On “word callers” and vowel sounds

by Erica L. Meltzer | May 21, 2021 | Phonics

One of my colleague Richard McManus’s favorite stories about reading instruction involves a young girl who was brought to his tutoring center to prepare for a private-school admissions exam. She was clearly very bright, and when she was asked to read aloud, she did so quite fluently. Richard assumed that she’d ace the test with little trouble. When he told this to her mother, however, the woman’s response was that her daughter sounded fine but in fact understood almost nothing of what she read.

Richard was baffled. Luckily, though his friend Jean Tucker—a speech-language pathologist, reading specialist, and creator of the Spell of Language program—happened to be visiting that day and overheard the child reading. She turned to Richard and promptly announced, “That little girl can’t hear vowel sounds.”

Richard tested the girl and discovered that Jean’s diagnosis was in fact correct: the girl could not distinguish between vowel sounds. As a result, she was unable to tell many words apart and thus could not comprehend what she read. Richard put the girl on an intensive diet of vowel sounds, and sure enough, her comprehension began to improve.

I’ve heard Richard tell this story a number of times, but it was only very recently that I connected it to the so-called “word caller” phenomenon, in which children decode with apparent ease but are unable to make sense out of what they are reading.

To back up for just a second, the Simple View (Gough and Tunmer, 1986) states that Reading Ability = Decoding Ability x Aural Comprehension. This is the generally accepted model of how people are (un)able to make sense out of written texts, and when you think about it, it’s quite logical.

Essentially, a person can be able to decode a text perfectly, but if the language involves vocabulary, syntax, and background knowledge above the level of their aural comprehension, they will not be able to understand it. This corresponds to basic observable reality: most skilled adults readers could decode an advanced engineering textbook with relative ease, but few would have the vocabulary or mathematical knowledge to understand what the text actually meant. For all intents and purposes, they would simply be reciting words.

When I was tutoring, I had several students who didn’t exactly fall into the “word caller” category, but whose decoding skills significantly outstripped their general language ability. In every case of this I can recall, the student had a diagnosed learning disability and had received significant intervention on the decoding side; their comprehension difficulties resulted from the fact that they had weak vocabularies, understanding of syntax, and general knowledge. The same issues would surface if someone read them a text.

Given that experience, when I started learning about early reading instruction and discovered the pejorative “word caller” epithet, I assumed that the issue stemmed primarily from the aural comprehension/general language development piece of the reading equation. But now when I think about Richard’s story, I realize that there can also be hidden decoding issues. It seems possible that some children who appear to be reading competently—or even well—don’t actually perceive vowels correctly and, as a result, don’t realize they’re saying words they know.

I realize this might sound a little wacky, but hear me out.

In my experience, very few learners experience difficulties that are truly unique to them; if one student is missing a particular skill, chances are lots of others lack it as well. It just might be that no one has ever noticed, or even thought to check. And this is particularly true for reading.

When it comes to learning to decode phonetically, it is almost impossible to overstate the importance of being able to perceive vowel sounds accurately—and in particular vowel sounds in the middle of words. Lots and lots of words are written with the same sets of consonants; changes in meaning are conveyed through vowels. Think bat, bait, bet, bit, bite, bought, but, boot, bout. A person who is not aware of the differences between those vowel sounds will struggle to connect them to specific letters/combinations of letters and thus to read words containing them.

The absolutely key thing to understand here is that this is not a speech or hearing problem but rather an issue of conscious aural discrimination. It is entirely possible for someone to be able to automatically pronounce vowel sounds correctly within words, and to be able to easily distinguish between spoken words that differ by only one vowel sound; and yet on a conscious level struggle to recognize, produce, and differentiate among those same vowel sounds in isolation. 

This might seem very counterintuitive, but reality is not required to follow the rules of logic.

I confess that when Richard first showed me the Fluency Factory’s vowel chart (covering long vowels, short vowels, and diphthongs) and referred to “training” people on it, I was slightly baffled: it struck me as pretty much self-explanatory. The idea that someone could be able to say a word like book perfectly and automatically but then be unable to isolate the short “oo” vowel sound and say it on its own did not occur to me. It also did not occur to me that a person could find this exercise in any way awkward or unnatural or unpleasant. To me, sounds just were what they were, regardless of whether they happened to be part of a word or not.

