Will the Science of Reading Movement Become a Frankenstein's Monster?

An overwhelming focus on problems with phonics instruction and a shallow understanding of cognitive science have led to some unintended consequences.

Frankenstein's monster reading a book

Is the Science of Reading becoming a Frankenstein’s monster, patched together from reanimated research and stalking the land, capable of wreaking havoc in ways its creators failed to anticipate? And is there anything we can do to rein it in?

The “SoR” movement, sparked by Emily Hanford’s audio journalism, has been a monster of a success, at least when it comes to creating a sense of urgency and spurring action. The number of states that have now passed or adopted policies aimed at bringing reading instruction in line with scientific evidence has reached 37—plus D.C.—with 17 of them coming on board just last year, and more in the pipeline. And the publishers of the instructional materials Hanford criticized are rapidly losing market share.

The question is whether these developments will result in improved literacy outcomes. Even Hanford is expressing reservations on that point. In the most recent installment of her “Sold a Story” podcast, Episode 10, she notes that schools and districts are spending lots of money on new curricula and materials but asks, “What’s the evidence they will lead to better results?” Her reporting has focused on problems with the standard approach to teaching kids to read words rather than on detailed recommendations on what to do instead.

In the same episode, cognitive psychologist Mark Seidenberg, who has figured prominently in Hanford’s work, says he’s concerned that educators and policymakers are trying to apply a few semi-digested ideas from the research while overlooking other important findings. He’s seen teachers pursue some individual components of skilled reading—as identified by models like the “five pillars of early literacy”—as though they were ends in themselves rather than the means to proficient reading.

In a blog post last year, Seidenberg held up an example: he heard a teacher ask whether she should continue teaching phonemic awareness (“PA”) to a child who already has good decoding skills. (PA is the ability to hear and manipulate the individual sounds in words.) It should be clear, Seidenberg wrote, that if a student is decoding words well, “instruction can focus on the many more things that need to be learned to become a skilled reader.”  But, Seidenberg wrote, well-intentioned teachers with limited knowledge of cognitive science are following others who are in a similar position—and who often recommend flawed approaches to schools and districts.

“Fact-Checking the Science of Reading”

If advocates of systematic phonics instruction like Hanford and Seidenberg are having doubts about the direction taken by the SoR movement, you can imagine what’s being said by those who are skeptical of or hostile to it. Actually, you don’t have to imagine. Recently, P. David Pearson—an eminence grise of reading instruction—joined forces with another emeritus education professor, Robert J. Tierney, to produce a 188-page monograph titled “Fact-Checking the Science of Reading.”

Some of this “fact-checking” needs to be, well, fact-checked. For example in defending the “three-cueing” method, Pearson and Tierney themselves overlook a key fact: a couple of years ago, a federally funded study of Reading Recovery, which uses three-cueing, found that by grades three or four, students who got that approach in first grade scored lower on state reading tests than a comparison group.

There’s a lot more in the monograph that could be picked apart, but I’ll leave that to others. I’ll just say that the thrust of the argument is that teaching phonics to children is not enough to turn them into proficient readers. That’s certainly true. But is that what SoR advocates, including Hanford, have been arguing?

I’d say the answer is complicated. SoR advocates have never reduced reading to phonics. One thing they have in common with the Balanced Literacy folks on the other side is their reverence for the “five pillars of early literacy” model, which includes not just PA and phonics but also fluency, vocabulary, and comprehension.

Here’s where the “reanimation” of old research comes in, to return to the Frankenstein’s monster analogy: The “five pillars” formulation is over 20 years old. The same thing is true of two other now-ubiquitous SoR models, the Simple View of Reading and Scarborough’s Reading Rope. They were all basically lying dormant until a few years ago, when the SoR movement brought them back to life with a jolt of electricity.

Relying on these models to define the “science of reading” is one reason the SoR movement is threatening to turn into a monster. The models were created to show that there’s more than one thing going on in reading. They weren’t designed to be guides to instruction. And when they’re treated that way, you get a checklist approach: teachers try to teach the first pillar or strand to “mastery,” then the next, and so on. That’s what Seidenberg is worried about.

I’m not saying that Hanford herself held up these models as instructional guides, but the education world is sometimes like a game of telephone, where the message gets increasingly simplified or distorted as it spreads. It can be hard to prevent that from happening.

