A couple of weeks ago, I attended a conference on the science of reading held by John Gabrieli’s lab at MIT. It was, if nothing else, an eye-opening experience—not always in good ways, but certainly in ways that laid bare the problems involved in implementing broad changes to how reading is taught in the United States.
At the reception after the conference, I happened to be introduced to Nancy Duggan, one of the founders of the Massachusetts chapter of Decoding Dyslexia, an organization that advocates for screening and support for dyslexic students. In the course of our conversation, I mentioned one of the stranger reading problems I’d seen among my students—namely, that they had trouble making their eyes follow a line of text from left to right. Instead, their gaze seemed to dart randomly around the page. “Oh,” Nancy said promptly, “that’s the three-cueing system. Kids are supposed to look at just the beginning of the word and then look at the pictures for context clues.”
I was vaguely familiar with the term, but I had never made the connection between it and the reading difficulties I had witnessed in teenagers. I also confess that I did not know that children were actually taught to read in quite this bizarre a manner, or that the three-cueing system had anything to do with it.
Of course I knew that students were encouraged to look at the pictures or at other parts of the text for clues, but I had somehow always imagined that they were instructed to do these things if they could not figure out what a word meant—naively, I assumed that the initial focus was still on all the letters in the word, and that students were taught to look elsewhere afterward. It literally did not occur to me that children could be told to look at just the first letter or two in a word, and then be directed to immediately look somewhere else, before they had even demonstrated any difficulty making the connection between the letters and the word as a whole. That was too far outside the realm of my experience. It was just too illogical. And, quite frankly, too dumb. Who in their right mind would teach someone to read that way?
But it fit exactly with the kind of crazily disorganized looping around the page I used to see. Finally, after a decade of puzzling over that phenomenon, I had a plausible explanation.
That conversation sent me back to an article on the three-cueing system by the reading researcher Marilyn Jaeger Adams. I had first encountered the piece while doing some research on phonics, but I hadn’t really grasped the article’s significance. It was interesting, sure, but at that point it also seemed kind of tangential to my interests. Now, however, I realized that it was far more important than I had realized, and I really wanted to understand her critique.
For anyone interested in why so many students struggle to learn to read, I highly recommend the whole piece. It gets a bit dense at points (I’ve read the whole thing three times so far and am only now getting a good handle on it), but it really makes clear the extent to which some very basic ideas about reading have become warped.
Essentially, Adams describes how she was puzzled to discover that a group of teachers to whom she was presenting understood the “three-cueing” system—the idea that reading results from the overlap of orthographic (spelling), syntactic (word-order), and pragmatic (essentially, getting the gist) factors—in a way that was fundamentally different from how she and other researchers understood it. The system was typically presented in a Venn diagram, like the one below:
Furthermore, the influence of the graphic in ed circles was much stronger than she had realized. As Adams describes:
When I asked these people what they meant by the three-cueing system, they looked at me as though I were from Mars. They were at least as embarrassed as I. For indeed, how could I not know? How could I present myself to them as an expert on early literacy and not know. What was at issue here was clearly not any general notion of the interplay of syntax, semantics, and graphophonemics but, rather, some particular, specific version of this notion-one with which I was frankly unfamiliar.
That experience led her on a quest for the graphic’s origins and eventually to the startling insight that what had begun as a way of depicting the interlocking systems used to making *meaning* out of a text were instead being used as a guide to figuring out what a text literally said—that is, to decoding.
Teachers were not only missing the point; they had absorbed an approach that was exactly the opposite of what was originally intended and, moreover, that flew in the face of decades worth of research.
Whereas researchers originally understood the system as a tool used to represent the commonsense point that students have a variety of methods available to them to make sense out of the literal words on the page, accepted wisdom had, through a sort of perverse game of telephone, come to dictate that students should avoid using the letters in a word to determine what it literally said:
If the intended message of the three-cueing system was originally that teachers should take care not to overemphasize phonics to the neglect of comprehension, its received message has broadly become that teachers should minimize attention to phonics lest it compete with comprehension. If the original premise of the three-cueing system was that the reason for reading the words is to understand the text, it has since been oddly converted such that, in effect, the reason for understanding the text is in order to figure out the words.
Furthermore, a significant part of the problem hinged on the misunderstanding/misuse of Venn diagrams, with the orthographic sector being presented below the other two systems in order to indicate that paying attention to the letters in a word (i.e., the phonetic approach) is less important than using context clues—often understood to mean pictures—or overall meaning as a guide. As Adams points out, however:
The position of the circles is formally of no significance to the logic of a Venn diagram. What matters is only whether they overlap partly, totally, or not at all with each other and the outcome of interest.
The cueing schematic [with orthography at the bottom] is sometimes presented as rationale for subordinating the value of the graphophonemic information to syntax and semantics and, by extension, for minimizing and even eschewing attention to the teaching, learning, and use of the graphophonemic system. This interpretation directly contradicts the logical import of the Venn diagram which, by virtue of its structure, asserts that productive reading depends on the inter-working of all three systems.
Ironically, this type of misunderstanding would seem to point to the need for better math education(!) for elementary school teachers. Or perhaps just better general education. This is a really striking example of how a general knowledge gap in one area can translate into a misunderstanding in another, and inadvertently create chaos on a wide scale.
