by Erica L. Meltzer | Jun 23, 2020 | Phonics, Reading Wars, The Science of Reading
One of the most serious, and most persistent, misconceptions in the world of reading early reading instruction involves the use of context clues. Regardless of whether they are explicitly taught an incorrect interpretation of three/multi-cueing system or simply absorb its tenets in graduate school or via professional development, many teachers of beginning readers erroneously learn that children should focus primarily on beginning/ending letters and then use a variety of guess-and-check methods (e.g., picture clues, other information in the text) to make educated guesses about unfamiliar words.
If you’re not familiar with the research, a reliance on context clues has been identified as a compensatory strategy for weak decoding skills (Nicholson, 1992; Stanovich, 1986); as children become more proficient decoders, they spend less time looking at contextual information.
Louise Spear-Swerling, professor of Special Education at Southern Connecticut State University, sums the findings up as follows:
Skilled readers do not need to rely on pictures or sentence context in word identification, because they can read most words automatically, and they have the phonics skills to decode occasional unknown words rapidly. Rather, it is the unskilled readers who tend to be dependent on context to compensate for poor word identification. Furthermore, many struggling readers are disposed to guess at words rather than to look carefully at them, a tendency that may be reinforced by frequent encouragement to use context.
In her 1998 article on the misinterpretation of the three-cueing system, Marilyn Jäger Adams furthermore makes the point that while skilled readers do in fact make use a combination of orthographic, syntactic, and semantic clues, they do so in order to construct meaning rather than to literally decode words. The misinterpretation of the graphic that has filtered down into many elementary-school classrooms is based on a confusion between “reading as extracting meaning from text” (which presumes solid decoding) and “reading as turning squiggles on a page into words.”
To be clear, using a combination of first letters and pictures, or other parts of a text, and making educated guesses based on “what would make sense” may indeed result in beginning readers coming up with correct words, or a generally accurate understanding of a particular scenario. But that’s beside the point.
Short-Term vs. Long-Term Thinking
To have a serious discussion about what types of strategies should be taught, and when, and why, it is necessary to distinguish between short-term and long-term thinking.
In short-term thinking, the focus is on getting the child to understand the particular text in front of them, at that particular moment. Exactly how that happens is not overly important.
And after all, the purpose of reading is to understand the text (or, to put it in edu-parlance to “construct meaning”). Obviously.
Understandably, this is a very appealing viewpoint, and one that seems to make intuitive sense: learning to read can be very challenging for certain children, and if especially one is working with a slow, struggling reader, the impulse to have them glean whatever they can, however they can, is entirely understandable. Any tools they can use to figure out what’s going on are helpful, right?
In the long term… No, actually.
The problem is that the (cueing) strategies that allow a beginning reader to get the gist of simple pictures books fall very, very far short when applied to the far more challenging texts students will be expected to read only a few years later. A kindergartner or first-grader who appears to be basically on track may in fact be missing very fundamental key skills.
Moreover, the habit of guessing at unfamiliar words is not one that children naturally outgrow—once established, it is often extraordinarily difficult to break.
Combine that with persistent decoding issues, and you end up with a burgeoning middle-schooler who’s just old enough to really push back when someone tries to intervene but not quite mature enough to appreciate why it’s so important for her to learn to sound out multisyllabic words phonetically (as Richard McManus will currently attest).
All that said, the use of context clues does in fact have a place in early reading instruction. But the key piece is that context must be used to support phonetic decoding (and thus orthographic mapping), not replace it.
Practically speaking, this involves the decoding of words that are only partly phonetic, or whose exact pronunciation cannot be determined from the way they are written.
As Tunmer (1990; see pp. 112-113) has explained, phonetic knowledge and the ability to use context combine to create a positive feedback loop in which context is used to actually strengthen phonetic understanding and facilitate orthographic mapping—the process by which words get stored in the brain as sound-syllable correspondences and made available for automatic retrieval.