When I observed Richard work with various students, however, I discovered just how much I was taking for granted. Differentiating between a short “oo” (as in book, look, cook) and an “uh” sound (as in but, cut, luck) turns out to be extremely challenging for some individuals, both children and adults. I did ultimately get some inkling of what this is like during a debate with Richard over a particular variant of short “o”. His usage was based on a pronunciation common in New England that never made its way into my speech, despite my having grown up in Boston, and it took me months to figure out that the sound was actually the pure “ah” sound used in father or avocado.

During the first Breaking the Code workshop, I briefly worked one-on-one with a kindergarten teacher who whizzed through the vowel chart, saying each sound  flawlessly and without the slightest hesitation. When I complimented her, she shrugged and said that she was just looking at the example words on the side of the chart. For her, plucking a vowel sound from a word and saying on its own was completely effortless.

I realized then that I didn’t know how to impress upon her—or indeed anyone who can perceive and isolate vowels easily—just how difficult that skill can be, or what an utter, long-term train wreck it can cause if it isn’t developed. I tried to convey to her that some people simply cannot connect vowels in the context of words to vowels on their own, but I suspect that it’s a problem that needs to be seen in action to fully be grasped by someone to whom vowel perception comes naturally.

English has 18 vowel sounds (compared to just five in Spanish)! They comprise both pure phonemes and diphthongs and are spelled in many different ways—some very common and others less so. Keeping the spellings straight is hard enough; for someone who isn’t even sure of the sounds themselves, it’s basically impossible.

Furthermore, vowel perception in one’s native language develops naturally in a window from birth to 18 months; it can be explicitly taught after that, but if a child grows up in a non-English-speaking home, or has repeated ear infections that interfere with hearing, it becomes much harder. This is in part why ESL students struggle so much with English reading—and things only get harder if their English-resource teachers are not native speakers themselves. (In New York State at least, ESL teachers are required to have proficiency in a foreign language, which sharply limits the number of people who qualify for certification). A teacher whose own ability to perceive and demonstrate English vowel sounds is compromised simply cannot teach that skill effectively.

Again, speech and aural discrimination are two different things: it is possible for a student who is immersed in an English speaking environment at a young enough age to learn to speak without an accent, but that does not necessarily mean they will develop an awareness of the sounds they are producing and/or how they correspond to symbols written on a page. On the other hand, a student who begins learning English at an older age may retain an accent but still learn to aurally distinguish between similar English vowel sounds well enough to easily match the language they hear to written words.

But that said, I’m going to end this post with a slightly off-color anecdote recounted to me by fellow Breaking the Code member Valerie Mitchell, who teaches high-school French.

In the course of a recent texting exchange with another foreign-language teacher, a native Japanese speaker, Valerie happened to refer to a student named Inès.

The Japanese teacher wrote that she felt terribly sorry for the girl, with a name like that. Valerie was baffled. “But why?” she asked. “Inès is a lovely name.” Then it occurred to her to ask whether the sounds in the name meant something vulgar in Japanese. “No, no,” her colleague replied.

Valerie wasn’t convinced. Assuming that the woman was simply too embarrassed to say what the word meant, she gently prodded until her colleague’s reaction became clear: the Japanese teacher interpreted the letter I in the girl’s name as making an “ay” sound—something it never does—and the e as making an “uh” sound (sort of fair, since short vowels are sometimes reduce to a schwa in names).

As a result, she assumed the name was pronounced… anus.

True, she was not an ESL teacher, but she was a foreign-language teacher: someone whose job entailed paying close attention to sounds.

She still could not reliably identify match English vowels to some of their most common phonemes.

And she had been living in the United States for 20 years.

Just something to think about.

Teaching reading isn’t rocket science

by Erica L. Meltzer | Apr 6, 2021 | Phonics

In the summer of 2020, the well-known researcher Louisa Moats published an article in AFT magazine entitled “Teaching Reading Is Rocket Science,” in which she laid out the daunting series of challenges involved in turning children into skilled readers.

For the record, I have immense respect for Dr. Moats; and from what I know of her LETRS program, it seems very well done (if a little heavy on theory). She’s done more to advance SoR-based approaches and to train teachers than almost anyone else around.

And yet, I have to disagree with her on this one—or at least with her (or perhaps an editor’s?) title.

To be fair, the type of reading Moats discusses in the article goes far beyond early-elementary decoding instruction and into “reading” in the full sense of the word. I would never dispute that becoming a strong adult-level reader involves developing an understanding of vocabulary and syntax as well as acquiring background knowledge in a wide variety of subjects—a process that is indisputably quite complex and that is, essentially, the ultimate purpose of school.

