Sunday, November 20, 2011

Junk Food is Bad For You, but Junk Reading is Good for You, Stephen Krashen and Joanne Ujiie

Junk Food is Bad For You, but Junk Reading is Good for You
Stephen Krashen and Joanne Ujiie  

The language teaching profession has assumed that students should be restricted to “quality” literature: Advanced, and often intermediate students are required to read the classics, and are rarely introduced to bestsellers, series novels, magazines, or comic books. 

The goal of reading the classics is a worthy goal: The study of great literature is the study of philosophy, covering both “ethics” (how are we supposed to live?) and “metaphysics” (what are we doing here?) in a way that it often difficult to do otherwise. We thus do not disagree at all with the goals of a language program that aims at literature. We disagree with the means used to reach that goal.

In this paper we suggest that an early diet of classical and “quality” literature may be the wrong way to facilitate the eventual reading of classical and quality literature, that encouraging light reading for intermediate students can create the background knowledge, linguistic competence, and desire to read more “serious” literature.

The evidence comes from several sources: We first present studies suggesting that light reading has a positive impact on language and literacy development. We then examine evidence that shows that light reading can serve as a conduit to heavier reading; those who read light literature do not typically remain on this diet, but go on to “heavier” reading.  In the final section, we discuss the question of just what “quality literature” is.

Light Reading Promotes Literacy Development

If comics prevent literacy development, as some people fear, we would expect more comic book reading to result in lower literacy scores. This is not what we find.  Elley (1994) investigated the relationship between comic book reading and reading achievement by 9 and 10 year olds in 27 countries (Elley, 1994, table 3.1, page 67).  Children were asked two questions about comic book reading: How often they read comics per week, on a scale of zero to 6 (the average for all countries was 1.87), and whether they read comics for fun the previous week. The measures were positively correlated (r = .78) but did not give identical results.

Elley reported a correlation of r = .36 between reading proficiency and the percentage of children in each country who read comics the week before.  (The correlation between reading proficiency and the percentage of children who said they read a book the week before was lower, r = .26).  The correlation of comic reading with reading proficiency using the other measure, number of times comics were read during the week, was also positive, r = .24).

Slightly different results were obtained by considering gender. Elley reported, as have others, that boys were more avid comic book readers than girls. We correlated the relationship between comic book reading and reading achievement for boys and girls separately (see tables 3.4, p. 72, in Elley, 1994, and table 4.7, page 105 in Purves and Elley, 1994). For boys, the correlation between the amount of weekly comic book reading and reading proficiency was r = .33, for girls it was r = .13 (Elley did not present separate data for boys and girls for the percentage who read comics last week.)  Comic book reading thus seems to be a better predictor of reading proficiency for boys than for girls.

These correlations range from small to modest, but they confirm that comics can have a positive effect and they are certainly counter to view that comics are harmful.

Teen romances

There is evidence that teen romances can have a positive impact on adult second language acquisition. Kyung-Sook Cho (Cho and Krashen, 1994, 1995a, 1995b) worked with a group of women in their 30’s, who, despite years of formal (grammar-based) EFL study in Korea and considerable residence in the United States, had made little progress in English.  Introduced to the Sweet Valley series, her subjects began with the Sweet Valley Kids series (written for 7 year olds), progressed through Sweet Valley Twins (for readers 8 to 12), and Sweet Valley High (teen-agers), and eventually moved on to adult Harlequins, making substantial gains in vocabulary knowledge.


Rucker (1982) provided junior high school students with two free magazine subscriptions relating to their personal interests for periods of a year and a year and a half. Those who received the magazines made superior gains on standardized tests of reading (but not on a test of “language,” i.e. mechanics and spelling).  A reasonable interpretation of these results is that the magazines themselves served as valuable input and that they stimulated even more reading. As Rucker points out, magazines are the most “reader interest specific” of all mass media and “may thus consequently be the most valuable as stimuli to reading” (p. 33).

Light Reading as a Conduit

Many people are fearful that if children engage in “light reading,” if they read comics and magazines they will stay with this kind of reading forever, that they will never go on to more “serious” reading.  The opposite appears to be the case. The evidence suggests that light reading provides the competence and motivation to continue reading and to read more demanding texts.

