When Should Reading Instruction Begin?

Tim Shanahan

Blast from the Past: This entry first dropped on October 26, 2019, and was reissued on January 24, 2026. The only change is an update to the NAEYC statement (it was in draft form when originally cited).

Teacher question: 

What does research say about early literacy and when to begin? I am aware that kids may reach the stage of development where they're ready for reading at different times. What does the research say about the "window" for when a kid can learn to read? What are the consequences if they haven't started reading past that time? 

Shanahan response:

Oh, fun. The kind of question that generates strong scholarly-sounding opinion, with no real data to go on.

The advocates on both sides will bloviate about windows of opportunity, developmentally appropriate practice, potential harms of early or later starts, and how kids in Finland are doing.

Despite the impressive citations that appear in the Washington Post, Huffington Post, and various blogs, the truth is that there is no definitive research on this issue.

The meager handful of supposedly direct comparisons between starting earlier versus later are so ham-handed that I’m surprised they were even published.

One example is a longitudinal study that followed kids for six years… after either a dose of academically- or play-focused preschool. The research claimed that the kids taught early ended up with lower later achievement (Marcon, 2002).

That sounds horrible until you look closely at the analysis, when it becomes evident that the comparisons were questionable and the statistics spurious (Lonigan, 2003). More of the play-group kids were retained along the way, so the final comparison—the one that finally found the difference the researcher was seeking—wasn’t between the same samples as at the beginning. The researcher’s response to this criticism suggests that the samples weren’t equivalent at the start either, further highlighting that this study couldn’t possibly reveal whether early teaching was helpful, hurtful, or not an issue at all.

I can provide examples going in the other direction, too. Since graduate school, I have been told that young children are especially able learners and that the earlier we start teaching, the better the odds that we’ll catch kids during that “portal of receptivity” (Gross, 2016).  

The evidence behind that argument seems mainly based on the fact that from about 18 months to 5-years-of-age children learn an amazing amount of vocabulary, and that so-called vocabulary spurt is a real one. However, the idea that everything, or even everything involving language, is learned easily during those years, is where the leap of faith comes in.

Reading development certainly does depend upon vocabulary, but there is much more to learning to read, and there is no convincing evidence that four-year-olds will learn to read more quickly or easily than would be the case a year or three later. Just because youngsters learn spoken words fast, doesn’t mean that they can perceive the sounds within words (phonological awareness), or that they'll be able to master the names and sounds of the letters (the beginnings of decoding), especially easily.

READ MORE...