The Core Knowledge Reading Program:

Where We Are Now, and What You Can Do in the Meantime

---

Outline:

1. The Simple View of Reading: R = D x C
2. More about D
3. More about C
4. An Ideal Program Will Build D and C
5. How to Build D
6. How to Build C
7. The Core Knowledge Reading Program
8. What Not to Do in the Meantime
9. What to Do in the Meantime

I. The Simple View of Reading: R = D x C

Although you came to hear about reading, not math, I will begin with an equation — an equation which provides a good starting point for understanding the Core Knowledge Foundation’s philosophy of reading. Here’s the equation:

R = D x C
where
R = Reading comprehension ability, expressed as a numerical value beween 0 and 1
D = Decoding skills, expressed as a numerical value beween 0 and 1
C = Language comprehension ability, expressed as a numerical value beween 0 and 1

This equation is a mathematical version of what is sometimes called “The Simple View of Reading.” The Simple View is based on research by Philip Gough and other reading researchers.

According to the Simple View, there are two elements that are equally important to reading comprehension — decoding skills (D) and language comprehension ability (C). To achieve reading comprehension a child needs not only to be able to decode the words but also to be able to make sense of those words. The first is made possible by D, the second by C.

The Simple View holds that a child’s ability to read and understand text can be predicted if you know the child’s ability to decode words (D) and his or her ability to understand spoken language (C).  If a child is a rapid and fluent decoder and also able to understand rich oral language, for instance, a story read aloud or a book on tape, then it’s a safe bet the child will not have difficulty with independent reading comprehension.

The Simple View can also account for reading deficits. Research has shown repeatedly that when a student has deficits in reading comprehension, that student always has deficits in either language comprehension or decoding skills, or both. There are some children whose reading comprehension suffers because, while they are very good at understanding spoken language (C), they are poor at decoding text (D). These children are called true dyslexics.  There are other children (although these are very rare) whose reading comprehension suffers because, while they are good at decoding text (D), they have difficulties understanding spoken language (C). These children are sometimes called hyperlexics.  Most commonly, when a child has difficulty with reading comprehension, the child is having difficulty with both decoding (D) and language comprehension (C). Children with deficits in both areas are described as having garden variety reading difficulties.

Look back at the equation. Notice that the variables can range in value from 0 to 1, where zero represents no ability whatsoever and 1 represents perfection. (Obviously most people fall somewhere between these extremes.) Since any number multipled by zero is zero, as a child’s decoding skills (D) approach zero, so also will his or her reading comprehension (R). But the same is true of language comprehension. As C approaches zero, so to does R. If you can’t understand a set of words read aloud, you won’t be able to understand the same words written down, no matter how good your decoding skills.

An example can make the “simple view” concrete: the example of Milton and his daughters. The English poet John Milton went blind late in life. Since Braille had not yet been invented, this meant he could not read for himself. But he had friends and relatives read books aloud for him. But Milton couldn’t always find a scholar who had free time to read to him in Hebrew and other ancient languages. The solution? Milton taught his daughters to decode Hebrew and other ancient languages so they could read books in those languages aloud to him. But Milton did not teach his daughters the actual languages — the words and meanings. He only taught the phonemic rules they would need to turn letters into sounds. Thus, his daughters had solid decoding skills for these languages (D>0) but would have scored a zero in language comprehension (C=0). They could turn symbols into sounds but had no idea what the words meant. Milton, on the other hand, on account of his blindness, had no functional decoding skills (D=0) but (by virtue of his great learning) a very high value for C. Between Milton and his daughters there was reading comprehension (R), but the younger generation brought the D and the patriarch brought the C. It must have been a miserable experience for the daughters to have to decode without ever being able to understand.

This is a rather freakish example meant to show the separability of D and C. Of course, the goal of this program is to avoid all such splits, to build D and C in all children, male and female, black and white, etc.

