There is little evidence that systematic phonics is better than whole language and other common methods used in schools (and no, this is not evidence for whole language)
The title of this post will no doubt be surprising to many. Proponents of phonics can point to multiple meta-analyses and 1000s of papers that strongly endorse systematic phonics. For example, the National Reading Panel or NRP (2000) that has been cited over 20,000 times concludes:
Students taught systematic phonics outperformed students who were taught a variety of nonsystematic or non-phonics programs, including basal programs, whole language approaches, and whole word programs. (NRP, 2000, p. 2-134)”.
Similarly, The Rose Report that led to the legal requirement to teach systematic phonics in English state schools concludes:
Having considered a wide range of evidence, the review has concluded that the case for systematic phonic work is overwhelming …” (Rose, 2006, p. 20).
Strong claims are routinely made in academic articles, such as:
It should be clear that I am advocating here a strong ‘phonics’ approach to teaching, and against a whole-word or whole-language approach… theoretical and laboratory-based arguments converge with school-based studies that prove the inferiority of the whole-word approach in bringing about fast improvements in reading acquisition. (Dehaene, 2011, p. 26).
Similarly strong claims are made by leading researchers in popular books (e.g., Seidenberg, 2017; Willingham, 2017). For example, in “Reading at the Speed of Sight: How We Read, Why So Many Can’t, and What Can Be Done About It”, Mark Seidenberg (2017) writes:
The phonological pathway requires knowing how print relates to sound, the focus on “phonics” instruction… For reading scientists the evidence that the phonological pathway is used in reading and especially important in beginning reading is about as close to conclusive as research on complex human behavior can get. (p. 124)
Given all this, how can I claim that there is little evidence in support of systematic phonics? Why would I even question claim?
Why question the evidence for systematic phonics?
Let me start with the second question first: what motivated me to question this claim? I started looking more carefully at evidence in response to the difficulties I had in publishing a paper co-authored with my brother that detailed an alternative to systematic phonics instruction called Structured Word Inquiry or SWI (Bowers and Bowers, 2017; see previous blogpost for brief summary of the main arguments in this paper). I lost count of the number of places the paper was rejected, and what made the process of publishing this paper particularly frustrating is that reviewers repeatedly mischaracterized many of our claims. For example, reviewers consistently criticized us for rejecting fundamental role that phonology plays in reading despite the fact we clearly wrote the opposite, as in the following quote:
SWI is motivated by a fundamental insight from linguistics, namely, the English spelling system makes sense when the sublexical constraints of morphology, etymology, and phonology are considered in combination… Consistent with phonics, SWI agrees that it is important to teach sublexical grapheme–phoneme correspondences, but it emphasizes that English spellings are organized around the interrelation of morphology, etymology, and phonology and that it is not possible to accurately characterize grapheme–phoneme correspondences in isolation of these other sublexical constraints
Our impression was that many reviewers did not understand our claims because they did not even try. We still find it hard to get many researchers to engage with the ideas now that they are published. Why? Because so many researchers assume that the case for phonics is settled — the “reading war” is over, phonics won, and it is almost irresponsible of us to question the efficacy of systematic phonics given the continuing resistance to phonics in many schools.
It was this absolute commitment to systematic phonics that led me to start looking more carefully at the evidence for phonics. And with a critical eye, it does not take long to discover that the case for systematic phonics is weak at best. The case is built on a house of cards, with 1000s of researches citing the NRP (2000), Rose (2006), and multiple additional meta-analyses without understanding what the reports actually show. Indeed, the authors of these reports and meta-analyses often misunderstand their own findings. The disconnect between the claims and evidence are so striking that I ended up writing the long and detailed paper you can download here.
What is wrong with the evidence with systematic phonics?
Briefly, I’ll summarize two key reasons why the current evidence does not support phonics. First, the findings reported in the meta-analyses are often poorly summarized in the abstracts and conclusions, with small and mixed results described as providing evidence for systematic phonics. These conclusions are then echoed and amplified in subsequent papers that cite these reports. However, as I detail in my review, if you carefully read the result sections (rather than the abstracts) of the meta-analyses, and accept the findings at face value, then there is no reason to be a strong advocate of systematic phonics. A more appropriate conclusion would be something like this: Systematic phonics supports small short-term effects on a pseudoword and word naming, the evidence for long-term effects on pseudoword and word naming is mixed, and that no evidence to suggest that systematic phonics has long-term effects on reading fluency, vocabulary, or reading comprehension (the aspects of reading we should care about most). This is a far cry from the standard claim that the evidence for systematic phonics is overwhelming.
Second, and more importantly, you should not accept the findings at face value. There is also a fundamental conceptual problem with the design with most of the meta-analyses that undermines even these modest conclusions. In two important but largely ignored papers, Camilli et al. (2003, 2006) showed that the NRP (2000) did not even test the hypothesis that systematic phonics is better than whole language and other common methods used in schools. This motivated their new analyses, with Camilli et al. (2003) showing that the effect of systematic phonics is roughly half the size as reported by the NRP (2000), and with an improved analysis, Camilli et al. (2006) failed to observe a significant benefit of systematic compared to non-systematic phonics (as practiced in whole language for example). In my review I show that this same flaw applies to all subsequent meta-analyses, as well as point out multiple additional problems that further undermine the conclusions that are commonly drawn.
What do I mean that the NRP (2000) and subsequent meta-analyses have not even compared systematic phonics to whole language?
