Why you understand more than you can say

This is so common that it barely registers as a puzzle. But it is a puzzle. If you know the words well enough to recognise them instantly, why can't you produce them? The answer is that understanding and speaking are not the same skill run in two directions. They are two different skills that happen to share a vocabulary.

Note 1. Listening lets you cheat

the key idea!When you listen, context does half the work. You hear "...train... late... sorry..." and you fill in the rest. You can understand a sentence while missing a third of it. When you speak, there is nothing to fill in for you. You need every word, in the right form, in the right order, right now. Speaking exposes exactly the gaps that listening lets you skip over.

In 1985 the Canadian linguist Merrill Swain made a version of this argument, which became known as the output hypothesis. She had studied children in French immersion programmes who had years of rich listening and reading in French, and who understood it very well, but whose spoken and written French still had persistent errors. Her suggestion was that producing language forces you to notice what you do not know. Trying to say something, and finding the hole, is itself a kind of learning.

Note 2. The real block is usually fear

Ask learners why they don't speak more and few say "I don't know enough words." They say "I'm embarrassed," "I'll sound stupid," "I don't want to waste their time."

Researchers in language education have a name for the thing that varies here: willingness to communicate. Two learners with the same level of skill can differ hugely in how often they actually open their mouths, and the one who speaks more improves faster. Confidence is not a reward for being good. It is closer to a precondition for getting good.

Note 3. Mistakes are the material

A conversation where you make no mistakes taught you very little. A conversation where you got stuck, reached for a word, made something up, and were gently corrected, that one rewired something. The goal is not to avoid errors. It is to make them out loud, where they can be noticed.

One useful habit: keep a small list of things you tried to say and couldn't. After each conversation, look up three of them. Those words will stick better than anything from a textbook, because you needed them.

Note 4. A four-week routine

• Week 1Shadowing, ten minutes a day. Play a short clip and speak along with it, a fraction behind, copying rhythm and melody rather than worrying about meaning.
• Week 2Talk to yourself. Narrate what you are doing while cooking or walking. Nobody is listening, so nobody can judge.
• Week 3Three short real conversations. Five minutes each is enough. Write down the gaps afterwards.
• Week 4One longer conversation on a topic you care about. Notice how much of week three's gap list you used without thinking.

Note 5. Who to practise with

Not just teachers. A patient native speaker is great, but so is another learner, who will be slower and more forgiving. Strangers have a particular advantage: if the conversation goes badly, you never have to see them again. Plenty of learners find a five-minute chat with someone they will never meet less frightening than one with a classmate.

One caution: a random conversation partner is not a tutor. They may correct you wrongly, or not at all, and accents and slang vary enormously. Treat them as practice, not as an authority.

Note 6. An open question

Children learning their first language spend roughly a year listening before they say much at all, and nobody expects them to speak correctly. Adults give themselves weeks, then decide they are bad at languages. Is the difference really ability, or is it just that adults are far less willing to sound like beginners?

→ next entry: the words I couldn't find this week.