In October 2022, Mississippi reported some exhilarating news. After the state started using phonics to teach reading in 2013, its fourth-grade reading scores jumped from 49th to 29th in the nation, according to the Urban Institute’s National Assessment of Educational Progress. Unreported was the fact that in the 29th position, the state’s 4th graders were now on par with the rest of the nation in that roughly 65 percent of them had only what NAEP called basic skills. Mississippi had gone from the bottom of the pile to being just as bad as everywhere else. Some educators claim that these figures are misleading because the Basic Level (the level below proficiency) isn't bad at all. Even were that to be the case, the below Basic Level figures still yield the disturbing information that approximately four out of ten fourth-grade students lack the ability to read.
Biennial testing through NAEP consistently shows that two thirds of U.S. children are unable to read with proficiency. An astounding 40 percent are essentially nonreaders. Most are taught through phonics—a system of instruction based on sounding out letters that is mandated in at least 32 states and the District of Columbia. The phonics method of converting each letter to a particular sound is totally unsuited to the English language. As but one example, e, the most common letter in print, has 11 different pronunciations (end, eat, vein, eye, etc.), including its role as the much-taught “silent e” (tape, cute, fine, etc.). This failure has been endemic from the early days of the country when Benjamin Franklin fought against phonics. The steady expansion of this mode of instruction will not fix the situation.
The teaching of reading has largely been framed as a battle between phonics and whole language. In contrast to the dissected units taught in phonics, whole language presents children with complete books. Its premise is that learning to read, like learning to speak, comes naturally via exposure to good material. In the battle between these two systems, phonics has triumphed. For example, in May 2022, Lucy Calkins, an education professor and leading advocate of balanced literacy, announced a major retreat. She acknowledged that phonics achieved superior outcomes, and her unit at Columbia University was closed down. The victories phonics has enjoyed have led us to overlook the high rate of failure it generates as well as the research that supports the creation of alternative systems.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
As a psychologist specializing in childhood literacy, I have spent much of my career working to develop those systems. I served as co-director of the Developmental Neuropsychiatry Program at Columbia University, where I created a literary program that helped nonverbal autistic children learn to read. I also create language-based, gamelike reading programs for use with all children. In 1990, the Software Publishers Association awarded my computer program the Special Education Software Award. Then, as now, few schools used it because it didn’t meet their phonics requirements.*
Breaking phonics into its components lays bare its shortcomings. First, in the initial instruction, months are spent teaching children how to decode a minute set of simple, three-letter words like man,cat and bus that can be sounded out. Those words bear little resemblance to the content of any book, leaving children unprepared for actual reading. For example, even in Dr. Seuss’ uniquely crafted The Cat in the Hat, only 40 of the 236 words, or 17 percent, fit the three-letter pattern the children have spent months learning. Second, for all the other words, children must memorize dozens of complex rules. They need three rules just to handle the e’s in eleven. Many children hate reading because the rules turn decoding into a complex, laborious, failure-filled process. Third, about 60 percent of the words on any page are “noncontent words”—words like the,so and what that cannot be decoded by sounding out. Ironically, three of the five words in the The Cat in The Hat title are noncontent words. Characterizing them as “renegades,” phonics accords them minimal teaching time. The result is that children have no solid tools for dealing with the most frequent words they see on every page.
Although alternative methods have received little attention, there are ways of teaching reading that do not rely on phonics. One that guides my work incorporates two long-recognized but neglected behaviors: the skill of automaticity, or the instant recognition of words; and the technique of overlearning, or repeatedly practicing a skill until it is applied automatically. Together, these behaviors open up options in a field that has been fossilized for decades.
Automaticity is what you are doing now, and that’s what successful readers do within a few months of starting to read. About 25–30 percent of readers “crack the code,” allowing rapid, accurate decoding of words—including words they’ve never seen. Automaticity can occur no matter what system is used: phonics, whole language, balanced reading or watching Sesame Street. This allows the various programs to claim success. What these programs do not acknowledge is the ability of children to take the limited input of any reading program and create an understanding of the total system.
Children with reading difficulties struggle with automaticity. Yet no attempts are made to teach them the skill because of the misguided notion that it can be developed only after extensive phonics practice. Overlearning to Automaticity (OtA) is the key to less-skilled children mastering reading. It requires multiple exposures to a word: typically 20 to 30 exposures on each of about 200 carefully selected words from both the content and noncontent domains. Through varied, engaging exercises, children learn all aspects of a word —its spelling, meaning, sounds, and role in sentences. This establishes the base needed for automaticity. Even if the teaching is limited to only one word a day, I’ve seen that children can achieve this base in just over a year, compared to the three to four years spent on phonics.
In current classrooms, teachers do not have time for the sustained one-to-one instruction that OtA requires. Computers, however, have no such limitation. That’s why I developed the software program cited above. Since then, improvements in technology have made computer-based reading instruction even more appealing and more effective.
Research offers sound ideas for alternative methods of teaching reading—methods we must explore for the health of our children, our nation and our scientific integrity. Phonics should not have monopolistic power. Even if phonics were effective, it would be important to study alternatives, since they might be faster, cheaper or otherwise more appealing. But given the rates of failure that have plagued phonics for generations, alternatives are not only desirable. They are vital.
This is an opinion and analysis article, and the views expressed by the author or authors are not necessarily those of Scientific American.
*Editor’s Note (9/26/23): This sentence was edited after posting to correct the description of few schools using the author’s computer program in 1990.