Narrow Viewing: The Vocabulary in Related Television Programs

Narrow Viewing: The Vocabulary in Related Television Programs MICHAEL P. H. RODGERS Fukuoka University Fukuoka, Japan STUART WEBB Victoria University of Wellington Wellington, New Zealand In this study, the scripts of 288 television episodes were analyzed to determine the extent to which vocabulary reoccurs in related and unrelated television programs, and the potential for incidental vocabulary learning through watching one season (approximately 24 episodes) of television programs. The scripts consisted of 1,330,268 running words and had a total running time of 203 hours and 49 minutes with a mean running time of 42 minutes. The vocabulary from a single season of six individual television programs (142 episodes) was compared with six sets of random television programs (146 episodes). The results indicated that, when there are an equivalent number of running words, related television programs are likely to contain fewer word families than unrelated programs. The findings also indicated that word families from the 4,000–14,000 levels were more likely to reoccur in a complete season of a television program than in random television programs. The percentage of lowfrequency word families encountered 10 or more times was higher, and the percentage of word families encountered once was fewer in all six programs than in the random television programs. doi: 10.5054/tq.2011.268062 n the English as a foreign language (EFL) setting, there has long been a challenge to provide learners with authentic input. The recent growth of extensive reading programs has sought to remedy this by having language learners consume a large quantity of books. These reading programs do little, however, to satisfy the need for aural input and, as a result, there has been recent research into authentic listening in an effort to provide suitable sources of aural input to EFL learners. Authentic aural texts as well as adapted and graded texts provide learners with opportunities to improve listening skills, build vocabulary, I TESOL QUARTERLY Vol. 45, No. 4, December 2011 689 and, like extensive reading programs, prove to be motivating for language learners (Vandergrift, 2007). A source of input that has the same authentic and motivating properties as those in reading and listening programs is television. Television, however, has not been widely researched as a source of input for EFL learners, despite its indisputable worldwide popularity. In a recent Organization for Economic Cooperation and Development (OECD, 2007) report on communications in its member countries, the television viewing habits of the world’s most industrialized countries were shown to be on the increase. In the United States, the leading viewer of television, households watched an average of 8.18 hours a day, whereas the lowest average daily viewing was found in Sweden at 2.43 hours a day per household (OECD, 2007). Adding in recorded television and the recently popularized mediums of Internet and mobile phone video, the amount of television watched increases, especially for younger audiences (Holmes, 2008). It is safe to say that television is a popular source of information and entertainment. This study identifies the television programs for use in narrow viewing and the books used in extensive reading as having the positive attribute of authenticity. The television programs considered, though, are of a scripted nature, which means that they are not made up of spontaneously and naturally occurring dialogue, and the graded readers, often part of extensive reading programs, have vocabulary loads modified to the level of the reader. Because authentic language is often considered to be language that is not artificial (Crawford, 2002) or where no concessions are made to foreign speakers (Harmer, 2007), this may mean that both television and graded readers might not be deemed authentic by some. The television programs analyzed here, however, are not made with foreign language speakers in mind but rather for English speaking U.S. audiences. They are, in that way, authentically representative of the input English speakers regularly come into contact with. The dialogue is also modelled on natural authentic language, and its linguistic features have been shown to share the same core features as natural conversation (Quaglio, 2009). Graded readers also contain authentic lexical features (Allan, 2009) and present authentic representations of the modality and Television is described differently in American English and British English. A television program (United States) or programme (United Kingdom) refers to something that people watch on television, and is often, and in this case, part of a periodically recurring run of episodes. A short succession of episodes, lasting usually less than a year, is referred to by North Americans as a season and in the United Kingdom and some other parts of the world as a series. Some confusion can occur because, for North Americans the term series refers to all episodes of a particular program across time (Grey’s Anatomy is the name of the series, and the episodes broadcast from fall 2009 to spring 2010 are part of the sixth season). Because the television episodes analyzed in this study are of U.S. origin, the U.S. English nomenclature is used throughout. 