Study: Children as young as 3 attempt to discern rules of reading, writing

A new study, published this month in the journal Child Development, provides new evidence that children start to learn about some aspects of reading and writing at a very early age.

When I first began covering education, there was an assumption schools were on a solo mission to educate the youth of America; parents dropped off their kids and teachers did the rest.

It didn’t matter whether the child was raised in a home with hundreds of books or just the telephone book: schools were supposed to be the great equalizer. We thought getting children at age 6 or 7 in a classroom was sufficient to remedy any early learning deficiencies and that they could catch up with more school-ready peers through routine classroom instruction.

Thirty-some years later, science has now shown us schools play a smaller role in academic outcomes than we thought and parents and homes play a larger role. In the last decade, research has revealed how what happens — or doesn’t happen — from birth to age 5 influences a child’s language skills and overall readiness to learn. What the child has seen and heard before that first day of school are major contributors to academic success.

Here is a release on a new study that will interest pre-k and early education teachers. I found it fascinating. It deals with the early efforts of children to spell. It is from Washington University in St. Louis.

Even the proudest of parents may struggle to find some semblance of meaning behind the seemingly random mish-mash of letters that often emerge from a toddler’s first scribbled and scrawled attempts at putting words on paper.

But new research from Washington University in St. Louis suggests that children as young as 3 already are beginning to recognize and follow important rules and patterns governing how letters in the English language fit together to make words.

The study, published this month in the journal Child Development, provides new evidence that children start to learn about some aspects of reading and writing at a very early age.

“Our results show that children begin to learn about the statistics of written language, for example about which letters often appear together and which letters appear together less often, before they learn how letters represent the sounds of a language,” said study co-author Rebecca Treiman, a professor of psychological and brain sciences in Arts & Sciences.

An important part of learning to read and spell is learning about how the letters in written words reflect the sounds in spoken words. Children often begin to show this knowledge around 5 or 6 years of age when they produce spellings such as BO or BLO for “blow.”

We tend to think that learning to spell doesn’t really begin until children start inventing spellings that reflect the sounds in spoken words — spellings like C or KI for “climb.” These early invented spellings may not represent all of the sounds in a word, but children are clearly listening to the word and trying to use letters to symbolize some of the words within it, Treiman said.

As children get older, these sound-based spellings improve. For example, children may move from something like KI for “climb” to something like KLIM.

“Many studies have examined how children’s invented spellings improve as they get older, but no previous studies have asked whether children’s spellings improve even before they are able to produce spellings that represent the sounds in words,” Treiman said. “Our study found improvements over this period, with spellings becoming more wordlike in appearance over the preschool years in a group of children who did not yet use letters to stand for sounds.”

Treiman’s study analyzed the spellings of 179 children from the United States (age 3 years, 2 months to 5 years, 6 months) who were prephonological spellers. That is, when asked to try to write words, the children used letters that did not reflect the sounds in the words they were asked to spell, which is common and normal at this age.

On a variety of measures, the older prephonological spellers showed more knowledge about English letter patterns than did the younger prephonological spellers. When the researchers asked adults to rate the children’s productions for how much they looked like English words, they found that the adults gave higher ratings, on average, to the productions of older prephonological spellers than to the productions of younger prephonological spellers.

The productions of older prephonological spellers also were more word-like on several objective measures, including length, use of different letters within words, and combinations of letters. For example:

A 5-year-old who writes “fepiri” when asked to write the word “touch” might seem to know nothing about spelling, but this attempt looks more like a word than “fpbczs” as produced by a 4-year-old.

“While neither spelling makes sense as an attempt to represent sounds, the older child’s effort shows that he or she knows more about the appearance of English words,” Treiman said.

The findings are important, Treiman said, because they show that exposure to written words during the 3-to-5-year age range may be important in getting children off to a strong start with their reading, writing and spelling skills.

