A Corpus-based Study of Vocabulary in the New Concept English Textbook Series

Abstract

Growing a large vocabulary is one of the most important tasks of language learners and yet research reports on low levels of lexical knowledge for English as a foreign language (EFL) learners. In EFL contexts, textbooks are a vital source of exposure to English and the words that are worth learning. The New Concept English textbook series is widely used in China by large numbers of learners in secondary schools and cram schools, and in preparation for university studies. This study finds that while learners will encounter mostly high-frequency vocabulary in these books, they will need to know between 3000 and 6000 word families to cope with these texts, and Book 4 requires 1000 more word families than Book 3. Science-related texts in the books contribute to the vocabulary load because they contain a large proportion of low-frequency words used only once and have a higher vocabulary load than humanities texts. There are opportunities for learning mid-frequency vocabulary in these textbooks. Limitations of current research and suggestions for future research are also presented.

Keywords

Vocabulary corpus EFL textbooks New Concept English coverage

Introduction

English textbooks play an essential role in learners’ vocabulary development in a foreign language learning context, where the textbook in essence is the curriculum (Macalister, 2016). In the China context, one of the most popular textbooks is the New Concept English (NCE) textbook series. It was first published in London as Longman’s New Concept English in 1967, and then revised by Qishen He specifically for Chinese learners in 1997 (Alexander and He, 1997). The NCE series is regarded as classic and has influenced many Chinese secondary school and university students since its publication (Song, 2019; Wang, 2018). The series has been reported as one of the main textbooks in almost 90% cram schools in Shanghai (Wei, 2012). The series has four books, ranging from beginner to advanced learners – that is, from elementary to university students. For the current study, Book 3 (Developing Skills) and Book 4 (Fluency in English) were chosen because they are meant for intermediate and above learners in high schools and at universities, and aim to help learners increase their vocabulary. The books contain reading texts on a range of topics, such as philosophy, arts, politics, education and astronomy. These texts are probably used for intensive reading in class, which Macalister (2014: 392) points out usually involves ‘. . .using a text that learners would be unlikely to read successfully without assistance’.

The NCE textbooks have been investigated in relation to cultural content (Song, 2019) and intercultural communication (Song, 2019; Wang, 2018), but it seems that few studies have focused on the vocabulary in the series. Ma (2013) investigated the vocabulary in Book 2, Book 3 and Book 4 of the NCE textbook series using the Range program (Heatley et al., 2002). The focus of Ma’s (2013) study was high-frequency vocabulary from West’s (1953) General Service List of English Words (GSL). This list contains 2000 word families (a word family contains a head word such as cover, and includes derivations and inflections, such as covered, covering, covers and coverage). Ma (2013) also looked into academic vocabulary from Coxhead’s (2000) Academic Word List (AWL). This list contains 570 word families and excluded the GSL. Ma (2013) showed the GSL covered a large proportion of the three textbooks, as would be expected because high-frequency vocabulary represents the majority of words in any text (Nation, 2013). The coverage of the GSL gradually decreased from Book 2 (92.35%) to Book 4 (86.2%). The coverage of AWL was low in Book 2 (1%) and increased to over 4% in Book 4. Book 3 showed a different pattern from Book 2 and Book 4 in that coverage of the GSL and AWL was quite different across the units in Book 3. Ma (2013) did not examine the frequency of vocabulary in the textbooks in detail. The increased percentage of word types could have been caused by the increase in text lengths from Unit 1 to Unit 3 in Book 3. One of the aims of the present study is to look more closely at the vocabulary in units in the textbooks.

The present study is, to the best of our knowledge, the first to examine two textbooks to find out more about what vocabulary learners would encounter. The results of this study will be useful for textbook designers and teachers, and ultimately students, in their quest to develop their vocabulary knowledge in English.

Vocabulary and English as a Foreign Language (EFL) Textbooks

There are several key concepts in vocabulary research which are important for the present study in relation to EFL textbooks. These concepts are frequency, vocabulary load and repetition.

