Where did Phonemic Awareness training come from?

Our story begins with the report of the National Reading Panel (2001), which listed phonemic awareness as a teachable component of reading with beneficial effects. They reviewed studies to that date that included conditions in which children engaged in activities for developing phonemic awareness (PA). These activities varied a great deal. Some only involved spoken language; some involved both speech and print; some included several activities. A meta-analysis showed that engaging in these activities was associated with increased reading skill. Thus PA was listed as one of the teachable components of reading for which there was substantive empirical support. The SoR movement has run with this recommendation. Phonemic awareness instruction is now a major phase in learning to read, the prerequisite to phonics. Many of the state SoR laws specifically mandate including phonemic awareness instruction. 

This approach is based on a misunderstanding of the research literature and is an unwarranted extrapolation from the NRP’s findings and recommendations. It is the area where the SoR approach has gone farthest off the rails, in my view, because so much time is being spent on an activity for which there is so little justification. Here I summarize the rationale behind PA instruction and then consider the practice in light of what is known about phonemes and reading, as discussed in previous chapters of this book and elsewhere. 

The rationale for PA instruction is:

Beginning readers need to learn how print (orthography) represents spoken words (phonology). 

The units used in print are called graphemes, which are individual letters such as S and H, and letter combinations such as SH. 

Spoken words consist of units called phonemes; they are to speech what letters are to print. 

Although phonemes are used in comprehending and producing spoken words, we do this without conscious awareness.

Learning to read requires gaining awareness of phonemes, which is the purpose of phonemic awareness instruction.

The beginning reader is taught the 44 phonemes of English, by analogy to being taught 26 letters (plus digraphs such as SH). 

They can then learn correspondences between graphemes and phonemes: phonics. 

Using those correspondences to pronounce words requires blending a sequence of phonemes.

Phonemes that are pronounced correctly can be blended more easily. 

Following this reasoning, the logical sequence for reading instruction is to start with teaching phonemes. If phonemic awareness is important for reading, knowing more of them will be better, and teaching all 44 leaves no phoneme behind. Phonemes can be taught using oral language activities such as hearing a word and then saying it with one of the phonemes deleted. A child who exhibits the requisite level of phonemic awareness as measured by performance on such tasks is then ready to move on to phonics.

Here are some facts about phonemes:

Spoken words do not consist of discrete phonemes that are “blended” together. 

Using speech doesn’t require knowledge of phonemes, conscious or unconscious.

Rather than being literally pronounced, phonemes are an abstraction: a way of thinking about spoken words. We treat words as if they consisted of discrete sounds.

Learning to read doesn’t require having conscious awareness of phonemes, learning their “correct” pronunciations, or performing at a prescribed level on phonemic awareness tasks. Rather, children need to learn to treat spoken words as if they consist of discrete sounds. 

How is this abstraction achieved? Reading comes with its own solution to the problem: it develops through activities in which print and sound are paired, such as reading aloud and spelling to dictation. Knowledge of phonemes is tacit–something the brain learns without telling us. 

Rather than being the precursor to reading, “phonemic awareness” results from progress in learning to read alphabetic writing. It emerges over time as spelling changes the neural representations of spoken words.

From this perspective many of the activities associated with phonemic awareness instruction seem downright peculiar. Learning phonemes is said to be essential for reading, but how did anyone learn to read before they began to be taught? English has 44 phonemes? Phonemes are an abstraction and so the exact number depends on which phonological theory one is using; you can’t just listen to words and count them. Teachers are being told they need to learn the correct pronunciations of phonemes in isolation, but there aren’t any. The way a phoneme (abstract unit) is realized in speech (articulation) depends on properties of the surrounding phonemes.  What people are practicing are weird, unnatural blips of sound rather than naturalistic segments of spoken words. 

My account of phonemic awareness differs greatly from the one in the SoR. Aside from delving into the research literature (not for the faint-hearted), how is a person to know which is correct? As a scientist I certainly do not recommend relying on the say-so of any single authority, including me. You need some additional evidence to go on. Here are two bits to consider.

The first is that the world happens to have conducted a highly relevant experiment. Consider how reading is taught in England (and much of the UK). Leaving aside minor but entertaining differences in spelling and vocabulary, learning to read in England is very similar to learning to read in the US. They too have issues about reading achievement, achievement gaps, economic circumstances, variability in language background, and so on. They too had adopted a Whole Language-like approach (the “searchlight” model), and have turned to methods that include basic skills instruction, among other things. 

The interesting difference is that they have not adopted phonemic awareness instruction. They have emphasized the importance of early language experience, and learning the alphabet and a sound for each letter (e.g., T and “t” as in “top”) for reading-readiness (they call this “sounding”, a useful term). The child then goes straight into learning to read, which includes phonics, reading aloud, spelling, and other familiar elements. British children manage to learn to read without phonemic awareness instruction and they do so about as well our students. The British experiment suggest that PA instruction isn’t necessary and doesn’t confer special benefits. 

Phonemic awareness was adopted in the US but not the UK for two reasons.

In another parallel to the US, the UK conducted its own review of the scientific evidence about learning to read. The Independent Review of the Teaching of Early Reading, known as the Rose Report (after Sir Jim Rose, who chaired it), was published in 2006. It reviewed much of the same literature as the NRP and considered most of the same topics, including phonics, vocabulary, and language development. The report discussed phonemic awareness and the importance of being able to “hear, identify, segment and blend phonemes in words.” The emphasis was somewhat different, however: PA was discussed specifically in the context of the alphabetic principle, that graphemes correspond to phonemes. That might seem like a small difference, but framing it this way emphasizes the importance of orthography–the alphabetic part–more than the term phonemic awareness. Unlike the NRP, phonemic awareness wasn’t included on a checklist of components of reading to focus teaching on. These differences between the reports seem to be related to the differences in subsequent practices.

