The Potential Use of the Lord of the Rings Book Series for Incidental Vocabulary
Acquisition
著者 Shaffer Jeffrey D.
journal or
publication title
Journal of Shizuoka University Education
volume 16
page range 1‑19
year 2020‑03‑23
出版者 静岡大学大学教育センター
URL http://doi.org/10.14945/00027255
The Potential Use of the “Lord of the Rings”
Book Series for Incidental Vocabulary Acquisition
JEFFREY D. SHAFFER (Education Development Center)
ABSTRACT
This paper is the second in a series of papers which aim to examine the lexical contents of popular works, especially longer ones, and assess their potential use for the incidental acquisition of vocabulary. These papers look at the types of words used within stories that span over multiple books, such as “Harry Potter” and the “Lord of the Rings” (this current paper). By analyzing the words that appear within a text we are thereby able to evaluate the lexical difficulty of the work. The long-term desire of this series of papers is to determine the overall difficultly of a large number of texts in the hope that such a list of books with their estimated “difficulty” level would be of service to learners of English as a second or foreign language.
The previous paper examined the lexical contents of the “Harry Potter” series, and the current paper will employ the same tools and methods to examine the “Lord of the Rings” series of books. The results of the lexical analysis are discussed, and a comparison between the two book series is also considered.
Keywords: Vocabulary Acquisition, Extensive Reading, Narrow Reading, Lord of the Rings, Lexical Analysis
1. INTRODUCTION 論文
This paper is the second in a series of papers which aim to examine the lexical contents of popular works, especially longer ones, and assess their potential use for the incidental acquisition of vocabulary. These papers look at the types of words used within stories that span over multiple books, such as “Harry Potter” (Shaffer, 2018) and the “Lord of the Rings” (this current paper). By analyzing the words that appear within a text we are thereby able to evaluate the lexical difficulty of the work.
The long-term desire of this series of papers is to determine the overall difficultly of a large number of texts in the hope that such a list of books with their estimated “difficulty” level would be of service to learners of English as a second or foreign language.
The previous paper (Shaffer, 2018) examined the lexical contents of the “Harry Potter” series, and the current paper will employ the same tools and methods to examine the “Lord of the Rings”
series of books. The results of the lexical analysis are discussed, and a comparison between the two book series is also considered.
2. INCIDENTAL VOCABULARY ACQUISITION
“Incidental vocabulary acquisition” is a technical term for what we and our children do all the time – we pick up new words unintentionally as we go through life. Obviously, we pick up new words from the people and the world around us, but research has shown that reading is one of the fastest ways in which to pick up new words (Nation, 2001). This fact may not come as a very large surprise as it is well known that people who read regularly have, in general, a larger vocabulary than those who do not (Nagy, Herman, & Anderson, 1985) and reading has historically been recommended as a quick way to pick up vocabulary in a foreign language as well. (See Lewis 2009 for a few of such examples).
But reading is not just something we do in order to improve our vocabulary, nor is all reading
done merely to gain “information.” We actually enjoy books. We enjoy diving into new worlds and experiencing other lives. And thus reading stories can be seen as not only a useful activity (at least as far as learning new words is concerned), but also a pleasurable activity: i.e., learning can be fun!
While this, too, is not a new discovery, more recently the idea of intentionally reading a large quantity of comprehensible books has been seen as an even faster way to build one’s vocabulary. The two important keys here are the “comprehensible” part, which means the reader should know almost every word on the page, and the “large quantity” part, which is the same as discussed at the beginning of this paper. This type of reading is called Extensive Reading and was first proposed by Krashen in 1981 and subsequently followed up with the work of many others, such as Nation (1997) and Waring (1997).
My own previous work in this field has been within an extension of the Extensive Reading idea based on the concept that if reading a lot helps one to improve their vocabulary quickly, then reading a lot about the same subject should improve it even faster (Shaffer, 2004, 2005, 2010, 2015, 2018). This type of focused reading is called Narrow Reading. Not only would Narrow Reading help a reader to improve their subject-specific vocabulary at an increased rate, but it is also thought that it will help them advance to more challenging texts at a faster rate (Cho & Krashen, 1994; Krashen, 1996).