690 TESOL QUARTERLY medium of the books on which they are based. In essence, authenticity is a multifaceted concept in which there are degrees of authenticity as well as situational authenticity. The statistics above reflect first language (L1) viewing of television, but there are no data indicating the amount of time viewers watch second language (L2) television. In the EFL environment, L2 television may not be as accessible as in an English as a second language (ESL) situation, where language learners have an abundant supply of L2 programs to choose from. However, with the rise of DVDs and their ability to easily provide full seasons of television programs, the increasing availability of satellite television and its abundance of English language channels, and the recent innovation of Internet pay-per-view television, inaccessibility should be becoming less of a factor for language learners in EFL situations. This increased availability of English language television could be a particularly valuable source of L2 input, particularly in the EFL setting where aural input is not as readily accessible as in the ESL context. In addition to providing sought after aural input, increased access to input in the form of television may also be a motivating factor. It has been found that EFL learners are motivated to study English using films (Chapple & Curtis, 2000; Colwell & Ipince Braschi, 2006; King, 2002), and because foreign language television has become as comparably available as foreign language films, it would stand to reason that EFL learners would be similarly motivated to learn from television programs. Television may also be a source of incidental vocabulary learning if learners acquire vocabulary from television in the same way they do from reading. Extensive reading done by language learners as a supplement to their language classes has been shown to lead to gains in vocabulary (Horst, 2005; Schmitt, 2008; Waring & Nation, 2004). An extensive reading program exposes learners to texts that are at a level which allows learners to read and learn with ease. Texts are available at a variety of levels suited to the ability levels of learners. Level appropriacy is important for incidental vocabulary learning, eases possible frustration that can result from too difficult a text, and allows for sufficient lexical coverage to help deal with unknown vocabulary. Learners are given the freedom to read at their pace and to select texts that they find engaging and interesting. For an extensive reading program to be effective, it needs to engage the learners’ attention and encourage reading in quantities large enough to allow for repeated encounters of new vocabulary. Vocabulary learning can only come about if the extensive reading program is part of a substantial and sustained program. To supplement incidental learning, explicit vocabulary teaching in a variety of formats is recommended (Waring & Nation, 2004). Generally, in these programs, learners are free to choose books suitable to their VOCABULARY IN RELATED TELEVISION PROGRAMS 691 language ability and are encouraged to read as many as possible (Horst, 2005). Should television episodes be used in a similar manner to books in an extensive reading program, the episodes available to the learners would need to satisfy the aforementioned requirements. Although serial book series like J. K. Rowling’s seven books featuring Harry Potter are common in children’s literature, novels like Clive Cussler’s 20 books featuring Dirk Pitt are common in certain genres of adult fiction, and serialized graded readers are available for extensive reading, they do not occur with the frequency that serials do in television. Television programs, by and large, do not appear in isolation but as part of a series made up of seasons, which are in turn made up of individual episodes. Television seasons vary in length and scope, but normally there is a consistent story arc or arcs linking the episodes in a season. Of course, television viewers do not always watch every episode of a program in their intended order, often failing to watch some episodes altogether. As part of a language program designed to expose language learners to authentic aural input, it may be better to take a more systematic approach to viewing. This approach may involve watching complete seasons of television in sequential order. The serial nature of a television series with its reoccurring characters, themes, and story lines allows for a viewer to build familiarity with the program, in contrast to watching random episodes of different television programs, each having different plots and characters. Also, the reoccurring attributes of a series may make certain vocabulary reoccur more frequently. Increased exposure to this vocabulary throughout a television season could have the potential to reduce vocabulary load and increase vocabulary acquisition. With these ideas in mind, the aim of this study is to determine the potential for vocabulary learning through watching related and unrelated television dramas and to establish the lexical composition of seasons