Frequency

Vocabulary learning is one of the important goals in the language classroom (Nation, 2013). To help learners make their effort the most efficient, Nation (2006) breaks vocabulary into three categories based on their frequency levels in the British National Corpus/Corpus of Contemporary American English (BNC/COCA): high-frequency, mid-frequency and low-frequency words. High-frequency words, consisting of the first and second 1000 word families, contain function words (e.g. of, be, an) and many content words (e.g. present, example and almost). High-frequency words make up the highest percentage of words in any text in English. Mid-frequency vocabulary consists of 7000 word families from the third to the ninth 1000 word families (e.g. prompt, appliance, recycle). Developing knowledge of mid-frequency words should follow high-frequency words and will help with reading authentic texts and fluency. Low-frequency words cover only a small proportion of texts, although it is a large group ranging from the 10th 1000 word onward (Nation, 2013). It consists of some technical vocabulary like zoomorphic, neurotoxic and rarely used words like plummet and adipose. Laufer (2013) suggests that low-frequency words can be ignored if they are not crucial for comprehension. Finding out more about the lexical profile of textbooks is useful because it can uncover possible reasons why a text or textbook might be difficult for learners (see, for example, Sun and Dang, 2020; Coxhead et al., 2010; Coxhead & Boutourwick, 2018), and what vocabulary learners are exposed to in text. This is important because it allows us to see, for example, the amount of high-frequency vocabulary in the textbooks, and whether there is mid-frequency vocabulary which learners can focus on after learning high-frequency lexis.

Vocabulary Load

The vocabulary load of the text (Webb and Nation, 2008) is important for an evaluation of whether a text is at the right level for learning. Vocabulary load relies on the concept of coverage, meaning the percentage of tokens (the occurrence of every word) in the text that might be known to learners (Nation, 2013), which is important for answering the question of how much vocabulary in a text learners need to know for adequate comprehension. Two thresholds for coverage are suggested by research: 95% and 98%. Laufer (1987) found that 95% was required for a pass score on a comprehension test. Hu and Nation (2000) suggested that 98% is needed for learners to gain adequate comprehension of a fiction text without assistance. Laufer and Ravenhorst-Kalovski (2010) revisited lexical threshold and confirmed that 95% should be the minimal threshold, which enables learner to reach acceptable comprehension with assistance, and 98% as optimal for making learners independent readers. Coverage is determined using word lists, as we have already seen in Ma’s (2013) research above with the GSL and AWL. For vocabulary load, Nation’s BNC/COCA frequency lists are used, as well as supplementary lists of abbreviations, proper nouns, marginal words used in speaking (e.g. um, ah) and transparent compounds (see Nation, 2016).

If learners’ comprehension of a text reaches 95%, it means that one word in every 20 words would be unknown to learners. Nation (2013) is clear that this is still a heavy load of unknown vocabulary because learners will not be able to guess the unknown words from the context and may need to refer to a dictionary or other sources to reach a better comprehension of the text. Instead, 98% coverage would be optimal because the density of unknown words is much lower than that of 95%.

The vocabulary load of textbooks has been explored in a range of studies. Alsaif and Milton (2012) analysed 22 textbooks used in public middle and high schools in an EFL program in Saudi Arabia, again using Range (Heatley et al., 2002) with 15 1000-word lists made from the BNC (Nation, 2006). Alsaif and Milton (2012) found that these books covered more than half of the most frequent 5000 words. Although Alsaif and Milton (2012) examined the number of new words in each book, they did not look into the distribution of the new words at different word levels. Nguyen (2020) examined the lexical features of the texts in 30 units of the English textbooks for high-school students in Vietnam using Lextutor (Cobb, 2019) with the 25 1000-word lists of Nation’s BNC/COCA frequency lists (Nation, 2016). This study showed that learners needed the knowledge of the first 3000 and first 5000 word families to comprehend 95% and 98% of most of the texts in the book, which was beyond the learners’ vocabulary knowledge of the most frequent 2000 word families. Moreover, the textbooks did not provide enough opportunities for learners to meet novel words repeatedly. Nguyen (2020) did not examine all the words that appeared in the textbook (e.g. proper nouns, marginal words and abbreviations), which needs to be taken into consideration for the lexical coverage.

In the Chinese context, Sun and Dang (2020) examined the vocabulary load of a high-school textbook series published by Yilin Press. They compared the vocabulary load with measures of vocabulary knowledge of 265 high-school students who use the textbooks. Sun and Dang (2020) used the Updated Vocabulary Levels Test (UVLT) (Webb et al., 2017), which was based on Nation’s BNC/COCA lists. Sun and Dang’s (2020) study showed that students needed 3000 word families and 9000 word families to comprehend 95% and 98% of the textbooks, respectively. This vocabulary load was far beyond the level of vocabulary knowledge of the test takers, based on their vocabulary test results. Only five of the 265 students had mastered the first 3000 words families, and most test takers had not mastered the most frequent 1000 word families. Webb and Chang (2012) also reported low levels of vocabulary knowledge in the Taiwanese context. For textbooks, it is essential that writers select texts at an appropriate level for effective reading (Webb and Nation, 2008).