Another advantage of reading highly similar texts is that there is a noticeable increase in the repetition of vocabulary within the texts (Shaffer, 2004, 2005, 2010, 2015, 2018) which decreases the burden of acquiring new vocabulary items (Nation, 2001). This, in turn, increases the learner’s change of acquiring new vocabulary (Nagy & Herman, 1985). An increase in word repetition also means a potential increase in word generation, which are the variations in how a single word is used (i.e., jump, jumps, and jumped). Frequent encounters with word generation are believed to improve learners'
understanding of a text and help improve their depth of knowledge of grammatical behaviors, collocations, register constraints, and alternative meanings (Schmitt & Carter, 2000). An additional advantage of using narrow reading is that it allows language learners to obtain a high degree of comprehension through the accelerated acquisition of context-specific vocabulary -- lexical items that are specific to the subject matter. The relevance of context-specific vocabulary has been demonstrated by Ward (1999), who showed that a carefully constructed list of 2,000 context-specific words could account for 95% of every word found within engineering texts.
This type of rapid vocabulary acquisition is further explained by Nagy and Herman's Vocabulary Learning Hypothesis (1985), which states that most vocabulary is learned gradually through repeated exposure to new and known words in various contexts. They estimated that when a learner encounters a new word, they have only a 5-10% chance of acquiring that word and that it may take 10-12 encounters before a word is acquired. This suggests that input materials with less vocabulary repetition may be less comprehensible to lower level learners than similar materials with greater vocabulary repetition (Schmitt & Carter, 2000). One implication of the Vocabulary Learning Hypothesis, therefore, is that input with a high degree of vocabulary repetition is preferred, and this is precisely what narrow reading provides.
A learner’s ability to comprehend the texts they have chosen to read is also of extreme importance. Hirsh and Nation (1992) calculated that at least a 95% vocabulary comprehension rate would be necessary for a learner to read unsimplified texts without any aid, though they suggest that a 97-98% vocabulary comprehension rate would be more ideal. They estimated that in order to reach such a high comprehension rate with academic texts, the reader must know approximately 5,000 word families. In 1996, Hazenberg and Hulstijn, working with learners of Dutch as a second language, estimated that 10,000 words or more might be necessary to comprehend academic texts. However, as
learners wishing to read academic materials in English may not be able to simply memorize 5,000 to 10,000 words, Nation (2001) has suggested that a well-designed language program should use no more than 25% of instruction time for the explicit instruction of vocabulary, while the remaining 75% or more should be dedicated to incidental vocabulary acquisition by such means as communicative output activities, fluency development tasks, and reading activities such as narrow reading.
Schmitt and Carter (2000) looked at the lexical advantages of narrow reading by comparing two sets of newspaper stories, one set comprised of an 8-day running story, and the other a collection of unrelated stories. By looking at the lexical content of these two sets of stories, they found that the increase in degree of similarity among the running story texts also lead to an increase in overlap, and this increase in overlap lead to an increase in both vocabulary repetition and word generation, which in turn improved vocabulary acquisition and depth of word knowledge, respectively. Schmitt and Carter's study did indeed show the use of related stories, such as those commonly used for narrow reading, to be advantageous.
Previous research, such as that done by Schmitt and Carter and by myself (Shaffer, 2004, 2005, 2010, 2015, 2018) have examined the use of texts for incidental vocabulary acquisition. My earliest work looked at how the selection of input texts, such as TV news as opposed to newspaper articles, can affect word repetition and word generation, and my most recent word looked at how the narrow reading of longer works affected incidental vocabulary acquisition. This current paper aims to examine another series of long works, namely the "Lord of the Rings" three-book series and to compare the results with that of my previous work.
