What is the million word gap?

It is an estimate from a 2019 study at Ohio State University of how many words children hear from storybooks before they start school. Researchers found a gap of about 1.4 million words between a child read five books a day and a child never read to. The figure is a calculation from book word counts and reported reading frequency, not a recording of what individual children heard.

How many words does a child hear if we read one book a night?

Around 296,660 by age five, on the study's figures. That compares with 4,662 for a child who is never read to, 63,570 for one or two sessions a week, and 169,520 for three to five a week. The jump from nothing to something is far larger than the jump from something to a lot.

Do I need to read five books a day?

No. The five-books-a-day figure is the top of the range the researchers modelled, not a recommendation, and the 1.4 million headline comes from it. One book a night already puts a child about 64 times ahead of a child who is read to never, which is where almost all of the benefit sits.

How did the researchers work out the numbers?

They asked the Columbus Metropolitan Library which board books and picture books were borrowed most, took 30 at random from each list, and counted the words. Board books averaged 140 words and picture books 228. They then assumed board books up to age three and picture books at four and five, one book per session, and multiplied by reading frequency.

Is the estimate too high or too low?

The lead author, Jessica Logan, said the gap is likely to be conservative, because parents talk around a book as well as reading it and that conversation is not counted. Against that, the model assumes exactly one book per session and uses average word counts, so any individual family's real number could sit either side of it.

Why do words from books matter more than words from talking?

Books use vocabulary that conversation does not. Logan's point was that the words children hear from books may have a special importance in learning to read, because a picture book will happily use words like gleaming, sulking or ancient that rarely come up while making toast.

What if my child only wants the same book over and over?

That is not a problem for word exposure, and it has its own advantage for learning the words. Repetition helps children retain new vocabulary better than variety does, so a favourite read forty times is doing work that forty different books would not.

Raising readers

The million word gap, and what the number actually counts

A child read to daily hears around 296,660 words from books before school. A child never read to hears 4,662. Where those figures come from.

David White Founder, Taleium

9 October 2026 About 5 min read

Four thousand, six hundred and sixty two.

That is the number of words a child who is never read to is estimated to hear from books before they start school. Not four thousand a year. Four thousand across five years, which is roughly the length of one longish magazine article spread over the whole of early childhood.

A child read one book a night over the same period hears around 296,660. The ratio between those two figures is about 64 to 1, and it is the part of this research that deserves attention rather than the headline.

The figures

The study came out of Ohio State University in 2019, led by Jessica Logan with Laura Justice, Leydi Johana Chaparro-Moreno and Melike Yumuş, and was published in the Journal of Developmental and Behavioral Pediatrics. It estimated cumulative word exposure from storybooks across the five years before school.

How often a child is read to	Words heard from books by age five
Never	4,662
One to two times a week	63,570
Three to five times a week	169,520
Daily	296,660
Five books a day	1,483,300

The million word gap ↗ of the title is the distance between the top and bottom rows. Logan called it striking, and it is.

It is also the least useful row in the table.

What the number is, and what it is not

This is a model, not a measurement. Nobody followed children around counting the words they heard.

What the researchers did was ask the Columbus Metropolitan Library which board books and picture books were borrowed most often, take thirty at random from each list, and count the words in them. Board books came out at an average of 140 words. Picture books averaged 228. They then assumed board books up to age three, picture books at four and five, one book per reading session, and multiplied by how often parents in the data said they read.

So every figure in that table is an average book multiplied by a reported frequency. Your own house will not match it, because your books are not the average book and some nights you read three and some nights nobody reads anything.

That matters for how you should read the table. The absolute numbers are illustrative. The ratios between the rows are the finding.

Logan was straightforward about the direction of the error, saying the gap is likely to be conservative, because parents talk around a book as well as reading the words on the page and none of that conversation is counted. Which is true, and cuts both ways: a family that reads without talking gets the printed number, and a family that stops every other page to discuss the picture gets rather more than it.

The row that actually matters

Look at where the big jump is.