II. More About D

If D and C are the critical components of reading, it pays to know a little more about them. Let’s start with D. Here are just a few of the skills and subjects that are closely related to Decoding Skills. Some of them might be thought of as components of D:

Knowledge of letters of the alphabet
Book and print awareness
Phonemic awareness
Phonics (as an optimal means of teaching decoding)
Decoding speed
Decoding accuracy
Decoding pseudo-words
Fluency

III. More About C

What about C? Here are some areas of knowledge that can be thought of as elements contributing to our ability to understand language.

Knowledge of grammar and syntax
Vocabulary
Background Knowledge
Cultural literacy
Knowledge of key subjects, e.g. history, science, the arts
Knowledge of the world

IV. An Ideal Curriculum Will Build Both D and C

Since reading depends crucially on both D and C, an ideal school curriculum will teach both D and C. How can each of these goals be achieved?

V. How to Build D

The great debate in decoding has been between “whole language” and “phonics.” Phonics instruction introduces letter-sound correspondences explicitly and (typically) provides a great deal of rehearsal and practice with decodable text that uses the letter-sound correspondences studied. Whole language discourages breaking words up into meaningless sounds and looks to immerse the child in meaningful language experiences. The assumption is that the child will begin to recognize whole words and will eventually learn to read “naturally” by inferring how the English system works, provided the atmosphere is sufficiently rich.

Core Knowledge believes that, among cognitive scientists and reading researchers who pay attention to results, there is a sturdy consensus — growing stronger with each passing year — that systematic phonics instruction is the superior way to teach decoding skills (D). This has been shown by numerous researchers and also by the review of research done by the National Reading Panel, whose reports are available online:

http://www.nationalreadingpanel.org/

For more information on reading and the superiority of phonics approaches, consult any of the following books (which cover a lot of the same ground):

• Susan Hall and Louisa Moats, Straight Talk About Reading
• Jean Chall, Learning to Read: The Great Debate
• Marilyn Jager Adams, Beginning to Read: Thinking and Learning About Print
• Diane McGuinness, Why Our Children Can’t Read
• Catherine E. Snow, M. Susan Burns, and Peg Griffin, Preventing Reading Difficulties in Young Children
• Keith Stanovich, Progress in Understanding Reading

So, as far as decoding is concerned, CKF is firmly on the phonics side, not the whole language side.

However, saying that phonics instruction is the best way to teach decoding is not the same as saying that phonics alone is enough to ensure children become good readers. Remember the Simple View:

R = D x C

Even the best decoding program will not turn children into strong comprehenders by itself. Schools need to address not only decoding (D) but also language comprehension (C).

VI. How to Build C

During the early years, when children are still learning to read on their own, they gain almost all of their knowledge about language from listening. Even once children master the basics of decoding, they remain better listeners than readers. In fact, reading comprehension abilities do not catch up to listening comprehension abilities on most measures until seventh grade!

Therefore, the best way to build language and knowledge and vocabulary (all components of C) is through frequent reading aloud. We recommend that teachers aim to include at least two read-aloud experiences per classroom day, 150 days a year. By a read-aloud experience we mean the sharing of a poem, story, or nonfictional excerpt, lasting 5-15 minutes, preceded, followed, and (where appropriate) interrupted by discussion.

Of course, children soak up words from an early age simply by listening to people speaking around them. So you might think just listening to everyday talk would be sufficient to build C. However, Hart and Risley have shown that home environments differ dramatically in the amount of language exposure they provide. A child in a language-rich home may hear millions more words in the first few years of life than a child in a language-poor household. All of this exposure will tend to boost C for the student in the language-rich home. But C is likely to lag for students in the language-poor household. Thus, the achievement gap takes its rise in the home. If this gap is to be narrowed, schools must provide exposure to rich language of the sort found in printed materials.

Research has shown that books contain a significantly richer vocabulary than everyday speech. When we speak we use a small subset of the words in English. Writers use a larger set. Even children’s books contain significantly more rare words per thousand than the recorded conversation of college professors. This is one reason why it’s important for kids to hear printed text as well as oral speech.

It’s important to read a wide range of materials, including lots of nonfiction. When a parent or teacher reads aloud a nonfiction story about George Washington, students are exposed to a different set of words than they are exposed to when they hear about Cinderella. Kids need to be exposed to a wide range of subjects, words, and genres.