I propose to do this by using the three-book series “Lord of the Rings” to build a corpus from which I will address the following questions: (1) How difficult, lexically, are the source materials?, (2) How suitable is this book series for incidental vocabulary acquisition as determined by the degree
of vocabulary repetition and word generation found within (a) the individual “Lord of the Rings”
books, and (b) within sets of books (book 1 vs. books 1-2 vs. books 1-3)?, (3) How does the overall length of a text affect the potential for incidental vocabulary acquisition?, and (4) How do these results compare to the results previously found by analyzing the “Harry Potter” seven-book series? which I will attempt to address within the analysis of the first three questions.
3. METHOD
As in the preceding paper in this research series, another extremely popular book series was chosen: J.R.R. Tolkien’s “Lord of the Rings.” This three-volume story is not only widely available, but is also continues to be immensely popular. According to Wikipedia (2018), “Lord of the Rings” is the number one best-selling book with over 150 million copies sold. Therefore, these books should be relatively easy for a reader to obtain, making it an ideal subject for the current paper.
As shown in Table 1, these books, taken singly or as a whole, are of considerable length, making them a desirable corpus for the study of incidental vocabulary acquisition through narrow reading. After obtaining the three books that comprise the “Lord of the Rings” series (Tolkien, 1954, 1954, 1955), the lengthy appendixes at the end of the book three were removed to allow the current analysis to focus purely on “story / narrative” and to avoid the grammatical and linguistic discussions they contained. A plain-text version of each book was then arranged into different datasets which were then analyzed using the AntWordProfiler software program (Anthony, 2014).
Based on the original lexical analysis software program Range (Heatley & Nation, 1996), AntWordProfiler reads in one or more plain-text files and then returns not only the number, but also the frequency of occurrence for word tokens (running words), word types (the first occurrence of a word), and word families (clusters of words based upon a common root). These resulting lexical data is presented across four vocabulary levels – the first 1000 high frequency vocabulary words and the second 1000 high frequency vocabulary words (which together comprise the 2000 most frequently occurring word of English as noted in the General Service List (West, 1953)), academic words (which is based on Coxhead’s (2000) Academic Words List), and low frequency words (which is all other words.) AntWordProfiler was chosen over Range (which was used exclusively in my previous work) as it boasts several improvements over the original, most notably in that it counts hyphenated words (such as “book-lover”) as two separate high frequency words instead of as a single low frequency word.
4. RESULTS Q1 How difficult, lexically, are the source materials?
It is possible to assess the overall lexical difficulty of a text by looking at the distribution of vocabulary as found among the three base word levels (high frequency, academic, and low frequency.) Nation (2001) states that for academic texts 80% of all word tokens (every instance of word) fall under high frequency vocabulary, with an additional 10% falling under academic vocabulary, and the remaining 10% being classified as low frequency vocabulary. While the Harry Potter series of books is obvious not academic in nature, West’s General Service List (1953) from which Nation (2001) partially based his work, also states that 80% of average texts, such as novels, should be covered by high frequency vocabulary.
Upon analysis of the “Lord of the Rings” corpus, it was found that the first 1000 high
frequency words provided 83% coverage and the second 1000 high frequency words provided an additional 6% coverage of the text (see Table 2 below), which means the most commonly occurring 2000 English words cover a total of 89% of the text. This is considerably greater than the 80% coverage proposed by Nation (2001) for these same words, which suggests that at least lexically speaking the
“Lord of the Ring” series may be easier to read and comprehend. Analysis also shows that less than 1% of the text is provided by academic words, which is far below Nation’s proposal of 10% (2001).
These initial results strongly suggest that the lexical makeup of the “Lord of the Rings” is far more comprehensible than that of academic texts, are more comprehensible than TV news (Shaffer, 2015) where the first 1000 high frequency words were found to provide 79.8% coverage, and the current results even appear to show that “Lord of the Rings” is lexically even more comprehensible than the
“Harry Potter” series of books (Shaffer, 2018) where the first 1000 words provided 77.2% coverage, the second 1000 words gave 6.42% coverage, and the academic word list gave 1.02% for a total of 84.6% coverage of the text.
Overall, the current lexical analysis indicates that the large coverage provided by the first 2000 high frequency words suggests that “Lord of the Rings” may, indeed, be useful for the incidental acquisition of high frequency vocabulary.