of these television programs originally produced for broadcast in the United States. BACKGROUND Research indicates that viewers can incidentally learn L1 and L2 words through watching television. Rice and Woodsmall (1988) found that children who watched a 12-minute cartoon with a narration which included 20 unknown L1 target words had higher scores on a test measuring knowledge of those items than children who watched the cartoon with a narration which did not include the target words. Rice and Woodsmall suggest that repeated encounters with target words in the cartoon contributed to learning. Oetting, Rice, and Swank (1995) replicated Rice and Woodsmall’s study with normally developed and 692 TESOL QUARTERLY specific-language-impaired children. The results supported the earlier findings; both groups of learners demonstrated greater gains in vocabulary knowledge on the test than a control group. The results also showed that the oldest learners with normal language development had significantly larger gains than the specific-language-impaired learners. They attributed the amount of learning to the age and aptitude of the learners. Together, the two studies indicate that young viewers are able to learn unknown L1 words through watching television. L2 studies have repeatedly found that viewers can incidentally learn L2 words through watching L2 television programs (d’Ydewalle & Pavakanun, 1997; d’Ydewalle & Van de Poel, 1999; Koolstra & Beentjes, 1999; Neuman & Koskinen, 1992; Pavakanun & d’Ydewalle, 1992). Several useful findings have emerged from the L2 studies. First, research indicates that watching L2 programs with subtitles leads to greater incidental vocabulary learning than watching without subtitles (d’Ydewalle & Pavakanun, 1995; Koolstra & Beentjes, 1999; Neuman & Koskinen, 1992; Pavakanun & d’Ydewalle, 1992). Second, the extent to which learners have watched L2 television programs may also have an effect on vocabulary learning. Koolstra and Beentjes (1999) found that learners who had previously watched more L2 television programs had higher incidental vocabulary learning gains than those who watched fewer L2 programs. Third, research indicates that incidental learning gains through watching television may be as substantial as through reading. Neuman and Koskinen (1992) found that watching television with subtitles led to greater vocabulary learning than reading the script of the program. Scores were also higher for viewers who watched the program without subtitles than for those who read the transcripts; however, the statistical comparison between the groups was not reported. If incidental learning gains though watching television are as significant as they are through reading, television may be a valuable resource for language learning. Taken as a whole, the L1 and L2 studies justify the use of L2 television viewing in language learning. In an analysis of the vocabulary in 88 English language television programs, Webb and Rodgers (2009b) suggest that a vocabulary size of 3,000 word families plus knowledge of proper nouns and marginal words may be sufficient for comprehension. This provides a target vocabulary size from which point it may be useful to learn through watching television. This vocabulary size provided 95% coverage of television programs. This benchmark for coverage was derived from reviewing the reading and listening comprehension research, because there are no L1 or L2 studies examining the relationship between coverage and television comprehension. Studies have varied in their estimates of adequate comprehension. For reading, the following estimates have been made: 95% for reasonable comprehension of a text VOCABULARY IN RELATED TELEVISION PROGRAMS 693 (Laufer, 1989), 98% for reading for pleasure (Hirsh & Nation, 1992), 98% for adequate unassisted reading comprehension (Hu & Nation, 2000), and 98% for ideal comprehension of written text (Nation, 2006). For listening, the following estimates have been made: below 95% for short texts (Bonk, 2000), greater than 98% for spoken language (Nation, 2006), and 99% for adequate comprehension when listening to graded readers (Donkaewbua, 2007). Although 98% coverage of television was considered ideal in Webb and Rodgers’ (2009b) study, 95% coverage was considered to be adequate for comprehension, on the basis of research indicating that the combination of visual and aural input might make comprehension of television programs easier than listening comprehension (Hanley, Herron, & Cole, 1995; Mueller, 1980). Webb and Rodgers (2009b) also suggest that, if learners with a vocabulary size of 3,000 word families plus knowledge of proper nouns and marginal words watch an hour of television each day, it is likely to result in significant incidental vocabulary learning. The findings are useful, because they highlight an approach to incidental vocabulary learning which is rarely discussed in the literature. Webb and Rodgers go on to suggest that one way in which comprehension and vocabulary learning may be increased is to watch programs in which topics and story lines may be similar. Corpus-driven research on narrow reading indicates that it may be more effective to learn vocabulary in related text