The new “science of reading” movement, explained

For over a century, a pendulum has swung back and forth on the best way to teach kids to read. On one side are advocates of phonics, an instructional approach focused on the relationships between sounds and letters; on the other, various methods, championed by educators, that emphasize word recognition and context clues.

This fight has occasionally grown so vicious it’s been dubbed the Reading Wars.

But in the last five years, one camp has scored a victory that, if not permanent, is at least decisive.

Since 2019, 45 states and Washington, DC, have passed at least one bill related to reforming reading instruction. The new rules apply to areas like school curriculum, professional development for teachers, screenings for dyslexic students, and requirements for testing. New York City — the largest public school system in the nation — has also ordered change for its 700 elementary schools.

These new policies — met with a mix of excitement and skepticism — are typically accompanied by bold promises. “These changes to our education system will actually educate our kids better in the future than we did before the pandemic,” Tennessee’s Republican Gov. Bill Lee declared. “This is the beginning of a massive turnaround,” said the New York City Schools chancellor.

A life with poor literacy skills is a hard life. Not being able to read means countless social, economic, and physical challenges, making the stakes of these new reading reforms tremendously high. Coming out of the pandemic, as students still struggle with learning loss, policymakers are especially aware of the need to get these new literacy interventions right. Though some critics have claimed we have no national reading crisis at all, for the millions of kids and adults who struggle to read in America, there is no question that each day brings about new isolating, draining, and avoidable challenges.

Momentum for these reading laws traces its roots to four main factors.

The first came five years ago, when a journalist named Emily Hanford released an influential radio series that looked critically at reading instruction, and specifically how popular strategies employed by American teachers conflicted with decades-old research in neuroscience and cognitive psychology about how kids best learn to read. Hanford’s reporting helped launch a movement dedicated to advancing the “science of reading” — a shorthand for applying these brain-focused studies to instruction.

The second factor came a year later, when Mississippi, a high-poverty state that overhauled its own reading policies back in 2013, soared in national test score rankings. Mississippi fourth graders went from being 49th in the nation for reading on a major exam known as the National Assessment of Educational Progress (NAEP), to 29th, and one of the reforms Mississippi had embraced in 2013 was a focus on the “science of reading.”

The third factor was the emergence in the 2010s of national grassroots networks of parents of children with dyslexia, who have more recently brought their organizing prowess to bear for new state policies they felt could better serve their children’s needs.

And lastly, but no less important, was the pandemic, which fueled a major drop in student achievement and sparked an infusion of new federal funds for schools.

This is not the first or even second time our country has quickly moved to revamp literacy instruction, and some worry that history is about to repeat itself with failed or misguided policy changes.

“I’ve been doing this for more than 50 years and I’ve been through three of these,” said Tim Shanahan, who once served as president of the International Literacy Association and on a federal study into the scientific evidence behind reading instruction.

Reading Wars veterans like Shanahan say this wave of reform has some new elements. Still, some problems that have undermined past waves threaten these new laws too, like a lack of funding and the challenge of maintaining political support if (and likely when) some standardized test scores go down.

Indeed, not all teachers and school administrators are on board. Many have resented the pendulum that has swung back and forth over decades, with competing rules and ideas and sometimes teacher-bashing to boot. Some educators reject the idea that they’ve been teaching reading incorrectly and note that the new materials they’re being told to use now lack a strong track record, too.

One of the most prominent cognitive scientists studying how kids learn to read has emerged as a vocal critic of many of the curriculum guides now being marketed as adhering to the “science of reading.”

Even the more enthusiastic champions of the new reading reforms recognize that passing laws is only the first step, and there’s a long road ahead to making these changes lasting and successful, as schools deal with other intense challenges including chronic absenteeism, a student mental health crisis, and unusually high staff turnover.

The science behind the science of reading

The definition of “science of reading” — a term first used in reading instruction in the 1830s — has evolved over the decades. Today, it primarily refers to cognitive research that pertains to what happens in the brains of strong and poor readers, and is frequently a shorthand for deploying more classroom time on phonics, with its emphasis on learning to sound out words. Past research has found that most kids need systematic phonics instruction when they’re young to ultimately read well.

Sometimes this phonics research gets overstated by proponents. The best studies have shown that teaching phonics is moderately better than other approaches — translating to about a few months of extra school instruction. Experts caution against forgetting that at its core, phonics is a means to an end: “Programs that focus too much on the teaching of letter-sounds relations and not enough on putting them to use are unlikely to be very effective,” stated the National Reading Panel report in 2000, a seminal federally funded investigation into science-based reading instruction.

And not everyone needs explicit phonics instruction at school to read well; indeed tens of millions of children have acquired strong literacy skills without it for generations, either learning through osmosis at home or through other classroom approaches that have waxed and waned in popularity like “whole language” and “balanced literacy.”

But what reading experts generally agree on is that most students — roughly 60 percent of children — will benefit from more systematic phonics instruction, especially in kindergarten, first grade, and second grade. A school’s average reading performance will likely go up, and it could be especially helpful for students from disadvantaged backgrounds or those struggling with dyslexia.