Vocabulary load research has also been the focus of analysis of textbooks, which are used by first- and second-language speakers of English in first language contexts. Coxhead et al. (2010) examined science textbooks used in New Zealand secondary schools and found that learners needed to know more than 4000 word families with proper nouns to reach 95%, and more than 14,000 word families with proper nouns to reach 98%. In an EFL context (Taiwan), Hsu (2014) examined the vocabulary load of 100 engineering textbooks for undergraduate students and found that learners needed the most frequent 5000 word families plus proper nouns, abbreviations and compound words to reach 95%. This heavier vocabulary load in Science is also reported in Coxhead’s (2017) research on teacher talk in an international school in Germany. One of the reasons for this heavier load is technical vocabulary related to science topics and the use of low-frequency vocabulary (see Macalister, 2016; Nation, 2013).

These studies suggest learners need a large vocabulary to be able to understand the textbooks. The current research will analyse the vocabulary load of the texts using both 95% and 98% coverage, which may provide a reference for book designers and book users to check if the design and use of the books are appropriate.

Repetition

Nation (2014) states that repetition is the major factor affecting vocabulary learning from reading. Vidal (2011) found that two to three repetitive meetings with vocabulary would lead to the greatest increase in learning. Webb (2007) found that more than 10 repetitions would be needed to the development of full knowledge of a word. Repetition has found to be lacking in studies of EFL textbooks. Matsuoka and Hirsh (2010) found that 66.4% of the less frequent words in the New Headway English textbooks only occurred once, and only 12.1% occurred five times or more. Sun and Dang (2020) found that 17% of the vocabulary beyond the first 3000 word families only occurred once in the Chinese textbooks in their analysis for high-school students, which lead to insufficient conditions for the learning of these words.

Matsuoka and Hirsh (2010) analysed New Headway for upper-intermediate learners to investigate the vocabulary learning opportunities using the Range program (Heatley et al., 2002). Matsuoka and Hirsh (2010) found that the book could help learners deepen the knowledge of the 2000 words from West’s (1953) GSL and Coxhead’s (2000) AWL. They also noted that New Headway provided limited opportunities for learners to extend their knowledge to another lexis, such as proper nouns (Korean, Korea), technical words (comma, vocabulary), textual words (italics, bracket) and technology words (software, online).

Since textbooks provide a framework or guidance for learning both inside and outside the classroom (Hutchinson and Torres, 1994), it is necessary to know whether textbooks help learners build their vocabulary knowledge gradually. Therefore, examining how and whether the vocabulary load changes across different parts of the book and different books is important. Also, in an EFL context, learners will have limited opportunities to use or meet the vocabulary learned in class after school. It would be useful to check whether the textbooks provide opportunities for learners to meet the target vocabulary repeatedly. The current research will examine whether the mid-frequency vocabulary in NCE textbook series occurs repeatedly to facilitate learning of these words.

Research Questions

The current study focuses on the following three questions:

RQ1: How many word families do learners need to know to reach the coverage of 95% and 98% of NCE Book 3 and 4?

RQ2: Do the vocabulary loads gradually increase as the learning progresses within each book? If so, how? If not, why?

RQ3: To what extent do the books facilitate the learning of mid-frequency vocabulary?

Methodology

Materials

Book 3 and Book 4 contain 60 lessons and 48 lessons, respectively. The lessons in each book are evenly divided into several units (Table 1). Each lesson in the unit starts with a reading and listening passage. Note that Book 4 contains much longer passages in the final three units than in the first three. There are 22,786 running words for all the passages in Book 3 and 18,109 for those in Book 4.

Table 1.

The approximate length of the passages in each unit of Book 3 and Book 4.

	Book 3	Book 4
Unit 1	250 words	250–300 words
Unit 2	350 words	250–300 words
Unit 3	530 words	300–350 words
Unit 4	N/A	350–400 words
Unit 5	N/A	400–500 words
Unit 6	N/A	550–700 words

Preparing the Textbooks for Analysis

The PDF versions of the passages and translation of NCE Book 3 and Book 4, excluding exercises and glossaries, were downloaded from the Internet. The PDF files were then converted to text format (.txt). Any Chinese translations, exercises and glossaries were removed.

Data Analysis

This research adopts the BNC/COCA lists’ 25,000 words plus the word lists of proper nouns, marginal words, compound words and abbreviations (Nation, 2017) using the Range program 1.0.0 (Heatley et al., 2002). These lists were adapted to ensure that all the vocabulary in the textbooks was accounted for through the following process:

Proper nouns, such as ‘Qomolangma’ and ‘Ferdinand’, were added to the proper nouns list of BNC/COCA in Range.

Single words that were derived from a base word in the frequency-based BNC/COCA lists were added to the corresponding BNC/COCA word family. For example, ‘disunited’ was added to the word family of ‘unite’, and ‘viperine’ was added to the ‘viper’ family.