One other guideline is important. Teachers should make an effort to group stories into clusters, or domains. By assembling a set of read-alouds on the American Revolution, teachers can increase the chances that children will get multiple exposures to key words from this domain, like revolution, taxation, and colony.

In addition, research shows that students learn more and comprehend better when they encounter a book or story on a subject about which they already know something. Relevant prior knowledge helps them learn more from the book or story. By grouping stories into domains, teachers can create a situation in which the early stories in the domain provide a familiar context for the later stories. Thus, in the later stages of the domain, students will be listening with more prior knowledge and will therefore be more likely to understand and learn from the experience. 

There is an additional reason for grouping stories into domains. Domain knowledge has been shown to compensate for low overall verbal skills. Generally speaking, if you read a story to children of different ability levels, the high-verbal children will tend to learn more from the experience than the low-verbal children. This is a specific example of the so-called “Matthew Effect.” If the high-verbal children learn more from each classroom experience than the low-verbal children, then the achievement gap will tend to widen over time. This is a very difficult problem to overcome — and it is a problem for all sorts of instruction, not just for reading aloud. However, there is one situation in which low-verbal children have been found to perform at levels equal to high-verbal children, and that is when the low-verbal children have relevant knowledge of the domain.

An important experiment by Recht and Leslie (successfully replicated by subsequent researchers) looked at the impact of baseball knowledge on comprehension of a baseball-related text. The authors arranged a clever experiment in which students were divided into four groups according to reading ability (high or low) and domain knowledge concerning baseball (high or low). All of the students were asked to read a story and answer questions about the story. Recht and Leslie found that students with high domain knowledge performed better on all assessment tasks, and that students with low reading skills were able to perform as well as the better readers when they had high domain knowledge. The authors wrote, “It appears that knowledge of a content domain is a powerful determinant of the amount and quality of the information recalled, powerful enough for poor readers to compensate for their generally low reading ability.”

Since domain knowledge can compensate for low verbal skills and allow low-aptitude learners to perform like high-aptitude learners, it may be possible to overcome, or at least mitigate, the Matthew Effect associated with listening to stories read aloud if students can be given some basic domain knowledge that will enable low-verbal children to listen at a level more nearly equivalent to that of their high-verbal peers. The early stories in each domain should be chosen to be good introductory materials and provide scaffolding for enhanced learning later in the domain. In this way the domain structure should help to ensure that the read-aloud experiences are optimal learning experiences for disadvantaged as well as advantaged students.

A good way to learn more about the importance of background knowledge in reading is to read some of E. D. Hirsch’s recent articles. Here are some links:

• http://www.aft.org/pubs-reports/american_educator/summer2001/lang_gap_hirsch.html
• http://www.aft.org/pubs-reports/american_educator/spring2003/AE_SPRNG.pdf
• http://www.coreknowledge.org/CK/about/articles/index.htm

VII. The Core Knowledge Reading Program

Core Knowledge is currently working on developing a reading program. Our reason for doing this is because we feel that most reading programs take a formalist view of reading. That is, they think of reading as more or less the same as decoding. There is a tendency to think of reading as purely a “skill,” and a skill that does not depend very heavily on background knowledge. We think this is wrong.

Remember: the Simple View of reading holds that C is just as important as D. It is our feeling that, while a growing number of existing program do a good job teaching decoding skills (D), none does an exceptionally good job at building comprehension ability (C) and its components — background knowledge, vocabulary, cultural literacy, etc.  We hope to create a program that combines optimal skills (D) instruction with optimal language instruction (C).

Initially we hope to produce materials for grades K-3. Then we may add a preschool component or materials for later grades. The program will consist of two strands, each of which will occupy appx. 50 minutes a day.