Essentially, when children with a solid phonetic understanding of the English code encounter irregularly spelled/pronounced words, they may use context clues to bootstrap themselves into an understanding of how those words are pronounced. That reinforces an understanding of more complex phonetic patterns and allows challenging language to be read more easily in the future.
Some children may figure this out their own, but there is no reason that others cannot be taught this strategy explicitly.
This phenomenon also supports the finding that children with good decoding skills can often infer the pronunciations of moderately irregularly word (Groff, 1987), something that directly contradicts the notion that English is too irregular for phonetic knowledge to be effective.
What the Heck is a MOSS-kwih-toe?
I’m going to illustrate this with a personal anecdote involving one of my earliest reading-related memories.
It happened when I was in first grade, and it involved the word mosquito.
Although this word is spelled in a way that is not entirely unrelated to its pronunciation, there are a couple of notable irregularities.
- First, the “qu” makes a “k” sound, as opposed to its usual “kw” sound.
- Second, there’s no obvious information about which syllable should be stressed.
So when I encountered this word in print at the age of six, I initially read it as MOSS-kwih-toe.
“Huh?” I remember thinking. “What the heck is a MOSSkwitoe?”
I knew that didn’t make sense, and I knew I wasn’t reading the word correctly, but I couldn’t fathom what it might actually be.
So I kept on plugging along, and a couple of minutes later, I suddenly had a lightbulb moment.
“Oh!” I thought. “Of course. The word is pronounced muskeetoe!”
I was really quite astonished to have figured it out. I had a huge vocabulary but was by no means an exceptional decoder. You know those kids who teach themselves to read at three? Well, I wasn’t one of them. I wasn’t even in the top reading group! But the incident made such an extraordinary impression on me that I still remember my thought process with almost total clarity more than 30 years later.
So let’s review what I did:
1) I used the specific sequence of letters and my knowledge of the sounds they made to take me as far as I could possibly go toward identifying the word.
2) Once I realized the story involved some sort of insect that flew and buzzed around during the summer, I drew the logical conclusion about the word’s pronunciation.
3) I got a lesson in the fact the “qu” can make a “k” sound in certain circumstances—an alternate phonetic pattern that helped me identify other challenging words quickly in the future (and eventually facilitated my ability to grasp French pronunciation).
4) The sound-spelling correspondence of the whole word was from then on etched into my mind in metaphorical granite (i.e., orthographically mapped).
5) I learned that even if words weren’t straightforwardly phonetic, I could combine my sound-spelling knowledge with my other skills and figure things out. That was an immensely powerful realization.
The interesting part is that technically speaking, my process was a stellar example of the three-cueing model. It involved the construction of (literal) meaning using the interplay among clues based on orthography (sequence of letters), syntax (the word had to be a noun based on its position in the sentence), and semantics (the story must have involved a tiny, buzzing insect).
The key piece, however, is that my close attention to sound-spelling correspondences underlay my ability to engage the other systems effectively. I did not look at the first letter(s) and make a semi-random guess based on context (the way I later saw many of my own, much older students do). If I had, in the absence of any solid contextual information at that point, I might have come up with something like mouse.
Rather, I paid close attention to all the letters in the word—beginning, middle, and end—in an attempt to sound it out, and only when that wasn’t enough did I move to thinking about the larger context of the story to in order to make the leap from that baffling collection of letters to certainty about a real word.
That leads me to my next point, namely that the episode was also consistent with the finding that children who learn to read phonetically are able to produce nonsense words/pronunciations—something that children who are taught via whole language do not do (Barr, 1974-5; click here for a discussion of the findings). Had I jumped to plug in a word I already knew how to read, I would have missed out on a significant learning opportunity. And even if I had managed to correctly guess mosquito, I would have missed out on an important phonetic lesson and would not have been able to carry that new knowledge forward into other words.
As Ehri points out, children who exhibit this type of phonetic non-word decoding in first grade move more quickly into the full alphabetic phase, in which “beginners become able to form connections between all of the graphemes in spellings and the phonemes in pronunciations to remember how to read words.” Indeed, by second grade, I was basically a totally fluent reader. Once things clicked, I never looked back.
