Listening mode: hear a word in a sentence before you say it

30-second version: Asking someone to pronounce a word they've never heard is asking them to guess — and English spelling makes that guess wrong often enough to matter. sayit's listening exercises fix the order: multiple-choice comprehension questions built from real TTS audio put unfamiliar vocabulary in your ear, inside a full sentence, before any speaking is asked of you.

There's a sequencing mistake baked into a lot of vocabulary practice: show a written word, ask the learner to say it, then reveal how it's actually pronounced. That order guarantees a wrong first attempt on anything genuinely unfamiliar, because English spelling is a poor predictor of sound — you're testing a guess, not a skill. Flip the order — hear it first, in context, then attempt it — and the same word becomes something you're reproducing, not inventing.

What listening mode actually does

Listen includes comprehension-style listening exercises: a question is generated from real TTS audio, you're given multiple-choice options, and you answer based on what you heard — not what you read. Because the audio is synthesized speech (an American voice) rather than a static recording, the same underlying content can also be replayed at a slower rate when a sentence doesn't land the first time, letting you isolate whether a miss was a vocabulary gap or a raw-speed issue.

The exercises put vocabulary inside full sentences rather than as isolated words, which matters for a reason beyond realism: a word's sound can shift slightly depending on what surrounds it — a stressed word in isolation is not identical to the same word reduced inside a fast sentence. Hearing vocabulary in sentence context is closer to what you'll actually need to recognize in real conversation than hearing the same words read one at a time.

Why hearing-before-saying matters specifically for vocabulary

Three concrete failure modes this order prevents:

• Wrong stress locked in early. Say a five-syllable word based on a spelling guess, get it wrong, and that wrong stress pattern is now the version your mouth defaults to — harder to unlearn than to have gotten right the first time.
• Vowel quality guessed instead of heard. English vowel letters don't map consistently to sounds; hearing the word removes the guess entirely.
• No sense of the word's natural rhythm in a sentence. A word said correctly in isolation can still sound foreign inside a sentence if you've never heard how it compresses or connects to its neighbors — see connected speech and why vocabulary sounds different in sentences.

A routine that uses the sequencing on purpose

1. 1.Run a listening set before you attempt new vocabulary out loud. If you're about to work through a curated list like business English or academic vocabulary, a quick listening pass on unfamiliar-looking words first means your first spoken attempt is a reproduction, not a guess.
2. 2.Slow down anything you miss, rather than moving straight to a second guess at full speed — isolating the vocabulary gap from the speed problem tells you which one you actually have.
3. 3.Follow a listening pass with production, immediately, while the sound is fresh — read the same words aloud in Read & listen right after, closing the loop from hearing to saying within the same session.
4. 4.Use shadowing for the sentence-rhythm layer. Once individual words are solid, shadowing trains the connected, real-time version of hearing-then-producing, at full sentence speed.

The relationship to dictation

Listening mode and dictation test adjacent but different things — listening mode checks comprehension (did you understand what was said), dictation checks transcription (can you reproduce it in writing). Both precede speaking in the same useful way: hear it correctly before you're asked to produce it.

Try it: open Listen, run a listening set, and notice which words you'd have mispronounced if asked to say them cold, before you ever heard them.