Should You Correct Your Child's English Mistakes? What Actually Helps (Ages 1–5)

A mother sent me a voice note last winter that I have thought about many times since. Her son, four, had run into the kitchen and announced: "Mama, I want apple!" She had said, gently and with the best intentions, "An apple, sweetheart. AN apple."

He said nothing else in English for the rest of the day.

She was not doing anything cruel or unusual. She was doing what almost every parent does, what most of us were taught was helpful, and what feels like the responsible thing: hearing an error and fixing it. And yet the exchange cost her something she could not get back that afternoon, and she knew it the moment it happened.

This article is about that decision — the one you make several times a day, in half a second, with no time to think. Your child says something in English that is not quite right. Do you correct it?

The answer is more interesting than yes or no, and it depends on something most advice never mentions: which language your child's brain is borrowing the rule from.

The short answer — and the trade-off behind it

Do not correct your child directly. Say the correct version back inside a normal reply, and carry on.

That is the whole instruction, and it takes about four seconds to learn. What takes longer is believing it, because it runs against a strong instinct: if you can see the error and you say nothing, surely you are letting it harden?

Here is the trade-off, stated plainly. Direct correction does buy you a small, immediate improvement in accuracy — your child may well produce the right form on the next attempt. What it costs is output. A child who is corrected produces less language, takes fewer risks, and increasingly chooses the safe sentence over the interesting one. And at ages one to five, output is the entire engine. Accuracy is a by-product of volume; volume is not a by-product of accuracy.

So the question is not "which method fixes the error faster". It is "which method leaves my child still talking in six months". Those have different answers, and only one of them matters at this age.

Why your child makes these particular mistakes

Open almost any article about children's English mistakes and you will find a list of errors treated as random noise to be cleaned up. That framing is wrong, and it is the reason most correction advice backfires.

Your child's errors are not random. They are systematic, and they are almost all traceable to one source: the language they already speak. Researchers call this transfer — the learner applies the rules of a known language to the new one. When a child from a language without articles says "I want apple", that is not sloppiness. It is a rule being applied faithfully. It happens to be the wrong rule for English, but the machinery is working exactly as it should.

This matters enormously for how you respond. If you believe the error is carelessness, correction feels like the fix. If you understand it is a rule collision, you can see that the child does not need to be told they are wrong — they need more evidence of how English does it. Those two conclusions lead to completely different behaviour in your kitchen.

It also means the errors are predictable. Because Ukrainian and English differ in a small number of specific ways, you can know in advance almost exactly which mistakes your child will make — which means you can prepare your replies instead of improvising them.

Six "mistakes" that are really Ukrainian showing through

Here are the six you will hear most, why each one happens, and the reply that does the work without the correction.

Read the right-hand column again and notice what is missing from every single reply: the word "no", the word "wrong", and any instruction to repeat. Your child hears the correct form immediately after their own version, attached to the same meaning, in a conversation that keeps moving. That adjacency is the entire mechanism.

The errors that are actually good news

One category deserves its own section, because parents almost always read it backwards.

At some point your child will say "goed", or "comed", or "foots". Very often this happens after a period of saying "went" and "feet" correctly. It looks unmistakably like going backwards, and mothers write to me convinced that something has broken.

The opposite is true. Early on, a child says "went" because they memorised it as a lump, the way they memorised "byebye". Later, they extract the rule — English marks the past with -ed — and immediately apply it everywhere, including where it does not belong. "Goed" is the sound of a rule being discovered. It is a genuine leap in linguistic competence, and it is the reason researchers describe this pattern as U-shaped: correct, then apparently wrong, then correct again on a deeper basis.

If you correct "goed", you are correcting the single most impressive thing your child has done that month. Recast it — "You went to the park! Who did you see?" — and let the machinery finish its work.

What a recast is

A recast is repeating what your child said back to them in the corrected form, without flagging it as a correction. The term comes from language-acquisition research, and it describes something most attentive parents already do occasionally by accident.

Child: "I want apple."
You: "You want an apple? Here's an apple."

That is the whole technique. What makes it powerful is what it does not do. It does not interrupt. It does not require your child to repeat anything. It does not mark the child's sentence as failed. From the child's point of view, they said something, they were understood, and the conversation continued — which is exactly the experience that makes a child speak again tomorrow.

What is happening underneath is a comparison. Your child holds their own version in mind for a second or two, and hears yours immediately afterwards attached to the same meaning. Recasts work through that contrast, and their effectiveness depends on staying implicit — the moment a recast starts to feel like a correction, it behaves like one.

One caveat I want to give you honestly, because you may read otherwise elsewhere. There is a body of research showing that explicit correction outperforms implicit feedback for second-language learners. That work is almost entirely about older learners in classrooms — teenagers and adults, in lessons, who can hold a grammar rule in conscious memory and apply it. It does not describe a three-year-old acquiring English naturalistically at the kitchen table, and applying it there is a category error.