The other reason is that phonemic awareness instruction was promoted by educational authorities in the US who didn’t have the same influence in the UK. Each of the following resources is either recommended or required as part of the professional development stipulated in science of reading legislation in multiple states in the US:

David Kilpatrick’s books discuss methods for teaching reading based on his understanding of reading research.

“Remember: phonemes are oral and letters are written. Phoneme awareness has to do with sounds in spoken words. It has nothing directly to do with letters. It is an awareness of the sounds in spoken language.” Equipped for Reading Success, p. 15

“A good way to remember the difference between [phonics and phonemic awareness] is that you can do phoneme awareness with your eyes closed but you cannot do phonics with your eyes closed.”  p. 16

Heggerty is a popular multi-year phonemic awareness and phonics curriculum. 

“Phonemic awareness: main focus is on phonemes or sounds; deals with spoken language; lessons are auditory; students work with manipulating sounds in words.”  p. i

Torgesen and Moats are the authors of the LETRS teacher education program.

“Phonemic awareness activities [unlike phonics] do not involve print. They are listening and speaking activities; they can be done in the dark or with a blindfold on."  p. 93 3rd edition, unit 2.

Treating phonemes as if they are units of spoken words with correct pronunciations is a mistake, and teaching phonemes without letters is the real “unnatural act” in learning to read.

I said there would be a second bit of evidence to consider about the best way to develop phonemic awareness; that would be the report of the National Reading Panel, which said:

“Teaching PA with letters helps students acquire PA more effectively than teaching without letters.”

“Teaching students to manipulate phonemes with letters yields larger effects than teaching students without letters, not surprisingly because letters help children make the connection between PA and its application to reading.” p. 2-41 

The panel did not recommend teaching 44 or any other number of phonemes before moving on to print knowledge. None of the studies they reviewed suggested such an approach, nor has any study to date shown that such instruction is beneficial, or that it is more beneficial than teaching a smaller number of phonemes, or more beneficial than teaching spelling-sound mappings for reading aloud and spelling instead. The panel’s discussion of phonemic awareness was relatively nuanced. Unfortunately, they also summarized their findings by listing phonemic awareness as one of the teachable components of reading, which was a reductio ad absurdum of their own discussion. Influential educators took this as license to round up to teaching all the phonemes first. 

Is there any direct evidence about the effectiveness of phonemic awareness training, as implemented in the SoR? Not much, because until recently teaching phonemes wasn’t on the instructional agenda. Studies are beginning to appear, however. Coyne et al. examined phonemic awareness instruction as implemented in the popular Heggerty curriculum. They found that such instruction improved children’s performance on the PA awareness activities that were used, but had no measurable impact on children’s reading. The study is not definitive, but it is a cause for alarm, given the many other concerns about PA instruction that I’ve raised.

In my view, US educators should adopt the British approach: children learn letters, letter names, and letter sounds and then start learning to read, beginning with pronouncing simple words aloud. The time that is freed up by skipping phonemic awareness instruction can be used for learning other things.

The recommendation to go straight from letter sounds to phonics runs contrary to the received wisdom in the SoR. It contradicts almost every major authority, and the many phonemic awareness zealots on social media. In some states it is against the law. It could also penalize children who have to demonstrate proficiency on phonemic awareness tasks used in assessments of student progress. What is a teacher to do?

I can’t countenance breaking laws or risking one’s job. I think what is needed is the same kind of grassroots activism that led to the adoption of PA training in the first place. As a teacher whose school district has adopted a curriculum that involves extensive instruction about identifying and pronouncing phonemes, I would ask: why? What is the justification for this practice? Why aren’t we using this time to teach reading? The answers to these questions can’t be, because it’s part of the science of reading approach, or because it is in a commercial curriculum, or because it was recommended by the NRP. I hope I’ve provided information that can be included in such discussions. Because laws can be changed. Merely making PA training optional would be a step forward.

I want to conclude by acknowledging that researchers’ use of terms such as “phonemic awareness” and “phoneme” has contributed to confusion about them in education. They started out as technical, theory-dependent terms. In Liberman and colleagues’ original usage, “phonemic awareness” referred to tacit (implicit) rather than conscious (explicit) knowledge of phonemes.That means behaving, automatically and unconsciously, as if words consist of phonemes. As in the Rose Report, they discussed PA in the context of the “alphabetic principle,” which is also tacit rather conscious. In everyday language, “awareness” is strongly associated with consciousness, however, and as the term circulated in education, the goal of inculcating conscious awareness of phonemes came to the forefront. 

The statement that “phonemes are the minimal units of speech” is repeated endlessly, and not just by educators. It is just a little too telegraphic to be useful at this point. Phonemes are the minimal units of speech according to phonological theories that abstract away from how words are actually articulated (which is covered in a different field, articulatory phonetics). For linguists as for beginning readers, making this phonemic abstraction depends on exposure to alphabetic writing. In this regard, linguists benefit from learning the IPA, an alphabet for representing phonemes in the world’s languages. Saying that “phonemes are the minimal units of speech if you can read an alphabet but aren’t actually used in speaking” is accurate, if unwieldy.