Stephen Krashen argues that self-selected pleasure reading of fiction is key to complex text comprehension for first- and second-language learners

I focus on one aspect of the fiction-nonfiction debate here: self-selected pleasure reading of fiction in which there is little or no accountability. Readers are not tested on what they have read, do not have to write book reports, do not have to finish every book they start, and can select, within reason, what they want to read. In addition to books, they can read newspapers, magazines, manga, comic books, and articles from the internet. Also, self-selected pleasure reading can be, and usually is, narrow. Good readers typically focus on one author or genre at a time.

There is little support in American schools for this kind of reading. The reasons for not encouraging self-selected fiction reading include the following:

1. The language of fiction is not “academic,” not the kind of language students need to succeed in school or in the workplace.

2. Fiction does not provide the kind of knowledge students need to succeed in school or in the workplace.

3. Self-selected fiction does not stretch or challenge the mind. It does not develop the habits of thought needed for school and career success.

4. If students are allowed to select their own reading, they will stick with easy books, and not progress to harder material.

I present here a proposal for the development of academic, or specialized reading competence, and the role of self-selected pleasure reading, followed by a case history. I then respond to the concerns listed above.

The proposal presented here is based on the comprehension hypothesis, the idea that we acquire language and develop literacy when we understand spoken and written messages—that is, when we receive comprehensible input (Krashen, 2003, 2004a).

Comprehensible input is most potent when it is highly interesting, or when it is “compelling,” so interesting that only the story or message exists for us: we are in a state of “flow” in which our sense of time and even our sense of self is diminished (Csikszentmihalyi, 1990; Krashen, 2015).

Academic Language
Academic language, or more properly “language for special purposes,” is the ability to use language for more than everyday functions. We develop academic language, it is proposed, in three stages. Each of the stages is a result of receiving compelling comprehensible input.

Stage 1: Hearing stories and books read aloud. Stories and read-alouds provide knowledge of vocabulary, grammar, and how texts are constructed, which makes the eventual reading of texts more comprehensible. Stories and read-alouds also stimulate an interest in reading (Trelease, 2013).

Stage 2: Free voluntary reading. Free voluntary reading (Krashen, 2011a) is self-selected reading, generally fiction, of material of great interest to the reader. This reading does not bring the reader to the highest levels of literacy development, but it provides the competence and knowledge that makes reading at the next stage more comprehensible. It is the bridge between conversational knowledge of language and academic knowledge. This idea is confirmed by data from Biber (1988), who analyzed texts in terms of linguistic complexity, and reported that fiction fell about midway between conversation and academic texts (abstracts of technical journal papers).

Stage 3: Specialized reading. Specialized reading refers to the reading of specialized texts in an area of great interest to the reader.

I provide a case history illustrating the three stages. The case history comes from English as first language, but my claim is that the three stages hypothesis is valid for both first- and second-language development. The important points I wish to make in providing this case history are these:

1. The reading done in stages two and three is compelling. It is also self-selected and narrow.

2. The stages are claimed to be universal, but within each stage there is individual variation.

I present my own progress as a reader, suggesting that my journey is similar to the stages others have gone through.

Stage 1 
I cannot present a heroic case of overcoming great odds and obstacles. I grew up in a middle-class home in Chicago, Illinois, in a very supportive and very functional middle-class family that fully recognized the importance of stories and reading. We had a rich supply of books in our home, and both of my parents and my older sister were readers. In addition to being read to at home, my sister introduced me to radio stories: we eagerly listened to The Long Ranger, Captain Midnight, The Cisco Kid, The Green Hornet, Sky King, and The Cinnamon Bear during the Christmas season.

I attended a well-funded middle-class school with an excellent school library, and with teachers who supported reading and storytelling.