The recast in four moves

If you want a repeatable version you can run without thinking, this is it.

1. Answer the meaning first. Before anything else, respond to what your child was actually trying to do. They wanted the apple. Give them the apple. The message being received is the reward that keeps them talking.
2. Say their sentence back, corrected, as part of your reply. Not as a separate statement — folded into a normal answer. "You want an apple? Here's an apple."
3. Stress the missing piece very lightly. A touch more weight on "an", nothing theatrical. Enough for the ear to catch, not enough to feel like a lesson.
4. Move on immediately. No pause for repetition, no expectant look, no "can you say it?" The moment you wait for them to produce the corrected form, you have converted a recast into a test.

Four moves, about two seconds, and it fits inside conversations you are already having. There is no extra time cost — which is exactly why it survives on the days when nothing else does.

When direct correction is the right call

I do not want to leave you with an absolute rule, because absolutes are how good advice turns into dogma. There are moments when saying it plainly is right:

• Your child asks. "How do you say it?" is an explicit request, and it deserves an explicit answer. A child who asks is ready to hear it.
• The child is older and asking for accuracy. Around school age, some children begin to want to get it right and will say so. Follow that lead when it appears — but let it come from them.
• A misunderstanding actually matters. If the confusion is about something that affects safety or a real outcome, clarity beats delicacy.
• The same fossilised form has persisted for months despite plenty of correct input, and your child is four or five and secure in speaking. Then one light, cheerful, non-repeating comment is reasonable — once, not as a campaign.

Notice that three of the four are led by the child. That is the reliable signal: correction lands well when the child has asked for it, and badly when they have not.

What correcting actually costs

It is worth being concrete about the price, because "it might discourage them" is too vague to change anyone's behaviour in the moment.

The bottom row is worth pausing on. Whatever your instinct in the moment, correcting a child's English in front of a grandparent, a sibling or another parent costs several times what the same correction costs alone in the kitchen. If you take away one prohibition from this entire article, make it that one.

Ready-made recasts for the moments you will hit

The reason correction wins in real life is that it is the fastest thing to say when you have not planned anything. The fix is to have the replies ready. These are the situations that generate the most errors in a normal day.

Read them aloud once and you will notice they all sound like ordinary conversation, because they are. Nobody listening from the next room would be able to tell that any teaching just happened. That is the standard to aim for.

Sounds are not grammar — different rules

A short but important distinction, because parents often apply one policy to both.

Grammar errors are worth recasting, gently and often. Pronunciation at this age is worth leaving alone almost entirely. A three-year-old's mouth is still developing, and many sounds that look like errors — the English th, the difference between the vowels in ship and sheep — are not yet physically or perceptually settled even in children growing up monolingual.

Correcting a sound your child cannot yet produce teaches only that their mouth is a problem. Model it clearly in your own speech, let the ear do the work over months, and read which pronunciation mistakes actually matter for the small number that genuinely do — including the ones you may be making yourself.

Why children end up correcting themselves

Here is the part that makes the whole approach feel less like passivity.

Children revise their own grammar. Not because they are told, but because the correct form keeps arriving, and the mismatch between what they say and what they hear eventually becomes impossible to ignore. The mechanism is comparison, and comparison needs one thing above all: a high volume of correct input attached to meanings the child cares about.

Which reframes your job entirely. You are not the error-detection system. You are the supply of correct forms — and every recast is one more delivery, arriving at the precise moment the child is most primed to notice it, because they have just tried to express that exact meaning themselves.

You will hear it happen. A child says "I goed", stops, and says "I went" without anyone intervening. That self-repair is what you are building toward, and it is far more durable than a form produced because an adult asked for it.

What to expect at 2, 3, 4 and 5

The same policy applies throughout, but what you will be hearing changes.

If your child is at the age where they produce very little English at all, the correction question is not yet the one to solve — understands but won't speak and only repeats after me deal with the stages that come first.

A one-week plan to change your reflex

Knowing the technique changes nothing on its own. The correction reflex is fast and automatic, and it needs to be replaced rather than suppressed.

1. Days one and two: only notice. Correct as much as you normally would, but keep a count on your phone. Most parents are surprised — the honest number is usually somewhere between eight and twenty a day.
2. Day three: pick your two most common errors from the table above and prepare exactly one recast for each. Two, not six. You are building a reflex, not a repertoire.
3. Days four and five: run those two only. Everything else, let go entirely — even errors that bother you. One fortnight of ignoring an error costs nothing; one fortnight of correcting it can cost months of willingness.
4. Day six: add the pause. After your child speaks, count one full second before you reply. That gap is where the correction reflex used to live, and it is also where children often self-correct.
5. Day seven: count again. Compare with day one. What you are looking for is not zero corrections — it is more recasts than corrections, and more talking from your child than last week.

Frequently asked questions

These come up in nearly every conversation I have about correction. The thread running through them is the same: at this age your child's willingness to speak is the scarce resource, and accuracy is the thing that arrives on its own if willingness survives.