Comic books

Krashen (2004) presented case histories of individuals, some very prominent who give comics the credit being a conduit to literacy. Bishop Desmond Tutu described his father as “very patriarchal,” but tells us that “One of the things I am most grateful to him for is that, contrary to educational principles, he allowed me to read comics. I think that is how I developed my love for English and for reading.”  Jim Trelease (2001) points out that anybody concerned about a possible connection between comic book reading and juvenile delinquency should consider Bishop TuTu’s experience.

More and more cases like this are coming to light: Children’s book writer Jack Gantos noted, in an article published in USA today (“Teachers are getting graphic, May 3, 2005) that Jean-Paul Sartre “started off reading comic books as a child and that if it wasn’t for comic books, he never would have stuck with books.”  And in a letter to the editor in response to the USA Today article, children’s book author Tina McElroy Ansa relates that between the ages of 7 and 11, she “spent the afternoons and summer days immersed in the world of comics, from Lulu and Tubby to Superman, from Little Lotta to Archie and Jughead” and tells us that she knows that “reading comics encourages creativity, imagination, curiosity, more reading – and sometimes writing” (Ansa, 2005).

An empirical study

Ujiie and Krashen (1996) asked seventh grade boys about their comic book reading, overall reading, book reading and attitude toward reading. Table 1 shows that those who reported more comic book reading also reported for pleasure reading in general. The results were similar for middle class children and for those who came from low-income families.

Table One: How often do you read for pleasure?
heavy comic reader
54% (19)
34% (12)
11% (4)
occasional reader
40% (32)
28% (23)
32% (26)
non comic reader
16% (4)
20% (5)
64% (16)

middle class

heavy comic reader
65% (17)
27% (7)
8% (2)
occasional reader
35% (31)
35% (31)
30% (27)
non comic reader
33% (8)
17% (4)
50% (12)

From: Ujiie and Krashen, 1996

Similar results were reported for book reading, and for attitude toward reading, with more comic book reading associated with greater enjoyment of reading.  What is especially interesting is that although the middle class boys tend to read more in general, undoubtedly related to the fact that they have far more access to books (Neuman and Celano, 1999), heavy comic book readers from low-income families reported more overall reading than the occasional and non-comic book reading middle class boys.

An intervention using comics

Dorrell and Carroll (1981) demonstrated that comic books can be used to stimulate additional reading. They placed comic books in a junior high school library, but did not allow them to circulate; students had to come to the library to read the comics. Dorrell and Carroll then compared the circulation of non-comic book material and total library use during the 74 days comics were in the library, and the 57 days before they were available. The presence of comics resulted in a dramatic 82 percent increase in voluntary library use, from about 273 visits per day to nearly 500, and a 30 percent increase in circulation of non-comic material, from about 77 volumes per day to just over 100.

What do children choose on their own?

Reading professionals take prizewinning books very seriously. Winners of annual awards, such as the Newbery or Coldecott, are announced in reading journals and newsletters, and the books are often put on display at libraries.

Several studies tell us that young readers do not have a strong interest in reading these books and generally ignore what critics regard as “good literature.”  We examine three recent studies showing this.

Lamme (1976) examined reading records of 65 middle school children (grades 4-6) over a three year period. She reported that the children “read few Caldecott or Newbury medal winning books and few books on a standard list of good literature ….  Only in the sixth grade was even 5 percent of their reading in medal winning books …. It appears that when these children freely select books, titles considered to be “good” do not comprise a large portion of the selections …” (p. 24).  Of great interest to us, Lamme found no correlation between what children read and their reading test scores: Those who selected “quality” books did not read any better.

Nilson, Peterson and Searfoss (1980) assembled a list of books “highly acclaimed by critics” (p. 530) from the years 1951 to 1975, books that were on various lists of “quality literature” as determined by adults (eg. the list of the Best Books of the Year complied by the School Library Journal, winners of the Newbery and Caldecott awards). Added to this list were books were selected by a librarian.

Children’s preferences were determined by ten children’s librarians who were asked to rate the popularity of each book, judging each as a “popular” (book checked out regularly, given two points) or “unpopular” (“I can hardly remember the book.” zero points.)  Nilsen et. al. then assembled lists for each year, from 1951 to 1975, containing books published that year along with rankings based on popularity scores. 

We present below one of their lists (table 2), containing books that were rated as popular with children, published in 1970.  Following each book is the “popularity rating,” from the most popular to the least.  Note that the “acclaimed” books are closer to the bottom of the list.