About the Three-Cueing System…
Let me conclude by saying that writing (and re-writing) this piece has actually been a rather enlightening process for me, not least because it allowed me to revisit a memorable childhood experience from an adult perspective and—entirely to my surprise—be able to analyze it in light of theories with which I’ve become acquainted only relatively recently. When I first began to write about the episode, I was unaware of just how clearly it embodied key findings about how children learn to read; it was only as I began to really probe it that I realized how illustrative it actually was.
It also made me develop a more nuanced understanding of the three-cueing system, as well as a better understanding of how things went so badly awry. What I’ve described here certainly isn’t the “look at the first letter and the picture and think about what would make sense” approach, but it also isn’t quite the mature use of textual cues employed by skilled readers to, say, determine correct definitions of multi-meaning words. Instead, it’s somewhere in the middle.
It’s precisely that in-between place that makes things so tricky, and I think it points to the overwhelming importance of precise language when discussing techniques for reading instruction. Indeed, so many conversations about this topic devolve into free-for-alls simply because various parties cannot agree on what central terms mean. (E.g., for researchers, the term “sight word” refers to a word that has been orthographically mapped and can be read instantaneously, whereas for teachers it generally refers to the high-frequency words on the Dolch/Fry lists that beginning readers are expected to memorize more or less by rote.)
It’s very easy to imagine how a cautious assertion like, “In some instances, children can use context clues to help them identify unknown words” could get transformed into, “Let’s look at the first letter and the picture and ask ourselves what the word might be.” Those two statements might not seem terribly different on the surface, but in fact they’re worlds apart.
The cognitive scientist Daniel Willingham has warned that inconclusive (or, I would add, poorly understood) theoretical models can easily get translated into ineffective classrooms practices, and I think the three-cueing system is the poster child for that. It’s not enough to say, for example, that skilled reading involves a complex interplay of systems; inevitably, given the history of reading instruction in the United States, that will be used to promote harmful ideas about the unimportance of phonetic decoding. When the devil is in the details, every word counts.
by Erica L. Meltzer | Oct 4, 2019 | Issues in Education, Phonics, Reading Wars, The Science of Reading
A couple of days ago, I got curious about the state of phonics instruction in New York City schools and started googling away. I learned all sorts of fascinating things about the respective reigns of Joel Klein and Carmen Farina, and about the ongoing and pernicious influence of Lucy Calkins/Reading Workshop and Columbia Teachers College.
I also came across a well-intentioned an article in Chalkbeat about the struggles of some Brooklyn parents to get their dyslexic children into appropriate programs. The content of the article was disheartening but fairly predictable—what I found more interesting was the semantic confusion the writers displayed in a discussion of balanced literacy vs. phonics, and it got me thinking about how the standard reading-war rhetorical tropes get wielded.
Here is the article’s more-or-less typical spiel:
Balanced literacy advocates believe learning to read is a natural process, that enjoyment of reading is paramount, and that helping students understand the meaning of text is more important than teaching them the mechanics of reading.
But experts question whether it’s the best method for struggling readers.
The opposing approach, rooted in phonics-based instruction, emphasizes decoding how letters correspond to sound and how the sounds connect to form words. It can be dry and repetitive, but many experts say the method establishes a firm base that benefits a wide array of students.
Many students can learn to read through a balanced literacy approach, particularly when it’s complemented with additional reading and vocabulary work at home, the experts say. But for those who struggle — including those with dyslexia — balanced literacy can feel like being thrown into the deep end of a pool.
There are a few things to notice here:
First of all, the (conclusively debunked) belief that learning to read is “a natural process” is primarily associated with the Whole Language movement (although it is likely held by many proponents of Balanced Literacy as well); Balanced Literacy, in contrast, represents the attempt to bring an end to the reading wars by reconciling phonics and whole language.
Although in practice it often favors the latter at the expense of the former, in theory it pays lip service to both. But if education reporters for a publication dedicated to education cannot even keep their terminology straight, how can the public at large be expected to do so?