Q2 How suitable is this book series for incidental vocabulary acquisition as determined by the degree of vocabulary repetition and word generation found within (a) the individual “Lord of
the Rings” books, and (b) within sets of books (book 1 vs. books 1-2 vs. books 1-3)?
The “Lord of the Rings” corpus was analyzed according to various datasets in order to more clearly assess how word repetition and word generation change as more texts are included. Three datasets were used for analysis: the first comprised simply of the first book, the second containing the first two books, and the third dataset containing all three books.
Word repetition can be indicated by a type-token ratio (see Table 3 below). As noted above, word tokens are counted as every occurrence of a word whereas word types are counted as only the first occurrence of a word in a text. No matter how often the same word appears throughout the text, it is only counted as one type. For example, if the word “dream” occurs 20 times within a text, then it is counted as 20 word tokens, but only as one word type. Thus, a type-token ratio represents how often words are repeated within a text, with the calculation being simply (word types / word tokens). Thus, if a word occurs only once, then the type-token ratio is (1 type / 1 token =) 1.00. If a word occurs twice, (1 / 2 =) 0.50. To simplify and summarize, as word repetition increases the type-token ratio (TTR) decreases.
Looking individually at each book, we find an extremely high rate of repetition, the least of which being found in Book 3 with a TTR of 0.0521 (though this is still a notable amount of repetition) and the greatest about being found in Book 1 with a TTR of 0.0460 (see Table 3 below). To help give an idea of what is meant by “a high rate of repetition”, my previous analysis of the “Harry Potter” book series has a low TTR of 0.0772 found in Book 2 and a high TTR of 0.0449 found in Book 5. However, only “Harry Potter” Books 5 and 5 have greater repetition than all three books in the “Lord of the Rings” series, which suggests that, comparatively speaking, “Lord of the Rings” would be more idea for picking up new vocabulary incidentally.
Looking at the results of the AntWordProfiler analysis for the different sets of books (see Table 4 below), we find that as the number of books is increased the TTRs decrease steadily. Thus, from a reader’s perspective, reading only the first book the reader would encounter a TTR of 0.0460, but if they were to read all three books, they would encounter a TTR of 0.0265. Thus, by reading the entire series, the reader would encounter 1.7 times greater word repetition than by reading only the first book in the series. This is considerably less than the amount of repetition given by the “Harry Potter” series where a reader taking in all seven books would encounter a TTR of 0.0184. Perhaps this suggests that a reader would do better to read the “Harry Potter” series (as far as incidental vocabulary acquisition is concerned), but the only real conclusion that can be made is that the more one reads the more repetition one encounters: a 1.7 times gain for the three books in “Lord of the Rings” and a 3.9 times gain for the seven books in “Harry Potter”.
Next, a deeper analysis was made of the “Lord of the Rings” series to discover how much repetition occurs within each of the four vocabulary levels. Looking first at the TTRs found in the individual books, we find an extremely high amount of repetition among the first 1000 high frequency
words (HF 1-1000), with the greatest amount of repetition occurring in Book 1 which has a TTR of 0.015 (see Table 5 below). The least amount of repetition at this vocabulary level is in Book 3 with a TTR of 0.018 (For reference, the amount of repetition seen in the “Harry Potter” books at this High Frequency 1000 level ranges from 0.013 to 0.030).
Looking at the next vocabulary level, the second 1000 high frequency words (HF 1001- 2000), we still see a large amount of repetition occurring with TTRs ranging from 0.151 to 0.190 (with
“Harry Potter” again ranging from 0.222 to 0.162).
The academic word list (AWL) level shows a marked decrease in repetition with a range from 0.337 to 0.363, though this may be caused by the very small amount of AWL vocabulary found within this corpus – Table 2 above shows that only 0.44% of the entire three-book series is comprised of words at this level. Again, for reference, the “Harry Potter” corpus at this level ranged from 0.255 to .0418).
Last, the low frequency vocabulary (LF) level shows less repetition than the HF 1001-2000 level, ranging from TTFs of 0.2218 to 0.1616.