Table Two: Books popular with children, published in 1970
1.Are you there God, it’s me, Margaret (Blume). Score = 20.
2.Runaway Ralph (Cleary). Score = 19.
3.A Bargain for Frances (Hoban). Score = 17
4.Animals Should Definitely Not Wear Clothing (Barrett). Score = 17
5.* Frog and Toad are Friends (Lobel). Score = 16.
6.The Snake that Sneezed (Leydenfrost). Score = 16.
7.Summer of the Swans (Byars). Score = 9
8.* The Trumpet of the Swan (White). Score = 8.
9.* In the Night Kitchen (Sendak). Score = 7.
10.* Sing Down the Moon (O’Dell). Score = 3
11.* The Marvelous Misadventures of Sebastian (Alexander). Score = 2.
12.* Knee Knock Rise (Babbitt). Score = 1
= acclaimed book
from: Nilsen et. al., 1980.

Ujiie and Krashen (2005a) performed a “secondary analysis” of this data using statistical tests and confirmed that Nilsen et. al. were correct: The average rank of the “popular books” (those on the popularity lists but not “acclaimed” by adults) for each year from 1951 to 1975 was higher than the average rank of “acclaimed” books for each year except 1962, that is, for 24 years out of 25. And the difference in the one exceptional year was small (table 3).

Table Three: Mean Rank for Popular and Acclaimed Books

Range = 0 to 20

Application of a statistical test (sign test) told us that the difference in ranks was statistically significant. This test controlled for the year of publication. We also did a t-test comparing ranks for all popular and all acclaimed books for all years combined (table 4).

Table Four: Mean popularity scores


The mean popularity scores were of course significantly different (t = 110.7, df = 260, far beyond the .0001 level of significance, confirming the results of the sign test, and confirming Nilsen et al’s claim that adult judgments of quality differ from children’s tastes. 

Ujiie and Krashen (2005b) examined children’s actual behavior, probing to what extent acclaimed books are taken out of public libraries.

Our list of “acclaimed” books consisted of winners of the Caldecott and Newbery Awards of 2003 and 2004 as well as the runner-ups, known as “honor books.” Interestingly, there was no overlap between the lists.

A list of popular books was obtained from, which provided records of bestsellers from bookstores.  Three lists of the top 15 bestsellers were consulted for use in this study: Bestsellers for the month ending January 9, 2004, May 27, 2004, and December 16, 2004. We found that very few, and sometimes no award winners were on any of the bestseller lists.

For each of the bestsellers on the January and May lists, and for each of the prizewinner books, circulation and inventory data was gathered from six Southern California library systems consisting in total of 127 separate libraries.

Table 5 presents the mean number of bestsellers and prizewinning books checked out from the six public library systems combined. The results from the January and May bestseller lists were nearly identical. Far more bestsellers were checked out than prizewinners. On the average, about 200 copies of the bestsellers were taken out, but only about 35 copies of the average prizewinner were checked out from all six library systems.

Table Five: Mean number of books checked out

total taken out
bestsellers, 1/04, n = 15
bestsellers, 5/04, n = 15
Caldecott winners, n = 8
Newbery winners, n = 8

sd = standard deviation

Why are some books more popular?  The answer is clearly not readability. The mean prizewinner readability (Flesch-Kincaid Readability Formula), in fact, was actually lower than the readability level of popular books.

A possible implication of these results is that children don’t know what is best for them. Another is that Newbery and Caldecott judges have different standards than the real audience of children’s and adolescent literature.

In a third study aimed at revealing children’s preferences, Ujiie and Krashen (2002) asked fourth and fifth graders if they had ever had a “home run” book experience, a reading experience that got them interested in reading. 1 All 266 children attended a school in which 74% were considered low income and received free or reduced price lunch. All were native speakers of English or considered fluent in English. The question asked was simple:  Was there one book or experience that first interested you in reading? If the answer was “yes” we asked the children to give the title of the book or tell us about the experience.  In agreement with previous studies (Von Sprecken, Kim and Krashen,, 2000; Kim and Krashen, 2000),  most children (82%) identified a "home run" book.