than in unrelated text. Two studies looked at the vocabulary in related newspaper stories versus unrelated stories. Hwang and Nation (1989) analyzed the vocabulary load of running stories, that is a story and its subsequent follow-up stories, versus unrelated stories from newspapers. The analysis was performed on 20 sequences of four related stories and 20 groups of four unrelated stories and looked at the effect that reading related stories had on repetitions of words outside the 2,000 most frequent words. Selecting related stories showed a major effect on repetitions outside the 2,000 most frequent words, which reduced the vocabulary load and provided better conditions for acquiring low-frequency words. Similarly, Schmitt and Carter (2000) analyzed the vocabulary from a series of nine newspaper stories related to the death of Princess Diana and compared it with the vocabulary in nine unrelated stories from the same newspapers. Both sets of articles contained the same number of running words (7,843). Schmitt and Carter found that the Diana-related stories contained 156 fewer types and that words were repeated more often than in the unrelated stories. The researchers concluded that the reading of related stories lowers the lexical load for L2 learners, which could allow for earlier contact with authentic reading materials. Sutarsyah, Nation, and Kennedy (1994) compared the vocabulary in an Economics text consisting of 295,294 running words and a corpus of 160 694 TESOL QUARTERLY selections from random academic texts. The 160 short academic texts were each about 2,000 words long and together totalled 311,768 running words. The results showed that there were 9,469 word types and 5,438 word families which occurred in the Economics text, and 21,399 word types and 12,744 word families which occurred in the random academic texts. The analysis also showed that there were more encounters with the most frequently occurring words in the Economics text than in the random texts. The findings indicated that readers are likely to need a larger vocabulary size to understand random texts than texts about a single topic. The results also indicated that there may be more encounters with technical words in text focused on a single topic than in random texts. In a study focusing on texts used in extensive reading, Gardner (2004) analyzed the lexical differences between expository and narrative texts used by fifth grade elementary students. Twenty-eight narrative and 28 expository texts containing 1,443,336 tokens were analyzed. In addition to being categorized by discourse mode, the texts were also divided into thematically related and thematically unrelated. Although the bulk of the analysis centered on the suitability of expository versus narrative texts for incidental vocabulary learning in a reading program, Gardner maintains that choosing theme-related texts can have a profound effect on the type of words learnt, the number of encounters with words, and the amount of prior vocabulary knowledge needed to learn new words during extensive reading. The studies on narrow reading have important implications for language learning. They indicate that texts which relate to the same topic are likely to have a smaller vocabulary load than random texts, because the number of different words encountered is likely to be less. Thus, learners need to know fewer words to understand a single topic or related topics than to understand unrelated topics. The research indicates that teachers may wish to organize courses around particular topics to reduce the vocabulary demands of learning and to allow learners to focus their attention on skill development. The results also provide some evidence that texts which focus on the same topic are likely to have more encounters with the technical vocabulary of that topic than in unrelated texts. This is an important finding, because research has consistently shown that the more often unknown words are encountered in context, the more likely they are to be learned (Horst, Cobb, & Meara, 1998; Jenkins, Stein, & Wysocki, 1984; Rott, 1999; Saragi, Nation, & Meister, 1978; Waring & Takaki, 2003; Webb, 2007). The corpus-driven findings suggest that vocabulary is more likely to be learned incidentally in written discourse which focuses on a single topic rather than multiple topics.1 1 For a more detailed summary of the literature on learner comprehension, vocabulary demands, vocabulary acquisition and retention, and the role of subtitles in television, see Webb and Rodgers (2009b). VOCABULARY IN RELATED TELEVISION PROGRAMS 695 The research presented above highlights three points. First, in both L1 and L2 settings, vocabulary can be incidentally learned through watching television. However, we do not know to what extent vocabulary learning may occur through watching television. Second, narrow reading studies indicate that reading texts with similar topics and story lines may facilitate vocabulary learning and assist comprehension. Television programs, in the form of serialized episodes comprising seasons, have characteristics well suited to take advantage of these points. Previous research, however, has not focused on the potential for vocabulary learning through