Sometimes “the science of reading” is invoked as a way to signal what’s not considered scientific. The most prominent example of that today is a popular reading instruction strategy known as “three-cueing” or MSV, an acronym for meaning, syntax, and visual information. Popularized in the 1960s and ’70s, these methods encourage students to look at context clues to guess the meaning of unknown words; for example, a picture of a house in a story about the country might help a struggling reader guess the word “cottage.”

But studies have shown that these sorts of predictive approaches are ineffective in helping students actually learn to read specific words. Strong readers, rather than guessing a word’s meaning based on adjacent clues, will work to decode letters in the unfamiliar word itself. Have students who rely on cueing generally understood the meaning of a passage? Yes. But they haven’t read the passage well.

These debates around phonics instruction can grow contentious, especially over the last few years. Many academics argue there should be a broader, more inclusive definition of “science of reading” that encompasses research fields beyond just the brain.

“There isn’t shared language with thinking and talking about this with any precision,” said Rachael Gabriel, a professor of literacy education at the University of Connecticut. “A lot of the language is dog whistle-y, and everyone gets really deep in the weeds and then confuses each other.”

For example, though some reading experts defend context-based strategies like cueing, pointing to experimental studies showing they could be complementary to phonics, others believe it’s a mistake to encourage teaching the strategies of poor readers at all.

“Being a reader isn’t like being pregnant where you either are or you aren’t,” said Shanahan, who supports the move away from cueing. “We care how good a reader are you, and we know that strong readers don’t try and guess the meaning.”

In 2019, a nationally representative EdWeek Research Center survey found 75 percent of K–2 teachers said they teach students to read using the three-cueing method, and 65 percent of postsecondary instructors who teach courses on reading instruction said they did too. As the “science of reading” movement has picked up steam, popular school curriculum providers who endorse these strategies have come under intense scrutiny, blasted for teaching young children strategies that could delay or derail their reading.

The tide is seemingly turning. Three states — Indiana, Arkansas, and Louisiana have recently gone so far as to ban “three-cueing” in legislation, and ExcelInEd, a national advocacy group founded by former Florida Gov. Jeb Bush, holds eliminating three-cueing as essential for any model reading policy. Last year Lucy Calkins, a longtime proponent of the “balanced literacy” approach to reading instruction, revamped her popular school curriculum program to include class-wide daily phonics lessons. “All of us are imperfect,” Calkins told the New York Times. “The last two or three years, what I’ve learned from the science of reading work has been transformational.”

The problem is we don’t know how to teach the “science of reading” well yet, either

Though states have moved to replace poorly rated curricula or ban weak instructional approaches, figuring out exactly what schools should do instead is harder. Politicians have falsely declared that experts know what science-backed alternatives teachers should use, and education companies have rushed to meet the demand for new materials, selling school districts their own questionable curricula and teacher trainings.

“We think we have some sort of tentative answers, that various kinds of instructional approaches work better than others, but that evidence is definitely far from ironclad,” said Morgan Polikoff, an education professor at the University of Southern California and an expert in K–12 curriculum standards. “Generally speaking there is not a lot of good research on the effectiveness of core curriculum materials, and that’s true both in reading and for other subjects.”

For example, LETRS, an acronym for Language Essentials for Teachers of Reading and Spelling, has become one of the most popular training courses claiming to adhere to the “science of reading.” Mississippi embraced it beginning in 2014 as part of its statewide reading reforms, and following Mississippi’s impressive gains on the 2019 NAEP exam, other states rushed to follow. (It’s a pricey professional development course that can take teachers up to 160 hours over two years to complete.)

Despite its close associations with the “science of reading” — LETRS has its own middling track record of effectiveness. One experimental study found teachers who were trained by LETRS did improve in their knowledge of reading science, but their students did not have statistically higher differences in achievement than teachers in the control group. (A spokesperson for LETRS told Vox, “we know it has been proven to drive positive student outcomes when coupled with other educational interventions.”)

Likewise, studies of reading interventions known as Orton-Gillingham, which have been promoted by many literacy reform advocates, have yielded mixed results so far. One 2021 meta-analysis found the approach did not generate statistically significant improvements in foundational skill areas like phonics. Even for students with some forms of reading disability, Orton-Gillingham wasn’t found to significantly boost comprehension or vocabulary.

Generally reading experts say the policies included in the new state reading laws are a “real mixed bag.” Some laws incorporate more research-backed ideas like coaching, while other endorsed approaches are more suspect. There is no clear amount of time that research shows should be spent daily on phonics, no established curriculum for the “science of reading” and studies on so-called decodable books — strongly endorsed by some phonics advocates to help young students practice letter-sound combinations — have their own mixed research track record.

“Often phonics advocates promote the use of research since research supports phonics instruction … but when it comes to specific prescriptions about how phonics should be taught, they make all kinds of claims that come down to: do it my way, I know best, don’t worry about what the science has to say,” said Shanahan. “There are features of effective phonics instruction suggested by the research and there are aspects on which there is either no research or the research rejects the advocate’s prescription.”

