Reading wars: Content vs. strategies

Round 2 of the reading wars pits content knowledge against reading strategies, writes Dan Willingham on Britannica Blog. (Round 1 was phonics vs. whole language.) Both sides have some merit, but he gives the nod to content.

Most of us think about reading in a way that is fundamentally incorrect. We think of it as transferable, meaning that once you acquire the ability to read, you can read anything. That is true for only part of what it takes to read. It’s true for decoding—the ability to translate written symbols into sounds. Once gained, that ability can be applied to any string of characters, including unfamiliar words like operculum, pronounceable non-words like slint, and letter strings like ctpaqw, which you readily identify as non- pronounceable.

But being able to decode letter strings fluently is only half of reading. In order to understand what you’re reading, you need to know something about the subject matter. And that doesn’t just mean that you need to know the vocabulary — you need to have the right knowledge of the world.

. . . Research findings consistently show that students who are identified as “poor readers” suddenly look quite good when they read passages on familiar subjects.

Schools that try to boost test scores by spending tons of time on reading skills are doomed to failure, he writes. Students need to learn about the world — including history, geography and science — to understand what they read.

About Joanne


  1. There is, of course, a bit of a paradox here. In school, students learn much of what they know about the world through reading (especially at the secondary level), and if they struggle with reading then they will struggle with learning the content knowledge that they need to improve their reading. Top this off with the fact that many science and history textbooks are written at levels above our struggling readers, and it becomes even more complicated.

    What I would love to see? More interactive textbooks that have information written at multiple reading levels, so that struggling readers can access the same basic content but at their own level.


  2. Ah, Parry–your dream has come true (nearly). In fact, the field of “assistive technology” has been responsible for fostering Universal Design for Learning. They wisely realize that, in ways similar to building ramps and curb cuts that benefit people beyond those in wheelchairs–and ought to be considered an implicit part of all new public building, so the production of electronic texts with all kinds of capabilities for interaction and differentiation benefits many students–not just those who are “disabled.” Most districts and classrooms have far more of this available to them than they are currently using.

    E-books are available that can be set to 1) read text aloud, highlighting each word as it is read; 2) read text aloud without highlights; 3) read words aloud when clicked; 4) include hyperlinked text with definitions or further meanings–and other things that I am just not familiar with.

    Basically the computer is able to provide the kind of differentiation that is provided by a one-on-one tutor (explaining, providing unfamiliar words, reading out loud, etc). Apparently this kind of assistance has been found to increase reading ability as well.

    Why isn’t this in use in every classroom? Teachers are not adequately trained, they tend to think of it as being only for kids with disabilities, they don’t do computers, they don’t know about it. Reality is, this is pretty revolutionary stuff. I have faced some real struggles trying to bring about use of tools that would be helpful in teachiing my son. But it takes an ability to look at things differently–and to really accept that it is possible for kids of many ability levels to learn within the same classroom.

  3. Margo–bear in mind that if you can decode, then a program that merely reads aloud won’t help at all. (In adults who decode well, listening comprehension and reading comprehension are correlated around .8 or .9, that is, they are essentially the same.)
    Parry–there are lots of ways of learning other than reading–substantive read alouds, for example–that I don’t think are being fully exploited.

  4. Miller Smith says:

    I could care less about the reading method used but rather care more about students who cannot read but were graduated forward year after year. Kind of a cover-up for the method of teaching reading.

  5. I don’t understand how there ever could have been people who thought that reading strategies could make up for lack of content knowledge.

    I’m a voracious reader, but there were some books which took 2-3 hours to get through 2 pages: college math books, particularly advanced calculus, modern math, and physics. Engineering texts were not as bad–only 1-2 hours to get through 4 pages. By comparison, fiction novels take me about 1-3 minutes per page, unless it’s really light reading.

    James Joyce takes about 10 minutes per page, and even then, I don’t understand most of it. And poems, forget it. I can spend 30 minutes on 10 lines and still not understand it.

    So content knowledge obviously matters a lot.

  6. My assumption — I don’t know this for sure — is that reading strategies, “child-centered” teaching and differentiated instruction are joined at the hip. When you have heterogeneously grouped classes, all reading self-selected books based on individual interest and reading level, then instruction can no longer be about the text. It has to be about a generalized skill that can be applied to any text. Or any child.

  7. One wonders what the reading scores for students upon whom the “one never has to learn anything because one can always look it up” mantra has been inflicted are like…

  8. This sure seems like one of those either or choices that shouldn’t be made.

  9. I have to lean toward reading strategies being the most important. In my unlearned opinion one of the most important skills for a reader is the ability to find meaning from context, even when the familiar context is sparse. In essence one must first build a working context from whatever is available, even if it take multiple passes or a few side trips to ask an adult or look at a dictionary.