watching seasons of television. Third, although Webb and Rodgers’ (2009b) study indicated that 3,000 word families may be sufficient for comprehension of television programs, there is not any research on viewing strategies that learners can use to help boost comprehension. Comparing television programs with related and unrelated content may indicate that fewer types and word families are encountered in different episodes of the same program. This would indicate that watching different episodes of the same program is an effective strategy for increasing comprehension. Taking into consideration these three points, the aim of the present study is to investigate the lexical composition of series of television programs and the potential for vocabulary learning through watching television programs with related and unrelated content. The scripts of 288 television programs (142 from programs with related content and 146 random television programs) were analyzed using RANGE (Nation & Heatley, 2002) to determine the percentage of words (coverage) at each 1,000-word level using Nation’s (2004, 2006) British National Corpus (BNC) word lists, and the number of encounters with word families. The analysis may provide direction towards an effective method of using television for language learning, by demonstrating a method of reducing the lexical burden of understanding television programs and optimizing vocabulary learning. RESEARCH QUESTIONS The present study seeks to address the following four research questions: 1. 2. 3. 4. 696 Are there fewer word types and families in related television programs than in unrelated programs? Does vocabulary reoccur more often in different episodes of the same television program than in random television programs? What is the potential for vocabulary learning through watching a full season (approximately 24 episodes) of one television program or an equivalent number of random television programs? What is the vocabulary size necessary to reach 95% coverage of the different television programs? TESOL QUARTERLY METHOD Materials The scripts of 288 television episodes were downloaded from the Internet and analyzed in this study. The episodes had a total running time of 203 hours and 49 minutes and a mean running time of 42 minutes. The running time for all but seven of the programs was 1 hour including commercials; however, commercials were not included in the analysis. All the television programs (for the purposes of this study, the terms television program and television series are used synonymously) used in the study were produced for broadcast on U.S. television stations. It is important to note that Internet-available television scripts do not always reflect dialogue with total accuracy. However, the scripts should provide a reliable assessment of the vocabulary in television programs. To determine the extent to which vocabulary reoccurs within a related program over time, 142 episodes from six different television programs and 146 episodes from random television programs were analyzed in the study. The six different television series were 24, Alias, Crossing Jordan, CSI, Grey’s Anatomy, and House, and the episodes for each of the programs comprised a single season. A season of a television series consists of the number of episodes broadcast in the same programming year, whereas an episode is a single instance of a series. The programs were selected according to their availability, running time, date when first aired, and place of origin. All of the episodes from these programs first aired between 2001 and 2006 and had an average running time of 43 minutes. The genre of all of these programs was classified as drama. Although drama is a very broad classification, it is the common way to categorize a series in which episodes last 1 hour (including commercials). The 146 randomly selected television episodes were used for comparison with the episodes in the six programs. The random television programs were also all of U.S. origin, originally aired between 1963 and 2009, and had running times of approximately 22 minutes (broadcast as a 30-minute program including commercials) or 44 minutes (broadcast as a 60-minute program). Six sets of different random television programs were created for comparison with the six series. Programs from a wide range of genres were used in each set of comparison programs, and no set contained multiple episodes of the same program. It should be noted that this does not ensure that all of the programs had completely unrelated content, because two shows from the same genre may have been included. However, the degree of VOCABULARY IN RELATED TELEVISION PROGRAMS 697 overlap in content was likely to be considerably less than in the episodes in one season of a single program. Each set of random television programs was made up of the same number of running words as one of the six seasons. To ensure that the sets had the same number of running words as each of the six programs, one program in each of the sets did not include all of its running words. The sets of random television programs were created with an aim of including as many running words as possible from the final program in the set. Comparing related and