Hyphens in lexical items such as ‘cat-like’ and ‘horse-riding’ were replaced with [space] hyphen [space] since they can also occur as separate non-hyphenated words. Some hyphens are considered as parts of words (Nation, 2016), such as ‘echo-location’ and ‘pot-holer’, so these hyphens were deleted, and the words were added to the compound word list.

After processing the texts of the two books using the Range program, the result files were generated, and they showed the coverage of tokens at each word level. These result files were then used as the basis for calculation of the accumulated coverage of tokens to answer research questions about vocabulary load. Then, the data of the lexical coverage of different books, different units within the book and also different lessons within a specific unit was compared and analysed. Subsequently, the frequency of the mid- and low-frequency words in the books was checked.

Results and Discussion

How Many Word Families Do Learners Need to Know to Reach the Coverage of 95% and 98% of NCE Book 3 and Book 4?

The cumulative coverage of tokens in the fourth column of Table 2 shows that 3000 word families plus the proper nouns (e.g. Bruce, Italy), marginal words (e.g. eh, oh), compound words (e.g. boatman, worldwide) and abbreviations (e.g. IBM, CD) (referred to as ‘supplementary lists’ from now on) are needed to reach 95% coverage of Book 3. To reach coverage of 98%, 5000 word families plus supplementary lists are needed (see bold numbers in Table 2). If the supplementary lists are not included, then the vocabulary load for 95% is 5000 word families and for 98% is 25,000 word families. This figure illustrates the importance of the supplementary word lists, in particular the proper nouns and compounds, in this vocabulary load analysis.

Table 2.

Vocabulary load of Book 3 and Book 4 without and with the supplementary lists.

Word list	Book 3			Book 4
Word list	Coverage (%)	Without supplementary lists	With	Coverage (%)	Without supplementary lists	With
1st 1000	81.95	81.95	84.39	78.25	78.25	79.85
2nd 1000	8.15	90.1	92.54	9.61	87.86	89.46
3rd 1000	3.05	93.15	95.59	4.92	92.78	94.38
4th 1000	1.57	94.72	97.16	2.13	94.91	96.51
5th 1000	0.92	95.64	98.08	1.16	96.07	97.67
6th 1000	0.53	96.17	98.61	0.63	96.7	98.3
7th 1000	0.46	96.63	99.07	0.46	97.16	98.76
8th 1000	0.25	96.88	99.32	0.34	97.5	99.1
9th 1000	0.22	97.1	99.54	0.2	97.7	99.3
10th 1000	0.21	97.31	99.75	0.19	97.89	99.49
11th–25th 1000	0.24	97.55	99.99	0.52	98.41	100
Proper nouns	1.69			1.16
Marginal words	0.11			0.15
Compound words	0.61			0.26
Abbreviation	0.03			0.03

To reach 95% coverage for Book 4, 4000 word families with supplementary lists are needed and to reach 98%, 6000 word families with supplementary lists are needed. This coverage is slightly higher than the vocabulary load of Book 3, but lower than the 9000 word families needed to reach 98% for the Yilin textbook series in Sun and Dang’s (2020) study. The increased vocabulary load from Book 3 to Book 4 seems reasonable, since Book 3 is for intermediate level learners, and Book 4 is for advanced learners. However, if we use the UVLT results in Sun and Dang’s (2020) study as a baseline and take into account that high-school learners in this context may not have mastered the first 1000 word families in English, the vocabulary load of Book 3 and Book 4 will be well out of reach for these learners. The increased vocabulary load of 1000 word families from Book 3 to Book 4 represents another challenge, as Webb and Chang (2012) show that some EFL dedicated learners could acquire up to 400 word families per year, but many do not make such progress.

Do the Vocabulary Loads Gradually Increase as the Learning Progresses Within Each Book? If so, How? If Not, Why?

The 95% and 98% coverage across the three units in Book 3 increases gradually, from 3000 (95%) and 5000 (98%) plus supplementary lists in the first two units to 4000 (95%) and 6000 (98%) in Unit 3 (see Table 3). This is an increase of 1000 word families from the first two units and means that the vocabulary load in Unit 3 is heavier than the first two units. This finding is consistent with Ma’s (2013) finding that the vocabulary of Unit 3 was not the same as other units in the book.

Table 3.

Vocabulary load of each unit in Book 3 (% tokens).