    I can remember finding an old copy of Guadalcanal Diary when I was eight or nine, and reading it in bed by flashlight. As I recall the text was no problem, although there were plenty of unfamiliar words. The context I acquired gradually as I read along, but I didn’t have to get very far into it before being hooked. It ended up being my most vivid childhood reading experience.

    Maybe the best thing to do, once the basic skills are learned, is to make some adult-level books available and hint that they’re not really appropriate for children.

  10. I didn’t receive any reading instruction at any level – like Joanne, if I remember her old blog posts properly, I picked it up on my own.

    I do pretty well. I come from a long line of voracious, capable readers, none of them too famous for formal education. I guess it’s a modern miracle that so many generations can know so much without formal lessons in “skimming” and “chunking.”

    I’d say [unscientifically] that content matters a great deal more than weakly-theorized, weakly-tested reading methods implemented by a corps of teachers who topped out at 430 on the verbal section of the GRE.

    So, content matters the most, or my clan has bucked normalcy and proven ourselves super-geniuses over the last 100 years. Choosing either option will brighten my day.

  11. Mark Roulo says:

    I’d say [unscientifically] that content matters a great deal more than weakly-theorized, weakly-tested reading methods implemented by a corps of teachers who topped out at 430 on the verbal section of the GRE.


    You are aware of the difference between an average
    and a maximum, right?

    -Mark Roulo

  12. Dick Schutz says:

    Hey, everyone. Round 1 of Whole Language vs Phonics is far from over. But it’s being won by Whole Language, under the mask of Balanced Literacy.

    Little is to be gained by raising the matters of “content” and “strategies” as a “versus.” The tis/taint battle would go on ad infinitum. We should learn from Round 1. Had “Whole Language” been up against the “Alphabetic Code” rather than “phonics” the battle would have been short lived. The Alphabetic Code has all the technical standing of the Genetic Code and the Table of Periodic Elements. “Phonics” has lost all meaning. “Balanced Literacy” accepts “phonics,” but Whole Language Rules.

    What Willingham has said elsewhere regarding “Readiing Strategies” is compelling. The “strategies” that have been identified can be taught readily. And this “how to” should be taught. But after that, the term “strategy” is an empty metaphor.

    Re “content.” It’s unarguable that you won’t understand anything when you lack the requisite background information. And the utility of reading is to enable an individual to expand and enlarge personal background information via text.

    But there are important matters of personal choice, and the content that is best embedded in personal memory rather than accessed from Internet memory to consider. These matters are fogged by setting up a spurious “Strategies vs. Content” pseudo-battle.

    Currently students are mis-served in both matters. The “strategies” being taught are of limited utility and very little thought is being given to “content matters.” (CK, of course excepted)

  13. Hey Tabor, my method achieves actual measurable results with students across a wide spectrum of backgrounds and abilities (including those who are mentally retarded). How about yours? I learned to read easily, too, but I’m not egocentric enough to assume everybody is just like me.

  14. > I learned to read easily, too, but I’m not egocentric enough to assume everybody is just like me.

    And yet we all do share common qualities so maybe it isn’t quite so egocentric to assume that a relatively simple skill like reading has one, best method of instruction and deviating from that one, best method results in higher rates of illiteracy.

  15. Dick Schutz says:

    You’re getting close Allen. The history of language and the history of the English language in particular has produced one Alphabetic Code. The Code is our most treasured heritage, but it is generally maligned rather than honored. The Code is what permits English speaking people to be commonly understood when speaking with very different national and regional dialects. At the same time it allows us to read as one. That’s quite a trick when you stop to think about it, isn’t it?

    As a result of openness to accept new elements over the centuries, the Code is a bit complex. The complexity is an inescapable consequence in an alphabetic-based language that has 26 letters and 40ish sounds (the number of sounds depending on which linguist is counting) But ALL of the complexity can be accounted for by about 175 letter/sound correspondences. And the difficult correspondences can be reduced to 30ish by instructional sequencing.

    Teaching kids how to handle the Alphabetic Code in order to read is a very feasible instructional endeavor. However, the Code is fogged with educational mush-talk surrounding the term “phonics”–a term that has lost all meaning because it is used in so many different ways.

    There is one best and only Alphabetic Code. There are a handful or so instructional ways to go about teaching young children how to handle the Code in order to read. And there has to be tolerance for variation in children’s rate of learning due to individual differences in general ability.

    Alternative product/protocols for teaching reading expertise DO result in higher rates of illiteracy, but almost no attention is devoted to alternative instructional product/protocols. Prevailing “wisdom” is that qualified teachers can make any program work and that instructional failures are due to student deficits. That’s just flat out wrong-headed thinking, but it’s deeply and widely embedded.

