What Can Readers Read After Graded Readers?

Reading in a Foreign LanguageISSN 1539-0578April 2016, Volume 28, No. 1pp. 63–78What can readers read after graded readers?Jeff McQuillanCenter for Educational DevelopmentUnited StatesAbstractNation (2014) concluded that most of the vocabulary one needs to read challenging textsin English can be acquired incidentally through voluminous reading. This study examinespossible texts that second language (L2) readers can use to move from controlledvocabulary materials such as graded readers, which go up through approximately the4,000-word-family level, to more challenging texts such as newspapers, classic novels,and academic texts, at the 9,000-word-family level. An analysis of a set of popular fictionseries books found that such books can provide a sufficient amount of input, with 98%vocabulary coverage, so as to serve as one possible “bridge” to more challenging texts.Keywords: extensive reading, graded readers, vocabulary acquisition, text coverage,comprehensible inputStudies in both first and second language acquisition have shown that new vocabulary can beacquired incidentally through the reception of comprehensible input via reading (Krashen,2004a). While the importance of reading in vocabulary acquisition is generally acknowledgedamong second language (L2) researchers, there has been disagreement as to whether thevocabulary one can acquire solely through reading can “take you all the way,” to a point whereyou acquire a sufficient number of words to understand more challenging texts, including classicnovels, newspapers, and academic writing.Cobb (2007, 2008) held that reading alone cannot provide enough input to allow L2 readers toacquire enough words to handle challenging texts within a reasonable amount of time.McQuillan and Krashen (2008), however, argued that L2 acquirers can indeed get enough inputto acquire most of the vocabulary they need through voluminous, self-selected reading, such asthat provided by extensive reading programs (Day & Bamford, 1998; Mason, 2013).Acquiring Sufficient Vocabulary to Read Challenging TextsNation (2014) attempted to settle this debate through a corpus analysis. His approach to thequestion was based upon three assumptions:(a) In order to have “adequate comprehension” of text, one needs to know at least 98% of thehttp://nflrc.hawaii.edu/rfl

McQuillan: What Can Readers Read after Graded Readers?64words in that text (Hu & Nation, 2000; Schmitt, Jiang, & Grabe, 2011);(b) To achieve this 98% vocabulary coverage for challenging texts, one must know the 9,000most frequently occurring word families in English (Nation, 2006); and(c) To have a reasonable chance of acquiring an unknown word family, one must encounter itat least 12 times in text.The logic Nation presents here is straightforward: If we know how many words one must read toencounter the first 9,000 word families at least 12 times, we can provide estimates of the amountof text and time L2 readers need to acquire a sufficient vocabulary to handle challenging texts.Are Nation’s assumptions reasonable? Several researchers (Hu & Nation, 2000; Laufer &Ravenhorst-Kalovski, 2010; Schmitt et al., 2011) have argued that “adequate comprehension” oftext requires somewhere between 95% and 98% vocabulary coverage. While vocabularyknowledge is not the sole factor in determining reading comprehension, it has clearly beenshown to be an important one in both the first language (L1) and L2 research (Anderson &Freebody, 1981; Hu & Nation, 2000). Laufer and Ravenhorst-Kalovski (2010), for example,found that vocabulary knowledge accounted for 64% of the variance in reading comprehensionscores.The exact percentage of words a reader needs to know to understand a text depends on how“adequate” comprehension is defined. Laufer and Ravenhorst-Kalovski suggest that 95% is the“minimum” coverage needed, with 98% (or more) being “optimum.” In choosing the higherfigure of 98%, Nation attempted to provide a conservative estimate of the percentage of words areader needs to understand text independently.1The other key assumption made by Nation – that 12 exposures to an unknown word are sufficientto acquire the word – is based on previous studies that produced differing estimates both aboveand below that figure (e.g., Brown, Waring, & Donkaewbua, 2008; Pellicer-Sanchez & Schmitt,2010; Waring & Takaki, 2003). These analyses attempted to determine the number of exposuresto a word needed to, in Nation’s words, “develop something approaching rich knowledge” of aword (p. 2). In Pellicer-Sanchez and Schmitt (2010), for example, unknown words that occurredat least 10 times in the text were acquired 80% of the time, as measured by a meaningrecognition test (Table 1, p. 41). In Waring and Takaki (2003), at least 15 repetitions wererequired for a similar level of success (72%).Based on these and other similar studies, Nation’s use of 12 occurrences as a threshold foracquisition appears to be an attempt to find a middle ground between competing estimates.Nation and other researchers acknowledge that acquisition of vocabulary depends on more thanjust the number of exposures to the acquired word, and any estimate depends on one’s criteria fordetermining the “depth” of knowledge as well as its breadth (Wesche & Paribakht, 1996).Nation (2014) analyzed a corpus comprised of 25 novels taken from Project Gutenberg(http://www.gutenberg.org). He also provided estimates of how long it would take a reader toread that amount of text, assuming a reading speed of 150 words per minute.Table 1 shows Nation’s results for the number of words that one would need to read in order toReading in a Foreign Language 28(1)