Therefore, comparatively speaking, it appears that, on the whole, the “Lord of the Rings”
series of books has a greater amount of word repetition than that of “Harry Potter”, and thus would appear to be a better choice for learning vocabulary incidentally.
Next, we look at the repetition among the four vocabulary levels as found in the datasets of
increasing size (see Table 6 below). As might be expected, we find a very large increase in repetition at the HF 0-1000 level with TTRs ranging from 0.007 to 0.015 (“Harry Potter” ranges from 0.0037 to 0.0175). The HF 1001-2000 level sees a similar increase in repetition with TTRs ranging from 0.078 to 0.151 (“Harry Potter” ranges from 0.1434 to 0.0369). At the AWL level we find a smaller increase in repetition with TTRs ranging from 0.220 to 0.363 (“Harry Potter” having 0.1084 to 0.3120), and at the low frequency level we find TTRs ranging from 0.146 to 0.229 (and “Harry Potter” has 0.0784 to 0.1700). Once again comparing results with my previous study on the “Harry Potter” books, we see more repetition occurring within the “Harry Potter” book-sets this time whereas with the individual books “Lord of the Rings” showed greater repetition.
The most interesting feature of the TTRs across the four vocabulary levels, as seen clearly in Table 6, is the steady increase in repetition among all vocabulary levels as the size of the input (word tokens) increases. Though as similar results were noted in the “Harry Potter” study, perhaps this is to be expected of a series of books.
Next, our analysis considered word generation within the “Lord of the Rings” corpus. Word generation is when a word is used in grammatically and semantically different ways. For example, we can have round circles, well-rounded students, rounds of golf, and doctor's rounds. As noted above, apart from word repetition, word generation is another key factor in incidental vocabulary acquisition as it promotes depth of word knowledge. Word generation can be measured as a family-type ratio (FTR) which represents the relation between the number of word families and the number of word types that appear within a text. For example, the word types “jump”, “jumped”, and “jumping” all belong to a single word family called “jump.” Similar to TTRs, lower FTRs indicate higher levels of word generation, and are therefore more desirable for acquiring depth of word knowledge. When looking at FTRs there is a problem in that a master list of all possible word families does not yet exist,
even though such lists of families do exist for the two high frequency vocabulary lists and for the academic vocabulary list. This means that programs like Range and AntWordProfiler cannot accurately assess the FTRs for the low frequency vocabulary level. Range simply returns a blank whereas AntWordProfiler returns the number of families as the same number of types it discovers.
This is obviously not an accurate portrayal of low frequency word families. Thus, the current study will only be looking at word generation as it is found in the first three vocabulary levels for the sake of accuracy and clarity.
By looking at the word family and word type results for the first three vocabulary levels of the “Lord of the Rings” series as seen as individual books (see Table 7 below), we notice a fairly linear decrease in FTRs from book to book, which means a steady increase in word generation in the high frequency and academic vocabulary levels from Book 1’s FTR of 0.7133 to Book 3’s FTR of 0.7328.
(This amount of word generation is a little more than half that found in the “Harry Potter” corpus which had an FTR range of 0.4980 to 0.4157). The semi-linear increase in generation is a curious effect seen in both the current corpus and the previous “Harry Potter” corpus. One might expect, this type of results in the datasets of increasing size and not as a feature of the individual books, themselves. Seeing as this same effect is found in two major book series, it would definitely warrant further study by looking at even more series of books. Perhaps it is a feature of a long narrative where there are many
“introductions” and “new things” at the start of a story and a repeat of characters and settings through the middle and ending. Only further research can say.
Looking at the results from an analysis of the datasets of increasing size (see Table 8 below), we find that word generation (as measured by FTRs) gradually decreases as the number of word types increase, as one might expect. With the “Lord of the Ring” corpus we find a very smooth linear increase in word generation as more books are added, starting with an FTR of 0.7133 (again, considering only the first three vocabulary levels) and decreasing steadily to an FTR of 0.7328 for all three books inclusively. (For reference, the “Harry Potter” corpus saw a decrease in FTRs from 0.4491 to 0.3334 over the seven-book series).