Thanks to this background, “learning to read” was effortless. Reading instruction was “look-say,” and the beginning stories we read in class were mundane, but instruction was simply a test that I passed, thanks to the rich experiences with stories I had had earlier and my familiarity with the alphabet (“Basic Phonics,” see Krashen, 2004b). Reading class included Round Robin reading, in which each child read aloud in turn. I, of course, was never ready for my turn, because I had been reading ahead. But I was not an especially good reader compared to the others. In the third grade, I was placed in the low reader group. The cure came in stage two.

Stage 2
Stage two for me was divided into three substages. (My claim is that stage two is universal, but the details of the substages are not.) It was my father who put me on this path. He was not pleased that I had been placed in the low reader group, and he knew what to do about it. Years ahead of his time, he brought home comic books, and encouraged me to read as many as I liked. I loved the comics, and the effect was nearly immediate. In a short time, I moved up to the intermediate reading group, and within a few months I was in the most advanced group. My dad continued to fully support my comic book habit for the next few years. I was the envy of my friends.

My comic book reading was nearly entirely the superhero type, largely DC’s Superman and Batman, and the rival Captain Marvel, with some peripheral reading of Superboy, Captain Marvel Jr., and Mary Marvel. This occurred during the “golden age” of comics, in the late 1940s and early 1950s, when nearly 90% of American children my age were comic book readers (Krashen, 2004a).

In my view, comic books and graphic novels today are far better than those available during the “golden age,” thanks to Stan Lee and the Marvel comic book company. Marvel introduced superheroes far more interesting than the two-dimensional ones of the 1940s and 1950s. Marvel’s superheroes have problems and face hard ethical decisions. Nevertheless, the comics of the ’40s and ’50s were much more interesting than anything else available to me at the time.

The second substage of stage two consisted of sports fiction. From ages ten to about 14, I devoured sports novels, mostly baseball stories, especially those authored by John R. Tunis, who chronicled the struggles of a mythical Brooklyn Dodgers team over a decade. The excitement was the game itself, of course, but also the players’ personalities and the problems they faced in their daily lives.

Here is one example: In the last book of Tunis’ series, World Series, the final chapter is, of course, the last game of the World Series. Each team has won three games, and the team that wins the last game is the champion. It is the last inning, and the fate of the entire series depends on the next play. The pitcher for one team is the father, the opposing batter is his son, and they haven’t spoken to each other for 15 years because of a long-standing family conflict. I won’t spoil the ending for you—used copies of the book are still available. As an adult, I reread the first book in Tunis’ baseball series, The Kid from Tomkinsville, and it still had all the excitement and drama it did when I read it when I was twelve years old.

The third substage of stage two was science fiction. My reading remained narrow; I read nearly exclusively the work of a few authors: Isaac Asimov, Ray Bradbury, Robert Heinlein, and Arthur C. Clarke. These authors were very prolific, and I read nearly everything they wrote.

In middle school and high school, the comics and books I read were my true language arts curriculum.

In middle school and high school, the comics and books I read were my true language arts curriculum. I read the novels assigned for language arts class and passed the tests, but I don’t remember any of the books. The content of the assigned textbooks I read in subject-matter classes is also long forgotten. But I remember nearly all of the books I read on my own during this time.

Stage 3
The assigned reading I did as an undergraduate made little impression on me, and my voluntary reading had nothing to do with the subjects I took—it was still the kind of science fiction I had enjoyed in high school. I had not yet found my real interests.

All this changed in graduate school. I discovered something close to my real interests—linguistics, an important and fascinating step on the route to my eventual specialization in language and literacy development. My second course in linguistics was Syntactic Theory, and the text was Aspects of the Theory of Syntax, by Noam Chomsky. At first, it was completely incomprehensible to me. It could have been written in Bulgarian. But I found a strategy that worked: I read Chomsky’s first publication, Syntactic Structures, which I found very comprehensible, as it was written for a readership with no background in his approach. I then read everything Chomsky had written since then in chronological order, up to Aspects of the Theory of Syntax. When I tried to read Aspects again, it was completely comprehensible.