As in other studies, children named a wide variety of home run books. Very few titles were selected by more than a handful of students. The champion home run book was Harry Potter (19), followed by Goosebumps (11), the Three Little Pigs (11), Dr. Seuss (6), Animorphs (5), Scary Stories (5) and Winnie the Pooh (5). What is of interest here is that none of these home run books ever won a Newbery, Caldecott or BlueBonnet (Texas) award. In fact, three of the children’s home run books (Harry Potter, Goosebumps, and Scary Stories) were on the list of the 100 most challenged books of 1990-1999 (see

These consistent results suggest that if push “literature” we will be fighting against readers’ natural tendencies, but if we facilitate light reading, we will be encouraging a tendency that is already there. 2  Unfortunately, for many potential readers, what they like to read is not easily available. 3

This paper has attempted to make the following points:

(1)Light reading promotes literacy in general
(2)Light reading leads to heavier reading, that is, light reading serves as a conduit for heavier reading
(3)Young readers tend to ignore books that adults think are “quality” literature.

Second language and foreign language education has made no serious attempts to encourage light reading. This is probably due to several barriers. One is a lack of access to such reading material. Another is that there is no obvious means of paying for them, other than from the teacher’s own pocket.  Still another barrier is the lack of an obvious mechanism to fit light reading into current programs.

We suggest establishing a firm place for light reading in the curriculum, This place, once established can also justify funds for the purchase of light reading material.  A sheltered popular literature class, to be taken after the beginning level but before the “serious study of literature” might be the place for light reading.

In a sheltered popular literature class, foreign and second language students would be introduced to “ordinary” and popular reading material (Krashen, 1998), presented as “literature,” that is, as a means of discussing philosophical issues as well as gaining a deeper familiarity with other cultures. A sheltered popular literature class will also familiarize students with what kinds of light reading are available, and will, we hope, encourage the establishment of a light reading habit, one that will continue after the class ends.

 We should point out, however, that while we predict progress from “light reading” and a transition to heavier reading, there is no guarantee that all readers will go on to what some people define as “quality” literature. Research, including our own, shows that officially designated “quality” literature is rarely popular: Award-winning books do not usually make bestseller lists. Nell (1988), in fact, has questioned the basis for adult judgments of literacy merit, reporting that judgments of literacy merit were positively correlated with judgments of passage difficulty or a measure of complexity (the “Fog index”). His conclusion was that for the judges in his study, “the best medicine tastes the worst” (p. 160).  The result held for several different groups, including librarians, university students and university teachers. 

We predict, however, that readers will arrive at books that are right for them, that they find interesting and that meet their needs.

Post-script: Series books

What struck us in examining lists of books that children truly like to read (bestseller lists, the homerun book list) is that a large percentage of books were “series” books, that is, books that were part of a continuing series with identical characters and a continuing storyline. On the January 9, 2004 bestseller list, 11 out of 15 were series books or at least part of a trilogy (Artemis Fowl, Harry Potter, Unfortunate Events, Captain Underpants, Sisterhood of the Traveling Pants, Amulet of Samarkand). On the May 27, 2004 list, five of the 15 were series books.  Note also that the 1970 bestseller list (table 1) contains a number of books by very popular authors (Judy Blume, Beverly Clearey, EB White).

Our evidence confirms the results of previous studies showing the value of series books:  As noted earlier, Cho (Cho and Krashen, 1994, 1995a, 1995b) reported great success with adult second language acquirers of English using the Sweet Valley Series, and Cho, Ahn and Krashen (2005) reported increased enthusiasm in English for fourth grade EFL students in Korea after reading books from the Clifford series.

Series books have obvious advantages, thanks to the familiar background knowledge, setting, characters, and the style of the writer. Series books are thus a form of “narrow reading.” Krashen (1981) has argued that narrow reading, reading focused on one topic, and/or by one author, 4 is very good for language acquisition, because texts have a good chance of being interesting and comprehensible. 

An obvious question that can be raised about narrow reading is whether it will allow students to develop the kind of competence they need to read several different kinds of academic texts. Is narrow reading, and light reading in general, enough? An interesting hypothesis is that enough reading in any genre will suffice to prepare a reader for demanding academic reading and for “serious” literature: Although there are clearly different styles of prose, there is also considerable overlap among styles (Biber, 1986): So-called narrative style has, for example, some, but not all of the characteristics of formal, expository prose. Thus, reading novels will not provide the reader with the ability to read all academic prose, but it will provide the reader with at least some of the features of this style, which will make reading academic prose more more comprehensible. Someone who has read 100 Goosebumps and Fear Street novels have a much easier time with a history text than someone who has not.


1. Trelease (2001) introduced the concept of a "home run" book, a reading experience that readers claim stimulated their initial interest in reading. The idea of a home run book comes from an observation made by Clifton Fadiman: "One's first book, kiss, home run, is always the best" (Trelease, 2001, p. 136).

