That New Yorker article
The New Yorker article “Alphabet Soup” (Dec 29/Jan 5 issue) did a good job of conveying the complexity of dyslexia, the frustrations of dyslexics and their families, and the promise of schools such as the Central Brooklyn Literacy Academy. It covered a lot of ground, and I know it wasn’t generated by AI because it was so well written. I am less enthusiastic about other aspects of the article. For better and worse, it accurately conveyed some beliefs about reading and dyslexia that are tenets of the “science of reading” approach but mistaken from my perspective as someone who both knows the research literature and recognizes the challenges of connecting basic research and educational practice. The article illustrates problems that arise from conflating the needs of dyslexics and nondyslexics, which differ in important ways. Because the New Yorker article will be so widely read, and because I feel strongly about the need to make better use of what is known from the research literature, I feel compelled to try to clarify the issues. They are discussed in greater detail here, here, here and in the new edition of my book that will appear later this year.
The New Yorker article reviews familiar elements of the “science of reading” story: educators had stuck with ineffective methods (the “theoretical zombies”) that weren’t consistent with facts about how reading works and children learn from extensive basic research; once this became broadly known, states passed “science of reading” legislation that allowed the adoption of research-based practices; the admirable Mississippi reading project shows that this approach is effective.(1,2)
Instruction in the SoR approach follows the precepts of “Structured Literacy,” in which elements of reading are taught systematically, explicitly, and incrementally. Structured Literacy was developed to meet the needs of dyslexics, but as an anonymous administrator at a school for children with reading disabilities told the author, “What works for our students actually works for everyone. It’s a matter of dosage.” I take issue with that claim. The instructional needs of dyslexics and nondyslexics are not the same. What is best for dyslexics, children who have conditions that interfere with learning to read, is not best for nondyslexics, who do not.
We know from basic research that all children need to learn the same things in order to read (leaving aside secondary differences due to properties of writing systems and languages). We also know that all children (except the ones who are already reading) need some explicit instruction about print and print-sound mappings at the outset. However, we also know that all children do not learn the same way. For a variety of reasons, dyslexics have more difficulty learning to read, meaning that they do not gain as much from standard instruction as nondyslexics. Breaking reading down into small components that can be explicitly taught over an extended period of time is meant to compensate for their impairments. With years of practice and the support that private schools such as Windward, the one mentioned in the article, can provide, many dyslexics become good readers.
Nondyslexic children do not require such extensive instruction because, well, they are not dyslexic. For one thing, they are much better able to learn using a mechanism called implicit learning. Once typically developing children have been taught enough to “crack the code,” they can begin learning from reading and other, related activities, relying less on explicit instruction. This idea has a long history in reading research, but has been overlooked in the SoR. (3)
What I’ve described is more than just a difference in “dosage:” it’s a difference in the goals of instruction. For typically developing children, the goal is “cracking the code,” which happens when they have learned what there is to learn (patterns linking print, sound, and meaning) and enough patterns and words to begin reading sentences and longer texts. The basic foundation provided by explicit instruction scaffolds implicit learning, which allows the child to assimilate additional patterns from a variety of activities including reading itself. Most of the knowledge that supports reading is acquired via this mechanism rather than explicit instruction. For dyslexics, instruction proceeds from the observation that this approach doesn’t work well enough for them. For many dyslexics, breaking reading down into smaller parts that can be explicitly taught and practiced over a long period of time is an effective alternative. (4)
In short, a curriculum that emphasizes building reading skill brick by brick via explicit instruction may be an effective way to accommodate the needs of children whose learning capacities are impaired. It is inappropriate for non-impaired children because it underutilizes their other strengths. Although the authorities in this field usually acknowledge the obvious point that nondyslexics do not need as much instruction as dyslexics, in practice the adoption of “structured literacy” has led to overteaching. Reading instruction is taking up more of the school day, and in many places it is starting earlier (in pre-K) and ending later (grades 3-4-5). In my view this wastes precious time that could be used for teaching and learning other things. Using this approach with typically developing children levels the playing field in a regrettable way, making learning to read as challenging for them as for dyslexics.
For these reasons, it was unfortunate that the article presented Structured Literacy as the evidence-based, scientifically-approved approach to reading instruction. The idea that systematic, explicit, extended instruction about the many components of reading is an effective way to teach reading is a hypothesis that is now being tested on a massive scale in American schools. There is exactly zero evidence about its effectiveness in teaching nondyslexics to read, or about the efficacy of the approach compared to other methods that do not rely as heavily on explicit instruction. In the absence of such evidence one has to look at the validity of the assumptions about reading and learning that it incorporates. Here there are numerous problems, starting the overemphasis on explicit instruction and the underrecognition of learning and language differences in reading development. Again, I’ve discussed these elsewhere and will continue to do so. The main conclusion is that, as a correction to the neglect of basic skills instruction in Whole Language, Structured Literacy, and with it the “science of reading” approach, have overshot the target.
Is Structured Literacy an effective approach for children with reading and learning impairments? That is a tricky question that requires more than a blog post to address. Extended instruction about components of reading is apparently effective for many dyslexic children. However, the empirical evidence for the effectiveness of specific programs is sparse. The New Yorker article highlights the Orton-Gillingham approach, which is widely used in remediating dyslexia and often presented as the gold-standard method. I do not know if the method is effective in treating dyslexia. I do know that it has been difficult to establish its effectiveness in systematic, controlled studies (see, e.g., Stevens et al., 2021). Whatever its role in helping dyslexics, I shudder at the idea that people are taking it as an “evidence-based” approach to teaching reading in general. The use of such methods for this purpose would be grossly inappropriate. (5)
In short, I wish the article had done a better job of recognizing how differences in children’s learning capacities affect the balance between what needs to be explicitly taught and what can be learned by other means. More importantly, it’s essential for the “science of reading” to do so. Although everyone requires explicit instruction at the outset, the needs of dyslexics and nondyslexics rapidly diverge.
Footnotes:
1. The Mississippi project was a remarkable effort that seems to have yielded real, lasting gains in literacy. It was a multi-pronged approach involving cooperation among numerous stakeholders as well as a significant infusion of funding from private philanthropy. We don’t know which aspects of this effort were crucial; I think it very likely that having a cadre of well-trained coaches advising teachers on site, in the schools, was a huge factor. Whether other states are providing this level of teacher support is unclear. In any case, the successes in Mississippi do not confer validity on Structured Literacy, which was not codified until well after that effort was underway.
2. The New Yorker’s version of this history is not accurate in detail. Scientists dating from Keith Stanovich in the 1980s and Marilyn Adams in 1990 had documented Whole Language practices that were contradicted by reading research (see my book for these and other references). I devoted a chapter to documenting how Kenneth Goodman and other whole language ideologues had promoted their discredited views. To everyone’s benefit, Emily Hanford then brought this story to a much wider audience in a compelling series of podcasts.
3. Gough & Hillinger, 1980. Here.
4. One of the continuing puzzles about dyslexia is what to do for the substantial number of children for whom these methods are not effective.
5. There are some unsettling parallels between methods such as Orton-Gillingham and Lindamood-Bell on the one hand, and Reading Recovery on the other. Reading Recovery is an intervention for 1st grade children who were doing poorly in Whole Language settings. Like the methods for remediating dyslexia, Reading Recovery was promoted by a national organization that provided certification for instructors who obtained specialized training in the approach. As with the other methods, practitioners were enthusiastic proponents of the approach despite the lack of evidence for its effectiveness.