Word list	Cumulative coverage (including supplementary lists)
Word list	Unit 1	Unit 2	Unit 3
1st 1000	85.81	85.82	82.66
2nd 1000	93.7	93.13	91.53
3rd 1000	96.39	96.15	94.79
4th 1000	97.66	97.38	96.75
5th 1000	98.5	98.07	97.87
6th 1000	98.71	98.65	98.53
7th 1000	99.19	99.05	99.02
8th 1000	99.5	99.35	99.21
9th 1000	99.62	99.49	99.53
10th–25th 1000	100	99.98	100

The first 1000 word families account for the largest proportion in all units. The comparison between each unit of Book 4 shows a fluctuation in the lexical coverage across the six units (Table 4). The 95% coverage remains at the same level (4000 word families) in the first four units and then drops to 3000 word families plus supplementary lists in the final two units. The coverage of 98% (8000 word families plus supplementary lists) is the highest for Unit 3 and drops to 7000 word families plus supplementary lists and even lower in the later three units. The coverage for Unit 3 in Book 4 (Table 4) is almost the same as the requirement for comprehending 98% of the texts in newspapers and novels (Nation, 2006).

Table 4.

Vocabulary load of each unit in Book 4 (% tokens).

Word list	Cumulative coverage (including supplementary lists)
Word list	Unit 1	Unit 2	Unit 3	Unit 4	Unit 5	Unit 6
1st 1000	82.64	77.86	77.64	78.77	81.3	80.49
2nd 1000	90.11	88.32	88.41	88.99	90.5	89.91
3rd 1000	94.22	93.88	92.9	94.15	95.43	94.99
4th 1000	96.37	96.33	95.22	96.22	97.39	97.00
5th 1000	98.05	97.41	96.64	97.39	98.14	98.08
6th 1000	98.75	98.21	97.31	97.95	98.89	98.55
7th 1000	99.17	98.63	97.83	98.6	99.33	98.91
8th 1000	99.54	99.15	98.13	98.94	99.61	99.21
9th 1000	99.63	99.29	98.47	99.19	99.73	99.42
10th 1000	99.77	99.38	98.77	99.44	99.85	99.61
11th–25th 1000	100	100	99.98	99.99	100	99.97

To find out why the vocabulary load varies between different units in Book 4, we checked the content of Unit 3. We found that this unit contained more science-related lessons (four out of eight lessons) and fewer humanities-based lessons. The science-based lessons require a larger vocabulary size to reach 98% (see Lessons 17, 18, 20 and 23 in Table 5). For example, to reach 98% of the humanities-based Lesson 22 Knowledge and progress, 4000 word families plus supplementary lists are needed, but for the science-based Lesson 18 Porpoise, 11,000 word families are needed. Unit 2 only contains two lessons related to science (Lesson 13 The search for oil and Lesson 14 The butterfly effect), which explains its lower vocabulary load (Table 4, column 3). This finding adds to evidence from Coxhead (2017) and Coxhead et al. (2010) that spoken and written science texts have a heavier vocabulary load than the other disciplines.

Table 5.

Vocabulary load of different lessons in Unit 3, Book 4 (including supplementary lists).

Lesson	Topic	Vocabulary load
Lesson	Topic	95% coverage(word family)	98% coverage(word family)
17	A man-made disease	5000	9000
18	Porpoise	7000	25,000
19	The stuff of dreams	3000	4000
20	Snake poison	6000	10,000
21	Early ‘Western’ Film	4000	6000
22	Knowledge and progress	3000	4000
23	Bird flight	4000	9000
24	Beauty	4000	6000

These science-related lessons contribute to the amount of low-frequency words in the lessons (Table 6), which increases the vocabulary load of the whole unit. Lesson 18 (Porpoises) and Lesson 20 (Snake poison) have the largest proportion of low-frequency words: 4.32% and 2.24% respectively. Note in Table 6 the proportion of high-frequency vocabulary compared to mid- and low-frequency vocabulary overall.

Table 6.

Vocabulary profile of each lesson in Unit 3, Book 4.

Lesson	17	18	19	20	21	22	23	24
Coverage	%	%	%	%	%	%	%	%
High-frequency words	84.32	85.14	90.96	89.33	80.44	89.91	89.12	87.22
Mid-frequency words	11.28	10.26	7.94	8.14	13.25	10.1	9.06	11.18
Low-frequency words	1.87	4.32	0.54	2.24	0.96	0	1.51	0.64
Proper nouns, compound words, etc.	2.51	0.27	0.55	0.28	5.37	0	0.3	0.96

When checking the occurrence of the low-frequency words in science-oriented lessons in Book 4, we found that many of these words only occurred once – for example, viperine (10th 1000), haemolytic (16th 1000) and strychnine (17th 1000). Some sentences may contain two or more low-frequency words. Below is the extract from Lesson 20 in Book 4, in which the BNC/COCA frequency first 1000 words are in normal type, and the frequency level for all other words is in brackets before each word. For example, {2} nerve indicates that nerve is in the second 1000 BNC/COCA list and {16}haemolytic means this word is in the 16th 1000 list (and is a low-frequency word).