It seems to me, however, that broadening the scope of the discussion so greatly, as well as casting it in such extreme terms, does a disservice to the efforts being made to improve the early teaching of reading in ways that are specific, concrete, and ultimately achievable.  Characterizing reading instruction as “rocket science” moves it from the realm of the pragmatic into the abstruse; from a field that can easily benefit from practical interventions requiring a moderate amount of training to one that can only be managed by an elite cadre of highly-trained specialists.

Yes, some of the neuroscience is quite complicated, but it’s questionable how much of the really hard stuff teachers actually need to know. Although the inconsistencies of English spelling make reading instruction a more involved process than it is in, say, Spanish or Finnish or even French, the fact is that millions of children do nevertheless become competent decoders every year.

There’s also the fact that thousands of parents without any specialized training have, for decades, managed to use books like Teach Your Child to Read in 100 Easy Lessons to successfully instruct their children at home.

So that whole rocket-science thing doesn’t really stand up.

Yes, a small minority of children will always struggle mightily to learn to match letters and sounds and words, but there are difficult cases in every field. There is no reason for reading instruction to be viewed as anything other than a regular skilled professional activity, something akin to speech pathology or physical therapy.

Then why the hyperbole and the hand-wringing?

Part of it, I suspect, is simply a reaction against the sheer entrenchment of practices like three-cueing/MSV in American schools. Rooting them out will essentially involve overhauling instruction in the entire unwieldy, decentralized, (non-)system, as well as breaking past the Romantic mindset that equates clear, direct, systematic instruction with the destruction of creativity and independent thought (a mindset that finds perhaps its fullest expression in the belief that teaching students to decode phonetically turns them into mindless word-callers as opposed to children who can, well, read).

Given that this has been the guiding vision of education in the United States for nearly a century, that part is arguably more difficult than rocket science!

To be clear, teaching reading well is not exactly easy, particularly if one is working with a class that includes children at many different levels. Teachers must have a thorough knowledge of the many English sound-spelling correspondences; know which ones represent standard patterns vs. less-common variations as well as when each group is used; and be able to convey them in a carefully sequenced way that does not inadvertently create confusion.

Their own aural discrimination skills must also be developed enough for them to notice, for example, when a child says “melk” instead of “milk,” recognize how that mispronunciation could lead to a reading problem, and know how to intervene so as to head it off.

They must understand the difference between blends and digraphs, and know enough about morphemes to be able to teach them as “chunks” as opposed to random sequences of letters that need to be re-sounded out from scratch every time.

But if these things do require some degree of skill, they are still not the equivalent of graduate-level physics. It is possible, however, to imagine how they might seem like it in comparison to the reflection papers and discussions of multiple literacies that appear to make up the bulk of ed-school curricula—surely the lack of real academic rigor contributes to the perception of moderately technical content as something impossibly hard and intimidating.

I think, though, there’s another issue here too.

In an interview with Children of the Code, the eminent reading scholar G. Reid Lyon pointed out that the resistance to research-backed reading instruction in the education community has traditionally been so extreme, and so disproportionate to the modesty and reasonableness of the proposals involved, that it can only be explained by the force of the adult professional and psychological issues at play.

These issues are of course responsible for the continued use of three cueing-based approaches, but I suspect that they are also involved in the current trend toward exaggerating the difficulty of implementing research-backed instruction—even by people who advocate it.

If you’ll allow me to play armchair psychologist for a moment, I’d like to propose that this is a sort of defense mechanism. The vast majority of people involved in early-elementary education love working with children and want desperately for them to succeed; and the realization that they have been using/promoting methods of reading instruction—for years, in some cases—that resulted in some children being left far behind is understandably quite difficult to accept. Even if there is no way they could have known, the notion that there might have been a relatively simple solution all along is not a palatable one. It is much more reassuring to subscribe to the belief that science-backed reading instruction is so complicated that no one could possibly figure out how to implement it without Very Serious Training. In the context of the AFT (American Federation of Teachers) article, this comes off as the profession’s way of letting itself off the hook.

