TCLD Page Banner

Bowers, J. S. (2020). Reconsidering the evidence that systematic phonics is more effective than alternative methods of reading instruction. Educational Psychology Review, 32, 681–705.

Summary by Dr. Jack M. Fletcher

Overview and Study Background

Bowers (2020) reviewed 12 meta-analyses addressing the efficacy of systematic phonics relative to other methods of teaching reading, such as whole language. He also reviewed national data in England that purport to support the efficacy of teaching systematic phonics. His starting point is a familiar refrain on the findings of the 2000 National Reading Panel report (NRP; NICHD, 2000), which found through meta-analysis that systematic phonics was an effective component of early reading instruction compared to some or no phonics instruction. Bowers cast this question as a comparison of phonics versus whole language, which was then characterized as a comparison of approaches to reading that emphasize systematic phonics relative to incidental or no phonics instruction. He argues that, although there is a strong consensus about the superiority of systematic phonics instruction to alternative approaches to teaching reading, the evidence has been exaggerated such that the effects are either less robust than reported or don’t support the effectiveness of systematic phonics instruction at all.


Bowers constructed the paper in three sections. First, he reviewed different methods of reading instruction and defined systematic phonics as a phonology-first approach:

... systematic phonics explicitly teaches children grapheme-phoneme correspondences prior to emphasizing the meanings of written words in text (as in whole language or balanced literacy instruction) or the meaning of written words in isolation (as in morphological instruction). That is, systematic phonics is committed to the ‘phonology first’ hypothesis ... It is called systematic because it teaches grapheme-phoneme correspondences in an organized sequence as opposed to incidentally or on a ‘when-needed’ basis. Several versions of systematic phonics exist (most notably synthetic and analytic), but they all adopt the phonology first hypothesis.

He then contrasted this approach with meaning-based approaches in which phonics instruction is incidental or absent. Second, Bowers reviewed 12 meta-analyses of experimental reading research that addressed the effectiveness of systematic phonics instruction. Third, he reviewed the results of what he described as large naturalistic experiments emanating from a mandate in the United Kingdom to teach systematic phonics, which included extensive professional development.

Key Finding

Bowers concludes that the evidence supporting the superiority of systematic phonics to what he describes as alternative reading methods is weak to nonexistent. He argues that any presumed scientific consensus on how reading should be taught is premature and based on unsettled science devoted to rhetoric on the superiority of systematic phonics. He concludes:

Despite the widespread support for systematic phonics within the research literature, there is little or no evidence that this approach is more effective than many of the most common alternative methods used in school, including whole language. This does not mean that learning grapheme-phoneme correspondences is unimportant, but it does mean that there is little or no empirical evidence that systematic phonics leads to better reading outcomes.

He also concludes, “‘The reading wars’ that pitted systematic phonics against whole language is best characterized as a draw.”


Drs. Robert Savage, Sharon Vaughn, and I responded to Bowers in A Commentary on Bowers (2020) and the Role of Phonics Instruction in Reading.

In our response, we observed five limitations of his conclusions:

  1. Definition issues
  2. What is the right question?
  3. The assumption of “phonics first”
  4. Systematic versus explicit phonics
  5. Misconstrual of evidence from meta-analyses

To summarize, there is replicable, positive evidence that supports explicitly teaching phonics. This instruction should not represent an isolated reading program, but is an essential part of comprehensive, integrated approaches to reading instruction. This instruction needs to be personalized to the learning needs of each child and will vary according to age and ability level. Phonics instruction should not occur in isolation of instruction in the meanings of words and how words are spelled. Rather, this instruction should occur concurrently to take advantage of the word and world knowledge the child brings to literacy learning. Learning to read requires multiple competencies. As such, the question of the superiority of phonics instruction versus other methods is outmoded. Rather, the focus of research and practice should be on combining evidence-based reading instruction to meet the learning needs of each child.

1. Definition issues. Bowers (2020) portrays phonics instruction as phonology first before meaning-based instruction, contrasting it to whole language and balanced literacy approaches as meaning before phonology. We argue that this definition of contemporary reading instruction is misleading. In addition, we question his definition of whole language and balanced literacy as typically including some phonics, but agree that the phonics instruction that is provided is incidental and rarely explicit. We note that advocates of whole language and balanced literacy see phonics instruction as a tool to be implemented when needed, usually in conjunction with the triple cuing approach to word recognition, widely recognized as inappropriate because it teaches children to guess at words they do not know instead of sounding them out. Recent reviews of balanced literacy programs implemented as core reading programs across the country echo these concerns.

2. What is the right question? Like many contemporary accounts of the best methods for reading instruction, Bowers frames his arguments as systematic phonics versus other reading methods. This contrast is an artificial dichotomy that cannot be completely addressed through empirical synthesis because few studies can be coded well enough to characterize all the components of reading instruction. Rarely is phonics the only method utilized in the intervention. Additional moderators, such as group size, other language and literacy activities, and the amount of time in general education and supplemental intervention, affect outcomes. Although extreme views of the need for isolated phonics instruction can be found, the NRP and other consensus statements view the development of literacy in terms of the integration of instruction in decoding through phonics, fluency through practice, and comprehension, the latter derived in part from oral language skills involving vocabulary, background knowledge, and strategic knowledge. Reid Lyon’s 1998 testimony to Congress preceded the NRP in describing reading as a multicomponent process in which explicit phonics instruction is necessary, but not sufficient, for many children.   

