The Short and the Long of It: Hooked on Phonics

During the period of Whole Language when classroom teachers had a seat at the policy table for a brief shining moment—a folding chair, but a seat, I was employed as an elementary reading specialist (1989-1991) and worked hand in glove for a six week period with an inspiring team of kindergarten and first grade teachers at an affluent suburban school and at a working-class school. My task was to help build instructional capacity in early writing pedagogy. I learned far more from them than they learned from me.

There was very little daylight between phonics and spelling in those classrooms. Unique to this historical moment, policymakers in California were more interested in a principled long range vision of teaching and learning than in mandates at the third and fourth degrees. These teachers taught their kiddos phonics every day, all day, directly and indirectly, systematically and serendipitously.

It didn’t matter whether class was about science or math or folklore, all content came wrapped in words, and all words were fair game for phonics. They taught reading and writing like breathing, and phonics was oxygen. Of course, to them Whole Language meant phonics, phonics, and more phonics since kids who can’t spell and decode naturally get hooked on phonics—any time a word needed sounding out to speak or spell got top priority.

Teachers often gathered the children on a rug for a read aloud or a write along. They would display large empty chart tablets or pocket charts in front of children to model how to write books using colorful markers then and there in front of children’s eyes, everyone working together to help the teacher to spell words. These books remained in the classroom, and each child got one to take home at the end of the year.

Zoo Phonics was a popular program. I’m not sure if it’s still legal. Does anyone know if it’s still legal? Maybe by waiver? Zoo Phonics helped children sound out words by associating each letter of the alphabet with an animal character. Children would act out movements associated with the animals, a kinetic-cognitive twofer.

Invented spelling was a crucial strategy classroom authors used to compose. Daily journal writing beginning with scribbles and scrawls in kindergarten morphed into full blown text by spring—seeds nourished by conversation. Are teachers still doing this?

These words belonged to the children, not to their teachers. The walls were made of words.

*****

Fast forward to 2024, the Science of Reading Era. The official view in every state in the United States and in Canada looks at Whole Language as a massive failure. The legacy of Whole Language that survived through the first back to basics winter during the Bush era is now illegal in the northern hemisphere. Reading Recovery, a professional pedagogy with empirical evidence galore supporting its efficacy (cf: my previous post), has been shelved—based on a theory of instructional delivery of phonics with ambiguous empirical support and serious logical errors.

In a whirlwind of research, the field of reading from the 1970s through the 1990s accomplished a goal set for the field at the turn of the century by Edmund Huey (1908) and then it experienced the whiplash of No Child Left Behind and the Science of Reading copyright ©️ . Early in the century Huey and his followers longed for a scientific understanding of the mystery of silent comprehension of text. The last third of the century produced a robust, empirically grounded theoretical framework for comprehension with models converging on the nature of reader-text interactions fueled by motivation, fluency, prior knowledge, language processes, and metacognition resulting in meaning-making.

The copyrighted science of reading position accepted by policy makers rejects mainstream literacy science in two profound ways. First, the mainstream position holds that reading comprehension is the outcome of orchestrated processes operating simultaneously, recursively, and reciprocally in real time. The copyright science holds that reading can be accomplished in the absence of comprehension and should be taught apart from comprehension processes from the beginning. Second, the mainstream position holds that growth in reading performance stems from pedagogy aimed at orchestration of subprocesses rather than fragmented skills. The copyrighted science insists that all children should receive instruction in an approved systematic phonics program with a gatekeeping test to determine when a child is ready to engage in orchestrated reading. Students are denied access to real texts when they most need real texts to learn.

In my last post I discussed Tierney and Pearson’s (2024) monograph fact-checking this break-away science. Now I want to discuss the scientific basis put forward by the copyright science published in support of a popular and legal phonics program called LETRs.

*****

The LETRs curriculum was written by Dr. Louisa Cook Moats, a classic example of a sanctioned phonics programs. The following document from which I excerpt is available online at lexialearning.com/letrs.

Note that this screenshot was excerpted from a three page pdf file. Tierney and Pearson’s (2024) fact-checking monograph, is 188 pages. Here is the overarching case made for the LETRs.

What Scientific Research Informs LETRS?

