Speaking Before You're Ready: The Case for Conversation Practice

Ask any adult French learner when they intend to start speaking and you get the same answer, phrased a hundred ways.

Once I've got a bit more vocabulary. Once I've finished this course. Once I don't sound like an idiot. Soon. Not yet.

It is a completely reasonable plan. It is also, empirically, the plan that produces people who have studied French for a decade and cannot speak it.

Why the plan fails

Because the readiness you are waiting for is manufactured by speaking. It does not accumulate quietly in the background while you study, waiting to be discovered at a sufficient threshold. There is no threshold. There is no moment at which the fear resolves on its own.

What actually happens is that the target moves. At two hundred words you tell yourself you need five hundred. At five hundred you decide you need the subjunctive. The bar rises to meet you, forever, because its real function is not preparation. Its real function is postponement.

What the research says about output

For a long time the dominant theory held that language is acquired chiefly through input — that if you understand enough, speech follows. It is a comfortable idea and it is incomplete.

The correction came from studying immersion students who had years of rich input and still spoke imperfectly. The proposal, the output hypothesis, is that something distinct happens when a learner is forced to produce language. Understanding lets you skate past what you do not know. Producing does not. The moment you must say the thing, you discover precisely which piece you are missing, and that discovery is what drives you to acquire it. Input builds knowledge; only output builds speech, and they are not the same skill.

Related work on interaction makes the same point from a different angle. What matters is not merely talking but being misunderstood, noticing it, and repairing it. That negotiation — the pause, the confusion, the second attempt — is where the language is actually built.

You cannot be misunderstood by a textbook. You have to be misunderstood by a person.

What speaking badly actually costs

Almost nothing, and it is worth being concrete about this, because the imagined cost is what keeps people silent.

• Nobody is judging you. The French person you are stumbling in front of is mostly relieved you tried at all, and is thinking about their own day.
• Errors are not permanent. There is no evidence that speaking imperfectly fossilises mistakes in a learner who is actively being corrected. This is a widely repeated fear with remarkably little behind it.
• The embarrassment is brief. Six weeks, roughly. Then it stops. Every adult who has crossed this describes the same short, uncomfortable, entirely survivable passage.
• The alternative cost is enormous. Ten more years of understanding French and not speaking it.

What actually helps

• Speak in your first lesson. Badly. On purpose. Before you have any right to.
• Find someone who will not rescue you. The instinct to switch to English at the first stumble is kind and it is fatal. A good teacher lets the silence sit while you hunt for the word.
• Practise in a group of people who are also bad at this. Half of the fear is social, and it evaporates in a room where everyone is struggling equally.
• Say the wrong word on purpose. Describe the thing you cannot name. That workaround — the circumlocution — is not a failure of vocabulary. It is the core skill of every fluent speaker alive.
• Measure minutes spoken, not hours studied. It is the only number that predicts whether you will ever hold a conversation.

You will not feel ready. That is not a signal to wait. That is the feeling of being exactly where every fluent speaker once stood, on the wrong side of about six weeks of discomfort.