Q3 How does the overall length of a text affect the potential for incidental vocabulary acquisition?
We have seen how the “Lord of the Rings” series, taken individually and in datasets of increasing size, show a large amount of word repetition and word generation, both of these being essential for incidental vocabulary acquisition. Next, we look at whether the length of a text (or collection of texts) directly affects both word repetition (TTR) and word generation (FTR) as proposed by the extensive reading and narrow reading models.
By arranging the texts in order of increasing word tokens, we find a clear (though not perfectly linear) increase in both word repetition and word generation (see Table 9 below). We see a gradual increase in word repetition from an initial TTR of 0.052 found in Book 3 down to a TTR of 0.026 for all three books taken as a whole (The “Harry Potter” books found a similar increase ranging
from 0.0709 for Book 1 to 0.0184 for the entire seven-book set). Similarly, we see an increase in word generation, though not quite so linear, from an initial FTR of 0.733 found in Book 3 down to an FTR of 0.713 for Book 1 then back up to an FTR of 0.744 for the entire series of book. (Comparatively, the
“Harry Potter” series remained fairly close to a linear increase with a range of 0.4980 for Book 1 to 0.3334 for the entire series).
The increase in word generation that is then lost as the number of tokens increases is a curious result, one not found in the “Harry Potter” corpus. One can only speculate as the its cause, though. One possibility might be the multi-threaded narrative employed in “Lord of the Rings” where no fewer than three simultaneous timelines are explored until nearly the end of the last book whereas
“Harry Potter” follows a fairly straight narrative timeline from beginning to end. Another possible cause might be the nature of the “Lord of the Rings” stories and how they continually reveal new cultures, new languages, and new “histories”. Book 1 focuses mostly upon the Hobbits, their home and culture, and then their journey to the Elves. Thus, we see mostly the Hobbit culture, some of the
“Men” culture at the town of Bree, and then some of the Elves’ culture at Rivendell. Whereas in Books 2 and 3 we introduce the cultures of Rohan, Gondor, and others lands in-between. Thus, perhaps it’s gradual increase in new subject matter in the later books that cause a decrease in word generation whereas in “Harry Potter” we find new creatures introduced as the series continues, but much of the same culture, language, and history, thus much of the “new subject matter” comes in the earlier books.
Again, however, this is mere speculation, and something that would be well to follow up upon in future research, perhaps by looking at individual chapters and their “subject matter” for comparison.
5. CONCLUSIONS
By taking a second corpus comprised of narrow reading materials into consideration, we once again note a large increase in word frequency, but less conclusive results regarding word generation. The increase in word frequency is idea for language learners who wish to increase their vocabulary through reading stories. This potential is seen in both the current “Lord of the Rings” corpus and in the “Harry Potter” corpus. As for word generation which leads to word depth-of-knowledge, the “Harry Potter” corpus shows an increase in word generation the more books that are read which suggests they are even more ideal for a language learner’s vocabulary acquisition, though perhaps the complex storytelling and background materials of the “Lord of the Rings” series makes those books somewhat less idea.
This presents a strange balance when both the word repetition (TTRs) and word generation (FTRs) are compared holistically as the “Lord the Rings” series has greater overall word repetition than the “Harry Potter” series, but less overall word generation. This would suggest that a reader will learn vocabulary faster by reading “Lord of the Rings”, but have a deeper knowledge of vocabulary by reading “Harry Potter”.
The results seen in the current study and in the previous “Harry Potter” study strongly suggest that a series of books based upon the same characters, the same “world”, or even the same overall story arc provide a very large degree of word repetition, and depending upon the series, word
generation as well. This makes such book series ideal for incidental vocabulary acquisition for L2 and L1 learners alike.
Areas left for further research would be to compare other “narrow” mediums such as a TV series or a movie series (again, lexically speaking) to measure their inherent degrees of word repetition and word generation. Similarly, it would be interesting to break down both the “Lord of the Rings”
series and the “Harry Potter” series by chapter and “subject matter” to discover where the variance in word generation comes from. And finally, and perhaps most importantly, future research needs to develop a means for accurately measuring both lexical complexity (a scale for comparing the lexical difficulty of a text) and grammatical complexity (a scale comparing not just simply average word length and average sentence length as is commonly used, but also takes part-of-speech into account.