Reading in chronological order, in the order in which the author wrote, made the texts far more comprehensible, and turned the reading into a kind of story, a narrative. This approach enabled me to see how grammatical theory had progressed and how Chomsky dealt with problems in the theory. It was as fascinating to me as John R. Tunis’ baseball stories. Unknown to me at the time was that in doing this I learned a great deal about how scientists think, and, relevant to this discussion, I acquired a substantial amount of the academic writing style.

Toward the end of my graduate career, my academic interests changed to brain and language, and my colleagues and I began to do experiments using a technique called dichotic listening, a method of determining which side of the brain is in use in listening to stimuli. I read the research literature on dichotic listening using the same method I had used when reading Chomsky: begin with earlier studies, working toward the present, and read narrowly. In this case, the central researcher was Doreen Kimura of the University of Western Ontario. Again, I read her papers, and those of her colleagues, in strict chronological order, and again I found the reading to be fascinating. Without knowing it, however, I learned the essentials of experimental design and applied statistics, and the careful and steady progress one can make through carrying out study after study, as well as the academic style of writing experimental reports.
Gaining academic linguistic proficiency was thus not the result of studying “English for academic purposes”: it was, rather, the result of self-selected, narrow reading for my own purposes.

Responding to the Concerns
We return now to the concerns listed at the beginning of this paper.

Concern 1: The language of fiction is not “academic,” not the kind of language students need to succeed in school or in the workplace.

I discuss here one argument that asserts that this concern is justified and present a re-analysis that comes to the opposite conclusion, followed by studies showing that Concern 1 is not supported by the research.

The Gardner argument
Gardner (2004) argued that fiction does not contain enough academic vocabulary to make the reading of academic texts comprehensible. Gardner analyzed narrative and expository texts written at the fifth-grade level for native speakers of English and reported on the number of “academic” words in each (sub-technical vocabulary; words such as academic, absorb, and abandon): only 7% of the academic words that appeared in the expository texts in his sample also appeared in the narrative texts at least ten times, the number of exposures considered to be frequent enough for acquisition to take place. Thus, he concluded, reading narrative does not prepare you to read expository texts.
Gardner regards this as a serious problem, but it is only a problem if we require that fiction read during a particular year will help readers understand a significant percentage of the words in academic texts that they encounter that same year.

But we don’t require this. The real issue is whether reading fiction puts readers in a position to start to understand academic texts, not only those which students have to read right away, but also those they will read eventually. This means that previous fiction reading counts, not just what is read in the current year. Gardner’s data actually shows that fiction does indeed do the job: there were 338 acquirable academic words in the narrative texts in Gardner’s sample. This is an impressive amount.

Gardner’s narrative sample contained one million words, about what the average middle-class English native-speaker fifth grader reads in one year. This suggests that a year of self-selected reading will result in the acquisition of about 338 academic words. That is a real contribution, whether or not these words also appear in the expository texts that the children might read that year.

Gardner (2008) presented a similar argument based on books written by a few authors or on a narrow range of themes. My analysis (Krashen, 2011b) showed that the narrow narrative texts Gardner analyzed would result in the acquisition of 783 words in one year, about double the figure estimated for academic words from the narrative texts in Gardner (2004), thus confirming the advantage of the narrow reading of fiction.

Thus, in both cases, Gardner’s attacks on fiction turn out to support fiction.

Studies of the impact of free voluntary reading
Studies of the impact of free voluntary reading consistently show positive results. We can interpret these results to be supportive of the role of fiction, because so much of our self-selected reading is fiction: Nell (1988) reported that fiction accounts for about three-quarters of what is borrowed from public libraries and 70% of mass-market paperbacks sold (p.19) (for recent confirming data, see According to the New York Times, all top ten young-adult bestsellers in August 2015 were fiction: 

I have reviewed the impact of free voluntary reading in previous publications. In brief, studies consistently show that those who read more show better development of reading ability, writing style, vocabulary, spelling, and grammar. Studies are of several kinds:

1. Sustained silent reading (SSR): One group receives traditional language arts or second-language instruction; the other has a similar treatment except that ten to 15 minutes of each period are set aside for self-selected reading, with no or very little accountability. SSR has been a consistent winner in these studies on tests of reading, vocabulary, and writing, as long as certain common-sense conditions are met (Krashen, 2004, 2011a, 2011c).

