The review of literature in chapter two centers around issues of vocabulary and textbook related to this study. It starts with the examination of the importance of vocabulary for second language learning, and then proceeds to explore second language (L2) learners’ vocabulary size from the perspective of lexical threshold for L2 reading comprehension, specifically targeting on reading academic texts. The review of vocabulary selection literature subsequently follows with discussions on vocabulary selection principles for English for General Purposes (EGP) as well as English for Academic Purposes (EAP). Concepts and word lists used for discussion on vocabulary selection, such as high-frequency words, academic vocabulary, General Service List (GSL), Academic Word List (AWL), are also introduced along with vocabulary selection literatures. The next part targets on the vocabulary in textbooks, with literature review of suggested arrangement and evaluation criteria for vocabulary in textbooks. The final part of literature review focuses on vocabulary in senior high school textbooks in the context of Taiwan. Related studies on vocabulary selection in the senior high school textbooks in Taiwan were also elaborated.
The Importance of Vocabulary
Vocabulary is an important component of for both language use and language learning. Linguist David Wilkins (1972) stated that “…while without grammar very little can be conveyed, without vocabulary nothing can be conveyed…” (p.111).
Laufer (1986) shared the same view and suggested that “without adequate lexis, there is no proper language competence or performance.” Nation (1990) also pointed out that inadequate vocabulary knowledge had resulted in many difficulties for L2 students’ both receptive and productive language use. Read (2000) indicated that
vocabulary knowledge is “a prerequisite for effective language use.” Vocabulary knowledge has been recognized as a good predictor of reading comprehension (Laufer, 1992; Nation, 2001& 2006; Qian, 1999 & 2002).
Numerous researches have also provided empirical data and demonstrated the importance of vocabulary in language learning by showing high correlations between vocabulary knowledge and various measures of language proficiency (Albrechtsen, et al, 2008, Alderson, 2005; Laufer, 1992). Vocabulary size of L2 students in Israel was found to correlate with reading comprehension at .50-.70 (Laufer, 1992). More
recently, Albrechtsen, Haastrup and Henriksen (2008) also revealed a high correlation of .73-.80 between L2 vocabulary size of Danish students and their L2 reading ability.
Alderson (2005) explored the relationship between vocabulary knowledge and language proficiency systematically in the DIALANG, a diagnostic test development project. Two vocabulary tests showed that vocabulary knowledge correlated with reading at .64, listening at .61-.65, writing at .70-.79 and grammar at .64. Vocabulary is proved to have a strong relationship with various language skills. The importance of vocabulary can be concluded by statement of Alderson (2005) that “the size of one’s vocabulary is relevant to one’s performance on any language tests. In other words, language ability is to quite a large extent a function of vocabulary size.” Since vocabulary knowledge or size plays such an important role in language learning, the questions that concerns most ESL/EFL teachers related to vocabulary learning would be: how many words does a learner need to know and what are the words that should be learn first?
Learners’ Vocabulary Size and Lexical Thresholds
The amount of words that learners should learn depends to a great extent on learners’ needs. Based on the logic of a communicative approach to vocabulary ability,
the discussion of vocabulary size of L2 learners would not be in an absolute sense, but in relation to particular contexts of use (Chappelle, 1994). For example, the questions raised would be how many words L2 learners need to know in order to read authentic novels or to read college textbooks in English. Since one of the objectives of the 2010 curriculum guidelines for senior high school in Taiwan is to cultivate students’
competence for higher education (MOE, 2009), and from recent surveys, the majority of university students in Taiwan need to read English-medium textbooks (Chen, et al., 2002; Lin & Kong, 2000; Yu & Zheng, 2010), the vocabulary size that senior high school students should develop could be examined through the concept of “lexical threshold,” the minimal vocabulary knowledge necessary for “adequate” reading comprehension (Alderson, 1984; Laufer, 1992 & 2010; Nation, 2006).
Lexical Thresholds for L2 Reading Comprehension
Reading in a foreign language involves not only content knowledge and reading skills as readers in L1 do but also L2 linguistic knowledge, which consists of both syntactic and lexical knowledge (Alderson, 1984; Laufer, 2010). Alderson (1984) raised the insightful question of reading in a foreign language: Is reading in a foreign language a reading problem or a language problem? To respond to the question, Alderson (1984) proposed “language threshold hypothesis,” which suggested L2 learners need to possess a certain amount of linguistic knowledge in order to function well in L2 tasks, such as reading or listening. With multiple regression analysis, Bosser (1991) showed that L2 linguistic knowledge level predicted L2 reading abilities four times better than L1 reading abilities except for learners at advanced level group. As for which component of linguistic knowledge playing a much more significant role for reading in a foreign language, lexical knowledge was found to contribute more to L2 academic reading success than syntactic knowledge
(Saville-Troike, 1984; Laufer& Sim, 1985). Most of the researches exploring language threshold for reading comprehension have focused more on lexical knowledge instead of syntactic knowledge. Findings from lexical knowledge for reading comprehension studies on lexical threshold are useful for second or foreign language education because teachers and course designers can set up the vocabulary learning goals or design lexical syllabi accordingly.