What is this flaw that Camilli et al. (2003, 2006) identified? The NRP (2000) assessed the following hypothesis:
…findings provided solid support for the conclusion that systematic phonics instruction makes a more significant contribution to children’s growth in reading than do alternative programs providing unsystematic or no phonics instruction [bold added] (NRP, 2000, p. 2-132).
The bold highlights the key point that systematic phonics was compared to control condition combined to separate conditions, namely (1) intervention studies that included unsystematic phonics and (2) intervention studies that included no phonics. Why is this important? Within the NRP analysis, most whole language interventions included non-systematic phonics, and accordingly, it is possible that the advantage of systematic phonics in the NRP analyses was due to the poor performance in the non-phonics condition, with children in the systematic and non-systematic phonics conditions doing similarly. Indeed, this motivated the Camilli et al. analyses, and when they compared systematic phonics compared to non-systematic phonics, the effects were greatly reduced (Camilli, 2003) or eliminated (Camilli et al., 2006).
This is not nitpicking. The “reading wars” was about whether systematic phonics is better than whole language, and the research community overwhelmingly take the NRP (2000) and many subsequent meta-analyses as providing evidence in support of systematic phonics. Indeed, this has led to the legal requirement in the UK to teach systematics phonics rather than whole language (and related methods) in English schools since 2007. But given that whole language includes non-systematic phonics, and given that there is little or no evidence that systematic phonics is better than non-systematic phonics, the conclusion does not follow. That is, the meta-analyses (other than Camilli et al., 2003, 2006) have not even tested the hypothesis that systematic phonics is better than whole language.
I expect many readers will claim that whole language rarely includes non-systematic phonics. But here is the description of whole language as practiced in the USA from the NRP (2000):
Whereas in the 1960s, it would have been easy to find a 1st grade reading program without any phonics instruction, in the 1980s and 1990s this would be rare… Whole language teachers typically provide some instruction in phonics, usually as part of invented spelling activities or through the use of graphophonemic prompts during reading (Routman, 1996). However, their approach is to teach it unsystematically and incidentally in context as the need arises. The whole language approach regards letter-sound correspondences, referred to as graphophonemics, as just one of three cueing systems (the others being semantic/meaning cues and syntactic/language cues) that are used to read and write text. Whole language teachers believe that phonics instruction should be integrated into meaningful reading, writing, listening, and speaking activities and taught incidentally when they perceive it is needed. As children attempt to use written language for communication, they will discover naturally that they need to know about letter-sound relationships and how letters function in reading and writing. When this need becomes evident, teachers are expected to respond by providing the instruction. (p. 1-202)
Similarly, in the UK, prior to the mandatory requirement to teach systematic phonics in the UK, Her Majesty’s Inspectorate (1990) reported on the teaching and learning of reading observed in 470 classes and over 2,000 children. They wrote:
…phonic skills were taught almost universally and usually to beneficial effect’’ (p. 2) and that ‘‘Successful teachers of reading and the majority of schools used a mix of methods each reinforcing the other as the children’s reading developed (p. 15).
So to summarize, apart from the Camilli (2003, 2006) meta-analyses, no meta-analysis even tested the hypothesis that systematic phonics is better than whole language. When the hypothesis was tested, the benefits of systematic phonics was no longer significant.
My review identifies many additional problems with the evidence taken to support systematic phonics (the evidence for systematic phonics is even weaker than Camilli et al. claim), but I’ll leave this for readers interested enough to download my paper.
What not to conclude
I am NOT claiming that grapheme-phoneme correspondences are unimportant. Indeed, Camilli’s reanalyzes of the NRP (2000) report showed that systematic phonics is better than a no phonics control condition. In addition, my review should NOT be taken as an excuse to continue to teach whole language. Camilli et al. (2006) and I have shown that there is no evidence that one approach is better than the other. It turns out that the reading wars was a draw. This contradicts the common view that the systematic phonics is best practice, but it does not provide any support for whole language.
What to conclude
I am claiming that researchers should be less committed to systematic phonics, and more open alternative methods. Indeed, this is the main message I want to get across with my detailed review of phonics – it is time to consider alternatives to whole language and systematic phonics
This brings me back to Structured Word Inquiry or SWI. A key feature of SWI is that is systematically teaches grapheme-phoneme correspondences (something that proponents of phonics should like) but within the context of morphology and etymology. As a consequence, SWI does a better job at describing grapheme-phoneme correspondences. At the same time, SWI teaches children systematic spelling-meaning correspondences with the goal of making sense of the English spelling system, and highlights the meaningful relations between words (something that critics of phonics might like). Bowers and Bowers (2017) emphasized the theoretical motivation for SWI, but we also reviewed some empirical evidence for this approach, including one study that showed the SWI was more effective than phonics in a randomized controlled study with 5 to 7 year-old children (Devonshire et al., 2013), and more generally, systematic reviews and meta-analyses found morphological instruction to be more effective in younger and struggling readers (Bowers et al., 2010; Carlisle, 2010; Goodwin et al., 2010, 2013). There is no evidence for the claim that SWI or morphological instruction should only be introduced in older children after systematic phonics (e.g., Rastle, 2018).
To avoid any conclusion, I want to emphasize that more studies are needed to assess the efficacy of SWI when introduced at the start of instruction or later. I hope a better appreciation of the (lack of) evidence for systematic phonics will help motivate more research. Indeed, given the lack of evidence that systematic phonics is better than whole language, there is every reason to assess new methods, including the hypothesis that children should be taught how their writing system works. For an easy introduction to SWI, I suggest going to the publication page on this website and reading Bowers and Bowes (in press), and for more detail, Bowers and Bowers (2017). And if you want even more information, I suggest you go to my brothers’ website here.
Any comments welcome!