LETRS is grounded in the science of reading. Teaching reading is rocket science, as stated by author Louisa C. Moats (2020a). The concepts and instructional approaches of LETRS are aligned with respected sources such as the Knowledge and Practice Standards for Teachers of Reading (Moats et al., 2010), the Elements of Effective Instruction (Florida Center for Reading Research, 2006), and Classroom Reading Instruction that Supports Struggling Readers: Key Components for Effective Teaching (Denton, n.d.). In addition, this course incorporates reading research conducted in neuroscience, cognitive development psychology, and linguistics so that educators have solid evidence on how to teach reading to benefit all students. Some notable research that informs LETRS is listed here.

As near as I can tell, the research base supporting the LETRs materials written by Louisa Moats includes a set of professional teaching standards also written by Moats. A second person is cited, a Denton, whose written matter is not dated and impossible to retrieve. Grounded in the science of reading, teaching reading is rocket science, according to the flyer. Isn’t it odd for a science to tout that its research has solid evidence? Is the implication that the other science does not?

As Tierney and Pearson (2024) documented in their fact-checking monograph, the LETRs claim, that the 3-cueing system in reading theory has been debunked, is mistaken. For curious readers here is the research advocating for LETRs which the copyright science claims debunks the 3-cueing system:

“The Four-Part Processing Model for Word Recognition (Seidenberg & McClelland, 1989; Seidenberg, 2017) is supported by modern brain science. In the Four-Part Processing Model, the phonological processing system is distinct from the orthographic processing system. Students need to be taught speech sounds and print patterns and then how orthography maps to speech. In contrast to this research-based model, the Three-Cueing Systems Model suggests readers depend on syntax and semantics to guess unknown words.”

Seidenberg and McClelland (1989¹) developed a model of subprocesses for word recognition. According to the LETRs interpretation of the 1989 paper, the theorists sever phonology (sounds) and orthography (spellings) from one another. In the paper, however, the authors are clear: Their interest is in parallel processing, not serial processes dumping output stepwise. In other words, Seidenberg et al. (1989) wanted to model how it’s possible to get from decoding a word to comprehending a text all at once: “Word recognition provides an interesting domain in which to explore the properties of the connectionist or parallel distributed processing approach to understanding perception, cognition, and learning” (p. 524). Later in the paper, the authors leave no question about whether the cueing systems operate serially and individually. The processes work simultaneously:

“The larger framework assumes that reading words involves the computation of three types of codes: orthographic, phonological, and semantic. Other codes are probably also computed (concerning, e.g., the syntactic and thematic functions of words), but we have not included them in the present model because they probably are more relevant to comprehension processes than to the recognition and pronunciation of monosyllabic words” (p.526).

It’s important to note the tentative modality in this excerpt, the statement of assumptions that are reasonable but not described as settled, the caution that syntactic and thematic functions of words “probably” are processed, that syntax was omitted because of relevance within the model. There is nothing in this article as near as I can see to warrant its citation as settled scientific evidence that debunks the 3-cueing system. On the contrary it supports the 3-cueing system.

*****

In their conclusion, Seidenberg et al. (1989) make clear that their model is not the last word on anything. They had nothing at all to say about 3-cueing pedagogy except that sentence structure, spellings, and word meanings must a priori work together in parallel. What fascinated them remained mysterious and open to future research. Moreover, this paper does not recommend in any way that the first ten weeks of reading instruction in first grade should consist of a daily diet of phonics in isolation or in nonsense books. Yet LETRs cited this paper as an important anchor of the copyright science.

I close here with the conclusion of their 1989 paper. Note the importance of the last sentence, cautioning that in this model the word reading process has been “artificially isolated,” precisely what LETRs is designed to to for beginning reading instruction—artificially isolate phonics from reading and then claim phonics is reading:

“Our basic claim is that the model [works] because of the close fit between the nature of the task (learning the structure of English orthography) and the capabilities of models of this type. English orthography is not strictly regular, and so it is not well captured by mechanisms involving systems of rules. Attempts to patch up this problem by proposing two routes (rules and lexical lookup) have been offered by others, but they have not been entirely successful. Our model, and others like it, offers an alternative that dispenses with this two-route view in favor of a single system that also seems to do a better job of accounting for the behavioral data. It remains for future research to establish whether the present approach can be successfully extended to longer words and to other aspects of word reading, and to integrate the word reading process, here artificially isolated, back into the process of understanding text” (p.564).

1. Psychological Review

1989, Vol. 96, No. 4,523-56