Suggested Vocabulary Size for L2Learners from Lexical Threshold Studies Ever since Alderson’s (1984) call for investigation of “language thresholds”
for L2 reading comprehension, several studies have been conducted to explore pedagogically usable thresholds (Laufer, 1989, 1992 & 2010; Laufer & Sim, 1985;
Hsu, 2011; Hu & Nation, 2000; Nation, 2001& 2006). Opinions and findings of these lexical thresholds vary as a result of different operationalized definitions of what it means to be “adequate” L2 reading comprehension and research approaches. Different levels of reading comprehension might be required in different contexts, and
subsequently result in different lexical thresholds. As for research approaches, the search for lexical thresholds for reading comprehension often proceeds in two complementary approaches: one is to examine the coverage that words of different frequency level provide to the texts; the other is to test learners on text comprehension and relate different reading scores to learners’ vocabulary size (Cobb & Horst, 2001 a).
In other words, the first approach places emphasis on the reading texts and the lexical coverage of the reading texts while the second approach involves the reader and search for the size of the readers’ sight vocabulary. Sight vocabulary refers to the words that learners are so familiar that they can be recognized and decoded
immediately without much cognitive effort when learners engage in a reading task (Laufer, 2010). The larger the size of sight vocabulary of a learner, the more lexical
coverage he or she has for the reading texts. The more lexical coverage the learner has for a text, the higher chances for the learner to have better reading comprehension of the text.
To tackle the issue of learners’ vocabulary size for adequate comprehension, Laufer (1989) first investigated lexical coverage of academic texts for “adequate”
reading comprehension, in which a score of 55% was set as the operationalized definition for “adequate” comprehension. The results showed that the learner group that scored 95% and above on lexical coverage had a significantly higher number of successful readers (scoring at least 55% or above on reading comprehension test) when compared to the 90% - 94% group. Some learners at different coverage level still received passing scores for their reading comprehension. This made Laufer (1989) conclude that the 95% text coverage is a lexical threshold of probabilistic nature.
“Adequate” comprehension might happen when students possess a vocabulary size lower than 95% of lexical coverage of texts, but the chance or probability is low.
Using two comprehension tests and adopting the score that most learners in the 100% coverage group receives as the passing score for “adequate” comprehension (scoring 87.5%), Hu and Nation (2000) searched the lexical coverage for “adequate”
reading comprehension on fiction texts. Four different lexical coverage groups (80%, 90%, 95%, and 100%) were created by replacing some words in the text with
non-words. Results showed that none of the students could reach “adequate”
comprehension at 80% of coverage while some students could at the 90% and 95%
coverage groups. Ninety-eight percent of lexical coverage of text is concluded to be the probabilistic threshold for reading fiction texts adequately while 80% lexical coverage of text is believed to be the all-or-nothing threshold.
The two different lexical coverage suggestions vary due to different definition of “adequate” comprehension yet both 95% and 98% lexical coverage
suggestions are acknowledged depending on what level of “adequate” comprehension is expected. In terms of lexical coverage of text for “adequate” reading
comprehension, it can be concluded by Nation’s (2001) statement that “The
probabilistic threshold is 98%. With this coverage, almost all learners have a chance of gaining adequate comprehension. If, instead of adequate comprehension, a standard of minimally acceptable comprehension is applied (as Laufer did in her study), then 95% coverage is likely to be the probabilistic threshold” (p. 147).
The lexical coverage of texts for “adequate” comprehension was identified, and the investigations on size of sight vocabulary for “adequate” comprehension followed subsequently from the two complimentary research approaches. To locate the threshold vocabulary level, Laufer (1992) took the reader approach in which the reading comprehension scores and the university students’ vocabulary size were examined. With “adequate” comprehension score set at 56%, 3,000 word-family level was found to be the lexical threshold where there are more readers than non-readers.
