My Child Says "Tat" Instead of "Cat": Which English Sounds Toddlers Get Wrong, and When It Matters

The message always contains a quotation. "She says 'tat' instead of 'cat'." "He calls it a 'yeyow' car." "Everything with an S at the end comes out as a T." And underneath the quotation, always the same fear, usually unwritten: have I broken something by adding English?

I have taught English to young children for over nine years, and my first degree is in speech therapy. That combination means I get this question more than almost any other — and it also means I can tell you something most articles on this subject cannot, because they are written either by speech therapists who assume one language, or by language teachers who never studied speech development.

Here it is: the great majority of "wrong" sounds in a two-, three- or four-year-old are not English problems at all. They are the ordinary, predictable, well-documented way that every child's mouth learns to work — and they would be happening in exactly the same form if English had never entered your home.

Three completely different reasons a sound comes out wrong

Before you can judge anything, you have to know which of three things you are listening to. They look identical from the outside and they need opposite responses.

Almost everything that worries parents is category 1 or 2. Category 3 is real and I will give you a clear boundary for it further down, but it is much rarer than the internet suggests, and — this is the part that matters — it is not caused by bilingualism.

Which sounds arrive when

Children do not acquire all consonants at once. They acquire them in a broadly predictable order, driven by how visible and how physically simple each sound is. Sounds you can see on a face come early. Sounds made invisibly at the back of the mouth, or requiring fine tongue control, come late.

Read that last row again, because it dissolves the single most common worry I receive. A four-year-old who cannot say "th" is not behind. Many English-speaking children in England are still saying "fink" for "think" at six. If your child says "sink" or "fink", you are hearing a sound that has not arrived yet — in any child, in any family.

The five error patterns, and their real names

Children's mistakes are not random. They follow a small number of well-described simplification patterns, and once you can name them they stop sounding alarming. Each one has an age by which it usually disappears.

If you recognise your child in this table — and you almost certainly will — then what you are hearing is a mouth doing exactly what mouths do. The pattern is the evidence that development is proceeding normally, not the evidence that something is wrong.

The sounds that are genuinely harder in a Ukrainian-speaking family

This is the part no general speech-development article will tell you, because it depends on your first language. English contains sounds that Ukrainian and Russian do not, and the ear tunes itself to its home language in the first year of life. When a sound has no slot in that system, the brain files it under the nearest neighbour it does have.

Two things follow from this table, and both are good news. First, these are accent features, not errors — millions of fluent, successful English speakers carry them for life and are perfectly understood. Second, they are the sounds most worth a little playful attention in your own speech, because your child is learning them from you. I have written a separate guide to which of your own pronunciation habits are worth adjusting and which genuinely do not matter.

What a "sound" actually is, and why order matters

It helps to know what your child is really doing, because it reframes everything that follows. A speech sound is not a piece of knowledge. It is a physical movement — a coordinated sequence involving the lips, the tongue tip, the tongue body, the soft palate, the vocal folds and the breath, executed in roughly a tenth of a second, and repeated accurately every time the word appears.

Two things follow. First, sounds are acquired in the order that the mouth can physically manage them, which is why the sequence is so similar across children and across languages. Lips are easy: your child can see them, and they are under gross motor control from birth. The tongue tip is harder. The tongue body, pressing against the soft palate to make /k/ and /g/, is harder still, and it is entirely invisible — nobody can show a two-year-old what that movement looks like.

Second, and this is the part that reassures most mothers: hearing the difference comes long before making it. A child who says "tat" can almost always tell you which picture is the cat and which is the hat. If you say "tat" back to them deliberately, many three-year-olds will object, sometimes indignantly — while still producing "tat" themselves in the very next breath. That gap between perception and production is not confusion. It is the normal shape of motor learning, and it is the reason correction is beside the point: the child already knows. The mouth is catching up.

What to expect, age by age

Why correcting the sound does not work

The instinct is universal and it is wrong. Your child says "tat", you say "no, say CAT, kuh-kuh-cat", and one of three things happens: nothing, or the child repeats it once and reverts within the hour, or — most often — the child stops saying the word at all.

Direct correction fails for a mechanical reason. A three-year-old who says "tat" is not choosing the wrong sound; the motor pattern for a back-of-mouth /k/ is not yet reliably available. Asking for it is like asking a child who cannot yet hop to hop. Repetition of a movement they cannot make does not build the movement; it builds reluctance to try.

What does work is far less effortful, and you may already be doing it without a name for it.

The recast. Your child says the word wrong. You say the word right, inside a normal, warm sentence, without stopping, without pointing it out, and without asking them to repeat it. That is the entire technique.

Notice what the right-hand column does. It confirms the child was understood, it delivers the correct form two or three times in a row inside meaningful language, and it costs the child nothing. No performance is demanded, so nothing is risked, so the child keeps talking — and talking is the only thing that ever fixes pronunciation.

The one measurement that actually matters

If you remember a single sentence from this article, make it this one: judge the child, not the language.

A speech-sound difficulty lives in the child, so it shows up wherever the child speaks. A language-specific quirk lives in the language, so it shows up in one and not the other. This single distinction separates almost every false alarm from every genuine one.

• Your child says "tat" in English but says кіт with a clean к → the /k/ is available. This is not a speech problem.
• Your child cannot produce /k/ in either language at four → worth a professional opinion.
• Your child says "sink" for "think" but has no difficulty in Ukrainian → interference, entirely expected.
• Your child drops the ends of words in both languages at four → worth asking about.

This is also why any assessment must be done across all of a child's languages. A monolingual assessment of a bilingual child measures one half of the system and calls the missing half a deficit — which is how perfectly typical bilingual children end up wrongly flagged. If a nursery or a clinician raises a concern, the first question to ask is a polite and specific one: "Which languages was my child assessed in?"