    How long will it take to change the belief system? Your guess is as good as mine. Unfortunately, we lack weapons of mass instruction. And we have to deal with the instructional weapons we have, not those that we’d like to have.

  16. As a former teacher for 25 years, I can’t see how any teacher today can teach reading using the published, state adopted programs that are out there for their use. These programs, which are not called Reading Programs, but called “Language Arts”, not only contain very little instruction in reading skills, but little what is provided has no research base for its scope and sequence.

    Most of the time teachers are left to their own devices as to what skills they can/want to teach and in which order. They are given ‘programs’ that are so unmanageable that it takes months of study to figure out how to use them. These programs contain so much material that it could NEVER be covered in 180 days….even if teachers had 180 full days of instruction in the school year to teach them.

    By the time a teacher does figure out how to implement the program and gets somewhat comfortable with it, a new Language Arts adoption is presented (usually about once every 3 years!). It’s no wonder children who cannot read before coming to school have little chance of learning to read from the instructional programs provided. If parents don’t help ouite A LOT, then you have many children left behind.

  17. Dick Schutz says:

    That’s about the size of it, Donna. The prevailing view is: A “qualified teacher” can make any “program” work. And efforts are then devoted to get “qualified teachers”–which is job- making for colleges and universities. Standardized achievement tests provide a fog that provides cover for the status quo. The tests are sensitive only to differences in SES, not to differences in instruction, so parents or “society” get blamed for shoddy instruction. Some kids learn without any instruction or despite mis-instruction. The rest are shuffled for to special ed, get by through social promotion, and/or drop out.

    Yet government and school authorities tell us each year “We’re making gains” and the media swallow the spin.

    That’s been the history. What the future will be remains to be seen.

  18. Mark,

    I’ll stick to Stats 101 content so the folks as dumb as I am re: math can keep up.

    Using the ETS’ GRE stats is unreliable for most fields. Test-takers indicate the discipline to which they’re applying; it isn’t necessarily a reflection of who actually enrolls. In certain disciplines, though, we can look at applicant data and know about enrollment data. Early Childhood Education is one of those disciplines because acceptance rates are astronomical [esp. at public universities, where Early Childhood Ed. is famously a bit of a cash cow]. The data that follows comes from test-takers between Oct 2002 and June 2004. [There’s a newer round of data, but it’s on my other computer.]

    The mean verbal score for Early Childhood Ed applicants is 418 [sample size of 1,418, so enough to draw conclusions] with a standard deviation of 83 points. 4.2% of applicants scored between 200-290 [remember, 200 is the minimum score]. 39.3% scored between 300-390. 39.1% scored between 400-490.

    82.% of Early Childhood Ed. applicants score below 500 on the verbal portion of the GRE. For the sake of comparison, 68.5% of applicants to Industrial Engineering graduate programs [a non-Humanities discipline which I chose randomly] score below 500. 38.4% in Political Science. 31.5% in History.

    Elementary Education fares little better with language than Early Childhood. 2.1% in 200-290, 30.0% in 300-390, 41.3% in 400-90 for a mean of 443 and a standard deviation of 88.

    73.4% of applicants to Elementary Education programs score below 500 on the verbal portion of the GRE.

    To say that our early/elementary teaching corps is populated with strong readers and writers would be a mistake. If we recognize that we often use flawed curricula [such as Whole Language and its cousins] implemented by those with limited verbal skills, we can begin to understand why we’re in the mess we’re in.

    Raising standards for education schools or certification is difficult financially/politically and it’ll take a long time to pay off. Committing to instructional methods/curricula like DI? Fairly cheap, quick, and effective.

  19. Lightly Seasoned,

    That was an awfully spicy comment given your handle. I’ll parse it.

    “…my method achieves actual measurable results with students across a wide spectrum of backgrounds and abilities (including those who are mentally retarded).”

    allen touched on this – it’s precisely that commonality, rather than egocentrism, on which my point rests. I don’t know if you’re a practitioner of DI, but Engelmann’s results with both gifted and mentally retarded [if I remember correctly?] groups, including measurement of longer-term information retention, are impressive suggestions that we all have a bit more in common re: cognition than we realize.

    “I learned to read easily, too, but I’m not egocentric enough to assume everybody is just like me.”

    I’d argue that I’m kind enough and compassionate enough to assume – and truly believe – that the vast majority of us have a great deal in common. The tremendous, demonstrable success across widely-varying populations that DI has shown would suggest, at worst, that I’m not entirely wrong.

  20. Andy Freeman says:

    The problem with this debate is that folks think that method matters in a way that it shouldn’t.

    The point is not whether a teacher uses “the right method”. The point is whether the student learns to read. If a student fails, it doesn’t matter what the teacher did – it wasn’t good enough. (Yes, I realize that some kids will fail no matter what the teacher does. Not all problems are solvable.)