With “adequate” comprehension score at 63%, the vocabulary knowledge of 4,000 word families was needed while the vocabulary knowledge of 5,000 word families would be required for “adequate” comprehension score set at 70%. The vocabulary knowledge of 3,000 word families is concluded to be the minimal requirement for reading unsimplified texts. The findings of this study offered practical implications for syllabus designers to set vocabulary learning goals. On the other hand, Hirsh and Nation (2001) adopted the text coverage approach in which they used General Service List (West, 1957) for the first 2,000 word families and the old Thorndike and Lorge’s teacher’s word list (1944) for words beyond 2,000 word families to analyze the lexical coverage of unsimplified teenager novels. To reach the 97%-98% text coverage for
“adequate” reading, the vocabulary size needed is suggested to be 5,000 word families.
With the intent of more accurate estimates for the number of word families that
learners need for “adequate” comprehension with various genres of texts, Nation (2006) developed word family lists from British National Corpus (BNC) frequency list (Leech, et al., 2001) and searched for more precise vocabulary size that learners need to comprehend different genres of texts. The findings revealed that 3,000 word families and proper nouns would give 95% coverage of the spoken texts. To get 98%
coverage for spoken texts, learners need to have a vocabulary size of 6,000-7,000 word families and the knowledge of proper nouns, which often accounts for about 4%-5% of text coverage. For written texts of newspapers or novels, learners need 3,000 word families and the knowledge of proper nouns to reach 95% of text coverage. However, to reach the 98% text coverage of written texts, learners would need a vocabulary size of 8,000-9,000 word families and the knowledge of proper nouns as required in Laufer (1989) and Hu and Nation (2000).
The two lines of lexical threshold investigations seem to find similar results when adopting the 95% text coverage as the probabilistic threshold despite different genres of written texts (Laufer focused mainly on academic texts while Nation aimed more at novels and newspapers): L2 learners need to have the vocabulary size of 3,000 word families and proper nouns to reach 95% of text coverage for “adequate”
reading comprehension. The lexical threshold of 3,000 word families has been
acknowledged and recommended by many researchers (Cobb & Horst, 2001a; Nation, 2000; Nation & Waring, 1997; Thornbury, 2002) to be an important vocabulary
learning goals for L2 learners in order to provide initial access to authentic texts.
Vocabulary size for English for Academic Purposes For L2 learners in the senior high school in Taiwan in preparation for university study, what concerns the instructors, textbook writers and syllabus designers more would be the lexical threshold for reading academic texts. Laufer
(2010) re-examined the lexical threshold for academic English with more rigorous research design and suggested two thresholds for reading academic texts in English:
the optimum one, (vocabulary knowledge of 8,000 word families plus knowledge of proper nouns resulting in the 98% text coverage) and the minimal one (vocabulary knowledge of 4,000-5,000 word families and knowledge of proper nouns leading to the 95% coverage of texts). Using BNC frequency word list, Hsu (2011) investigated the vocabulary threshold of English-medium textbooks and research articles in the field of Business for EFL learners. Business textbooks were discovered to reach 98%
text coverage at the 5,000 word-family level and to get to 95% text coverage at the 3,500 word-family level. Research articles in business discipline requires the vocabulary knowledge of 8,000 word families to provide a 98% text coverage and 5,000 word families to supply 98% coverage of texts. The results from these two lexical threshold studies for “adequate” comprehension of academic texts offer rich information for L2 learners and instructors to set up the vocabulary learning goals for academic studies.
Vocabulary Selection
While 3,000 word families are generally agreed to be a reasonable goal for L2 learners on the English for general purpose track and students in preparation for university studies should aim at around 3,500-8,000 word families depending on the language demands of different disciplines, L2 instructors, syllabus planners and students would have to deal with another important issue in vocabulary learning planning, that is, which 3,000 word families should be learned or what words should be learned first. This is the issue of vocabulary selection (Richards, 2001).
General Principles for Vocabulary Selection
The investigation of the issue of vocabulary selection can be traced back to
early works of Faucett, et al. (1936), Ogden (1930), Thorndike and Lorge (1940), and West (1957). Frequency of words counted from a large collection of texts plays an important role in the early works of vocabulary selection at the first half of the twentieth century (Richards, 1974 & 2001). Early lexicometrics specialists propose that word frequency counts offer the basis for a more scientific and objective
approach to vocabulary selection (Richards, 1974). This “frequency” principle is still acknowledged by numerous researchers of these days as an important criterion for vocabulary selection (Coady, et al., 1993; Cobb, 2013; Gairns & Redman, 1986;
Geothals, 2004; Nation, 2001& 2003; Nation & Waring, 1997; Richards, 1974 &
2001; Schmitt, 2000; Schmitt & Schmitt, 2012; Sinclair & Renouf, 1998). Word frequency refers to “how often the word occurs in normal use of the language”
(Nation & Waring, 1997). High-frequency words tend to give a much greater return in opportunities for language use than low-frequency words do (Nation, 2003).
