Shadowing in English: what it actually trains (and what it doesn't)

You repeat an English video half a second behind the speaker, and you feel like you're training to speak. That's how YouTube sells shadowing: as a shortcut to fluency. The research tells a narrower and more useful story.

Shadowing means repeating out loud what you hear almost at the same time as the speaker, as if you were their shadow. It is, first and foremost, an ear drill. It trains you to catch English sounds in real time and to cut the continuous stream into words. Only once that works does it start paying off for pronunciation, intonation and speech rate. On its own, it's neither a comprehension method nor a conversation method: repeating someone else's sentence doesn't teach you to build your own. And it works best with audio whose content you already know, with no script in front of you.

What shadowing trains: your ear before your mouth#

Shadowing didn't come from polyglot channels. It came from training simultaneous interpreters. Yo Hamada of Akita University, the method's leading researcher, traces it to a 1992 paper by Lambert and explains that the technique reached Japanese English classrooms in the 1990s. Lambert's definition, as quoted by Hamada, is "a paced, auditory tracking task which involves the immediate vocalization of auditorily presented stimuli."

Hamada warns against the most popular reading of the method. In a 2021 article in The Language Teacher, he writes that shadowing "is also optimistically misinterpreted as a speaking activity (Hamada, 2017). In fact, shadowing is used for bottom-up listening skill development," meaning you recognize sounds and words before meanings.

The difference from classic listen-and-repeat is where your attention goes. When you repeat after a pause, you process meaning. When you repeat almost simultaneously, most of your attention goes to the sounds, Hamada explains, citing Kadota. Your ear has to decide on the fly where one word ends and the next begins.

Sound perception is the gain that shows up most consistently. After a study of 43 Japanese university students at low and intermediate levels, Hamada summed it up: "Statistical analyses indicated that phoneme perception was enhanced in both groups, but only low-proficiency learners improved their scores for high-school level listening questions."

The comprehension gain was narrower than it looks. On Hamada's advanced-level test, neither group improved, as Ryotaro Hashizaki and colleagues point out in a 2025 study of 30 Japanese university students. Those authors also warn that "the consideration of a learner's pre-test score as a proficiency indicator could suggest an overestimation of shadowing effectiveness."

When it starts to help you speak#

Shadowing does improve pronunciation, but only once your ear is ready. Hamada puts it as a rule: move on to pronunciation only when you catch the sounds well, because otherwise the task overloads your cognitive capacity.

One of the few studies that measured spontaneous speech used advanced learners. Jennifer Foote and Kim McDonough had 16 ESL learners in Canada shadow short dialogues at least four times a week for eight weeks. Twenty-two English speakers rated their spontaneous speech: they sounded more comprehensible and more fluent, but not less accented.

The broadest review confirms the pattern and marks its limits. A systematic review of 44 studies, Benen Whitworth's master's dissertation at Oxford (2024), found positive effects on comprehensibility, fluency and prosody. On individual sounds, Whitworth writes that the research "was, however, inconclusive." And of 31 studies classified by task, 20 measured only controlled tasks, such as reading aloud.

Gains in free speech are small. Yuka Muraoka followed 16 female Japanese university students through 11 weeks of shadowing classes, with no control group. Of three fluency measures, only speech rate improved: from 82.75 to 91.88 syllables per minute, a small effect (d = .38). Two earlier studies she reviews found no significant gains. Shuhei Kadota's position, which argues in his 2019 book for effects on speech production, runs ahead of the data.

None of these studies shows that shadowing teaches you to compose what you want to say. You repeat other people's sentences, and in a conversation you have to build your own in seconds. Could you say "I'm late because I missed the train" in English right now, without thinking? That's a different skill, and it's the one Tabla's exercises train: you see the sentence in Spanish and say it out loud in English before time runs out.

What the research still doesn't know#

The evidence base is small and heavily Japanese. The Oxford review counted 20 studies from Asia, 10 of them in Japan, and in 17 of the 44 the participants' first language isn't even stated. The review's table of first languages lists no group of Spanish speakers. Whitworth calls for "research exploring shadowing amongst learners with a wider range of L1s, L2s, and nationalities."

The only study with Spanish speakers we found is modest. In Cuenca, Ecuador, about 46 A1 schoolgirls aged 12 to 14 were split into a shadowing group and a control group and measured with the listening section of Cambridge's A2 Key. The shadowing group's edge was significant in only two of the five parts, the short conversations, and not on the longer recordings.

The transcript is where popular advice and the research split. Many polyglot blogs tell you to read it while you repeat. The academic tradition says the opposite: Hamada insists that you shadow without the transcript in front of you and save it for intermediate steps.

How to shadow without overloading#

The main rule is to know the content before you repeat it. Hamada puts it this way: "If they are asked to shadow an unknown story, it is too demanding and difficult because multiple processes in their brains will lead to cognitive overload."

Audio difficulty is the other lever. The classic rule is to use material at or slightly below your level, but in a study of 59 Japanese university students, Hamada found that alternating easy and hard audio improved comprehension more than sticking to one level. Start easy, then stretch.

A five-step routine based on Hamada's rules looks like this. Listen to the audio. Read the transcript to understand the content. Repeat under your breath, almost mumbling. Shadow out loud without the text, checking the transcript only to review what you missed. Repeat until you're in sync: according to Shiki and colleagues, as cited by Hamada, five or six repetitions of the same passage are enough.

The dose is short: Hamada suggests two 10–15 minute sessions a week for a month. To understand more, repeat more. In the 2025 study, sound processing improved regardless of the number of repetitions, but comprehension rose with more, and the authors propose going past five when comprehension is the goal.

Use shadowing to sharpen your ear with audio you already understand, and train conversation separately.

Try this today: pick a one-minute clip you already understand, shadow it five times without the transcript, and record the fifth round so you can mark the words you swallowed in the text.