Second, note the the both-sides-ism (some people believe x, but others believe y), which clearly comes down on the side of “Balanced Literacy.” Even if the authors do mention grudgingly that “experts question” whether balanced literacy [sic] is “the best method for struggling readers,” nothing is mentioned about other readers—only that “many students can learn to read through a balanced literacy approach.”
Given that experts believe (and in fact have ample evidence to support) that phonics is the most effective means to teach decoding period, not just for children who struggle, this is a rather serious omission—especially since the writers link to Emily Hanford’s original piece on phonics for American Public Media, which makes that viewpoint abundantly clear. But it is as if their distaste for the premise is so ingrained that they are unwilling, or perhaps even unable, to acknowledge the actual argument being made. (Perhaps they are engaged in “making meaning”?)
Again: if education reporters cannot even lay out well-established views accurately, how can the public understand what is at stake?
Even when phonics is acknowledged to be helpful, it’s presented as the educational equivalent of spinach: good for you, perhaps, but also kind of icky. In contrast, whole language is something like a slice of pizza: gooey and tasty and oh-so-satisfying.
This is fundamentally a rhetorical problem.
Traditionally, the whole language/balanced literacy faction has played the rhetorical game much more effectively than the phonics people, appropriating the lexicon of romanticism (naturalness, wholesomeness, enjoyment) to their advantage and assigning the role of artificiality (rote learning, memorization, drill ’n kill) to their opponents, who are forced into the defensive position.
What’s interesting, though, is that the reality is exactly the opposite:
English contains about 250 graphemes—letters or letter-combinations that represent specific sounds—approximately 70 of which are commonly used. That might sound like a lot to memorize, but a child who masters these correspondences can sound out literally thousands of words and take a decent stab at many more that are not perfectly phonetic.
In contrast, a student who never masters sound-letter correspondences will essentially need to memorize, by rote, often in a dry and repetitive fashion, every new word as a random bunch of squiggles disconnected from the sounds they make, and over the course of their educations, they will need to do this thousands upon thousands of time.
Why have members of the phonics camp not seized on this fact? Why have they not shouted it from the rooftops (or from their Facebook accounts)? Perhaps because they take it as self-evident that if they just present the science, then people will listen , particularly if said people claim to be in favor of “critical thinking.”
They assume that if method x is shown to be more effective than method y, then of course schools will want to adopt it.
They also assume that because the logic behind teaching phonics seems so obvious—you can’t focus on meaning unless you can figure out what the words say—that there is no need to explain things further.
Those are reasonable assumptions, but they rest on the notions that 1) people are moved by logical arguments; and that 2) that the desire to actually solve problems outweighs institutional inertia and/or the instinct to cling to existing beliefs, no matter how misguided.
Effectiveness is beside the point here. (Besides, what is effectiveness anyway, and how can it truly be measured? Isn’t it more important that children learn to love reading than that they know how to break words into little pieces? Isn’t learning the point of education to inspire children to love learning so that they can become lifelong learners?)
I also suspect that proponents of phonics systematically underestimate the hold that romantic ideology has on the American classroom: anything presented as natural is assumed to be good; anything presented as artificial (sometimes including school itself) is presumed to be detrimental.
These assumptions are so deeply embedded in the discourse surrounding education that they must be acknowledged and dealt with directly if any headway is to be made.
So here is my modest proposal: if proponents of phonics want to make any progress with the general public, it is necessary to flip the existing narrative on its head and insist that PHONICS = CRITICAL THINKING (applying knowledge to novel situations) whereas WHOLE LANGUAGE/BALANCED LITERACY = ROTE MEMORIZATION (random squiggles disconnected from authentic language).
This narrative needs to get repeated over and over, ad nauseam. Don’t try to sound smart, don’t go on about science, or logic, or peer-reviewed journals, just “MEMORIZING WHOLE WORDS BY ROTE IS BORING AND UNNATURAL.”