Either way, it is hoped that the current research would encourage language learners to enjoy a good book, especially a good book series, with the pleasure of knowing they’re also building their vocabulary. As far as narrow reading is concerned, the pleasure of reading leads to the power of progress.
References
Anthony, L. (2014). AntWordProfiler (Version 1.4.1) [Computer Software]. Tokyo, Japan: Waseda University. Available from http://www.laurenceanthony.net/
Cho, K-S., & Krashen, S. (1994). Acquisition of vocabulary from the Sweet Valley Kids series: Adult ESL acquisition. Journal of Reading, 37(8), 662-667.
Coxhead, A. (2000). A new academic word list. TESOL Quarterly, 34(2), 213-238.
Dörnyei, Z. (2001) Teaching and researching motivation. Harlow, UK: Longman.
Hazenberg, S., & Hulstijn, J. H. (1996). Defining a minimal receptive second-language vocabulary for
non-native university students: An empirical investigation. Applied Linguistics, 17(2), 145-163.
Heatley, A., & Nation, P. (1996). Range [Computer Software]. Wellington, New Zealand: Victoria University of Wellington. (Available from http://www.vuw.ac.nz/lals).
Hirsh, D., & Nation, I. S. P. (1992). What vocabulary size is needed to read unsimplified texts for pleasure? Reading in a Foreign Language, 8(2), 689-696.
Krashen, S. (1981). The case for narrow reading. TESOL Newsletter, 15(6), 23.
Krashen, S. (1996). The case for narrow listening. System, 24(1), 97-100.
Lewis, C.S. (2009) The Collected Letters of CS Lewis, Volume 1: Family Letters (1905-1931), San Francisco: HarperCollins Publishers. (See page 270).
Locke, E. A., & Latham, G. P. (1994). Goal setting theory. In H. F. O'Neil, Jr., & M. Drillings, (Eds.), Motivation: Theory and research. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Nagy, W., & Herman, P. (1985). Incidental vs. instructional approaches to increasing reading vocabulary. Educational Perspectives, 23, 16-21.
Nagy W., Herman P., & Anderson R. C. (1985). Learning words from context. Reading Research Quarterly, 20(2), 233–253.
Nation, P. (1997). The language learning benefits of extensive reading. The Language Teacher, 21(5), 13-16.
Nation, I. S. P. (2001). Learning vocabulary in another language. Cambridge: Cambridge University Press.
Schmitt, N., & Carter, R. (2000) The lexical advantages of narrow reading for second language learners.
TESOL Journal, 9(1), 4-9.
Shaffer, J.D. (2004). Varying degrees of contextual narrowness in narrow reading and its potential impact on vocabulary acquisition. Temple University Japan Studies in Applied Linguistics:
Teaching Vocabulary to Second Language Learners. 41, 95-104.
Shaffer, J.D., (2005). Choosing narrow reading texts for incidental vocabulary acquisition. The Language Teacher. 29 (7), 21-27.
Shaffer, J.D. (2010). Factors influencing vocabulary acquisition in narrow listening. Journal of Shizuoka University Education. 2010 (6), 27-37.
Shaffer, J.D. (2015). Using video for incidental vocabulary acquisition. Journal of Shizuoka University Education. 2015 (11), 211-220.
Tolkien, J.R.R. (1954). The fellowship of the ring. London: George Allen & Unwin.
Tolkien, J.R.R. (1954). The two towers. London: George Allen & Unwin.
Tolkien, J.R.R. (1955). The return of the king. London: George Allen & Unwin.
Waring, R. (1997). (Ed.) Special edition on Extensive Reading. The Language Teacher, 21(5). 1997.
West, M. (1953). A general service list of English words. London: Longman, Green & Company.
Wikipedia. (n.d.). List of best-selling books. Retrieved from https://en.wikipedia.org/wiki/List_of_best-selling_books#List_of_best-selling_individual_books
