When it comes to talking about improving students’ reading, one of the factors that makes having a coherent conversation so challenging is that the word “reading” itself has two meanings: it can refer to decoding—that is, the literal process of matching squiggles on a page to their corresponding sounds in the English language—or it can refer to the much more sophisticated process of comprehension, which is also dependent on things like vocabulary, ability to navigate various types of syntax, and background knowledge. Although the same word is used to describe both of these abilities, the first meaning does not necessarily imply the second.

And as if that weren’t already complicated enough, there’s yet another factor that is often overlooked: listening.

A while back, I came across research indicating that elementary school students’ maximum level of reading comprehension is determined by their level of listening comprehension—that is, children cannot both decode and comprehend written texts at a level surpassing what they can understand aurally. This theory, laid out in a seminal 1986 paper by Philip Gough and William Tunmer, is known as the “simple model” of reading comprehension. While there may be aspects of it that have not been fully explored, it is accepted as generally accurate among reading researchers.

So while children may be able to comprehend texts that far exceed what they are able to decode, the opposite is not true: they cannot, under normal circumstances, comprehend written texts that they would be unable to understand by ear.

When you think about it, this is perfectly consistent with observable reality: plenty of children who are not yet fluent readers have no problem understanding, say, Harry Potter when the books are read aloud.

Conversely, a precocious nine-year old may be able to decode an economics textbook with near-perfect fluency, but without the slightest comprehension.

One implication of this model is that in order to ultimately become strong overall readers, students’ aural skills must progressively catch up to their decoding skills—that is, children must develop the ability to listen to gradually more complex forms of speech.

Typically, the two sets of skills do not converge until late middle school, when students finally begin to be able to read texts as complex as those that they can hear. But obviously, the texts that students are expected to read post-middle school continue to increase in complexity. In addition, as students progress through high school and then college, those texts become increasingly unrelated to everyday spoken language.

As has been very well documented by this point, American high school students are not exactly stellar readers. Between 1995, when scores for the old SAT were re-centered, and 2016, when the test was changed, reading scores declined from 504 to 494. (The fact that the average reading score on the new exam miraculously jumped nearly 40 points to 533, while the top score of 800 remained unchanged, strongly supports the thesis that the test was changed in part to deflect from the decline in reading scores. Indeed, today’s 533 Verbal corresponds to a pre-1995 Verbal score in the low 400s).

In addition, the NAEP (the national reading and math assessment administered to students in grades 4, 8, and 12 all over the United States every four years) shows a revealing trend. While reading scores among did improve by several points among 17-year olds in the mid-1990s, they have since declined. Furthermore:

The national trend in reading achievement shows improvement at ages 9 and 13, but not at age 17, between the early 1970s and 2012. The average scores for 9- and 13-year-olds in 2012 were higher than those in 1971 (13 and 8 points higher, respectively), but the average score for 17-year-olds in 2012 (287) was not measurably different from the score in 1971. (https://nces.ed.gov/programs/coe/pdf/coe_cnj.pdf)

Regardless of demographic trends that may be affecting scores, the data do seem to suggest that in American high schools, something is not quite working in terms of reading.

Perhaps it was because a friend who teaches high school had just explained to me that she could get punished on her evaluations if she was caught spending too much time talking (i.e., teaching) when administrators observed her classes, but when I came across an article on the connection between aural comprehension and reading comprehension by University of Toronto professor Andrew Biemiller, something leapt out at me.

Biemiller writes:

We often assume that children’s reading experiences contribute much to their increasing ability to comprehend language (e.g., Nagy & Herman, 1987; Sternberg, 1987). However, for many children, most language growth continues to come from non-print sources (parents, peers, teacher lectures, class discussions, television, etc.) throughout the elementary years. For many children, the skills necessary for reading printed English remain too poor for them to read texts that introduce new vocabulary and new conceptual structures.

Among people not familiar with the current state of the American classroom, it is more or less taken for granted that lectures are one of the pedagogical tools used in the classroom. But I’m not so sure that’s the case anymore (more on that in a little bit). And if that isn’t the case, then the implications in terms of reading are potentially quite serious.

Even though the simple model of reading applies to elementary school students, it seems plausible that some sort of link between aural comprehension and written comprehension remains. In order to read high-level texts, students must presumably be able to understand speech of approximately the same level. However, if students lack exposure to high-level speech, it stands to reason that they won’t (fully) develop the necessary aural skills, and their reading skills won’t improve either. Could the decline of direct instruction be contributing to poor high school reading scores?

Think of it this way: everyday spoken language bears little relationship to the kind of highly structured formal language found in academic writing—a kind of writing students generally aren’t exposed to until high school. (Virtually no one, even the most distinguished professors, speaks the way they write.) A lecture is essentially a bridge between the oral and the written. Though spoken aloud, it also contains linguistic features found more or less exclusively in writing: transitional words and phrases such as therefore and in fact; subordinate and relative clauses; concessions, e.g., words like although and despite; rhetorical questioning, etc.