It might be too late—but still, you just might stand a fighting chance.
by Erica L. Meltzer | Feb 19, 2019 | Phonics, The Science of Reading
In discussions about reading instruction, a commonly raised point is that students with reading disabilities—particularly dyslexia—suffer disproportionately when deprived of systematic instruction in phonics. In fact, this is virtually impossible to dispute—whereas many students in whole language classrooms do manage to figure out enough of the rules to become reasonably proficient readers, students who cannot make sense out of word/sound relationships have no way of keeping up. And if their difficulties are not noticed in time, or they lack access to competent reading specialists, either through their schools or privately, the consequences can indeed be extremely dire. (The percent of prison inmates with reading disabilities is, for example, astronomical.)
I’m saying this upfront because I do not want in any way to minimize the difficulties faced by these students and their parents. But what I’m interested in examining here is how some of the rhetoric surrounding reading pedagogy operates—how concepts like “normal” and “abnormal” are defined and how, in some cases, the recognition of the importance of phonics for students with reading disabilities like dyslexia can become a tool for reinforcing naturalistic ideas about reading.
According to one narrative I find particularly pernicious, a phonetic approach to teaching reading is acknowledged as acceptable, but only for children with reading disabilities; for “normal” children, it is implied, exposure to high-quality literature and participation in enjoyable activities that foster a love of reading are by themselves sufficient for reading (that is, decoding) to be learned automatically.
Incidentally, I was already planning to write this post when, by pure chance, a perfect specimen of this viewpoint appeared (as if by magic!) in my inbox, courtesy of Diane Ravitch’s blog.
In the course of a post about the insanity of subjecting kindergartners to standardized reading tests—a stance that, for the record, I agree with 100%—Ravitch cites Nancy Bailey, a former special-education teacher and Ph.D. in educational leadership, on the ways in which children learn to read. As Bailey describes:
Some children easily acquire reading skills without formal phonics instruction. They are curious about words and are able to sound letters out as they listen to and enjoy picture books. They may read well before they start school.“
Other children learn a little later. And some with disabilities may need extra assistance with a formal phonics program.
“Repeatedly testing young children to find out how they read at such an early age would be better spent reading out loud lovely, funny, engaging picture books, and letting children develop their language skills through play!”
What interests me in this passage is the way Bailey contrasts the needs of students with and without reading disabilities. Consider the phrasing: “Some children easily acquire reading skills without formal instruction.” Yes, this is true, although it only applies to a very small percentage. (Another 35% will learn to read with broad instruction.)
What Bailey says next is where things get problematic: “Other children learn a little later. And some with disabilities may need extra assistance with a formal phonics program.”
Look carefully at how this is phrased: not “other children need varying amounts of direct instruction to learn to read,” but “other children learn to read later.” Taken on its own (and in the absence of any pre-existing knowledge about Bailey’s stance on whole language), this could charitably be written off as an unintentional omission, but it is the next statement that gives the game away. By juxtaposing “other,” non-learning-disabled children and those “with disabilities, who may need extra assistance with a formal phonics program,” Bailey clearly implies that “normal” children will learn to read more slowly than children in the most exceptional group—but that they do not need formal phonics in order to acquire that skill. And that is simply not true.
Furthermore, the use of the word may suggests that even in cases where a learning disability is present, a phonetic approach is only one of many possible options. Are there other ways in which a dyslexic student might learn to read? Theoretically, yes—in the absolute sense, anything is possible. And presumably a very small percentage of students will continue to struggle despite receive intensive phonics instruction. But phonics is the only solution demonstrated to be consistently effective for students who struggle with decoding—to the best of my knowledge, there is no accepted mainstream research suggesting anything to the contrary. (And again, parents don’t pay for their children who are struggling with reading to go to whole language centers.) In light of Bailey’s other remarks, her wording here seems to imply that given enough time, some students with genuine reading disabilities will eventually just pick up the whole reading thing on their own, no intervention needed. Again, theoretically possible, but in reality… the result will probably be years of frustration.
Reading between the lines here, it is possible to catch of whiff of something vaguely… if not exactly condescending, then perhaps slightly pitying. The poor dears… They can’t learn to love reading naturally, through fun activities, like their peers. No, they need to sit with workbooks (the horror!) doing those terrible, awful, boring rote drills that won’t teach them to love reading at all.