    We’re never going to have decent education as long as we let teachers use “but I taught using {method name}” as an excuse, regardless of the method.

    I don’t care what method a teacher uses, even the “Harold Hill” method. I do care whether the student leads to read.

  21. Sorry, Tabor. Been locked in a room for a week writing a state assessment that tests, well, whether or not high school kids can read. I’m a bit crabby.

    My point is not that we are all intellectual islands, unable to learn anything the same way; my point is that learning comes easier for some than others. We label this ability intelligence. For those of us who are quite intelligent, the process is transparent because we grasp it quickly and easily. For those who are less intelligent or have disabilities, the process has to be made more explicit in order for them to grasp it. Skimming, chunking, and all these techniques are something we picked up easily and without too much effort, so we don’t notice them. They’re automatic. But children who struggle have to be taught these things.

    I’ve spent years learning how to break down stuff that I find quite simple so that others can understand it. The way I do it is not the be all and end all one true way. But it is working in my classroom and producing measurable results (I assess reading levels regularly), so why you would throw cheap shots at a method that actually teaches at risk students to read — what you say I should be doing — is a bit mystifying.

    PS. I never took the GRE, but scored in the 90th percentile on the LSAT. Am I verbal enough to teach high school English?

  22. Lightly Seasoned,

    You can call me Matt. I’m glad that you’ve had a great deal of success with the methods you’ve developed. It’s encouraging. Hopefully we can streamline and make more efficient those practices so they might be applied more broadly.


    Scoring in the 90th percentile of the GRE or LSAT shows that you’re a capable reader – an attribute I’d value in a literacy teacher.

  23. Dick Schutz says:

    “Been locked in a room for a week writing a state assessment that tests, well, whether or not high school kids can read. I’m a bit crabby.”

    Wow! You’ve got every right to be crabby–outraged in fact–if that’s how you spent your week. Determining if a kid can read should have been over and done with by the third grade at the latest.

    Betcha the test you constructed is confounded with background information matters that goes beyond reading and that makes the test a measure of general ability, not reading expertise.

    If that’s not the kind of test you constructed, tell us more about it. But that’s the kind of tests that prevail in alleged “reading tests” at the high school level. The test you constructed possibly broke out of the tradition.

  24. Dick Schutz says:

    “Raising standards for education schools or certification is difficult financially/politically and it’ll take a long time to pay off. Committing to instructional methods/curricula like DI? Fairly cheap, quick, and effective.”

    Why is it so hard to get people to accept and act on this point, Matt?

    The only explanations I can come up with are: One. Ed students are cash cows for colleges and universities. Since Ed departments are low on the academic totem pole, those at the upper levels of the pole don’t really care and are glad to have the Ed departments to look down on.

    Two. The fog of standardized achievement tests protects El Hi personnel and maintains the status quo irrespective of personnel characteristics or instructional program characteristics. Worse, no one examines instructional programs and if they do, they don’t know what to look for.

    Have you got a better explanation?

  25. I am writing to some pretty narrow criteria. I will say the questions are carefully aligned to state standards for the grade level and that each item is carefully screened to not require previous knowledge other than how to read, grade-level vocab, comprehension, literary technique, etc. Whether the test is effective as a measure of learning or not I’m not sure yet. In constructing the items, I have tried to think very carefully about how a student might think through answering it. Since I spent time in an inner-city school, I did try to think of the experiences of those students. OF COURSE you can’t separate background knowledge from comprehension. Is it possible to construct a test like that? These passages are killer-bland, though. I’ll probably blog about the whole experience in my own space next week when I’ve had time to process. I’m just a writer; I had nothing to do with its design — that’s DESE’s job.

  26. Dick,

    The economic part is important, especially at the public universities. They’d yelp from the rafters if these ed programs were reined in sensibly.

    I don’t think upper divisions enjoy looking down on the ed schools. My sense is that scholars wish the ed schools were more rigorous and had a stronger seriousness of purpose. Having said that, they love a real, solid education scholar [and there are lots of them – they’re just tough to find].

    The reason no one will give, though? That the current climate of education debate has rendered it nearly impossible to criticize the martyrdom of teachers and admins – or curricula and the other parts of the school day. There’s a reason why the most successful critical measures in local public education are almost always bond/infrastructure initiatives. Even then it’s hard, but there’s a little less taboo there.

    There’s plenty more to it, obviously, so I won’t hijack the interesting literacy debate going on in this thread. And, unfortunately, the issue will still be there tomorrow for us to discuss.

  27. Joanne, thanks for sharing this, and for giving your commentary on it. I’ve featured this post on my blog as one of The Cornerstone accolades for January 2009.

    See you on Twitter!


  1. […] from Robert Pondiscio of the Core Knowledge Blog on whether content knowledge is part of most school’s reading […]