1. The skills strand will build D using a systematic phonics approach, decodable text, and regular practice to build fluency. This part of the CK program will not differ dramatically from the decoding instruction in some of the better reading programs currently on the market.
2. The language strand will build C by reading aloud to students twice a day, 150 days a year. The read-alouds will be grouped into coherent domains to increase learning, and the program will include non-fictional as well as fictional readings. Virtually all of the topics covered will connect to the Core Knowledge Sequence in one way or another. For instance, a domain on “kings and queens” in kindergarten might contain selections from the Core Knowledge language arts standards, e.g. “King Midas,” “Cinderella,” “Old King Cole,” “Sing a Song of Sixpence,” and “Snow White.” A grade 1 domain on colonial America might include a series of biographies of important Americans named in the history section of the Sequence, e.g. George Washington, Ben Franklin, and Paul Revere. A grade 2 domain on ancient Greece might look at Greek mythology and also contain a simple but lively narrative of  the Persian Wars.

We’re going to need a lot of money and a few years to build the program. There’s no way the materials will come on the market prior to Fall of 2007 . . . which raises the question, what should teachers do in the meantime?

VIII. What Not to Do in the Meantime

Before commenting on what you should do in the meantime, I want to comment on one thing you should not do in the meantime.

In many schools these days history, science, and the arts are in danger of being pushed out of the early elementary curriculum because of a perceived need to focus on reading and math above all else.   Principals and administrators know that the federal government is requiring schools to make “adequate yearly progress” in reading and math. They want to ensure that their schools make adequate yearly progress, so they instruct teachers in the early grades to focus more on math and reading (which usually means the decoding or “skills” aspects of reading, or teaching “comprehension skills”), and they set aside larger and larger periods of time for instruction in these areas, even if this means neglecting science, history, and the arts. The justification given is that these other subjects “can wait” until reading skills are solidified.

If you’ve been following the argument in this presentation, you should be able to see what is wrong with this strategy. It is based on a formalist understanding of reading.  Remember that R = D x C. Reading comprehension requires not only good decoding skills but also stron language abilitys and a broad base of knowledge — and students who have not been exposed to history, science, and the arts in the early grades are not likely to have the knowledge and vocabulary they need to read and understand the texts they will encounter in the later grades.    Although the formalist strategy is proposed by many well-meaning people who hope to raise reading scores, it is likely in the long run to have the opposite effect.  Reading depends not only on decoding but also on language, vocabulary, and knowledge of key domains.  Therefore schools that neglect history and science in the early grades are in fact increasing the likelihood that students will encounter reading problems in grade 4 and beyond.   

If administrators at your school put forward formalist ideas of this sort — if they say that sections of the curriculum must be set aside or de-emphasized to allow more time to be spent on math and the decoding aspects of reading — you should sound the alarm. Remind them that reading comprehension depends on knowledge as well as decoding skills.

IX. What to Do in the Meantime

What can you do until the Core Knowledge Reading Program is developed? For starters, keep the simple view of reading in mind. Make sure that your school’s curriculum is addressing both D and C.

To address D, choose a reading program with a strong, systematic phonics component. Make sure students develop phonemic awareness. Give them plentiful practice to build fluency. Follow the recommendations in the books I mentioned under point VI.

To address C, make sure your school continues to teach history, geography, science, and the arts. Make reading aloud a part of your everyday curriculum. Look to read twice a day, 150 days a year. Choose read-alouds with rich language. Don’t be afraid to read books above the child’s reading level: remember that children can listen to texts they couldn’t read on their own. Try to group read-alouds into coherent domains. Include plenty of nonfiction.

If you teach at a Core Knowledge school, you already have a tremendous advantage when it comes to building C, because you have a curriculum which lays out specific subjects students should study. All you need to do is look for subjects in the Sequence that lend themselves to a read-aloud exposition, and then for books that you can share during read-aloud time.

Here are a few guidelines that will be useful to you as you select books and construct domains for reading aloud

