The Shadowing Technique: How to Do It Right (Free Tools)

If you could do only one daily exercise for your accent, shadowing would be a strong candidate. It’s free, it takes ten minutes, and it trains the things that are hardest to learn from books: rhythm, stress, melody and linking, all at once.

What shadowing is

Shadowing means listening to a native speaker and speaking along with them, a fraction of a second behind, like a shadow. You don’t wait for the sentence to finish and then repeat it. You copy it as it happens.

That’s the key difference from “listen and repeat”. When you repeat after a pause, your brain has time to switch back to reading mode and rebuild the sentence with your own rhythm. When you shadow, you’re forced to copy the speaker’s timing and melody, because there’s no time to do anything else.

Classroom research on shadowing, much of it with Japanese university learners of English, reports improvements mainly in listening comprehension ^[1]. For advanced speakers working on accent, it’s also a direct way to practice the prosodic features that most affect how easily listeners understand you ^{[2] [3]}.

What you need

• A short native recording with a transcript. 15 to 60 seconds is ideal. Longer is too much to master.
• Headphones, so you can hear the original over your own voice.
• A way to record yourself. Your phone’s voice recorder is enough. Audacity is free if you want to see the waveforms.

Good free sources of recordings with transcripts:

• TED Talks (interactive transcript, clear speakers)
• BBC 6 Minute English (British, conversational)
• VOA Learning English (American, slower)
• LibriVox audiobooks with the free text
• Any YouTube video with accurate captions

Choose a speaker whose accent you’d like to have and whose style matches how you need to speak. If you give presentations, choose a presenter. If you need small talk, choose a podcast conversation.

The method, step by step

Step 1: Listen without reading (2 times)

Just listen. Notice where the speaker speeds up, slows down, and which words jump out.

Step 2: Mark the transcript (5 minutes)

Print or copy the transcript. Mark:

• stressed words (underline or capitals),
• the focus word of each phrase (circle it),
• pauses (a slash /),
• links between words (a small arc ‿).

This step turns the invisible music into something you can see. Our guides on sentence stress, intonation and linking explain what to look for.

Step 3: Mumble along

Play the clip and hum or mumble along without real words, just copying the melody and beat. This is surprisingly hard at first, and surprisingly useful.

Step 4: Shadow with the transcript

Now speak the words along with the speaker, reading your marked transcript. Don’t worry about perfect sounds. Prioritize: timing first, melody second, sounds third.

Step 5: Shadow without the transcript

Once you can keep up, try it without reading. This is where it starts to feel like speaking rather than reading aloud.

Step 6: Record and compare

Record yourself saying the clip alone, without the audio. Listen to the original and then yours. Pick one difference to fix (for example, “my focus word is too weak” or “I separated turn it off”). Fix that one thing on the next round. Don’t try to fix everything at once.

Step 7: Repeat for several days

Stay with the same clip for three to five days until you can say it naturally. Depth beats variety. Then move on.

Common mistakes

• Clips that are too long. A five-minute clip turns into reading practice. Keep it short.
• Material that’s too fast or too hard. If you can’t follow the meaning, you can’t copy the melody. Choose something you understand at first listen, or almost.
• Copying sounds instead of rhythm. Beginners at shadowing obsess over the TH and lose the beat. The beat matters more.
• Never recording yourself. You can’t hear your own prosody while speaking. Recording is not optional.
• Not slowing down when needed. If a clip is too fast, slow it to 0.75x speed in YouTube settings or in Audacity. Keep the connections even at slow speed.

A 10-minute daily routine

1. Two listens without reading (1 minute)
2. Mark or review the transcript (2 minutes)
3. Three rounds of shadowing with transcript (3 minutes)
4. Two rounds without transcript (2 minutes)
5. One recording and comparison, one fix noted (2 minutes)

Do this five days a week and keep your recordings. After a month, listen to your first and latest recording of the same kind of material. The difference is usually obvious.

Tools that make it easier

• YouGlish for shadowing a single phrase said by many speakers.
• Language Reactor for looping lines in YouTube videos.
• Praat if you want to compare your pitch line with the original.
• Speechling for a human coach to check a few recordings a month for free.

Next step

Shadowing works best when you know what to listen for. If you haven’t yet, read why advanced speakers still sound foreign, and if your first language is Spanish, the pronunciation habits Spanish speakers carry into English.