The makers of the {2}nerve {2}poison {2}include the {19}mambas and the {8}cobras and their {7}venom is called {18}neurotoxic. {10}Vipers ({13}adders) and {33}rattlesnakes {3}manufacture the blood {2}poison, which is known as {16}haemolytic. Both {2}poisons are {2}unpleasant, but by far the more {2}unpleasant is the blood {2}poison. It is said that the {2}nerve {2}poison is the more {4}primitive of the two, that the blood {2}poison is, so to speak, a newer {2}product from an {2}improved {3}formula. Be that as it may, the {2}nerve {2}poison does its business with man far more quickly than the blood {2}poison. This, however, means nothing. {2}snakes did not {3}acquire their {2}poison for use against man but for use against {4}prey such as {2}rats and {2}mice, and the {2}effects on these of {10}viperine {2}poison is almost {2}immediate. (Alexander and He, 1997: 115)

Seven words in the example belong to low-frequency word lists, six of which occur in the first two sentences (i.e. 6/31 words in total). The density of the low-frequency words means that learners may have difficulty in fully understanding the meaning of this sentence since it is hard to guess meaning from the context when the unknown words only occur once (Nation, 2013). Planning for this vocabulary is important because learners who struggle will need support from peers and learning resources such as dictionaries to help with this lexis. The research carried out by Coxhead et al. (2010) on secondary science books also found a drop of the high-frequency vocabulary in the book from the first to the third year of high school in New Zealand, and a great increase of vocabulary at a lower-frequency level in textbooks for senior-level students. This point about low-frequency vocabulary is particularly important when we consider that the learning in today’s class should support tomorrow’s learning (see Macalister, 2011). The main question is: are these low-frequency words likely to be any use for tomorrow’s language learning? Having looked into high- and low-frequency vocabulary, the focus in the next section is on mid-frequency lexis in the textbooks.

To What Extent Do the Books Facilitate the Learning of Mid-Frequency Vocabulary?

Table 7 shows again that high-frequency words (e.g. topic, year, wide) cover more than 85% of Book 3 and Book 4, and mid-frequency (e.g. amber, despair, repetitive) lists cover fewer than 10%. As the coverage of high-frequency words drops, the proportion of mid-frequency coverage increases (i.e. from 7% in Book 3 to 9.84% in Book 4). The same pattern can also be seen from Unit 1 to Unit 3 in Book 3, from 5.92% in Unit 1 to 6.36% in Unit 2 and then to 8% in Unit 3 (see Table 3). It could be that the texts are longer in Unit 3 than in the other units, thereby giving more opportunity for mid-frequency vocabulary to occur. The mid-frequency words in Book 3 come mostly from the third and fourth 1000 word families, and then it extends to the fifth 1000 word families in Book 4 (all coverage is above 1%) (see Table 2). This means that the two textbooks provide opportunities for learners to develop their vocabulary knowledge to less-frequent vocabulary by exposing them to more mid-frequency words.

Table 7.

Frequency and supplementary list profile of Book 3 and Book 4.

	Book 3 (%)	Book 4 (%)
High-frequency words	90.1	87.86
Mid-frequency words	7	9.84
Low-frequency words	0.45	0.71
Supplementary lists	2.44	1.6
Total	99.99	100

Note: Book 3 total is subject to rounding.

Nation (2013) suggests that non-native English speakers should learn the third to the fifth 1000 word families in mid-frequency words explicitly after acquiring high-frequency words. It seems that the mid-frequency words in the Chinese textbooks are consistent with this guidance.

That said, while the textbook exposes learners to more mid-frequency words in the later stages of learning, these words lack repetition: 629 out of a total of 948 (66.35%) mid-frequency word families in Book 3 only occur once, as do 707 out of the total 1057 (66.89%) mid-frequency word families in Book 4. These figures are higher than those reported by Sun and Dang (2020), who noted that 17% of the vocabulary outside the first 3000 word families occurred only once in the Yilin series. Nation (2013) reminds us that it is common to find many words occurring only once in textbooks and in simplified texts (see Zipf, 1965). In our study, for instance, in Lesson 58 of Book 3 in the sentence ‘thereby proving that the burglars had either used skeleton keys or entered over the balcony’, balcony (fifth 1000) occurred only once in the passage and also only once in all the passages in the whole book. Taken together, both the current study and Sun and Dang’s (2020) research show that gaining enough repetition to learn words through the textbooks might be challenging for learners.

