Elfrieda H. Hiebert, University of Michigan
Why don't we have more guidelines today for selecting literature appropriate for different aspects of the literature-based instructional program for beginning readers? The answer to this question lies in the kinds of beginning reading materials that long dominated in the U.S.
For most of this century, literature was not used in the early reading texts that were used to instruct American children. Instead, texts for beginning readers were developed to feature one of the following:
Generations of American school children learned to read with texts that emphasized high frequency words, words such as prepositions, articles, and conjunctions (e.g., of, the, and) that serve critical functions in sentences. Texts of this type were based on suggestions from educational psychologists such as Thorndike (1921) and Gray (Elston & Gray, 1930) who helped create the early Scott-Foresman basal readers. For decades, these early programs were updated, but philosophically remained unchanged.
Selections with high-frequency words dominated the children's collection of stories in the basal readers. The story, Tiny (Rider, 1986) is illustrative of these texts and reflects a somewhat "modern" version that might still be found in today's classrooms. Other than the name of the main character (i.e., Tiny), 18 high-frequency words are combined to create a story about two children who need to capture an escaped baby pig. The scene in which the children solve the problem by scattering food to allure Tiny back into the pen is captured by, "I have it! Now we will get Tiny. Tiny will want to come." (p. 23). As is typical, little of the story is conveyed through the language of the text.
In the 1950s and 1960s, phonetically regular words were suggested as the basis for the texts of early reading instruction (Bloomfield & Barnhart, 1961; Chall, 1967/1982; Flesch, 1957). Developers of reading programs based on word regularity followed similar procedures as in the past. They created contrived texts to highlight a word's feature. For example, in the text, The Bad Fan (Rasmussen & Goldberg, 1964), a set of characters named Dad, Nan, and Dan engage in a series of actions that involve a fan. The denouement of the passage is expressed as "Dad had a bad fan." As was typical, the text conveys little of the story, instead focusing on the phoneme /an/. While such texts were neither used universally in American schools nor produced strong evidence that more children learned to read at higher levels as a result of their use (Bond & Dykstra, 1967), phonetically regular texts continue to be presented as a panacea for American reading problems (Adams, Bereiter, Hirschberg, Anderson, & Bernier, 1995).
Texts based on high-frequency words dominated instructional materials and assessments through readability formulas. The presence of words on high-frequency word lists was one of the factors used to establish the ease or difficulty of texts in most readability formulas (Klare, 1984). To ensure that a text was viewed to be readable, adjustments would have been of the following type (from a current favorite, Henkes's (1996) Lilly's purple plastic ): An uncommon word such as "permitted" might be substituted with a high-frequency word such as "let," "creations" with "things," or "jaunty" with "happy." In the mid-1980s, a number of researchers analyzed the effects of such changes in texts on children's comprehension (e.g., Beck, McKeown, Omanson, & Pople, 1984; Brennan, Bridge, & Winograd, 1986). They showed that children were able to understand original stories more successfully than the rewritten "easier" ones. The original stories had structures that fit the expectations of readers regarding stories. The original stories also had descriptions and actions that were memorable for readers and made sense. This research helped set the stage for shifting the focus away from words used to literary merit. However, this new criterion has presented new challenges for beginning reading instruction.
With the publication of Becoming a Nation of Readers (Anderson, Hiebert, Scott, & Wilkinson, 1985), the findings on the possible problems of contrived texts for meaningful reading were shared with literacy educators beyond the research community. This message resonated with what teachers had observed as they taught young children and educators at many levels shared the belief that different books were needed. They turned to trade bookswritten by authors for children's pleasure and learning, not for school instruction.
However, there were few guidelines for making decisions about what kinds of stories within the realm of high quality literature could be used to support students' literacy learning. Moreover, concerns about "basalizing" the literature and ruining the literary experience were raised. Despite these concerns, high-quality literature became the norm for teaching literacy to students of all ages and levels of readers.. For example, in California, tradebooks had been mandated in the California English/Language Arts Framework Committee (1987); in Texas, through the Texas Education Agency (1990). Because of the influence of these large, state-adoption states, basal reading programs replaced contrived text with literature selections. For example, in the series in which the story, Tiny, appeared, a new story featuring a pig was included. This selection, If I Had a Pig (Inkpen, 1988), uses a predictable pattern, narrated by a child describing a fanciful set of events that he and his pet pig might do: make a house, paint pictures, have fights, and so forth until it is time to "tell him a story and take him to bed."
Initial studies such as that by Hiebert et al. (1995) indicate that unless children have a substantial foundation in emergent literacy at the beginning of first grade, they struggle when literary quality is the sole or primary criterion for selecting passages. At the end of first grade, about 45% of the children in literature-based classrooms were not able to read trade books that were less difficult than If I Had a Pig. There are just too many unknown words with complex features for those children who are not already competent readers to learn. These findings contributed to my focus on features of texts that could potentially scaffold beginning readers' literacy learning. Further, I was interested in how particular features of text that support independent reading skills could be balanced with literary merit.