
From this comes the most useful observation of all: the filler takes the place of a pause. It isn't there for no reason; it's covering a gap that existed anyway. That's why trying to "eliminate filler words" rarely works on its own: if you don't replace them with something, the gap comes back.
Nerves matter too. The more a situation counts, the faster we speak, the less we breathe, and the more padding slips in. It's a recognizable loop: nerves accelerate, acceleration leaves gaps, gaps get filled.
The three kinds of padding
Not all padding sounds alike, and telling them apart helps you tackle them.
The sounds: "um," "uh," "mmm." These are the most automatic and the most noticeable on a recording. They almost always mark a search for a word.
The filler words: "like," "basically," "I mean," "sort of," "kind of." They sound more polished, but they carry no information. Their danger is that they pass for real language, which makes them harder to notice.
The tag questions: "you know?", "right?", "okay?", "does that make sense?" These seek confirmation from the listener. Used sparingly they connect; used constantly they signal insecurity and ask for approval in every sentence.
Silence is the tool, not the enemy
Here's the underlying shift. Most people pad because they find silence unbearable, and that discomfort is a distortion of perception: a two-second pause feels like an eternity from the inside and like a normal pause from the outside. The speaker always overestimates how long it was.
And the pause does things that padding doesn't:
- It buys thinking time without giving up your turn.
- It separates ideas, which is what lets the listener process them.
- It emphasizes: whatever comes right after silence gets heard more closely.
- It projects confidence. Someone who can stay quiet in front of others without getting flustered reads as being in control.
Good speakers don't talk faster or use more words. They talk with more space.
How to reduce them in practice
1. Record yourself. It's uncomfortable and it's essential. Three minutes talking about anything, listened to carefully, reveals more than any advice. Note your specific fillers: you can't fix what you haven't identified.
2. Pick just one. Attacking them all at once doesn't work. Start with the most frequent one and forget the rest for two weeks.
3. Replace, don't erase. Whenever you feel the "um" coming, close your mouth. Literally. Filler sounds need an open mouth; closing it cuts them off and leaves a pause in their place.
4. Slow down. A good share of padding disappears on its own when you speak a little more slowly, because thought stops lagging behind. It's the intervention with the best effort-to-result ratio.
5. Breathe through your nose between sentences. Besides giving you air, it marks a natural pause and prevents the rushed start. Breathing is the metronome of speech.
6. Prepare your openings. Many fillers cluster in the first second of each contribution, when you don't yet know how to begin. Having a clear first sentence removes a surprising share of the total.
7. Ask for a signal. In a recurring meeting, someone you trust can give you a discreet gesture each time your filler appears. Immediate feedback speeds learning enormously.
What not to do
Obsessing in real time. Monitoring every word while you speak consumes the attention you need for thinking, and the result is stiff, worse speech. The work happens in practice; in the real situation, you just speak.
Chasing zero. Speech with no hesitation at all sounds artificial, almost read aloud. Natural hesitations lend credibility and humanity. The goal is to turn down the volume on padding, not to eradicate it.
Correcting other people. Unless they ask, pointing out someone else's fillers is an efficient way to make them speak worse: they get nervous and produce more.
When it actually matters
In a conversation among friends, not at all. In a presentation, a job interview, a class, or a recording, it does: those are contexts where the listener is evaluating and where excessive padding gets read as not knowing your subject, even when you do. It's worth saving the effort for those occasions rather than monitoring yourself all day.
Final thought
Speaking well isn't about filling all the available time with sound. It's about saying what you mean and letting it be understood, and that requires space. Record yourself once, identify your dominant filler, slow down a little, and learn to stay quiet for two seconds without feeling the world is ending. That silence that feels eternal to you is, to the person listening, simply a sentence coming to an end.