In order to see whether the book provides any other chance for learners to meet the mid-frequency words repeatedly within the book, we checked the accompanying glossaries and exercises in Lesson 58 in Book 3. We found that 16 mid-frequency words were listed in the glossary, but only three of them occurred in exercises, which are discarded (fourth 1000), burglars (fifth 1000) and reprimand (eighth 1000). The opportunities to meet these mid-frequency words often within the book are quite limited and it may not be enough to help learners enhance the memory of these words. Webb’s (2007) study indicates that if learners encounter unknown words at least 10 times in reading, the possibility to learn these words increases greatly. Reading these texts alone might be insufficient for vocabulary gains to occur.

Implications for Textbook Writers and Teachers

There are two main implications for textbook writers and teachers from this study, and while they relate specifically to the two NCE textbooks, they can apply to other EFL textbooks as well. Firstly, textbook writers and teachers could analyse the vocabulary profile and load of the texts using Range to make adjustments to the reading texts before publication in the case of textbook writers and before using texts in class in the case of the teachers. If the vocabulary load for a text is too high and far beyond target learners’ current vocabulary knowledge level or vocabulary size, it may lead to learners’ incomprehension of the texts and learners will need extra support (Webb and Nation, 2008). The amount of low-frequency words would need to be controlled to make vocabulary learning more efficient. One way to do this would be to carefully consider the content of reading texts. For example, since science-related texts require more vocabulary knowledge than humanities-related texts, the proportion of science-related texts in textbooks could be reduced. Another approach would be to replace lower-frequency words with the higher-frequency ones if these words are not essential for language learning. The repetitions of target mid-frequency words could be increased in the textbooks.

Teachers could test the vocabulary size of their students to find out whether they have enough vocabulary to cope with the textbooks – for example, using the Webb et al. (2017) UVLT. Teachers can also ensure that learners reread texts to increase exposure to vocabulary (see Macalister, 2014), and use read and retell activities to encourage learners to use vocabulary from the texts in context, including mid-frequency words. They could also introduce some strategies inside or outside the classroom to guarantee that learners will have chances to meet the target words repeatedly to enhance their learning and help learners develop the fluency in using these words (Nation, 2013; Webb and Nation, 2008). For example, learners can be trained to make and use vocabulary cards (Nation, 2013). Students could be asked to reread texts from earlier lessons in class as fluency practice and to provide more opportunities to encounter target vocabulary again. Opportunities to use the target words in speaking and writing are also important for building vocabulary knowledge, memory for words and fluency.

Limitations and Future Research

There are several limitations in the present study. Firstly, it only contains two books and cannot present a comprehensive overview of the vocabulary profile of the whole series of NCE. Another limitation is that the current research cannot reveal a comprehensive profile of how the books scaffold vocabulary learning since the analysis mainly focused on the texts in the two books, excluding the glossaries and exercises following the texts. Furthermore, the current research did not investigate the vocabulary knowledge of learners who have studied the textbooks or are about to do so, which is also important to determine whether the textbook is at a suitable level for their learners (Webb and Nation, 2008).

Future research could include the full range of NCE book series and the exercises following the texts to gain a more comprehensive view of how the textbook writers have designed the books for vocabulary learning. Researchers can also examine the actual vocabulary size of the book users and compare the results with the lexical requirement of the book to see the extent to which they do, or do not, match. Finally, investigating the multiword units of textbooks would be a very useful further step.

Conclusion

The current research has focused on the vocabulary of the texts in Book 3 and Book 4 of the NCE textbook series. We used the Range program with the 25 1000 word-family lists of BNC/COCA. The result shows that vocabulary load gradually increases from Book 3 to Book 4, and also within Book 3 across different units. The analysis on the topics of one unit in Book 4 suggests that science-based texts require more vocabulary than the others do.

Footnotes

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Lu Yang

Averil Coxhead

References

Alexander

(1997) New Concept English. Beijing: Foreign Language Teaching and Research Press.

Alsaif

Milton

(2012) Vocabulary input from school textbooks as a potential contributor to the small vocabulary uptake gained by English as a foreign language learners in Saudi Arabia. The Language Learning Journal 40: 21–33.

Cobb

(2019) Range for Texts. 3rd ed. Available at: https://www.lextutor.ca/cgi-bin/range/texts/index.pl (accessed 1 June 2020).

Coxhead

(2000) A new academic word list. TESOL Quarterly 34: 213–238.

Coxhead

(2017) Academic vocabulary in teacher talk: Challenges and opportunities for pedagogy. Oslo Studies in Language 9: 29–44.

Coxhead

Boutorwick

(2018) Longitudinal vocabulary development in an EMI international school context: Learners and texts in EAL, maths, and science. TESOL Quarterly 52: 588–610.

Coxhead

Stevens

Tinkle

(2010) Why might secondary science textbooks be difficult to read? New Zealand Studies in Applied Linguistics 16: 37–52.

