This post is a response to Wesley Hoover’s comments about my previous post about the Simple View of Reading. I thank him for those comments and hope that this discussion helps to clarify and advance the issues we have raised.
Our concerns are largely different. Hoover is concerned about the scientific status of the Simple View: what the many studies it inspired have shown, whether they have corroborated the original proposal, and how the Simple View compares to later theories of reading. These are not my concerns, for a few reasons. The scope of the original Simple View articles was narrow. As he and I have noted, it is a snapshot of the conditions that exist at the onset of learning to read (child knows spoken language, needs to learn print, etc.) rather than an account of how children learn to read. It did not address questions about what the different types of information relevant to reading involve, how they develop and affect each other, their brain bases, and their use in reading that interest me and many others. Other theories developed in the 40 years since the Simple View was proposed that do address such questions are of greater interest to me, personally. Finally, studies of the Simple View like the ones Hoover has cited have very little relevance to educational issues, as, again, I think we agree. The results of such experiments are usually summarized using statistical models that indicate correlations among a large number of factors related to reading comprehension. Whatever the interest of such models to researchers, they have limited relevance to questions about what to teach, when, how, and so on. As someone interested in bringing cognitive science research to bear on education, I look to other types of research that are more directly relevant.
My discussion of the Simple View has focused on a different question, how it is being used in education. The Simple View is indeed used to support instructional practices that the theory did not address, for which the authors aren’t responsible. That said, there are aspects of the theory and how it was presented that have contributed to problematic tenets of the SoR approach and mistaken beliefs about what modern research has established about reading.
The Simple View has been very useful to teachers as a high-level framework for thinking about the goals of instruction. Reading comprehension does depend on knowledge of language and knowledge of print, broadly construed; helping children to gain and use these types of knowledge is an essential instructional goal. Taken at this level, the Simple View provides a useful touchstone. Nothing said here affects that use of it.
Problems arise from the treatment of print knowledge (aka “decoding”, “word recognition”) and language comprehension as separate components in the reading comprehension formula. In the SoR, the two are treated as different types of knowledge that are acquired in different ways. This is not consistent with a large body of theoretical, computational, and empirical research showing the deep dependencies between the two in learning to read and skilled reading (reviewed in my book and elsewhere). Given these dependencies, treating them as separable for instructional purposes is unwarranted and ill-advised; in the worst case, it could lead to “barking at print”. What happens in SoR classrooms is likely to vary a great deal, of course; thus the extent to which treating print and language as different objects of instruction also varies. I’ve described the several considerations that led me to air the concern. I assume that people can assess whether it is relevant to their particular circumstances.
Hoover has responded by saying that the Simple View isn’t an account of learning to read or how reading should be taught. The concerns I’ve raised therefore have no bearing on its validity, which is his major concern. Fair enough. The important take-home point for educators is that the Simple View didn’t sanction treating print and language as separable components learned in different ways, nor does other research. Setting this idea aside would make room for instructional practices that build on other evidence about how the two work together in learning to read.
Belief that the Simple View incorporated a strong distinction between print and language did, however, arise in part from the authors’ characterization of the theory, which included the following elements:
The Simple View equation treats print and language as separate variables in reading comprehension. The Simple View diagram graphically illustrates this distinction. (Hoover misattributes this diagram to me. It originated somewhere in the SoR community, and has been reproduced in many variants in teaching materials and products. I merely redrew it.)
Because it is a snapshot of conditions at the onset of reading rather than a developmental theory, the Simple View treats the two components as static elements. That was sufficient given the authors’ goals. In the context of teaching reading, however, it reinforces the idea that they are distinct, unlike other theories that examine how they develop and interact with each other over time.
In presenting the Simple View, the authors emphasized the fact that both components are necessary, describing what happens if one or the other component is missing. This frequently-repeated observation also conveys the idea that they are separable.
Gough and Tunmer (1986) and Hoover and Gough (1990) observed that the two types of knowledge are dissociable, mentioning hyperlexics (good at decoding words they do not comprehend), and non-readers (who comprehend spoken words without being able to read them). This pattern is called a “double dissociation” in neuropsychology. These dissociations were cited as evidence that the two types of knowledge are distinct components. These conditions do not show that the two components develop independently under normal conditions. Moreover, research on double dissociations has shown that they can arise from different types of impairments within a single system rather than two separate systems (Plaut, 1995).