Problems with Reading Instruction Go Beyond Phonics

But Hanford’s reporting did focus only on problems with phonics instruction. As a result, many have assumed that schools are handling the other pillars and strands just fine. Hence the overwhelming emphasis on just fixing phonics, which has led many schools and districts to keep everything else they’re doing in place while adding a “phonics patch.” But many of the other things they’re doing, especially with regard to comprehension, also conflict with scientific evidence—evidence the SoR movement has largely overlooked.

Nor does it work to spend the entire two hours or more of the daily reading block on phonics, which is another misinterpretation of Hanford’s message. That’s likely to exhaust and dismay both teachers and students. We do need to do something more to enable children to comprehend written text—but not the kind of comprehension instruction that the vast majority of schools have been providing.

The typical approach is to focus on comprehension “skills and strategies,” like making inferences. It’s assumed that these skills can be taught directly and applied generally, like decoding skills. But there’s lots of evidence from cognitive science that knowledge of the topic you’re reading about—or general academic knowledge and vocabulary—is far more important to comprehension than abstract “skill.”

In fact, there’s really no such thing as a general skill of something like “making inferences.” A reader might have no trouble making inferences about a topic she knows well but struggle to make inferences when reading about one that’s unfamiliar. Typical reading comprehension instruction, which takes up many hours every week in elementary and middle schools, actually denies kids access to the kind of knowledge they need to apply comprehension skills, especially to complex text.

As far as I can tell, Hanford has never delved into the conflict between comprehension instruction and the scientific evidence—and I’ve listened to every audio documentary she has produced, going back years before Sold a Story. She did do a one-hour piece on comprehension, in 2020, but it didn’t mention how schools try to teach comprehension or suggest there’s anything wrong with it.

In fact, the 2020 piece implied that if schools just did a better job of teaching decoding, students who struggle with comprehension—who are mostly from lower-income, less highly educated families—would be just fine. That comes pretty close to presenting phonics as the solution to all our reading problems. (If you’re looking for a podcast that does tackle the comprehension problem head-on, may I be so bold as to suggest the one I hosted, which you can find here.)

Some SoR advocates argue that we need to fix phonics first, before we move on to comprehension. But if we want to ensure that the gains being made in effective phonics instruction last—and there are gains, despite some of the misunderstandings out there—we need to start building the knowledge that fuels comprehension early, while kids are still learning to decode fluently. If we don’t, many kids will reach higher grade levels able to decode complex text but unable to understand it. And at that point phonics skeptics may say, as they have in the past, “You see? Phonics doesn’t work.”

What a Different Kind of “SoR” Might Look Like

I don’t know why Hanford hasn’t talked about problems with comprehension instruction. But I do wonder where we might be if she had. Maybe we would now have a more coherent approach to addressing the range of issues plaguing reading instruction—one that is less likely to divide people and more likely to last.

What would a more coherent approach look like? Maybe something like what has been laid out by the Scientific Advisory Committee of the Knowledge Matters Campaign. In providing model language for legislators, they call for a content-rich curriculum that begins building knowledge in the early grades and is logically sequenced. That language would certainly be more helpful than the “five pillars” formulation that has found its way into a lot of recent reading legislation.

But of course, the devil is in the details, and it probably doesn’t make sense for legislators to wade too far into them. For reading reform to be effective, teachers need a curriculum that covers both foundational reading skills and knowledge-building in a way backed by the evidence.

Teachers do need some familiarity with the science related to reading—including the role of knowledge in comprehension—but they shouldn’t be expected to translate that science into classroom practice themselves. A good curriculum can guide them, for example, to spend an appropriate amount of time on phonics—say, 20 or 30 minutes in the early grades—and an appropriate amount on the many other things that go into skilled reading.

State education agencies need to provide guidance to schools and districts about which curricula will be effective, but there are potential pitfalls there too. Officials need to go beyond just recommending the “all green” curricula on EdReports, which aren’t necessarily well-constructed or rich in the kind of content that enables reading comprehension. And even some content-rich curricula are hard for teachers to implement well, especially in challenging conditions. Teachers need good training and support to implement any knowledge-building curriculum effectively.

This message is a lot more complex than just “fix phonics,” and inevitably it’s harder to get across. But some states, schools, and districts are getting it nonetheless. I hope many others will follow suit soon, before the SoR movement sinks under the weight of misunderstandings about evidence-based instruction and the resulting backlash.

Frankenstein’s monster, by the way, starts out being a nice guy who is eager to learn—and who actually teaches himself to read. It’s only when people are repulsed by his appearance, and what they believe that says about who he is, that he becomes violent and destructive. Maybe if people had taken the time and effort to really understand him, his story would have had a happy ending.