Heatley

Nation

ISP

Coxhead

(2002) Range: A program for the analysis of vocabulary in texts [software]. Available at https://www.wgtn.ac.nz/lals/resources/paul-nations-resources/vocabulary-analysis-programs (accessed 13 March 2020).

Hsu

(2014) Measuring the vocabulary load of engineering textbooks for EFL undergraduates. English for Specific Purposes 33: 54–65.

10.

Nation

(2000) Unknown vocabulary density and reading comprehension. Reading in a Foreign Language 13: 403–430.

11.

Hutchinson

Torres

(1994) The textbook as agent of change. ELT Journal 48: 315–328.

12.

Laufer

(1987) What percentage of text-lexis is essential for comprehension. In: Laurén

Nordman

(eds) Special Language: From Humans Thinking to Thinking Machines. University of Vaasa: Multilingual Matters Ltd, 316–323.

13.

Laufer

(2013) Lexical thresholds for reading comprehension: What they are and how they can be used for teaching purposes. TESOL Quarterly 47: 867–872.

14.

Laufer

Ravenhorst-Kalovski

(2010) Lexical threshold revisited: Lexical text coverage, learners’ vocabulary size and reading comprehension. Reading in a Foreign Language 22: 15–30.

15.

(2013) An analysis of the vocabulary in New Concept English based on range. Journal of Qiannan Normal University for Nationalities 1: 60–62.

16.

Macalister

(2011) Today’s teaching, tomorrow’s text: Exploring the teaching of reading. ELT Journal 65(2): 161–169.

17.

Macalister

(2014) Teaching reading: Research into practice. Language Teaching 47(3): 387–397.

18.

Macalister

(2016) Applying language learning principles to coursebooks. In: Renandya

Widodo

(eds) English Language Teaching Today: Linking Theory and Practice. English Language Education, volume 5. Cham: Springer, 41–51.

19.

Matsuoka

Hirsh

(2010) Vocabulary learning through reading: Does an ELT course book provide good opportunities? Reading in a Foreign Language 22: 56–70.

20.

Nation

ISP

(2006) How large a vocabulary is needed for reading and listening? The Canadian Modern Language Review 63: 59–82.

21.

Nation

ISP

(2013) Learning Vocabulary in Another Language. Cambridge: Cambridge University Press.

22.

Nation

ISP

(2014) How much input do you need to learn the most frequent 9,000 words? Reading in a Foreign Language 26: 1–16.

23.

Nation

ISP

(2016) Making and Using Word Lists for Language Learning and Testing. Amsterdam: John Benjamins Publishing Company.

24.

Nation

ISP

(2017) The BNC/COCA Level 6 word family lists (version 1.0.0) [Data file]. Available at: http://www.victoria.ac.nz/lals/staff/paul-nation.aspx (accessed 13 March 2020).

25.

Nguyen

C-D

(2020) Lexical features of reading passages in English-language textbooks for Vietnamese high-school students: Do they foster both content and vocabulary gain? RELC Journal. Epub ahead of print 3 April 2020. DOI: 10.1177/0033688219895045

26.

Song

(2019) Exploring the cultural content in Chinese ELT textbooks from intercultural perspectives. The Journal of Asia TEFL 16: 267–278.

27.

Sun

Dang

TNY

(2020) Vocabulary in high-school EFL textbooks: Texts and learner knowledge. System 93: 102279. DOI: 10.1016/j.system.2020.102279

28.

Vidal

(2011) A comparison of the effects of reading and listening on incidental vocabulary acquisition. Language Learning 61: 219–258.

29.

Wang

(2018) Analysis of New Concept English from the perspective of cross-cultural communication–A case study in Book 2 (critical essay). Theory and Practice in Language Studies 8(10): 1270–1277.

30.

Webb

(2007) The effects of repetition on vocabulary knowledge. Applied Linguistics 28: 46–65.

31.

Webb

Chang

AC-S

(2012) Second language vocabulary growth. RELC Journal 43: 113–126.

32.

Webb

Nation

(2008) Evaluating the vocabulary load of written text. The TESOLANZ Journal 16: 1–10.

33.

Webb

Sasao

Ballance

(2017) The updated Vocabulary Levels Test: Developing and validating two new forms of the VLT. ITL – International Journal of Applied Linguistics 168: 33–69.

34.

Wei

(2012) New concept of English: Advantage, enlightenment and application in ethnic minority college students’ English teaching. Journal of Ningbo Polytechnic 16: 65–68.

35.

West

(1953) A General Service List of English Words: With Semantic Frequencies and a Supplementary Word-List for the Writing of Popular Science and Technology. Harlow, Essex: Longman.

36.

Zipf

(1965) The Psycho-Biology of Language: An Introduction to Dynamic Philology. Cambridge, MA: MIT Press.