Finally, authors of the Simple View emphasize differences in how the two types of knowledge are learned. Whereas language is acquired “naturally,” print is not (Gough & Hillinger, 1980). Language evolved in the species; reading is a technology of recent origin, etc. Everyone in the SoR is familiar with these arguments which, as explained in the previous post, are not wholly correct, a point with significant instructional implications.
These aspects of the Simple View led me to conclude that it got some things wrong. As I have noted, the discussion of the issues in the original articles was more nuanced than in the SoR, which essentially rounded their claims up for simplicity.
Aside from stating that the Simple View didn’t address issues about learning and instruction, Hoover has noted that in the equation, the relation between the two factors is multiplicative. Thus, he asserts that I was wrong to say that the theory didn’t address how they affect each other. He also cites a study (Lonigan et al., 2018) as providing evidence for the multiplicative effect. This point relates to properties of statistical models used to predict reading comprehension scores from measures related to print, language, and other factors. Lonigan et al. showed that including the interaction term (print x language) improved the fit of the model they were testing compared to when it was omitted. They viewed their results as supporting the Simple View. However, other studies (e.g., Kershaw & Schatschneider, 2012) found that including this term did not improve the fit and concluded that the results did not support the Simple View. These and other such studies are interesting for other reasons, but their relation to the Simple View is not straightforward. Such statistical models are far more complex than the Simple View equation, incorporating additional factors and exhibiting complex relations among the parts. Some support the claim that print and language account for nearly all the variance in reading comprehension, but others do not. This is frustrating, but the results of such studies depend on which measures are included (e.g., to assess print knowledge) and whether other potential factors (e.g., working memory, executive function) are included, as well as the type of statistical modeling methodology that is employed.
Finally, such models do not bear on questions about learning to read or instruction, as the authors usually acknowledge. When I describe the interactions between print and language in learning to read and skilled reading, I’m referring to a different class of theories that attempts to account for how these types of information are acquired, represented in memory, and used in reading and other activities.
In the SoR, the Simple View has been represented as providing the scientific evidence for many assertions about learning and instruction that the work did not address. The assertion that the Simple View is supported by “over 150 studies,” repeated many times in the SoR, has contributed to its misuse, in my judgment. It comes from Hoover and Tunmer’s (2021) assessment of their work. The many studies inspired by the Simple View moved into territory it did not address and into issues that couldn’t have been imagined when it was formulated, a reflection of the progress that the field has made. Later studies have yielded a huge variety of results and models, as could be expected given the sheer range of issues that have been investigated. What is common to them is that, like the Simple View, they incorporate knowledge of print and knowledge of language in some form. That isn’t an empirical claim or specific to the Simple View. It is true because of the definition of reading. It is nonetheless the only discernible basis for the claim that 150 studies support the Simple View. Unfortunately, this broad claim encouraged taking the Simple View as the evidence for other assumptions and practices.
What is a teacher to do in response to the concerns I’ve raised? In the short term, not much. I think that deeper understanding of the Simple View and how it is being used will be beneficial to teachers and, eventually, their students. In the past, some teachers have responded to my criticism of tenets of the Science of Reading as if it is criticism of their work–which it unequivocally is not. Consumers of educational workshops, websites, curricula, software, and other products should recognize that the Simple View doesn’t establish their validity or make them “evidence-based.” Authorities who use the Simple View this way are indeed selling another story.
I want to close by noting again the important role that the Simple View has played in reading research. It inspired an unusually large number of research studies that addressed many additional questions and advanced our understanding of reading and learning to read. The fact that we continue to discuss and debate it is a testament to the interest of the work and its lasting import.
References
Gough, P. B., & Hillinger, M. L. (1980). Learning to read: An unnatural act. Bulletin of the Orton Society, 30, 179-196.
Gough, P. B., & Tunmer, W. E. (1986). Decoding, reading, and reading disability. Remedial and special education, 7(1), 6-10.
Hoover, W. A., & Gough, P. B. (1990). The simple view of reading. Reading and writing, 2(2), 127-160.
Hoover, W. A., & Tunmer, W. E. (2021). The primacy of science in communicating advances in the science of reading. Reading Research Quarterly, 57(2), 399-408.
Kershaw, S., & Schatschneider, C. (2012). A latent variable approach to the simple view of reading. Reading and writing, 25(2), 433-464.
Lonigan, C. J., Burgess, S. R., & Schatschneider, C. (2018). Examining the simple view of reading with elementary school children: Still simple after all these years. Remedial and Special Education, 39(5), 260-273.
Plaut, David C. “Double dissociation without modularity: Evidence from connectionist neuropsychology.” Journal of Clinical and Experimental Neuropsychology 17.2 (1995): 291-321.