2. Case studies: A number of case studies of those who have engaged in self-selected reading have been published, and they consistently point to reading as the key factor in literacy growth and academic success (e.g., Krashen, 2004a; Krashen and Mason, in press).

3. Multivariate studies: These show that free voluntary reading and access to reading materials are consistent predictors of reading achievement, even after controlling for many other potential predictors. The most recent multivariate study of this kind is Sullivan and Brown (2014), who found that the amount of reading done at age 42 is a clear predictor of vocabulary test scores among native speakers of English in the UK, controlling for reading done earlier in life. They also reported that the reported frequency of reading high-quality fiction was a very strong predictor of vocabulary knowledge, and reading “middle brow” fiction was also a good predictor—slightly stronger, in fact, than reading nonfiction.

Clearly, there is a strong case that reading fiction does indeed build competence in “academic” language.

Concern 2: Fiction does not provide the kind of knowledge students need to succeed in school or in the workplace.

Self-selected reading has been shown to have a clear influence on knowledge of several different kinds. Ravitch and Finn (1987) found that American 17-year-olds who lived in a print-rich environment did better on tests of history and literature, and there was a clear relationship between the amount of self-selected leisure reading done and performance on the literature test. Stanovich and Cunningham (1992) used the same tests of history and literature and found that college students who read more did better, even when nonverbal ability factors were controlled.

West and Stanovich (1991) reported that those who reported more self-selected reading also did better on a test of cultural literacy, and those who had more exposure to print also did better on this test, even when factors such as SAT scores (West and Stanovich, 1991), age, education, exposure to television (West, Stanovich, and Mitchell, 1993), and nonverbal ability (Stanovich, West, and Harrison, 1995) were controlled.

The above results are probably no surprise to most readers of this article. But many people will be surprised to learn that Stanovich and Cunningham (1993) found similar results for a test of “practical knowledge,” as well as on a test of science and social studies.

Again, when we talk about self-selected voluntary reading, we are talking largely about fiction. These studies strongly suggest that fiction does indeed provide a considerable amount of the kind of knowledge students need to succeed in school and in the workplace.

Concern 3: Self-selected fiction does not stretch or challenge the mind. It does not develop the habits of thought needed for school and career success.

The results of several kinds of studies suggest that self-selected reading of fiction does indeed develop the habits of thought needed for school and career success. One crucial habit of thought needed for many kinds of career success is to understand others’ points of view and ways of thinking. A series of studies suggests that reading quality fiction develops an expanded “theory of mind,” defined as “the capacity to identify and understand others’ subjective states” (Kidd and Castano, 2013). In addition, the results of other research suggest that fiction readers also have more tolerance for vagueness—that is, they are better able to deal with uncertainty, which is important for problem solving (Djikic, Oatley, and Moldoveanu, 2013).

Additional case histories
Several case studies confirm that general reading is related to school and career success. The success could be due to superior literacy development, a superior knowledge base, and/or more developed habits of thought.

Simonton (1988) concluded that “omnivorous reading in childhood and adolescence correlates positively with ultimate adult success” (p. 11). Other reports speak to school success: several successful students who grew up in poverty and who nevertheless did well in school gave a great deal of credit to books and their reading habits, in one case hardly ever attending classes.

Geoffrey Canada, whose book Fist Stick Knife Gun (2010) describes dealing with poverty and violence while growing up in a high-poverty area of New York, managed to get access to books from a friend as well as his mother: “I loved reading, and my mother, who read voraciously too, allowed me to have her novels after she finished them. My strong reading background allowed me to have an easier time of it in most of my classes” (Canada, 1995).