Furthermore, a lecture—or even a 15-minute lesson—is also a model for how to think through an issue clearly and in a nuanced way. Ideas are clearly divided into discrete sections, with introductions, explanations, and analyses. In some cases, potential counterarguments and alternative explanations are also discussed. When students take notes, particularly by hand, they solidify and reinforce these skills and concepts.

But if students are not getting experience processing these key features aurally, they have considerably more limited means to comprehend them in writing. The gap between the spoken and the written becomes a chasm.

And increasingly, this type of formal speech is disappearing from the broader culture; if students aren’t exposed to it in school, they’re unlikely to be exposed to it elsewhere. Television news, for example, primarily consists of talking heads alternately spouting nonsense and hurling invective at one another – and at any rate, high school students are more likely to get their news from YouTube than from CNN. And it’s not exactly like most teenagers are watching PBS in their spare time. While they may be in a text- and media-saturated environment, it’s pretty safe to assume that most of the speech they’re being exposed to isn’t at a particularly high level.

Unfortunately, most people have no idea of the extent to which schools have actually changed over the past few decades, or that teachers spend increasingly little time at the front of the classroom, talking. And the standard narrative of course proclaims that nothing has changed at all, that students are still being educated according to a nineteenth century factory model, that radical, disruptive change is urgently needed to prepare students for the twenty-first century, and so on.

This excerpt, from an article by Prof. Elizabeth Sturtevant of George Mason University (aka Koch Bros. University) sums things up nicely:

Many American adults remember middle and high schools as places where teachers lectured absolutely endlessly—or, at least, for forty-two minutes. Students listened, took notes, and suffered through pop quizzes and multiple-choice tests. Occasionally, the teacher might ask a question. There was always a “right” answer, and success was dependent on how well the student could guess or remember yesterday’s lecture.

However deeply this picture of stultifying high school classrooms has worked its way into the American psyche, it bears an increasingly tenuous relationship to reality. (I also feel obligated to point out that it bears no relationship to my own school experiences in the mid-late 1990s, which were a reasonable mix of lectures, Q&A, and class discussions.)

What by and large predominates in ed schools today is almost slavish adherence to a small group of theorists (Dewey, Freire, Vygotsky, Gardner) along with an endless championing of child-centered learning and the power of technology in the classroom. The result is a bizarre mishmash of good intentions mixed with half-baked notions about differentiated instruction, active learning, and compulsive data collection.

Lest you think I’m exaggerating about all this, consider some typical lessons posted on the popular website Teachers Pay Teachers:

 

 

One teacher describes the prevailing atmosphere thus:

I was told that my students possessed multiple intelligences, and it was strongly hinted to me that the more technology I could accommodate into my lessons, the better their needs as digital natives would be met. My initial classroom design of rows and columns was frowned upon, and tables and horseshoes were recommended. And all because, I was told, the research confirmed each avenue.

And another (from the UK, where many of the worst practices of American education have taken root as well) describes how progressive pedagogy can itself become rigid and doctrinaire:

Nowadays, child-centred learning is an article of faith… If I question it at work I am met with bemusement at best, but usually righteous anger. Its principles pervade everything a new teacher hears about “best practice”: avoid chalk-and-talk; don’t point out a child’s mistakes (it will harm his self-esteem); never teach anything pupils may find boring; and never, on any account, organise the pupils’ desks in rows. Islands of desks where the pupils can “group learn” are dogmatically promoted. 

Moreover, the Danielson Framework, which was used as a basis for teacher evaluations (at least in New York) several years back, explicitly cited lecturing for 45 minutes as an example of “ineffective teaching.” Yes, of course, some teachers are terrible lecturers, guaranteed to put students to sleep within minutes. And it should go without saying that lectures of this length aren’t appropriate for young children. Other teachers, however, are stellar lecturers, for whom talking for 45 minutes is an extremely effective form of pedagogy. (Some of my favorite teachers in high school fell into that category—and yes, they always encouraged us to ask questions.) Yet schools would penalize them on strictly ideological grounds, without consideration of whether students are actually learning.

Now, to be clear, I am most certainly not arguing that teachers should do nothing but lecture, or that students should never discuss things with one another. I am, however, suggesting that there is a minimum point below which a lack of exposure to formal adult speech may begin to impede the comprehension of higher-level texts.

It would be fascinating to know, on average, the amount of time teachers spent talking in class a few decades ago compared to the amount of time they spend today. I doubt the statistics exist, but I would bet that there was a sharp drop-off sometime in the early-mid 2000s, when the oldest Baby Boomers began to retire in waves.

At any rate, what I want to know is this: Are students actually listening to structured adult speech in class? And if so, how much (or how little)? Or are they spending most of their time sitting in groups, listening to each other talk and thus never developing their aural comprehension past a teenage level? I suspect that the answers to those questions are respectively “not really,” “too little,” and “yes.” I didn’t realize this when I was tutoring, but in retrospect, it would explain a lot.