Obviously, social media plays a big role in amplifying this message as well. Facebook/Twitter discussions can get very, um, chaotic, with a lot of people slinging around technical-sounding terminology—not always correctly—like there’s no tomorrow. For someone not initiated into the nuances between, say, phonological awareness and phonemic awareness, or orthographic mapping vs. phoneme-grapheme mapping—the conversation can seem very intimidating. The occasional voice that pipes up to say that all they need is a whiteboard, some paper and pencils, and a good chunk of time tends to get lost in the shuffle. That makes it easy to lapse back into the default position of, “Well, everyone learns differently, so teachers should just pick and choose based on their individual preferences.” That’s a perfectly reasonable position if teachers have a solid knowledge base to work from, but it’s not fine if they don’t.

Moreover, in social media discussions there is often an assumption—sometimes implicit, sometimes explicit—that if teaching reading well were easy, then everyone would be doing it. That’s the logical conclusion, but it overlooks the possibility that perhaps things are so complicated because at some level, people want them to be that way.

Since schools of education were first established in the late nineteenth and early twentieth centuries, replacing normal schools (teaching-training colleges) and/or being absorbed into larger and more prestigious universities, they have suffered a persistent inferiority complex relative to, say, schools of law or business, and have been at continual pains to convince the rest of the academic world that they can do serious scholarship too. (Look at us, we’re just like the big kids… Really.)

In practice, that has often meant taking relatively straightforward processes and turning them into abstract theoretical constructs far removed from the day-to-day happenings of a real-life classroom. Thus, the teaching of five- and six-year-olds to decode simple words must either be presented as a complex, esoteric process and highly individualized process, or neglected because it is of insufficient scholarly interest (the serious neuroscientific component being tellingly ignored).

And now, despite all the rightful pushback against the three-cueing system and its ilk, what concerns me is the ease with which phonics-based reading instruction can disappear down the same rabbit hole of distorted definitions, misunderstandings, and overcomplications.

The cognitive scientist Mark Seidenberg (Language at the Speed of Sight) has, for example, pointed out that the separation of phonemic awareness and phonics in the 2000 National Reading Panel Report is often mistakenly interpreted to mean that the two skills must be taught separately and sequentially—even though the key to decoding is phoneme-grapheme correspondence, and there is no reason sound work cannot be done in conjunction with letter work.

That is not to imply that a separate phonemic-awareness program like Heggerty hasn’t been invaluable for many children, just that it makes sense to ask which groups of five-year-olds actually need to spend months practicing sounds in isolation. In a kindergarten class where most children come in with solid phonemic awareness, it would presumably make more sense to jump right to working with letters and to simply provide additional sound work for the few students who require it.

Moreover, the intense pressure to demonstrate that students are engaging in “higher-order thinking” has created a tendency to over-complicate simple tasks to the point of absurdity. Learning to decode can be challenging for some children, but it is fundamentally a rote skill, and failing to respect that reality can make it harder to acquire than necessary.

The other side of that coin is the creep-over from the special education world. Because so much of phonics-based instruction has traditionally been relegated to special-ed classrooms (where it is grudgingly viewed as acceptable because it only involves those children), special-ed teachers are generally most experienced with programs such as Orton Gillingham. But although children who have severe difficulty retaining letter shapes and sounds may require all sorts of sensory interventions, the average child can probably just start with printed letters on a page.

To reiterate, the over-arching goal should be to get students decoding as efficiently as possible so that they can start to focus on other aspects of reading, not to draw out the basic instructional processes longer than is necessary.

As Seidenberg also emphasizes, much of the translation to the classroom is an ongoing process. Cognitive scientists are not first-grade teachers and, for the time being at least, many of the pedagogical specifics are still being worked out. The fact that there exists a set of guiding principles does not mean that it is possible to create one single best program guaranteed to meet every child’s needs, and it is imperative that teachers have the autonomy to figure out just how much phonics is the “right” amount for their students.

All that said, let me conclude with an excerpt from a story in Michael Maloney’s (decodable) K-2 reader. After head-spinning Facebook threads about the benefits of Wilson vs OG vs. Lexia, it seems almost ridiculously, laughably simple—until you think about all the children who have actually learned to read this way:

…Jack’s father began to teach Jack how to read. 

First he taught him some letters and told him the sound that each letter makes. 

Soon Jack began to sound out words. 

Each day he sat learning with his father for an hour after dinner. 

Soon Jack could read a short story. 

But many words were still hard to read. And he could not read fast. 

His father said, “You are doing very well. Work hard every day. Soon you will be able to read out loud as fast as you can talk. Then you will be a good reader.” 

His father was right. 

In a little while Jack could read any book, even the big first-aid book. 

The end of the story