unrelated programs with an equal number of running words was necessary to determine the extent to which vocabulary reoccurs in television programs. All of the words which could not be heard in the programs, such as stage directions, setting features, and speakers’ names, were removed from the scripts. Only the spoken words from the programs were analyzed. Contractions and hyphenated words were changed to conform with the spellings used in Nation’s (2006) BNC word lists. Contractions accounted for 0.39% of the tokens in the study (0.28% of the tokens in the six programs and 0.49% of the tokens from the random television programs). For example, shoulda, d’you, c’mere, and ‘nother were changed to should have, do you, come here, and another, respectively. If the spellings were not changed, these items would have been classified as being less frequent than the most frequent 14,000 word families. However, it is important to note that knowing the changed spellings does not ensure that the original spellings will also be known. For example, learners may know come here, and another but they might not recognize c’mere, and ‘nother. Webb and Rodgers (2009a, 2009b) suggest the percentage of contractions may be a factor which may influence comprehension and vocabulary learning from spoken discourse. However, at present there is not any research which has investigated the effect that the percentage of contractions may have on comprehension and incidental learning. The relatively small percentage of items that were changed to conform to the spellings in the BNC lists would suggest that the effects of contractions, at least as indicated by the scripts, would be quite small. Analysis RANGE (Nation & Heatley, 2002) was used to analyze the modified scripts. RANGE is a computer program which lists the words that occur in a text according to their frequency. Nation’s 14 lists of 1,000-word lists were used with the RANGE software to determine the 1,000-word level (1,000–14,000) at which the words in the programs occurred. The 14 lists were based on the frequency and range of occurrence of words in the BNC. The word families in the lists were categorized as Level 6 698 TESOL QUARTERLY according to Bauer and Nation’s (1993) classification of word families. Level 6 word families include inflections (a variation in the form of a word to express grammatical categories like the plural -s and past tense -ed) and more than 80 derivational affixes (an affix that creates a new word, for example pain becomes painful). All word stems were free forms, not bound forms (free forms can stand alone, whereas bound forms cannot occur as separate words on their own; for example, the word reuse consists of the affix re- and the free form use, whereas the word repeat contains the bound form peat which cannot stand on its own). Less frequent words than the most frequent 14,000 word families were classified by the RANGE program as Proper Nouns, Marginal Words (interjections, exclamations, and hesitation procedures, for example, oh, uh, mmm, and ah), and Not in the Lists. Although the proper nouns list has over 13,000 entries, this will rarely account for all of the proper nouns in an analysis of a corpus, and a large number of proper nouns will be classified by RANGE as Not in the Lists (words less frequent than the most frequent 14,000 word families). Proper nouns found in the Not in the Lists were reclassified as proper nouns and added to the proper nouns totals. The RANGE program and the word lists can be downloaded from Paul Nation’s Web site (Nation, 2011). For more information about the word lists, see Nation (2004, 2006). It should be noted that several words such as bartender, blog, cheerleader, donut, email, fiancé, merci, momma, roommate, and t-shirt, which were classified as Not in the Lists, are likely to be known by learners with a vocabulary size of 3,000 word families. Words like these being classified as Not in the Lists can more than likely be seen as a function of the BNC being made up primarily of formal written texts as well as consisting of a relatively low percentage (10%) of spoken language (Nation, 2006). A learner’s knowledge of these low-frequency words may be a result of many factors; for example, if the English language learner’s first language were Japanese, then the word bartender (bātendā) would be known as an English loanword in Japanese (Kamiya, 1994). This suggests that the coverage figures may be slightly conservative. Another reason for potentially conservative coverage figures is that this study analyzes American television and the BNC consists primarily of British text (Nation, 2004). This is evident when looking at words such as bartender, cheerleader, donut, and roommate, which were classified as Not in the Lists. These words are likely to be more common in U.S. English than British English and contrast with words that might be considered more typically British such as bloke, fortnight, lorry, and rubbish and are in the 1,000- and 2,000- level BNC word lists. These words would likely occur at less frequent levels in an American corpus. This suggests that coverage of American television could be higher if it were measured with lists developed from a