A more useful yardstick than counting sounds

Rather than auditing individual consonants, clinicians often use a simpler, more robust guide: how much of what your child says can be understood by someone outside the family. Parents are unreliable judges here, because you have spent years learning your own child's private phonology and you decode it without noticing.

The rough clinical rule of thumb, widely used as a screening guide rather than a hard law, runs roughly like this: around half of a two-year-old's speech is intelligible to an unfamiliar listener, around three-quarters at three, and essentially all of it by four — even while individual sounds are still developing.

A practical version you can run at home: hand your phone to a friend or relative who does not live with you, play them a minute of your child talking, and ask them how much they caught. That single test tells you more than any list of sounds, and it removes your own trained ear from the equation.

When to ask a professional

I want to be precise here rather than reassuring, because vague reassurance is how real difficulties get missed. Seek an assessment — in all your child's languages — if any of the following is true.

• The same sound errors occur in both languages and persist beyond the ages in the table above.
• People outside the family regularly cannot understand your child at three or four.
• Your child is frustrated by not being understood, or has begun avoiding speaking.
• Your child lost words or sounds they previously had.
• There is no back-of-mouth sound at all (no k, no g) by about four.
• Something in your own instinct persists. That is a legitimate reason on its own.

None of these is a reason to drop English. Removing a language from a child who has a speech-sound difficulty does not make the difficulty smaller; it makes the child's world smaller. The evidence on this is consistent, and it is the single most common piece of bad advice these families receive.

What actually helps, and it is not drilling

Everything on this list is play. None of it is a lesson, and none of it requires your child to perform a target sound on demand.

1. Let them see your mouth. Sit face to face at eye level rather than side by side. Many "hard" sounds are hard mostly because they are invisible.
2. Use the sound, do not teach it. If /k/ is wobbly, spend a week where cups, cars, cats and kicking happen to come up a lot. Exposure inside meaning beats isolated repetition.
3. Slow down slightly and leave gaps. Not louder, not exaggerated — just a little slower, with a pause afterwards that invites a turn.
4. Sing. Songs stretch and repeat sounds naturally, and a child will produce in a song what they will not produce on request.
5. Play with mouths, not with words. Blowing bubbles, licking honey off the top lip, animal noises, silly faces in a mirror. This is motor practice with no pressure attached.
6. Recast, every time, without comment. The single highest-value habit in this article.
7. Keep talking to them as a person. Children fix their own pronunciation through volume of real conversation, not through correction.

Six mistakes I see most often

1. Dropping English "until the speech settles". The most common and the most costly. It removes input, removes turns, and removes the very practice that develops the mouth.
2. Making the child repeat a word three times. This converts a conversation into an examination. Repeated examinations produce silence.
3. Praising the sound instead of the message. "Good talking!" teaches a child that speech is a performance to be graded.
4. Comparing with a monolingual cousin. Different input, different distribution, and often a different total vocabulary spread across two systems.
5. Over-articulating. Exaggerated mouth shapes model something that is not real speech, and children copy what they hear.
6. Accepting a monolingual assessment. Always ask which languages were tested.

One family, eight months

A mother wrote to me about her son, then three years and two months. He had been hearing English at home since he was one — the usual rituals, morning and bath. He was speaking in short sentences in both languages. But every /k/ and /g/ came out at the front of his mouth. "Tome here." "Doe away." She had started to wonder whether the English had confused him, and her mother-in-law was certain it had.

Two questions settled it. Did it happen in Ukrainian too? Yes — кіт was "тіт". So it was not English; it was fronting, in both systems, exactly as the pattern predicts. And could strangers understand him? Mostly, yes. That put him inside the normal range for three, with one pattern still resolving.

We changed one thing. She stopped asking him to repeat words, and started recasting instead: he said "tome", she said "Yes, come here, come and see." She also moved their book time so they sat facing each other rather than side by side, which is a smaller change than it sounds and a bigger one than it looks.

The /k/ arrived on its own at about three years and ten months, first in Ukrainian, then in English within a few weeks. Nothing was taught. What changed was that he kept talking freely, because nothing he said was ever treated as a mistake — and volume of talking is the mechanism by which these patterns resolve.

I tell this story not because it is remarkable but because it is typical. The intervention that worked was the removal of an intervention.

Where the Mommy & Me English method fits

Almost everything above depends on one thing: a steady, unhurried, daily supply of real conversation with you. Not lessons — turns. The recast only works if there are hundreds of small exchanges for it to live inside.

That is exactly what my course is built to produce. It is a system for the mother: which sentences to use at which everyday moment — morning, meals, walks, bath — how to respond when a word comes out wrong, and how to keep going through the weeks when progress is invisible. It assumes you are not a native speaker, and it assumes your time is limited. You can see how the Mommy & Me English course is built.

My background is a Trinity College London CertTESOL, NILE Young Learners training, and Ukrainian specialist diplomas in speech therapy and primary education, alongside more than nine years teaching children. The speech-therapy half is why this article exists: the questions parents ask me about English are, more often than not, questions about ordinary speech development wearing an English costume.

The short version

Your child's "tat" is a stage, not a symptom. The late sounds — r, th, s, l, and the clusters — are supposed to be late. The sounds that are genuinely harder for a Ukrainian-speaking child are accent features shared by millions of fluent speakers. What deserves your attention is not a single consonant but a pattern across both languages, and how well a stranger understands your child at three or four.

And what you do about all of it is the same thing in every case: say the word back correctly, warmly, inside a real sentence, and carry on the conversation.