Elizabeth Murray also grew up in poverty. Her source of books was her father, who would get library cards from as many local libraries as he could, take out books, and never return them. Their house was filled with fugitive library books from all over New York. This provided Ms. Murray with access to books, which, she tells us, allowed her to pass the yearly exams with minimum attendance: “Any formal education I received came from the few days I spent in attendance, mixed with knowledge I absorbed from random readings of my or Daddy’s ever-growing supply of unreturned library books. And as long as I still showed up steadily the last few weeks of classes to take the standardized tests, I kept squeaking by from grade to grade” (Shanahan, 2010).

These two cases are consistent with the results of a study suggesting that the positive impact of access to books via libraries can balance the negative impact of poverty on reading achievement (Krashen, Lee, and McQuillan, 2012).

Concern 4: If students are allowed to select their own reading, they will stick with easy books and not progress to harder material.

There is good evidence that this isn’t a problem. A number of studies show that when children are allowed to select their own books, they typically select books that are at their reading level or are harder (Southgate, Arnold, and Johnson, 1981; Shin and Krashen, 2007).

A study done in 1958 (!!!) showed that as students mature, they select more complex books and select from a wider variety of genres (LaBrant, 1958). Our study (Krashen, Lee, and Lao, forthcoming) confirmed this result nearly 60 years later. We examined the reading choices made by elementary school children in Hefei, China, reading in Chinese, their first language. In agreement with LaBrant, we found that the books children selected as they progressed to higher grades were more complex, both in content and language.

Students don’t stick to easy books. They read what interests them.

Is There Another Route?
The case history presented here (my own), and the responses to the concerns, present a strong argument for the role of fiction in developing more complex language competence. My claim is that it is not only a pleasant path, it is the only path. There are no alternatives.

The development of academic language competence does not come from participation in classes, even heavily academic classes: Biber (2006) reported that classroom discourse is closer to conversational language than to academic language.

Academic language cannot be taught—the system to be acquired is far too complex to be analyzed, presented to students, and consciously learned. See, for example, the very complex descriptions of some aspects of vocabulary (Hyland, 1996), grammar, and text structure (Swales, 1990; Schleppegrell, Achugar, and Oteiza, 2004).

Also, the systems to be consciously learned are very large. Vocabulary is a good example: there are simply too many words to be acquired one at a time through study (e.g., draw a line from the word to the definition, write three sentences with every word). Esti­mates of adult vocabulary size in the first language alone range from about 40,000 to over 150,000 words (Smith, 1988; Krashen, 2004a).

A Short Conclusion
The case for fiction is very, very strong. It is, however, a product of evidence from different areas of research not usually in contact with each other. This paper is an attempt to improve the situation, and make life more pleasant for students worldwide.

Notes: *An earlier version of this paper was presented at the International Conference on School Librarianship of Asian Countries, National Taiwan Normal University, Tapei, October 15, 2015.

Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York: Harper Perennial.
Djikic, M., Oatley, K., & Moldoveanu, M. (2013). “Opening the closed mind: The effect of exposure to literature on the need for closure.” Creativity Research Journal, 25(2), 149-154.

Biber, D. (1988). Variation Across Speech and Writing. Cambridge: Cambridge University Press.

Biber, D. (2006). University Language: A corpus-based study of spoken and written registers. New York: John Benjamins.

Canada, G. (2010). Fist Stick Knife Gun: A personal history of violence. Boston: Beacon Press.

Gardner, D. (2004). “Vocabulary input through extensive reading: A comparison of words found in children’s narrative and expository reading materials.” Applied Linguistics, 25(1), 1-37.

Gardner, D. (2008). “Vocabulary recycling in children’s authentic reading materials: A corpus-based investigation of narrow reading.” Reading as a Foreign Language, 20(1), 92-122.

Hyland, K. (1996). “‘I don’t quite follow’: Making sense of a modifier.” Language Awareness, 5(2): 91-109.