U.S. corpus. Unfortunately, there currently are no VOCABULARY IN RELATED TELEVISION PROGRAMS 699 word frequency lists derived from a U.S. corpus using the same criteria Nation (2006) used to create the BNC lists. Because Webb and Rodgers (2009b) found that knowledge of the most frequent 3,000 word families plus proper nouns and marginal words provided 95% coverage of television programs, the low-frequency vocabulary found in the 4,000- to 14,000-word lists, which may be unknown to L2 learners, was examined to determine the potential for incidental vocabulary learning. The number of times word families from these word lists were encountered may provide some indication of whether unknown vocabulary is more likely to reoccur in related television programs than unrelated programs. RESULTS Table 1 shows the cumulative coverage of the complete corpus, the six seasons combined, and the random television programs. The corpus consisted of 1,330,268 tokens: 665,134 tokens were from the individual programs and 665,134 tokens were from the random television programs. The last two rows of the table show the number of word types and word families. The combined seasons consisted of 19,256 word types from 9,401 word families, and the random television programs consisted of 20,212 types and 10,511 word families. The combined seasons had approximately 5% fewer types and 12% fewer word families than the random television programs. This difference indicates that the vocabulary load of related television programs is likely to be lower than watching random programs, because a greater number of different words may be found in unrelated programs. Table 1 also shows that a vocabulary size of 3,000 word families plus proper nouns and marginal words provided over 95% coverage of all programs, the combined six seasons, and the random television programs. This finding supports Webb and Rodgers’ (2009b) earlier findings indicating that a vocabulary size of the most frequent 3,000 word families plus proper nouns is sufficient to reach 95% coverage of television programs. Table 2 presents the cumulative coverage of the six individual seasons and the random television programs with which they were matched, as well as the number of tokens, types, and families in each. The number of word types found in the individual programs and the matching random television programs was inconsistent. Although Table 1 showed that there were fewer word types found in the combined six seasons than in the random television programs, Table 2 reveals that there were fewer word types in three of the programs (24, Alias, Grey’s Anatomy) and more word types in three of the programs (Crossing Jordan, CSI, House). The 700 TESOL QUARTERLY TABLE 1 Cumulative Coverage Including Proper Nouns and Marginal Words for All Television Programs, Complete Seasons, and Random Matching Seasons Word list All programs Six seasons combined Random television programs 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 11,000 12,000 13,000 14,000 Proper nouns Marginal words Not in the lists Tokens Types Families 89.76 94.02 95.83a 96.89 97.52 97.96 98.22 98.45 98.62 98.78 98.91 99.03 99.10 99.16 2.62 0.71 0.85 1,330,268 28,290 13,592 89.16 93.56 95.45a 96.51 97.21 97.64 97.91 98.17 98.35 98.51 98.66 98.82 98.90 98.97 2.65 0.70 1.03 665,134 19,256 9,401 90.38 94.49 96.22a 97.27 97.83 98.27 98.53 98.73 98.90 99.05 99.16 99.24 99.29 99.33 2.60 0.73 0.67 665,134 20,212 10,511 Note. aReaching 95% coverage. results for word families were more consistent. The final row of Table 2 shows that there were fewer word families for five of the six individual programs than in the matching random television programs. Only Crossing Jordan contained more word families than its matching random television programs. Together the results suggest that watching different episodes of the same television program is likely to reduce the vocabulary load. However, this may vary from program to program. Programs which have the same characters in every episode and contain a story line which continues from episode to episode may be most likely to have a lower vocabulary load. For example, we believe that 24 was the program in which the characters and story line were most closely linked from episode to episode in this study. It contained 42% fewer types and 40% fewer word families than its matching random television programs, which was far less than any of the other programs. Table 3 shows the differences with respect to word types and families for 24 and its matching random television program, as well as for the other programs and their matching random television programs. Tables 4 and 5 show the number and percentage of encounters with low-frequency words (word families from the 4,000- to 14,000-word lists) which occurred in the six programs and the random matching programs. The results were consistent between the six programs and VOCABULARY IN RELATED TELEVISION PROGRAMS 701 702 TABLE 2 Cumulative Coverage Including Proper Nouns and Marginal Words for All Six Programs and Random Matching Programs Word list 24 RTP matching 24 Alias RTP matching Alias Crossing Jordan RTP matching Crossing Jordan CSI RTP matching CSI Grey’s Anatomy RTP matching Grey’s Anatomy House RTP matching