3. The assumption of phonics-first. Bowers concludes that proponents of systematic phonics indicate that phonics instruction should occur before any focus on larger units of words (e.g., morphology) or on meaning. We are unaware of researchers who believe that children should learn a full set of phoneme-grapheme correspondence rules before being introduced to morphology and meaning. Even when emphasizing phonics, reading instruction does not occur in a vacuum. Children are taught and exposed to vocabulary from the beginning of language development through reading development. They learn new things about the world and improve background knowledge. There should be a focus on strategies for understanding what is heard and read. Similarly, children learn to read print by accessing sublexical parts of words, linking what words sound and look like. This is true in a beginning reader and in an illiterate adult. What confuses us is the enormous variation in how easy or difficult this task can be for the novice reader. Their view of phonics instruction as phonics before meaning or morphology does not align with contemporary approaches to effective reading instruction.

4. Systematic versus explicit. Although the term “systematic phonics” is widely used, it is instructive to think about what “systematic” really means. In general, systematic typically refers to teaching of grapho-phoneme correspondence rules in a prescribed and often manualized sequence, with a defined scope and sequence. Bowers is correct in that the evidence in support of the need for a specific scope and sequence in which the child has to learn grapheme-phoneme rules as a prerequisite for decoding has a weak evidence base. There are many ways to teach phonics, not all of which are systematic. However, effective approaches are explicit, intentional, and teacher-led (Vaughn & Fletcher, in press). As we point out in our commentary, there are studies that show little difference in outcomes in comparisons of approaches in which phonics is taught with a rigid scope and sequence versus explicit and intention approaches that are embedded in the curriculum.

5. Misconstrual of evidence from meta-analyses. We found many interpretative issues in Bowers’ review of the 12 meta-analyses, and added two meta-analyses that are more recent. The bottom line is that regardless of Bowers’ reinterpretation, the evidence is consistently positive, indicating strong replicability of the meta-analytics findings. We conclude that the advantage of reading programs that include explicit phonics instruction is consistently reported for typical and struggling readers,  across multiple languages, in second-language learners and immigrants, and that the effects persist for at least 1 year postintervention.

Implications for Practitioners

There is a science of reading that addresses how children learn to read and why some fail. This science shows that learning to read is a multicomponent process involving multiple competencies. The science shows how children progress from accessing sublexical units of words and indirectly accessing meaning to becoming lexical readers who process orthographic relations rapidly with immediate access to meaning (i.e., language at the speed of sight; Seidenberg, 2017). This account does not show that children figure out words based on different cues or that learning to read is as natural as learning to speak.

We do not have a strong science of teaching children to read (Seidenberg, 2017). Although ideally we could break down the teaching of reading into a very precise package outlining the role of each component and the best way to teach it, unpacking these components is difficult to do and would vary across children. Posing comparisons of systematic phonics versus alternative reading approaches is artificial. As legislation is passed mandating certain approaches to reading instruction, and even worse, prescribing specific commercial programs for children identified with dyslexia, we would do well to ask new questions and embrace the complexity and multiple competency view of reading put forward by Reid Lyon in his 1998 testimony to Congress, which was the main message of the NRP. As Reid Lyon stated in his recent tribute to Duane Alexander, “As I observe the emergence of a new discussion about the best ways to teach children to read, I fear that the discussion will turn again to the age-old issue of phonics versus some type of immersion program.” Even in the area of phonics, there are many ways to help children access sublexical components of words and one approach does not work with all children, even those identified with dyslexia.

The correct approach is to move past the reading wars and the code-meaning pendulum and embrace comprehensive approaches that integrate, as opposed to balance, different components of reading instruction in ways that are responsive to students’ learning needs. For many children, these components need to be taught explicitly. In particular, children who have trouble accessing words because of problems with phonological processing need explicit code-based instruction. They also need to practice to build automaticity. Comprehension instruction should also be more explicit because children do not develop vocabulary, background knowledge, and comprehension strategies through osmosis or passive reading. A focus on how to integrate these activities in a comprehensive approach to reading instruction is vital to future progress in reducing the effects of poor literacy skills.

Related Reading

If you are interested in another response to Bowers’ article, please see Jennifer Buckingham’s blog post titled, "The Grass is Not Greener on Jeffrey Bowers’ Side of the Fence: Systematic Phonics Belongs in Evidence-Based Reading programs," as well as a revised version of this post, "Systematic Phonics Instruction Belongs in Evidence-Based Reading Programs: A Response to Bowers," published in The Educational and Developmental Psychologist. Bowers' response to Buckingham can be read on his website: "Responding to Buckingham’s post."


Bowers, J. S. (2020). Reconsidering the evidence that systematic phonics is more effective than alternative methods of reading instruction. Educational Psychology Review, 32, 681–705.

Fletcher, J. M., Savage, R., & Vaughn, S. (2020). A commentary on Bowers (2020) and the role of phonics instruction in reading. Educational Psychology Review.

National Institute of Child Health and Human Development. (2000). Report of the National Reading Panel. Teaching children to read: An evidence-based assessment of the scientific literature on reading and its implications for reading instruction (NIH Publication No. 00–4769). U.S. Government Printing Office.

Seidenberg, M. (2017). Language at the speed of sight: How we read, why so many cannot, and what can be done about it. Basic Books.

Vaughn, S., & Fletcher, J.M. (in press). Identifying and teaching students with significant reading problems. American Educator.