Frequency of a word can be an indicator of its usefulness (Goethals, 2004; Nation &
Waring, 1997) and its difficulty (Ryder & Hughes, 1985). Thus, the frequency of words is often regarded as an important criterion for vocabulary selection.
Word frequency count is acknowledged to be an essential component in planning word lists for language teaching yet, to ensure the usefulness of the selected words, the high-frequency words also have to occur across a wide range of different texts. “Range” of the words also plays an important role in the process of vocabulary selection. The selection of “range” would be determined based on learners’ needs and learning goals. For students with learning English for general purposes, words
appearing across a wide range of texts would be selected while learners of English of specific purposes, words from specific fields would be selected. Students’ needs and proficiency level are also recognized as important criteria for vocabulary selection in a course or curriculum (Allen, 1983; Gairns & Redman, 1986). When taking students’
needs and background into consideration, cultural factors should also be taken into consideration in the process of vocabulary selection as learners might have special cultural interests in L2 that might be distinct from native speakers (Gairns & Redman, 1986). In addition to the four criteria mentioned above, Richards (2001) also points out five other criteria that are useful for vocabulary selection: (1) teachability:
concrete vocabulary can be easily illustrated through pictures or by demonstration, (2) similarity: word items similar to words in learners’ native language, (3) availability:
less-frequent words yet are readily available when certain topics are presented, (4) coverage: words that cover the meaning of other words, and (5) defining power:
words that are useful in defining other words. The act of vocabulary selection would involve different criteria depending on the objectives of the language curriculum or programs or the purposes of the wordlists.
Vocabulary Selection for Learners in Preparation for Academic Study Allen (1983) pointed out that learners’ needs should be the most important criteria for selecting words. For L2 learners in preparation for academic study, such as the senior high school students in Taiwan, several researchers and vocabulary
specialists (Cobb & Horst, 2001a; Decarrico, 2001; Nation, 2001& 2003; Paquot, 2011; Schmitt, 2010; Thornbury, 2002) have recommended that L2 learners should be first equipped with a core vocabulary of 2,000 high-frequency words, then with study of academic words for higher education study.
High-frequency Words and General Service List
Nation (2001) categorized English vocabulary into four groups:
high-frequency words, academic words, technical words, and low-frequency words.
High-frequency words, also known as core words, or basic words (Paquot, 2010), provide a large amount of text coverage of both spoken and written texts and occur
highly frequently in all kinds of language uses. The best-known list of high-frequency words is the West’s (1953) General Service List of English Words (GSL), which is created from a five-million word corpus of written texts and contains about 2,000 headwords considered suitable for foreign language teaching (Decarrico, 2001; Nation, 2001; Richards, 2001). The list was created with the incorporation of findings from a major vocabulary selection study of that time: The Interim Report on Vocabulary Selection (Faucett, Palmer, West, and Thorndike, 1936). The criteria for GSL compilation include: word frequency, structural value, universality, subject range, definition words, word-building capacity and style (Howatt, 2004). Information on frequency of different meanings of each word is also available to teachers. GSL has been quite useful in second or foreign language education as it provides a reference in making decisions about what words to use in L2 learning materials: course books and graded readers and which meaning of the words to be taught first. GSL also claims to provide about 80% coverage of most written texts, and thus become quite influential on L2 learning (Carter & McCarthy, 1988). On the other hand, GSL has also received some critiques, mainly on its out-datedness, lack of spoken data, utility, and
availability (Richards, 1974 & 2001; Carter & McCarthy, 1998). Despite several disadvantages mentioned, GSL has continued to be useful as Nation and Hwang (1995) demonstrated quite large overlap between more recent frequency count and GSL. The GSL used in the study was a re-organized version of GSL, in which the original GSL was arranged in frequency order and concept of word families. Based on Nation and Bauer’s (1995) criteria of determining word families, the re-organized GSL contains 1,965 word families and has been adopted for numerous lexical profiling and
vocabulary selection studies (Coxhead, 1998 & 2000; Fan, 2004; Horst, 2005; Nation
& Hwang, 1995; Nation & Wang, 1999). Nation and Hwang (1995) discovered that replacing some words in GSL with some other words in the top 2000 frequency band
only resulted in 1% coverage difference. The first 1,000 words of GSL was also found to provide about 77% of text coverage while the second 1,000 words offer about 5%
only resulted in 1% coverage difference. The first 1,000 words of GSL was also found to provide about 77% of text coverage while the second 1,000 words offer about 5%