What shadowing is, in plain English
Shadowing is speaking along with a model audio just behind the speaker, copying timing, melody, and sounds. A script is a written version of that audio aligned line by line. Prosody means sentence music such as stress and intonation. Linking is smooth connection between word endings and beginnings. Chunking means breaking speech into short thought groups of 4 to 7 words.

Why a 2-3 minute script works
Short scripts let you finish a full cycle every day: listen, mark, shadow, record, review. Repetition builds muscle memory for sounds, stress, and pace without fatigue. With targets, you can measure progress in a week.

The 5-step daily loop (2-3 minutes of audio, 12-15 minutes total)

1. Select a clean clip at 0.9 to 1.1x speed from news or talks. Avoid heavy music or crowd noise.
2. Mark the text: add slashes for thought groups, caps on one focus word per group, arrows ↑ or ↓ for rises or falls.
3. Shadow cold once with the audio. Stay one beat behind. Do not pause the clip.
4. Shadow focused twice: first for sounds (problem phonemes), second for stress and linking.
5. Record and review one pass. Log two errors to fix tomorrow.

What to mark on the script

• Focus words: the single most important word per group.
• Tonic syllable: the syllable that carries the pitch change inside the focus word.
• Linking cues: mark vowel-vowel links (go‿on), consonant-vowel links (pick‿it), and stop releases (act).
• Breath points: only at group boundaries.

Two model scripts you can try today

Example 1 — Information tone
/ the BUS lanes ↑ / have CUT travel TIME / for many COMmuters ↓ / in the city CENTER ↓ /.
Coaching notes: rise on lanes to show more is coming, final falls to sound complete. Focus words BUS, CUT, COMmuters, CENTER. Link have‿cut, many‿commuters.

Example 2 — Balanced opinion
/ onLINE classes ↑ / are CONvenient for WORKers ↓ / but the FEEDback / is OFten THIN ↑ / in LARGE groups ↓ /.
Coaching notes: contrast marked by but. Keep one clear peak per group. Link convenient‿for, feedback‿is.

Micro-phoneme tuning for advanced users

• /θ/ vs /t/: think vs tink. Place the tongue lightly between teeth. Start with thin, growth, healthy.
• /ɪ/ vs /iː/: live vs leave. Shorten /ɪ/ and relax the jaw.
• Final stops /p t k b d g/: release softly before a vowel link, hold before pause. Practice map, fact, back, cab, road, bag.

Metrics that keep you honest

• Words per minute of your recording: aim for 110 to 150 with steady groups.
• Peaks per group: one peak only. Score yourself out of total groups.
• Link-rate: percent of marked links you actually make. Target 70 percent week 1, 85 percent week 2.
• Filler count: ums per minute. Target under 2, then under 1.

Mini case — Nawar in Dhaka
Starting point: flat melody, rushing ends, and 8 fillers per minute.
Plan: one 2-minute news clip for 14 days. Day 1 to 4 focus on chunking and peaks, day 5 to 8 on linking and weak forms, day 9 to 14 on contrastive stress.
Results after two weeks: average group length 5 to 6 words, peak-per-group accuracy 85 percent, fillers down to 2 per minute, clearer final falls in Part 2 answers. Mock speaking rose from 6.0 to 7.0.

Drills with time boxes and targets

• 30-second loop: shadow 30 seconds three times at 0.9x, 1.0x, 1.05x. Keep one-beat lag.
• Contrast pairs 2 minutes: GREEN cards vs green CARDS, onLINE vs ONline depending on noun or adjective use.
• List contour 1 minute: rise on items 1 to n minus 1, fall on last.
• Link ladder 2 minutes: read go‿on, take‿it, pick‿up, hand‿over, build‿in. Count correct links.

Common mistakes

• Chasing words, losing music: copying every consonant while ignoring stress. Fix by marking one focus per group.
• Breathing mid word: plan breaths at slashes only.
• Over-linking: connecting when a pause is needed for meaning.
• Flat endings: rising at every full stop. Add deliberate final falls on statements.
• Too hard input: accents or domain jargon beyond your range. Choose clear sources first.

Edge cases and safe fixes

• Fast speakers: start at 0.9x speed, then step up.
• Multiple accents: work one accent for a week before rotating.
• Weak forms in numbers: say three‿and‿a‿half with light and a.
• If you stumble: stay with the audio, skip a word, rejoin at the next group.

Tips and tricks

• Track with a pen. Tap once on the desk for each focus word.
• Smile on positive lines. It lifts pitch and improves resonance.
• Record in a quiet corner and listen with one earbud while speaking out loud.
• Keep a personal error bank: three sounds, three words, three links you miss often.

To avoid

• Memorising whole monologues without audio. Shadowing needs a live model.
• Overusing tongue twisters instead of connected speech.
• Choosing clips longer than 3 minutes. Completion beats ambition.
• Editing every tiny slip before finishing the pass. Flow first, surgery later.

Glossary
Shadowing: speaking with a model audio slightly behind it.
Prosody: melody and rhythm of speech such as stress and intonation.
Linking: smooth connection across word boundaries.
Thought group: a short chunk of 4 to 7 words that carries one idea.
Weak form: reduced pronunciation of function words like to, of, and.
Tonic syllable: the syllable that carries the main pitch change in a group.

Next steps
Pick one clear 2-minute news paragraph. Print or transcribe it. Mark slashes, focus words, and arrows. Do one cold pass, two focused passes, and one record-and-review. Log words per minute, peak accuracy, link-rate, and fillers. Repeat with the same script for three days, then switch to a new one.

1. Actionable closing — Cheatsheet

Daily loop (12-15 minutes total)

• 2 min listen and mark
• 6 min shadowing passes: cold, sounds, stress and linking
• 3 min record and review
• 1-2 min error log update

Marking guide

• One focus word per group
• ↑ for rise, ↓ for fall
• Link vowels and consonant-vowel pairs

Quality targets

• 110 to 150 wpm
• 1 peak per group
• 70 to 85 percent link-rate
• Fillers under 2 per minute

Fix-it cues

• Breath only at slashes
• If lost, rejoin at next group
• Shorten weak forms inside groups

Rotation plan

• Days 1-3 sounds
• Days 4-6 stress
• Days 7-9 linking
• Days 10-14 contrast and lists

CTA: Start a 14-day streak with one 2-minute script. Track words per minute, peak accuracy, link-rate, and fillers after each session. Aim to raise link-rate to 85 percent and cut fillers below 1 per minute by day 14.