The Mind, Explained episode 1, Memory, introduces readers to some disorienting facts about human memory, transported in the soothing and authoritative voice-over by Emma Stone.
The episode shares a 9-11 memory from a young woman, recalling sitting as a child in her classroom and watching the smoke from the Twin Tower collapse billowing past the window as she worried about her mother working in the city.
Her memory is vivid and compelling, but it also factually wrong—both the detail of the billowing smoke (the window didn’t face that direction and the proximity of the school would not have allowed that event to occur) and her mother was not in the city that day.
Memory, the episode reveals, is often deeply flawed, as much a construction by the person as any sort of accurate recall.
Watching this, I thought about one of the most misinterpreted poems commonly taught in schools, Robert Frost’s “The Road Not Taken.”
This poem, and how people almost universally misread it, is parallel to the problems with memory in that people tend to impose onto text what we predict or want that text to say; and the verbatim elements of a text, the raw decoding of words, also depends heavily on schema, what the reader knows and the correlations that reader makes with words and phrases.
Frost’s poem, by the way, is about the significance of choosing, in that when we choose we determine our path. But the poem literally states multiple times that the paths are the same; therefore, the poem is not some inspirational poster about making the right choice—although this is the sort of simplistic message many people want to read.
Since I have been wrestling with the recent rise of dyslexia advocates calling for intensive systematic phonics instruction and dyslexia screening for all students as well as the over-reaction to the 2019 reading data from NAEP, I believe the memory and poem interpretation phenomena help explain how and why the “science of reading” narrative is so effective while simultaneously being deeply misguided (and misleading).
The media and most people find a single explanation for reading problems compelling; the argument that more students have dyslexia than are being identified and that one program type (intensive systematic phonics, usually Orton-Gillingham–based) will cure the low reading achievement crisis matches what people want to hear.
The disturbing irony is that those oversimplifying reading challenges and solutions as “the science of reading” are themselves not being very scientific even as they idealize “scientific research.”
I have argued against this in education for many years, and have identified this broadly as technocratic, an over-reliance on narrow types of measurement in order to control the teaching/learning process in ways that are not realistic in real-world classrooms.
The call for reading instruction driven by the “science of reading,” then, comes against several problems. First, literacy acquisition and instruction are both inherently messy and chaotic. Despite our seeking efficient and effective methods, mandating that all students develop the same ways and at the same rates is futile, and harmful. Concurrent with that reality, highly structured teaching of literacy is something that is manageable but likely ineffective, and again, harmful.
Narrow expectations for “scientific” tend to include controlling for external factors and reaching generalizable conclusions—both of which can be inappropriate for guiding teaching real students in an actual classroom.
Should reading policy and practice be informed by scientific studies? Of course, but any teacher must frame that against the needs of each student, needs that may dictate practices outside the parameters of narrow research. And every teacher has another type of evidence—their practice.
We must also, recognize, however, that the science of reading, and the science around dyslexia, are both not as clearcut as some advocates seem to suggest.
When we ask the questions being posed now—Is there a reading crisis (distinct from the historical trends of reading test scores)? Is there a dyslexia crisis? Should all students be screened for dyslexia? Should all students receive intensive systematic phonics instruction?—the answers do not match the current media and advocacy frenzy.
The 2016 International Literacy Associations Reading Advisory on Dyslexia offers a much different framing of “scientific,” in fact:
Both informal and professional discussions about dyslexia often reflect emotional, conceptual, and economic commitments, and they are often not well informed by research.
First, some children, both boys and girls, have more difficulty than others in learning to read and write regardless of their levels of intelligence or creativity. When beginning literacy instruction is engaging and responsive to children’s needs, however, the percentage of school children having continuing difficulty is small (Vellutino et al., 1996; Vellutino, Scanlon, & Lyon, 2000).
Second, the nature and causes of dyslexia, and even the utility of the concept, are still under investigation [emphasis added]….
Third, dyslexia, or severe reading difficulties, do not result from visual problems producing letter and word reversals (Vellutino, 1979)….
Errors in reading and spelling made by children classified as dyslexic are not reliably different from those of younger children who are not classified as dyslexic. Rather, evidence suggests that readers with similar levels of competence make similar kinds of errors. This does not suggest a greater incidence of dyslexia, but instead that some difficulties in learning to work with sounds are normal [emphasis added]….
[I]nterventions that are appropriately responsive to individual needs have been shown to reduce the number of children with continuing difficulties in reading to below 2% of the population [emphasis added] (Vellutino et al., 2000).
As yet, there is no certifiably best method for teaching children who experience reading difficulty [emphasis added] (Mathes et al., 2005). For instance, research does not support the common belief that Orton-Gillingham–based approaches are necessary for students classified as dyslexic [emphasis added] (Ritchey & Goeke, 2007; Turner, 2008; Vaughn & Linan-Thompson, 2003)….
Some have advocated for an assessment process that determines who should and should not be classified as dyslexic, but this process has been shown to be highly variable across states and districts in the United States, of questionable validity, and too often resulting in empirically unsupported, one-sizefits-all program recommendations [emphasis added].
The science of reading actually suggests we should not see a reading crisis in our test data, and not conflate low test scores in reading with special needs such dyslexia.
And we must reject calls to adopt singular evaluation processes and reading programs that claim to address the natural development and challenges of 98% of students.
All general population students and students with special needs need rich and robust literacy instruction that helps teachers recognize their needs in order to foster their natural development over many years (resisting as well the false notion that 3rd grade is a magical moment for all students to attain the same mastery of literacy).
While there are certainly good intentions behind calls for identifying and serving students with dyslexia, the over-reaction to reading test scores and oversimplification of “scientific” are pathologizing and stigmatizing students, and eroding effective teaching and authentic learning.
Like memory and poems by Frost, teaching reading and becoming a reader are complicated. Many find that so unpleasant, they have retreated into a mantra that isn’t itself very well grounded in evidence, the hallmark of “scientific.”
The evidence we use on reading, test scores, has for at least a century shown us that there really is no normal development of reading at predictable benchmarks and that measurable reading achievement has always been and continues to be a powerful marker for the socio-economic status of the students tested.
Technocrats do not want evidence that is historical and sociological, however, preferring instead to impose a problem and a solution onto the data in ways that are as comforting as a detailed, though significantly flawed, memory.
See Also
dr. p.l. (paul) thomas 