Mark Seidenberg, a cognitive neuroscientist and author of Language at the Speed of Sight, has emerged as a critic of some of the new materials being promoted under the banner of the “science of reading.” He’s blasted “influencers” leading the movement whose own background in cognitive science is limited. (Seidenberg’s book on reading research heavily shaped Hanford’s reporting.)

In an interview with Vox, Seidenberg expounded on his criticism: “It’s a difficult situation because people want to adopt better practices, they understand the idea that what was done before was not really based on solid ideas … but now you have a huge demand for science-based practices pursued by advocacy groups and people who don’t have a great understanding of the science.”

Seidenberg believes that moving away from strategies like three-cueing is important. But he warned that a simplistic reliance on some of the foundational reading science research can lead to some misinformed instructional conclusions, like the idea that children should learn units of sound (or “phonemes” ) before letters, and letters before syllables and words.

“That’s a basic misunderstanding,” said Seidenberg. “Phonemes are abstract units that are results of being exposed to an alphabet, they’re not a precursor.” He also lamented that some leaders have incorrectly cited his research to suggest there’s no downside to teaching kids phonics in the early grades for too long. “There are big opportunity costs and the clock to fourth grade is ticking,” he said. “You only want to do a lot of instruction on these components enough to get off the ground.”

The UK offers a cautionary tale. Last year reading researchers published a major study that concluded England’s present emphasis on phonics instruction came at the expense of other needed literacy skills. They attributed the country’s heavy prioritization in part to a national phonics screening test introduced by the government in 2012, which incentivized in some cases up to three years of daily hour-long phonics lessons.

“There’s no question that phonics instruction is important,” Dominic Wyse, the study’s lead author and an education professor at the University of College London, told Vox. “But let’s be clear, there are risks to overdoing it. You’re wasting their time and damaging their time to develop reading comprehension.”

Getting implementation right

The success of the new reading instruction laws will depend primarily on how well they’re implemented on the ground in schools and individual classrooms.

This past spring researchers from Michigan State presented a study assessing the impact of the new reading reform laws on test scores. The research, which has not been peer-reviewed, found students in states with more “comprehensive” supports — like coaching for teachers, adequate funding, and summer tutoring — had larger gains on test scores and on the NAEP than students in states with less comprehensive reading laws.

One concern for advocates is the great variability in how states might hold schools accountable to the new laws. Some worry that without more proactive transparency and enforcement provisions, the reforms could wither on the vine. Others worry about unfunded mandates in a year or two once Covid-19 aid dries up.

Still, others point to the fact that many colleges of education are still teaching methods to prospective teachers like three-cueing, and the New York Times quoted Lucy Calkins at a conference in March telling educators they can join other school leaders in rejecting these new approaches. “You can say no,” Calkins had said. “And people all over the country are doing so.”

To encourage more successful implementation, Gabriel of the University of Connecticut and Sarah Woulfin, a professor of education policy at UT Austin, have been encouraging more focus on school systems, structures, and leadership, rather than a narrow focus on what individual teachers know about reading. “It is not enough to deliver new materials that may sit in boxes in a closet without strong leadership for their integration,” they write. “It is not enough to press teachers to use new tools, without providing high-quality opportunities.”

In some states, such as North Carolina, lawmakers have rolled out new policies but provided little support for educators to incorporate the new mandates into their existing workflow, fueling backlash.

Ginny Sharpless and Amanda Harrison, co-founders of Literacy Moms NC, a group for parents of students with reading disabilities, told Vox they’ve been deeply disappointed by the lack of investment lawmakers have made in building teacher buy-in for the new policies.

“If you’re already working crazy hours, already not getting paid enough, and then someone says, ‘Oh, we want you to do 200 hours of extra training on your own time,’ well, you’re going to be pissing off people who are already burnt out,” said Harrison.

“State lawmakers did absolutely nothing to try to change [educators’] views, you can’t just show people the data and expect them to be converted,” added Sharpless. “Leaders should have gone out there and had meeting after meeting to talk to teachers to bring them on board.”

Polikoff, of the University of Southern California, said the history of US education policy reform suggests that building teacher buy-in and avoiding overly punitive mandates will be important to helping the new “science of reading” laws actually stick. He also warned against hinging hopes to specific state test score gains. When scores trended down following the introduction of the federal Common Core standards, opponents used those declining results to then water down the policies.

“There’s the risk of setting yourself with unrealistic expectations, and states that go up in test scores also go down,” he said. “One of the states that could go down is Mississippi, which has a lot of conditions that make it likely to have educational difficulty.”

As teachers embark on the 2023–24 academic year, eager to educate students using the best reading materials available, they’ll have to wade through this morass of information, hoping to do the best job possible but waiting for more clarity on exactly how. And in the meantime, there are real disagreements over what instructional strategies to use in schools, what emphasis each should get, what counts as evidence, and who gets to decide.

“Policymakers don’t necessarily want to hear that the science is more complex and less certain than we thought,” said Amanda Goodwin, the co-editor of Reading Research Quarterly and a professor of literacy at Vanderbilt University. “It’s much easier to get public support for a handful of ‘proven’ practices.”