The whole mess also epitomizes of one of my favorite themes, namely the problem of referring to both the process of deciphering words (decoding) and the much more complex act of deriving meaning from a text as “reading.” The fact that the two skills are denoted by the same term provides another clue as to how the confusion could have come about. It’s easy to imagine how some not-terribly-astute literacy coaches or district consultants might have encountered the original diagram and made a slip from “reading as extracting meaning” to “reading as figuring out what words the letters make,” then, as Adams describes, propagated their confusion via workshops and conferences.
At any rate, my first reaction when I finally really got the gist of the whole piece was, “Oh my goodness, how many people outside the education world have *any* idea this is going on? And does anyone realize how widespread the problem has become?
Adams tells the story of the backlash that ensued in Massachusetts when an MIT professor discovered that his son’s school was discouraging a phonetic approach to decoding, but this was in Massachusetts and it involved an MIT professor. If that’s what it took to get state standards changed, then it’s really no surprise that most people are in the dark. Granted, I worked with teenagers, but I’ve been interested in reading pedagogy for a number of years now, and I had never heard of the three-cueing system until a couple of months ago. Even then, the encounter occurred only by chance, via a Google search.
Reading Adams’s article was like being led through a twisted hall of edu-mirrors. The reader is, for example, introduced to the figure of Regie Routman, a teacher-turned-literacy coach and the author of a series of books on reading pedagogy, who, based on the observation that only strong readers had a firm grasp of phonics, derived the backwards conclusion that these students could decode phonetically because they were strong readers—as opposed to the far more logical conclusion that they were strong readers because they knew phonics—and that overemphasizing a phonetic approach was therefore of limited value for weaker readers.
Here again is Adams:
[S]cientific research argues incontrovertibly that becoming a good reader depends on understanding and using spellings and spelling-sound correspondences and, conversely, that poorly developed knowledge or facility with spellings and spelling-sound correspondences is the most pervasive cause of reading delay or disability (Rack, Snowling, & Olson, 1992; Stanovich, 1986).
In a section on partnered reading in her book Reading Essentials, however, Routman includes the following list of strategies for children to follow if they are unable to identify a word:
- Give you partner time to think (wait time)
- Go back and reread
- Read past the tricky word and come back to it.
- Slide through it.
- Put in what makes sense.
- Sound it out with your partner.
- Cover part of a word and ask, “What does it say?”
- Look at the pictures.
- Ask, “Would you like me to help you?”
- Tell your partner what the word is.
Not only is the phonetic approach presented as one of many options, it is buried smack in the middle of the list, right below “put in what makes sense.” (Is there no possibility that more than one word might make sense?) Notably, Routman does not even advise that children attempt to sound out a word by themselves before asking their partner for help.
The message is clear: a little bit of phonics is ok now and then, but it should by no means be the go-to option, and there are lots of other strategies that are just as good.
Furthermore, the suggestion that students attempt to sound out full words is immediately undermined by the following, confusedly expressed point, which recommends covering part of a word and asking what it says. In a remarkable feat of pretzel logic, children are taught that reading means ignoring the letters that are actually on the page. One can easily imagine a six-year-old swapping “sold” for “stole,” or “picked” for “packed.” In fact, it is difficult to imagine a method better designed to facilitate misunderstandings.
The reality is that many children, particularly weak readers, lack the metacognitive skills to effectively monitor their comprehension, nor is it fair to expect them to do so beyond a very basic level—that’s why it’s so important that they be taught to read what words actually say. And given the largely non-supervised nature of the exercise, it is entirely likely that teachers would remain unaware of the extent of some students’ difficulties. (As Adams points out, Routman’s switch from classroom teacher to resource coach permitted her to avoid observing the longer-term effects of the methods she advocates.)
Adams’s article was published in 1998, which is right about when my first crop of students would have been learning to read. Between then and now, however, Routman has continued to work as a literacy coach, advising teachers and districts, and to publish several additional books. Barring evidence to the contrary, it seems reasonable to assume that many of the workaround, non-phonetic strategies she recommends are still being adopted for struggling readers. (Adams, by the way, is clear that she does not believe Routman herself is the source of the backwards interpretation of the three-cueing system; rather, she cites Routman’s work to illustrate the ways in which the error has been embraced and disseminated by influential figures in the field.)
I’m still trying to wrap my head around the fact that a misconception this significant has gained such widespread traction and that, moreover, it has managed to do so almost entirely below the radar of the research community—indeed, beyond any community outside the world of early elementary education. It also points the extent of the split between the mainstream research community and the K-12 one. They are, in effect, two parallel universes.
For me, the situation is fundamentally a cautionary tale about the way in which really bad ideas can insidiously gain a foothold and then spread throughout in a disorganized system populated by well-meaning but inadequately prepared individuals who either lack the knowledge—or, it also needs to be said, the common sense—to identify gaping problems in their training, or who find themselves so low on the totem pole that they have no clout to challenge the dominant approach. A colleague of mine pointed out that a lot of ed school types go out of their way to cow the newbies who might question them. (In that regard, teaching is exactly like any other field.) And once ideas get entrenched to a certain degree, practitioners can become extremely doctrinaire; power breeds obstinacy, not to mention defensiveness. If evidence to the contrary exists, the system will support their rejection of it, indeed ensure that they remain insulated from it entirely. After all, teachers are the real experts, not some stuffy scientists who don’t even work with children, know them as unique individuals, etc. And why would anyone pay attention to people who don’t really know what they’re talking about?
Note: I’ve done a lot more research since this piece was first published and wrote a list outlining 10 reasons three-cueing methods are ineffective. There’s a little bit of overlap with some of the issues discussed here, but a number of the more serious ones are not at all obvious, and I was unaware of them when I wrote this post.