“Our results show that there is change and improvement with age during this period before children produce spellings that make sense on the basis of sound.” Treiman said. “In many ways, the spellings produced during this period of time are more wordlike when children are older than when they are younger. That is, even though the spellings don’t represent the sounds of words, they start looking more like actual words.”

“This is pretty interesting, because it suggests that children are starting to learn about one aspect of spelling – what words look like – from an earlier point than we’d given them credit for,” she said. “It opens up the possibility that educators could get useful information from children’s early attempts to write– information that could help to show whether a child is on track for future success or whether there might be a problem.”

Other Washington University co-authors include Brett Kessler, a research scientist in psychological and brain sciences;  former Arts & Sciences undergraduates Hayley Clocksin and Zhengdao Chen;  and Kelly Boland, a former research assistant in Treiman’s reading lab who is now a psychology graduate student at the University of Missouri.


Reader Comments 0


While the premise of this study may be correct, I would also point out that younger children choose letters that are easier to form when they "write."  That might or might not be germane.

I recall talking to a kindergartener about a (non) word they had written.  I asked, "Why did you write it that way?" and the answer was, "That's the way I know HOW to write!"


How do we begin to fix the problem?  The schools are what's currently available. Remember the Stone Mountain "educators," both grown women, who recently got into a fistfight in the classroom? We owe these kids teachers who are, themselves a product of the culture that settles differences by fighting, stabbing or shooting. We can do that much.



Please don't generalize as to the merits of an entire "culture" based on the actions of two individuals.


I worked in the court system for decades. I'm familiar with the cultures of generational poverty and slavery.


I am delighted that you wrote such an erudite column about the process of reading and spelling and how early it begins.  I agree heartily with what you have penned, Maureen Downey.

For readers, I simply must state, once again, that all parents are not the same and that we must stop judging parents who are not doing well by their children and start giving them societal support.

Parents are really older children, and some of these parents are illiterate themselves, as well as poor, and even psychologically damaged, unable to hold jobs.  Until we as a society stop thinking in punitive ways and start thinking in enlightened ways, we will not help to make America literate and productive, across the board to every child and within every family.  The problems are often generational.


@MaryElizabethSings  "For readers, I simply must state, once again, that all parents are not the same and that we must stop judging parents who are not doing well by their children and start giving them societal support."

While that is true, it does point out that this is even MORE of a culture issue than previously thought. It points out that the early home environment is even more influential that previously thought, and that PARENTS are more important to success/failure than previously thought.     Of course, it has always been true and a lot of us know it. But how can we destroy the culture of poor performance without sounding judgmental to those who MUST change if their children are  to succeed,  and not become a burden on society.   

It has always been the hard nut to crack, and it always will be: how to get parents to realize they THEY are responsible for educating their children, and then to get them to do it.

In case anyone is interested, I have 4 kids.  Two have graduated college, two are young.  My 8 year old reads at a middle school level and my 6 year old, who isn't even in 1st grade, can read at a second grade level. This happened because WE, as parents, started working with them very young, and we worked with them all through prek, K, etc. That is something that I push nearly every single night of the week.   I KNOW that there are some kids who enter K  whose only exposure to the alphabet is singing the letters.   That just doesn't cut it. 


@rdh @MaryElizabethSings

You have been blessed, but others have not been so blessed as you.

My father often said, "It is better to give a man a slap on the back than a kick in the (crude word for one's 'behind')."  I believed him then, and I still believe that way of handling others to be the better way today, in my 70s (17 years after my father died).


@rdh @MaryElizabethSings Many kids in the inner city, and the poor suburbs, and poor rural areas can't sing the letters, or count... you have to work in a juvenile court for a while to see how bad it can be.


@Wascatlady @MaryElizabethSings @rdh

Cliched and too easy an answer, Wascatlady, although it sounds good on the surface.  As you and I both know, each student will require a different response from his/her teacher based on many factors unique to him/her.  I was speaking of an overall slant in working with people.

Thanks for the compliment, btw.