Consider now the broader assumptions lurking behind this perspective. Phonics can be written off as something that “other,” disordered children need. Normal people do not need to learn in a structured, sequential way, or to have concepts carefully and explicitly broken down for them. Given the right environment and right set of attitudes, they just… catch it. Like magic.
Furthermore, framing the problem in these terms allows the problem to be placed squarely on the students’ shoulders. Teachers need not question their own methods because a student who fails to learn naturally must be deficient in some way and can be passed off to a trained expert, thus preserving the status quo.
What makes this such a seductive fantasy is that children without obvious learning disabilities may in fact appear to do just fine in whole-language classrooms. The words they are asked to read are mostly simple and straightforward, and they can get pretty far on memorization—particularly if no one bothers to probe their true level of understanding or has an active interest in not assessing it. (In light of many teachers’ visceral loathing of memorization, I actually find this very bizarre.) But once reading becomes more challenging—once the books no longer have pictures, and the terms are too technical to be figured out without context clues—things start to fall apart.
Richard McManus at the Fluency Factory mentioned to me recently that he sees a lot of kids who seemed okay in elementary school but are about to run headlong into a huge amount of trouble in middle school because they just can’t compensate anymore. But by then, the damage has started to accumulate. With solid instruction, students may be able to scramble back up to grade level; without it, they fall increasingly behind.
Then there are the students who continue to manage to keep skating by through the first couple of years of high school; for many of these students, a low (P)SAT or ACT score is the first real sign that something is amiss, particularly if they spend most of their in-class time doing groupwork and have grades based more on projects and class participation than on tests or papers (or if grades at their school are exceptionally inflated; or if they rely heavily on tutors for help with schoolwork, as was the case for several of my students; or if their parents call to protest every time they receive anything lower than an A-…). Unfortunately, by that point it is exceedingly difficult to play catch-up. These were among the students I saw, the ones who read in ways so disorganized that at first I could barely wrap my head around it. Without exception, they were unable to score above the high 500s in Reading on the (old) SAT or the low-mid-20s on the ACT, regardless of how much tutoring their received.
I suspect that the proliferation of students in this category also contributed to the longterm decline in SAT reading scores, and thus to the recent elimination of the vocabulary portion of the SAT. A decent percentage of the words tested could be figured out—or at least reasonably guessed at—based on their component parts (roots, prefixes, suffixes). Indeed, that was in large part the point of the test! But students who learn to read through whole language, particularly if they attend schools that do not emphasize systematic vocabulary development in the later grades, are ill-prepared to break down words that way (although if they’re lucky, they might pick up something from foreign-language class).
A striking number of my students could not distinguish between “looks/sounds like” and “is related to,” for example assuming that defer meant something like differ because the two were pronounced similarly (I saw that one on multiple occasions). At the extreme end, they might not even really notice the difference in appearance between the two—or at least not register the difference in vowels as having any particular significance for their meaning. As a result, it was impossible to help them approach what was essentially a logic test in a logical manner.
Perversely, then, the practice of treating phonics as something reserved for students with reading disorders has the inadvertent effect of making better decoders out of students diagnosed with learning disabilities (they’re the ones who get the intervention), and with producing reading disorders in students who would not otherwise have them. And even if students do not have outright reading disabilities, they may become far weaker readers than necessary.
And perhaps the biggest irony of all: no matter how joyful their early experiences, students who have difficulty reading do not like to read—once the words get hard, those six-year-olds who were oh-so-excited to participate in “literacy-building” activities suddenly get a lot less enthusiastic. A student who consistently misreads words will find it difficult to figure out just what is going on in many texts, and/or may actively misinterpret their literal meanings (often, as I observed, in ways that do not make sense). Furthermore, if they are expending too much mental energy just trying to figure out what the words say, there will be no room left over to think about the meaning.
Who could possibly enjoy such a frustrating activity?
Who would want to engage in it at all?
