The “Too Much Phonics” Question

Is there too much phonics instruction in the “science of reading” approach? The question is now on the table, with articles such as these (EdWeek, EdNext), as well as predictably contentious exchanges on social media. It’s unfortunate that yet another debate about reading has devolved on phonics. The “science of reading” emphasizes the importance of extensive instruction about phonics and other “components of reading” (phonemes, morphemes, syllables, vocabulary, etc.). The broader question is about how much of any of this to teach, for how long, for which learners.

As a researcher with a deep interest in education, I have raised several concerns about the science of reading approach (here, here, this, that). One is that it is leading to overteaching: explicit instruction that is both unnecessary and costly because it eats into the time needed for other kinds of learning. These concerns are not directed at teachers and do not speak to what happens in any given classroom. Rather, they are directed at policies and practices as codified in “science of reading” legislation in many states, incorporated in new educational products (such as Heggerty, UFLI, and LETRS) and disseminated by teacher-facing NGOs (such as the Reading League and International Dyslexia Association). The version of the SoR they incorporate holds that children require extensive explicit instruction in many “components of reading,” one of which is phonics. Is this a good idea given (a) what is known about reading, language, and learning and (b) real-world considerations, such as what else might be done with limited instruction time?

I have argued that the SoR is an overcorrection to the inadequacy of basic skills instruction in the Whole Language approach. The SoR is the leave-no-stone-unturned alternative: if you teach the components of reading, children will be able to read. If you don’t teach them (as in Whole Language), they will struggle. This is wrong in my view because its assumptions about how children learn are too limited. They don’t only learn via explicit instruction. In fact, most of the knowledge that underlies basic skills such as recognizing and understanding words and sentences is acquired without explicit instruction, via implicit, “statistical” learning (overview of statistical learning here), while engaged in reading and language activities.

The SoR hasn’t incorporated this type of learning because it wasn’t mentioned in the paltry few resources on which the SoR is based: the Simple View of Reading, the Reading Rope, and the report of the National Reading Panel. The SoR’s view of learning comes from Structured Literacy, an approach based on the needs of dyslexics and other children with learning impairments. Many such children need extended instruction, with elements of print and language broken down into small pieces that can be taught and practiced. Applying the same approach to non-impaired children yields instruction that isn’t necessary because the information can be acquired through the implicit learning mechanism.

In my view, the purpose of explicit instruction is to enable implicit learning. The goal isn’t to teach as much about as many properties of print and language as time allows, appropriating more of the school day to get it all done. Rather, it is to enable the child to start acquiring the relevant knowledge as they engage in reading, writing, and spoken language activities. Explicit instruction is necessary to clue the child in to what there is to learn; they also learn enough basic facts (e.g., about spelling-sound correspondences) to allow new patterns to be assimilated with less reliance on explicit instruction. But the goal is to obviate the need for extensive instruction, freeing the child to focus on reading itself.

I have discussed the relations between explicit instruction and implicit learning in the sources I’ve already cited and in several talks (e.g., here and here). The main point is that the balance between them changes with development. Explicit instruction is important at the outset but if it is successful the balance shifts to implicit learning. Implicit learning depends more on feedback (about whether a word or sentence has been understood correctly) and less on instruction (about how to read a word or sentence).

The Science of Reading seems to condone overteaching, but is it an actual problem? Laws and regulations are one thing; how they are implemented is another. There isn’t reliable empirical evidence as yet about the impact of the “science of reading” on what is being taught or what children are learning. I have raised the concern because overteaching would result from implementing SoR regulations with fidelity. The fact that “science of reading” instruction is now starting in Pre-K and ending in grades 4 and beyond in some areas is also a red flag. My goal is to encourage educators to ask questions: what is the goal of explicit instruction? How much instruction is required to meet that goal? In areas where explicit instruction is necessary (e.g., phonics), are there offramps for children who are ready to move on?

Overteaching of phonics and other basic skills is clearly more of a concern in some contexts and for some children than others. It’s less of a concern for children with dyslexia or another learning disability, for whom an incremental, brick-by-brick approach may be required. For those learners the challenge is gaining access to quality remedial instruction of this kind.

Overteaching also isn’t the main issue in the many schools where Whole Language continues to prevail. For learners in those settings, obtaining sufficient quality basic skills instruction is the bigger problem. Many experienced teachers learned how to teach reading from curricula such as the Calkins “Units of Study,” and remain comfortable with it. New York State recently spent $10 million to develop a new PD course for reading teachers, and what did they come up with? Balanced Literacy. (I’ll have a post about this soon.)

Raising concerns about practices in the “science of reading” invariably brings out the people who think it is all a mistake because there wasn’t anything wrong with how reading was being taught. One of those people, Diane Ravitch, has recently declared “there is no science of reading.” She has not been paying attention. Of course there is a science of reading (and language and learning and development), but using it to improve teacher effectiveness and reading achievement is challenging because of the attitudes of educators such as Ravitch. The fact that the first version of the SoR didn’t get everything right is neither surprising nor damning. The obvious path forward is to build on what is in place, making adjustments as it becomes apparent they are needed.