Kidd, D. & Castano, E. (2013). “Reading literary fiction improves theory of mind.” Science, 342 (6156), 377-380.

Krashen, S. (2003). Explorations in Language Acquisition and Use: The Taipei lectures. Portsmouth, NH: Heinemann.

Krashen, S. (2004a). The Power of Reading. Portsmouth, NH: Heinemann and Santa Barbara: Libraries Unlimited (second edition).

Krashen, Stephen. (2004b). “Basic Phonics.” TexTESOL III Newsletter, November 2004, 2-4. Available at

Krashen, S. (2011a). Free Voluntary Reading. Westport: Libraries Unlimited.

Krashen, S. (2011b). “Academic proficiency (language and content) and the role of strategies.” TESOL Journal, 2(4), 381-393.

Krashen, S. (2011c). “Nonengagement in sustained silent reading: How extensive is it? What can it teach us?” Colorado Reading Council Journal, 22, 5-10.

Krashen, S. (2015). “The end of motivation.” New Routes, 55: 34-35.

Krashen, S., Lee, S. Y., & McQuillan, J. (2012). “Is the library important? Multivariate studies at the national and international level.” Journal of Language and Literacy Education, 8(1), 26-36.

Krashen, S., & Mason, B. “Can second language acquirers reach high levels of proficiency through self-selected reading? An attempt to confirm Nation’s (2014) results.” International Journal of Foreign Language Teaching, in press.

LaBrant, L. (1958). “An evaluation of free reading.” In Research in the Three Rs, ed. C. Hunnicutt and W. Iverson. New York: Harper and Brothers, pp. 154-161.

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

Ravitch, D. and Finn, C. (1987). What Do Our 17-Year-Olds Know? New York: Harper and Row.

Schleppegrell, M., Achugar, M., Oteiza, T. (2004). “The grammar of history: Enhancing content-based instruction through a functional focus on language.” TESOL Quarterly 38(1): 67-93.

Shanahan, K. (2010). “Review of Liz Murray, Breaking Night.” Retrieved Nov. 23, 2010, from Homeless/productreviews/0786868910/ref=cm_cr_pr_link_3?ie=UTF8&showViewpoints=0&page Number=3.

Shin, F. and Krashen, S. (2007). Summer Reading: Program and Evidence. New York: Allyn and Bacon.

Simonton, D. (1988). Scientific Genius: A psychology of science. Cambridge: Harvard University Press.
Smith, F. (1988). Joining the Literacy Club. Portsmouth, NH: Heinemann.

Swales, J. (1990). Genre Analysis: English in academic and research settings. Cambridge: Cambridge University Press.

Southgate, V., Arnold, H., and Johnson, S. (1981). Extending Beginning Reading. London: Heinemann Educational Books.

Stanovich, K. and Cunningham, A. (1992). “Studying the consequences of literacy within a literate society: the cognitive correlates of print exposure.” Memory and Cognition 20(1): 51-68.

Stanovich, K. and A. Cunningham. (1993). “Where does knowledge come from? Specific associations between print exposure and information acquisition.” Journal of Educational Psychology, 85(2): 211-229.

Stanovich, K., West, R., & Harrison, M. (1995). “Knowledge growth and maintenance across the life span: The role of print exposure.” Developmental Psychology, 31(5): 811-826.

Sullivan, A. & Brown, M. (2014). Vocabulary from Adolescence to Middle Age. London: Centre for Longitudinal Studies, University of London.

Trelease, J. (2013). The Read-Aloud Handbook. Penguin. Seventh Edition.

West, R., and Stanovich, K. (1991). “The incidental acquisition of information from reading.”Psychological Science 2: 325-330.

West, R., Stanovich, K., and Mitchell, H. (1993). “Reading in the real world and its correlates.” Reading Research Quarterly 28: 35-50.

Stephen Krashen is professor emeritus at University of Southern California.