When should a child be able to count objects, not just recite numbers?

Most children can count out a small pile of objects, and know that the last number is the total, somewhere between three and a half and four. Reciting the sequence comes a year or more earlier.

The test in this post costs nothing, and the best practice for it is a drawer of forks.

Two skills that look like one

When a two-year-old chants "one, two, three, four, five" up the stairs, they are reciting a list. It is a good list to know, and it comes first. Knowing that "four" means a particular amount, a quantity you can hand over, is a different skill, and it arrives much later. Children learn it one number at a time. A child will know exactly what "one" means for months while "two" still means "some". Then "two" settles, and "three" is still fuzzy. Then "three". Somewhere after that the whole thing clicks, and the child understands that the last word you say when counting a pile tells you how many are in it. Researchers call that the cardinal principle, and it is the thing you are actually asking about.

The trap is that a child in the middle of this looks like they can count. They point along a row, say the words in order, and reach the end. Ask "how many?" and they say the last word. That can be pure habit. Sarnecka and Carey found in 2008 that many children answer "how many?" with the last word they counted while not understanding how counting works at all. The words are right. The meaning is not there yet.

Rote counting and one-to-one correspondence

Rote counting is the list. One-to-one correspondence is touching each object exactly once and giving it exactly one word, no skipping, no double-tapping. It sits between the list and the real thing, and a child can have it for a while before the last word means anything.

There is a two-minute test for where your child actually is. Karen Wynn used it in her 1992 study, and researchers call it Give-a-Number. Put a pile of small identical things on the table, a dozen or so, and ask for exactly one. Then two. Then three, four, five, mixing the order so it is not a pattern. Do not count aloud, do not help, and do not react.

• Hands you one when asked for one, and a fistful for anything else. They know "one". Nothing else has a meaning yet. Wynn's youngest children started here.
• Gets one and two right, grabs for three. "Two" has landed and "three" has not. In Wynn's group this stage ran to months.
• Manages three or four, then guesses. Still working number by number.
• Counts out any number you ask for, and recounts when unsure. That recount is the tell. They understand that counting is how you find out. This is the cardinal principle.

Run it once a month if you like. Not every day; a quiz is the fastest way to make a three-year-old refuse to count anything.

What Wynn actually found

Wynn's study followed twenty children aged two to three for seven months, testing them repeatedly with a pile of twenty-seven toy animals. Some of the testing was done at the Cambridge Montessori School, a pleasing detail for a blog like this one and no help with the size. Twenty children is a small study, and the ages in it are averages from that group, not a chart your child has to hit.

What I keep coming back to is the pace. The number words really do land one at a time, and the gaps are long: on average roughly four to five months between knowing "one" and knowing "two", and a similar wait for "three". Children who already knew what "one" meant took close to a year to reach the point where counting itself made sense. In her group, that happened on average shortly after the third birthday, somewhere between three years one month and three years four months.

So the child who has been reciting to ten since they were two, and still cannot hand you three blocks at three and a half, is doing the slow part. The slow part takes about a year however good the recitation was.

Counting objects is not the same as reading numerals

A separate question gets tangled up in this one. "My child can count but doesn't recognise numbers" is usually about the written digits. Knowing that a squiggle shaped like 4 is called "four" is a third skill, and it has very little to do with the other two. A child who counts a pile perfectly can still be baffled by a written 4.

The order that matters is quantity first, then the word for it, then the symbol. Number puzzles and magnet digits teach the symbol. Do them later, and when a child looks blankly at a written 7, file it under reading, where it belongs. It is a small thing and it sorts itself out.

If your four-year-old is not there yet

Common, and usually not a problem on its own. Sarnecka and Carey's children ran from two to four years old, split between those still working number by number and those who had got there. Their paper sorts children by stage, and we have not pulled its counts for the four-year-olds, so I cannot tell you what share of that age is still climbing. What Wynn's group shows is that the average child got there shortly after turning three, which puts a four-year-old who has not at the later end of normal.

Worth a mention to the preschool teacher, or to your pediatrician at the next visit, if any of these are true alongside it.

• The list itself is not there. No rote counting to five by four, in any order at all.
• "One" has no meaning. Asked for one thing, they hand over a pile, every time, and have done for months.
• The rest of their language is late too. Counting is language and tends to arrive with the rest of it. If words in general are behind, that is the bigger question.

If it is only the counting, do the test again in two months and count real things in between.

What a counting toy can and cannot do

Here is the admission I promised. Nobody has shown that a particular toy moves a child from knowing "two" to knowing "three", or from any stage to the next. Neither study tested a toy, an app or a teaching method, and I have not found one that does. Our math and logic shelf was chosen on our reading of what those two studies describe. We ran into the same wall in Are Montessori toys actually better?, and the answer has not changed.

What I can say is which half of the job a toy is aimed at. Anything that sings or has the digits printed in a row is drilling the list, and the list is the half your child already has.

• Sets you can line up against each other. Three beads next to four, so "one more" is something you see.
• Pieces that only fit if the amount is right. A peg board with one hole per peg tells the child they are wrong without anyone saying so. That matters here, because "no, count again" ends the game.
• An abacus, for the sliding. Beads that move one at a time under a finger are one-to-one correspondence you can feel, which is also what the sensory shelf is for.

The free version beats all of it. Ask for four forks at dinner. Three cups. Every meal is a Give-a-Number test with a reason to get it right, and the reason is what holds a child's attention, which is the idea behind practical life work. The 3 year olds page has what suits this stage; by four and five the same beads start doing adding.

What to do at dinner tonight

Run the pile test once, quietly, and write down where your child is. Then stop testing and start asking for things by number, at the table, from the toy box, on the stairs. Do not correct the answer. Hand back the extra fork and ask again tomorrow.

Reciting to twenty is nice, and every grandparent will praise it. Real counting takes its year regardless, and nothing you buy shortens it. The forks are free, and they are the best counting toy in the house.