It is, I think, part of why my students used to have such a terrible time with tone questions on the old SAT. (Not coincidentally, these questions have been almost eliminated on the new exam.) Identifying tone requires some sense of connection between written and spoken word, particularly when things like sarcasm and irony are involved (concepts that are notably absent from the redesigned SAT as well). If those connections are not developed via speech of a certain level, they become much more difficult to recognize on the page.

It’s also part of why I’ve always found writing so incredibly difficult to teach: if students lack an intuitive sense of how phrases and sentences and thoughts are constructed—a sense that is almost impossible to develop without knowing how they should sound—then no amount of memorization can compensate.

For students with highly educated parents, what goes on—or doesn’t go on—in school is not such a big deal: they can just have things explained at home. Obviously, no one is going to monitor how much time their parents spend talking, or call them out for explaining things directly instead of making their children “discover” them naturally! This is why wealthier students can get away with a project or discussion-based curriculum; what isn’t covered in the classroom can be learned at home or outsourced to a tutor who does provide direct instruction.

And of course there will always be a small minority of truly gifted, motivated students who are in fact able to teach themselves to a high level—but the vast majority of students do not fall into this category. The exception should not be taken as the norm and/or used as the basis for policy.

When students do not get exposed to a high level of speech at home or elsewhere, then problems start to occur. But if the curriculum is sufficiently watered down or grades sufficiently inflated, then they might not be noticed until a student starts to prep for the SAT or ACT.

Incidentally, this is is not purely a socioeconomic issue: some of the wealthiest students I taught had reading issues that almost certainly stemmed in part from lack of exposure to higher-level speech.

Still, disadvantaged students are inevitably the ones who bear the brunt of this kind of pedagogy. If teachers do not explain things to them directly, not only are they unlikely to learn the actual subject matter—below a certain baseline level of knowledge, it is extraordinarily difficult to put things together on one’s own—but they will also be deprived of the experience of listening to what a clear, coherent explanation sounds like in an academic context, and of being exposed to important vocabulary, sentence structures, phrasings, idioms, etc.

Interestingly, research suggests that the amount of vocabulary students learn through explicit instruction is quite limited, maxing out at about 400 words/year. An enormous amount of vocabulary knowledge is gained passively, through repeated exposure to new words. As E.D. Hirsch describes:

All parties…agree that even when teachers spend up to thirty minutes a day in explicit word study, the maximum number of new words they can teach this way during a school year is about four hundred. Compare that to the average of a thousand to five thousand words per year that an advantaged child will have learned from age two to seventeen. It is clear from these ballpark figures that most of our word learning occurs indirectly, through hearing, reading, and understanding a lot of text and talk. The consensus of all teachers is that indirect, implicit learning is by far the main mode of increasing one’s vocabulary. (The Knowledge Deficit, 62)

If students are primarily listening to people their own age speak – people with similarly limited vocabularies—it stands very much to reason that their ability to acquire new vocabulary will be limited.

But in the vogue for project-based learning, these types of concerns go almost entirely unmentioned. And if test scores then demonstrate that students’ skills are in fact lagging, then the tests are promptly deemed useless as well. As exhibit A, consider the current SHSAT fiasco in NYC. Call it the “shoot the messenger” model of education.

Thus, even if a lack of direct instruction were in fact playing a role in low reading scores, that finding would almost certainly not be permitted to interfere with current practices. (As a side note, I suspect that at some level school reformers don’t actually want things to improve: persistently stagnant scores allow them to operate in continual crisis mode, which justifies the implementation of even more reforms, creating a never-ending cycle.)

As Paul Bruno, a doctoral fellow at USC’S Rossier School of Education has pointed out:

When students are mostly proficient or advanced, teachers, administrators, and parents tend to have plenty of independent verification that students are skilled; ambiguous, student-centered activities are not relied on for demonstrations of mastery. With lower-skilled students, adults are more likely to be worried about their students’ skills, because much of the available evidence (e.g., test scores, independent classwork) suggests those skills are absent or weak. When students engage in student-centered activities, they can easily give the illusion of proficiency—talking to one another, handling materials, and so on—especially if you don’t examine their work too closely or don’t know what you’re looking for.  And it’s easy to interpret ambiguous evidence of learning favorably if you really want to see proficiency (as most educators do).

As I’ve written about before, the goal is a sort of “performance” of active learning—the appearance of what teachers (or administrators) think learning should look like is given precedence over whether students are actually learning. Style wins out over substance.

But to come back to my original point here: if students are primarily sitting in groups and talking to one another, they are not hearing adult speech. If they do not hear adult speech, they will not develop more sophisticated aural capacities. If they do not develop more sophisticated aural capacities, they will not then be able to form relationships between spoken and written words. If they cannot form those relationships, their reading will persistently fail to improve.

And then everyone will end up… exactly where we are now.