When should you actually start speaking a new language?

Do you think starting to speak with native speakers too early, or any type of speaking practice, could harm your progress? I put that question to Hassan, the founder of Arabic All the Time, who learned Spanish almost entirely through listening. We disagree on some of it. Here's where each of us ended up.

Hassan's view: wait before you speak

Hassan's answer was that it can be bad, though not always. The danger is pronunciation. Every language has sounds we don't have in our native language. If you try to speak before your brain has acquired those sounds, it relies on the existing infrastructure, which is mainly your native language, to produce the new ones. That causes problems he believes are very hard to fix later.

Where speaking early can help is motivation. And sometimes we want or need to communicate, for example with a partner, or when travelling to a country where we're put in situations where we have to speak. But he's against forcing yourself to speak too early just because you think it will make you better in the language. It will help your speaking in the short term. If your goal is lasting fluency and you care how you'll sound in two or three years, it may be wiser to delay speaking a little.

On how long: some people in the comprehensible input world say you need thousands of hours of input before you speak. He thinks that's too extreme, because perfectionism risks losing motivation and quitting before there's anything to perfect. His own number is 400 to 500 hours of input, and he'd be against any speaking attempt before about 300, whatever the language.

He also said it matters even more for a language closely related to one you know. If you learn Portuguese after Spanish, your brain mixes the sounds and you end up pronouncing Portuguese in a Spanish way.

My view: speak as early as you feel comfortable

My opinion is that it's good to practise speaking as early as possible, even if it's grammatically bad.

After the first four weeks of learning a language, you lose motivation massively. If you start speaking in that period you can use it in real life, and I want evidence that what I'm learning is working. The best evidence I can think of is sitting in a restaurant, saying to the waiter in Spanish, "Could I have this, please? Is this spicy?" and they understand me.

When I was learning Italian to visit my partner's family in Italy, I wanted to communicate with them. Even if I said "I want to go to the park tomorrow" with really bad grammar, they understood me. We laughed about my bad Italian, but I felt part of their social circle.

I also found out the hard way about similar languages. I learned Spanish first, then Italian, and every other word out of my mouth was a Spanish word. My Italian teacher asked if I was learning Spanish too and said he wouldn't recommend it. I was like, "I'll show you." I learned that the hard way. I can see why he'd say input first in that situation.

Where we agree

We ended up realising we're talking about two different objectives.

• A trip or short-term goal. Hassan's point was that if you're planning a holiday in Italy and don't plan to make Italian part of your life, spending ages building comprehension has a low return on your time. You're better off memorising phrases and practising saying them. And ordering a coffee takes retrieval and speaking practice: you can have 1,000 hours of input and still not manage to order coffee if you've never tried to speak.
• Making the language part of who you are. If you want lasting fluency, patience pays off. Input is the cornerstone. Your output can't outperform your input. His rough split is 80 to 90% input and 10 to 20% speaking practice, to activate and retrieve what you've acquired.

He also pointed out that we're both speaking from the point of view of someone who doesn't live in the country, has a job and is learning on the side. For that person, this applies.

How do you know when you're ready?

Hassan's sign: if you can follow a conversation between two slow speakers, or a talking head, with little visual support, you're ready to give it a go. If you've built enough vocabulary to understand intermediate content, you've built enough to hold a basic conversation. He suggested building up gradually if you have the luxury.

His advice: follow your heart. If you feel like speaking, do it. Don't force yourself before you're ready, because that's where you become anxious. But if you've received lots of input and understand a lot and you're stuck and can't speak, that's where you need some forcing. Travel to a country where it's spoken, or hire a tutor who only speaks to you in the language. It's the only thing that gets you unstuck.

I'd add that your environment matters. If you put pressure on yourself to speak, and it backfires because the native speaker is in a rush or it's noisy, it can make you scared of practising again. In English I couldn't hear someone either if it was really noisy. Consider your environment when you start.

Two things we both agreed on

You'll always understand more than you can speak. This is true in every language, including your own. There are English words you understand or can guess but never use. Understanding requires key vocabulary. Producing the language requires conjugating correctly, being specific, using connectors. Let go of the pressure to match them.

Don't wait for perfect grammar. Even native speakers make grammatical mistakes. If you're ready to speak and the grammar isn't 100% there yet, don't stop yourself from practising, or you'll never start having conversations.

If you're scared of speaking with humans, you can practise conversations with Voco first.