1. Begin with short books. Especially in K and 1, it is important to begin with short books and readings, such as Aesop’s Fables, nursery rhymes, and simple picture books. For the first few weeks of the school year your readings should probably not last much longer than 5-6 minutes. Once children have become familiar with the listening experience, you can read books and stories that last 10-15 minutes.
2. Move quickly to strong verbal content.   For your first few readings, especially in the earliest grades, it is a good idea to select books that are generously illustrated — books in which the pictures help “tell the story.” This is helpful to students who are just learning to sit still and listen to stories read aloud. The pictures offer them a second chance to infer what is happening. After the first few weeks, however, you should move away from books which rely heavily on illustrations to tell the story and choose books in which the story is conveyed primarily by words, and the pictures are supplementary and/or occasional. Remember that the goal of all this reading aloud is to help children develop the ability to understand strings of words, without looking at the pictures.    
3. Include discussion and introductions. As noted above, stories should be preceded by, followed by — and in some cases interrupted by — discussion.
4. Include plenty of nonfiction. A read-aloud program of the sort we are describing can realize its full potential only if it moves beyond the kind of contemporary fiction found in most basal series to include poetry and significant quantities of various kinds of nonfiction, such as biography, historical narrative, and informational text.
5. Remember that reading aloud is great vocabulary work. Although there is a tendency to think of reading aloud and “vocabulary work” as separate activities, research indicates that children can and frequently do learn words to a testable level simply from hearing the words in a story read aloud — particularly if the word appears more than once in the story, or several times across a variety of stories. This is called “learning from context,” or “incidental vocabulary learning,” and it is probably our main channel for vocabulary acquisition. Therefore, reading aloud is itself a kind of vocabulary work.
6. Allow for some repetition.  Children like to hear favorite stories more than once, and re-reading gives them additional opportunities to acquire knowledge of the words in the story.   However, if you read every story twice, you cut in half the amount of material you can cover during a school year. The key is to strike a balance between the desire for repetition and the need to introduce lots of new material. We suggest that you read 3-8 new stories (or sections) and then pause for a repeated read-aloud. In some cases you may wish to choose the story to be repeated. In other cases you might allow students to choose among two or three stories they have heard by voting. This will encourage them to discuss the stories and will give them a feeling that they are participating in the classroom experience. It also provides an excellent way of modeling democratic behavior.

Notes on Constructing Domains:

Although constructing a domain may sound like a great deal of work, in some cases it is no more complicated than selecting a single appropriate book. Books like Winnie-the-Pooh, The Wind in the Willows, and Narrative of the Life of Frederick Douglass can be read aloud chapter by chapter, so as to constitute a domain. The early chapters in the book set the stage for the later chapters and give students the domain knowledge they need to listen and learn more effectively as the readings progress.

Collections and anthologies can also be utilized as ready-made domains. For example, a domain on heroes might be based on selected readings from Fifty American Heroes Every Kid Should Meet, by Dennis Denenberg and Lorraine Roscoe, or William Bennett’s Children’s Book of Heroes, or selections from both books. Likewise, a domain on Greek Mythology can be assembled from William Russell’s Classic Myths to Read Aloud and/or D’Aulaire’s Greek Myths, or from What Your Second Grader Needs to Know.

In other cases, the construction of a domain may be slightly more complicated and may involve locating more than one book and sequencing the books according to some logical principle (e.g. easiest books first, chronological order, etc.).

Here are some suggestions that will help you find the books you need to build domains.

• Ask librarians to recommend books for children on a particular topic.
• Ask other teachers to share titles of their favorite books on various subjects.
• Consult guidebooks on children’s literature such as Books to Build On, by E. D. Hirsch and John Holdren, The New York Times Parent’s Guide to the Best Books for Children,  by Eden Ross Lipson, etc.
• Utilize Amazon.com or other online bookstores with search features. Amazon’s “advanced search” feature allows you to search for books on a particular subject like “Insects” that are suitable for children of various age levels. Reviews and customer comments help you assess the quality of the books.
• Consult online booklists and databases, including “Resources to Build On,” developed by the Core Knowledge Foundation, the “Children’s Literature Web Guide” at the University of Calgary, the “Children's Picture Book Database”
  at Miami University, and “The Baldwin Project” online. All of these databases can be found using a search engine like Google.com. Simply type in the phrases in quotations above.
• Don’t forget about audiobooks, either. A large range of spoken word recordings are available from suppliers like Great Hall Productions, Recorded Books, Blackstone Audiobooks, Books on Tape. Again, use Google to locate these vendors.

If you missed the presentation at the conference, feel free to email your response to Matthew Davis.

Last updated: Fri, May 23 2008