2. For evidence that pushing the classics too early can result in potential readers losing the taste for reading, see Carlsen and Sherrill, 1988.

3. Worthy, Moorman, and Turner (1999) reported that school and classroom libraries typically did not carry much of what the children said they liked to read (comics and scary stories).  This is an especially serious problem for students from low-income families who often have no other source of reading material. Worthy et. al. reported that “teachers who had such materials usually used their own money to buy them or asked students to donate their used books” (p. 23).

4. Lamme (1974, 1976) reported a positive correlation between reading achievement and reading books by “known authors.”


Ansa, Tina McElroy. 2005. Comics in class can provide rich, creative education. USA Today, May 9, 2005, Letter to the editor. Page 12A.

Biber, D. 1986. Spoken and written textual dimensions in English. Language 62: 384-414.

Carlsen, G. R., and Sherrill, A. 1988. Voices of Readers: How We Come to Love Books. Urbana, IL: NCTE.

Cho, K.S., and S. Krashen.1994. Acquisition of vocabulary from the Sweet Valley High Kids series: Adult ESL acquisition. Journal of Reading 37, 662-667.

Cho, K.S., and S. Krashen 1995a. From Sweet Valley Kids to Harlequins in one year. California English 1 (1): 18-19.

Cho, K.S., and S. Krashen. 1995b. Becoming a dragon: Progress in English as a second language through narrow free voluntary reading. California Reader 29: 9-10.

Cho, KS., Ahn, KO, and Krashen, S. 2005. The effects of narrow reading of authentic texts on interest and reading ability in English as a foreign language. Reading Improvement 42,1: 58-63.

Dorrell, I. and E. Carroll, 1981. Spider-Man at the library. School Library Journal 27: 17-19.

Elley, W. 1994. Voluntary reading activities. In W. Elley (Ed.) The IEA Study of Reading Literacy: Achievement and Instruction in Thirty-Two School Systems. New York: Elsevier Science. pp. 65-88.

Kim, J. and Krashen, S. (2000). Another home run.  California English, 6(2): 25

Krashen, S. 1981. The case for narrow reading. TESOL Newsletter 15:23

Krashen, S. 1998. Foreign Language Education: The Easy Way. Taipei: Crane Publishing Company

Krashen, S. and Von Sprecken, D. 2002. Is there a decline in the reading romance? Knowledge Quest, 30,3, 11-17.

Lamme, L., 1974. Authors popular among fifth graders. Elementary English 51: 1008-1009.

Lamme, L. 1976. Are reading habits and abilities related? Reading Teacher 30: 21-27.

Nell,, V. 1988. Lost in a Book. New Haven: Yale University Press.

Neuman, S., and D. Celano. 2001. Access to print in low-income and middle-income communities. Reading Research Quarterly 36(1): 8-26.

Nilson, H., Peterson, R., and Searfoss, L. 1980. The adult as critic vs. the child as reader. Language Arts. 57(5): 530-539.

Purves, A. and Elley, W. 1994. The role of the home and student differences. Elley, W. 1994. Voluntary reading activities. In W. Elley (Ed.) The IEA Study of Reading Literacy: Achievement and Instruction in Thirty-Two School Systems. New York: Elsevier Science. pp. 89-122.

Rucker, B. 1982. Magazines and teenage reading skills: Two controlled field experiments. Journalism Quarterly   59: 28-33.

Trelease, J. 2001. The read-aloud handbook. New York: Penguin.Fourth edition.

Ujiie, J. and Krashen, S. 2002. Home Run Books and Reading Enjoyment, Knowledge Quest 31(1): 36-37, 2002

Ujiie, J. and Krashen, S. 2005. Is “Acclaimed” Children’s Literature Popular Among Children?: A Secondary Analysis of Nilson, Peterson, and Searfoss (1980). Knowledge Quest 34(1)

Ujiie, J. and Krashen, S. 2005. Are Prizewinning Books Popular Among Children?: An Analysis of Public Library Circulation. Knowledge Quest. 34 (3)

Von Sprecken, D., Kim, J. and Krashen, S. 2000. The home run book: Can one positive reading experience create a reader? California School Library Journal, 23 (2): 8-9.

Worthy, J., M. Moorman, and M. Turner. 1999. What Johnny likes to read is hard to find in school. Reading Research Quarterly 34(10): 12-27.