Never to once or twice a week is a jump of about 59,000 words, a factor of thirteen. Once or twice a week to daily is a jump of about 233,000 words, a factor of under five. Daily to five books a day is exactly five times, because that is literally what the model does to get there.

In other words, the steepest part of the curve is at the very bottom. Going from no reading to some reading does proportionally more than any amount of intensification after that, and the 1.4 million figure at the top is an arithmetic ceiling rather than a thing anybody should aim at.

The practical reading of this study is not “read more”. It is “read at all, on most days”, and then stop worrying about the number.

Why book words are different

The reason any of this matters is not the count itself. It is which words get counted.

Logan’s observation was that the words children hear from books may have a special importance in learning to read, and the mechanism is not mysterious. Conversation between a parent and a three-year-old runs on a small, practical vocabulary: shoes, dinner, careful, nearly. Books do not. A picture book will use gleaming, sulking, ancient, bellowed, reluctant, and it will use them in a sentence that makes the meaning obvious from context.

That is why 296,660 words from books is not interchangeable with 296,660 words of talking. The book words are the ones a child will later meet in print and need to recognise.

Our piece on why children ask for the same story every night is relevant here, because it looks at the other half of the question: not how many words a child hears, but whether they keep them. The answer turns out to favour repetition over variety, which means a child with a small shelf and a favourite is not at a disadvantage.

What to do with this

Very little, which is the honest answer.

If you read to your child most nights, you are already on the row of the table that does the work, and the gap between you and a hypothetical five-books-a-day household is not something to lose an evening over. How long a bedtime story should be is a more useful question than how many, and the answer is shorter than most people assume.

If reading has stalled because your child has decided books are boring, that is a different problem and the word count is not the lever. What to do when a child says reading is boring covers it.

And if you are starting from nothing, the table says something genuinely encouraging. The distance from zero to one book a night is the biggest step available, and it is also the easiest one.

The short version

A child read to daily hears somewhere around 296,660 words from books before school. A child never read to hears 4,662. Those numbers are modelled from average book lengths rather than measured, so treat them as a shape rather than a count.

The shape is what counts. Almost all of the benefit is in the move from never to usually, and the million in the headline comes from the one row nobody lives on.

Questions

What is the million word gap?: It is an estimate from a 2019 study at Ohio State University of how many words children hear from storybooks before they start school. Researchers found a gap of about 1.4 million words between a child read five books a day and a child never read to. The figure is a calculation from book word counts and reported reading frequency, not a recording of what individual children heard.
How many words does a child hear if we read one book a night?: Around 296,660 by age five, on the study's figures. That compares with 4,662 for a child who is never read to, 63,570 for one or two sessions a week, and 169,520 for three to five a week. The jump from nothing to something is far larger than the jump from something to a lot.
Do I need to read five books a day?: No. The five-books-a-day figure is the top of the range the researchers modelled, not a recommendation, and the 1.4 million headline comes from it. One book a night already puts a child about 64 times ahead of a child who is read to never, which is where almost all of the benefit sits.
How did the researchers work out the numbers?: They asked the Columbus Metropolitan Library which board books and picture books were borrowed most, took 30 at random from each list, and counted the words. Board books averaged 140 words and picture books 228. They then assumed board books up to age three and picture books at four and five, one book per session, and multiplied by reading frequency.
Is the estimate too high or too low?: The lead author, Jessica Logan, said the gap is likely to be conservative, because parents talk around a book as well as reading it and that conversation is not counted. Against that, the model assumes exactly one book per session and uses average word counts, so any individual family's real number could sit either side of it.
Why do words from books matter more than words from talking?: Books use vocabulary that conversation does not. Logan's point was that the words children hear from books may have a special importance in learning to read, because a picture book will happily use words like gleaming, sulking or ancient that rarely come up while making toast.
What if my child only wants the same book over and over?: That is not a problem for word exposure, and it has its own advantage for learning the words. Repetition helps children retain new vocabulary better than variety does, so a favourite read forty times is doing work that forty different books would not.