McQuillan: What Can Readers Read after Graded Readers?65encounter a word family at least 12 times in his chosen corpus of novels, and a calculation of thetime required to read them. Estimates are broken down by 1,000-word-family groups, from the2nd to the 9th 1,000-word-family levels.Table 1. Amount of input and time needed to acquire the 2nd through the 9th mostfrequently occurring 1,000-word families in EnglishHours Needed1,000- WordAmount toCumulativePer LevelLevel ListReadHours(@150 1,223Total11,000,0001,223Note. Data from Nation (2014), Table 4Assuming you know the 1,000 most frequently occurring words in English already, Nationestimated that you would need to read approximately 200,000 words in order to have areasonable chance of acquiring most of the words in the 2,000-word-family level. Afteracquiring most of the words in the 2,000-word-family level, you would then need to read another300,000 words in order to encounter most of the words in the 3,000-word-family level at least 12times, and so on.As shown in Table 1, one would need to read approximately 11,000,000 words to reach the9,000-word-family level, and that this feat would take about 1,200 hours to complete. At onehour per day, this represents a little over three years of reading, very doable for a motivated adultor adolescent acquirer. One hour per day of reading is in line with what is expected of universitystudents in the United States for out-of-class assignments. Nearly half of all American professorsexpect their students to do at least six hours of homework outside of school per week, or nearlyan hour per day (Sanoff, 2006).If Nation’s analysis and the assumptions behind that analysis are correct, it appears that freereading can indeed provide L2 readers with the opportunity to acquire the necessary vocabularyto handle challenging texts.Between Graded Readers and Challenging TextIf free reading is sufficient, the next step is to determine what sort of texts L2 acquirers shouldread. Nation (2014, p. 11) noted that controlled-vocabulary materials such as graded reader seriescan provide students with enough input to reach approximately the 4,000-word-family level. Butwhat should readers read after graded readers? How is this “gap” between the 4,000- and 9,000word-family levels to be filled? The problem can be summarized this way:Reading in a Foreign Language 28(1)

McQuillan: What Can Readers Read after Graded Readers?66Graded Readers ? Challenging TextsMid-Frequency ReadersNation (2014) proposed that the gap can be made up in part by the use of “mid-frequencyreaders.” Mid-frequency readers are adaptations of texts that meet the 98% vocabulary coveragecriterion at lower vocabulary levels than the texts were originally written. The texts are createdby substituting the less frequently occurring words in the stories and novels with more frequentlyoccurring synonyms, as well as by controlling the number of new word families the reader willencounter in the text (Nation & Anthony, 2013; Schmitt & Schmitt, 2014). Texts have beendeveloped by Nation and others at the 4,000-, 6,000-, and 8,000-word-family levels. Nation’sproposal to fill the gap is thus:Graded Readers 4K Readers 6K Readers 8K Readers Challenging TextsMid-frequency readers can be an important source of input for English as a Second Language(ESL) and English as a Foreign Language (EFL) acquirers. However, the number of such readersis still small, and for copyright reasons, the mid-frequency readers have thus far been limited toadaptation of works that are in the public domain. There is also the question of interest: not allL2 readers will find the texts chosen for adaptation to be sufficiently engaging to do the kind ofvoluminous reading required to read several million words. But it is one possible path, and givenenough adapted texts, one that could allow readers to acquire sufficient vocabulary to read morechallenging texts.Light Reading, Narrow ReadingKrashen has long advocated the use of self-selected “light reading” to bridge the gap betweenmodified texts such as graded readers and challenging, academic texts (2004a, 2010). Lightreading refers to the materials being read, and may include comic books, children’s books, youngadult fiction, popular adult fiction, and popular magazines. In particular, Krashen (2004b)advocates a specific approach to light reading called “narrow reading.” In narrow reading,readers read books by the same author or on the same topics. An example of narrow reading isthe use of series books, texts written by the same author and usually involving the same maincharacters in the same or similar settings (Hwang & Nation, 1989; Schmitt & Carter, 2000).Narrow reading of series books takes advantage of the powerful influence of prior knowledge oncomprehension (Eidswick, 2010). Once readers finish the first book or story in the series, theyhave considerable background knowledge about the characters and setting that in turn canfacilitate comprehension of subsequent stories.In narrow reading, readers also become familiar with the writer’s style and word choices, as wellas the proper nouns (character names, places). This in effect reduces the vocabulary loadrequired for reading additional novels in the series. This vocabulary “recycling” is particularlyReading in a Foreign Language 28(1)