House TESOL QUARTERLY 1,000 92.80 90.48 90.59 90.28 89.02 90.55 86.95 90.59 90.10 90.29 86.33 90.14 94.61 94.58 94.34 93.40 94.64 92.35 94.53 93.99 94.40 91.21 94.44 2,000 96.64a a a a a a a a a 96.10 96.02 95.31 96.30 94.66 96.20 95.89 96.13 93.73 96.30a 3,000 97.58 96.37 a a 4,000 98.42 97.28 97.04 97.27 96.40 97.35 96.00 97.20 96.75 97.18 95.03 97.36 5,000 98.89 97.90 97.65 97.82 97.32 97.92 96.68 97.70 97.35 97.76 95.90 97.90 6,000 99.23 98.39 98.01 98.23 97.78 98.29 97.15 98.18 97.70 98.28 96.47 98.29 7,000 99.32 98.63 98.20 98.45 98.06 98.55 97.51 98.50 97.97 98.50 96.83 98.58 8,000 99.40 98.85 98.41 98.62 98.34 98.76 97.83 98.71 98.16 98.71 97.24 98.76 9,000 99.45 99.03 98.57 98.80 98.49 98.93 98.08 98.87 98.34 98.85 97.51 98.94 10,000 99.51 99.16 98.75 99.00 98.62 99.05 98.28 99.04 98.48 98.97 97.77 99.13 11,000 99.58 99.26 98.84 99.10 98.76 99.16 98.45 99.14 98.64 99.08 97.99 99.24 12,000 99.61 99.34 98.96 99.17 99.02 99.23 98.73 99.23 98.73 99.18 98.16 99.32 13,000 99.64 99.40 99.04 99.23 99.10 99.28 98.83 99.28 98.81 99.23 98.24 99.38 14,000 99.66 99.43 99.07 99.26 99.16 99.34 98.95 99.34 98.87 99.26 98.38 99.41 3.61 2.36 3.58 2.62 2.61 2.76 2.90 2.80 2.63 2.54 1.19 2.52 Proper nouns Marginal 0.27 0.67 0.56 0.78 0.65 0.80 0.84 0.83 1.27 0.70 0.44 0.61 words Not in 0.36 0.56 0.93 0.75 0.84 0.67 1.04 0.66 1.15 0.74 1.63 0.60 the lists Tokens 93,038 93,038 89,582 89,582 117,614 117,614 98,044 98,044 138,561 138,561 128,295 128,295 Types 3,936 6,766 6,154 6,830 7,807 7,731 7,752 7,253 7,139 8,733 8,622 8,334 Families 2,506 4,172 3,727 4,234 4,668 4,697 4,536 4,476 4,021 5,197 4,598 5,014 Note. aReaching 95% coverage. RTP 5 random television programs. indicate that there is greater potential for incidental vocabulary learning through watching different episodes of the same program than watching random television programs. The percentage of word families encountered 10 or more times was higher in all six programs than in the matching random television programs. Moreover, the percentage of word families encountered once was less in all six programs than in the matching random television programs. For example, in House 51% of the word families were encountered once and 6% were encountered 10 or more times. In the random television programs matching House, 59% of the word families were encountered once and 2% were encountered 10 or more times. It is important to note that the number of lowfrequency tokens varied between programs and the random television programs, so it is necessary to look at the percentage of encounters rather than the number of encounters in this comparison. The total number of tokens and word families in encounters with 4,000- to 14,000level word families was less consistent than the overall token and word family totals. Three out of the six programs had fewer tokens and word families at these levels, with 24 having by far the greatest difference for tokens (48.5% less) and word families (151.2% less). Table 6 presents the differences, in both real terms and as a percentage, between the six programs and the random matching programs for both families and tokens in encounters with 4,000- to 14,000-level word families. DISCUSSION The present study expanded upon earlier corpus-driven research on television programs (Webb & Rodgers, 2009b) and narrow reading (Gardner, 2004; Hwang & Nation, 1989; Schmitt & Carter, 2000; Sutarsyah et al. 1994) by contrasting the vocabulary in six television programs with six sets of random television programs. This should provide an accurate assessment of the extent to which vocabulary is likely to reoccur within a program, unrelated programs, and the potential for incidental vocabulary learning within those sets. In answer to the first research question, the results indicate that related television programs are likely to have a lower vocabulary load than unrelated television programs. When the six programs were combined, there were 11% fewer word families and 5% fewer word types in the related programs than in the random television programs. In five of the six television programs, the number of word families was less than in the random television programs with which they were matched. The number of word types was less consistent. There were more types in three of the six programs than in the random matching programs. Taken as a whole, the analysis indicated that related programs VOCABULARY IN RELATED TELEVISION PROGRAMS 703 704 TABLE 3 Difference Between Word Types and Word Families in All Six Programs and Random Matching Programs Word list 24 Tokens 93,038 Types 3,936 Difference for 22,830 types Difference for 242 types (%) Families 2,506 Difference for 21,666 types Difference for 240 families (%) RTP matching 24 93,038 6,766 Alias 89,582 6,154 2676 RTP matching Alias 89,582 6,830 3,727 2507 212 TESOL QUARTERLY Note. RTP 5 random television programs. 117,614 7,807 +76 117,614 7,731 +1 210 4,172 Crossing Jordan RTP matching Crossing Jordan 4,234 4,668 229 21 CSI 98,044 7,752 +499 RTP matching CSI 98,044 7,253 +7 4,697 4,536 +60 +1 Grey’s Anatomy 138,561 7,139 21,594 RTP matching Grey’s Anatomy 138,561 8,733 4,021 21,176 223 128,295 8,622 +288 128,295 8,334 +3 218 4,476 House RTP matching House 5,197 4,598 2416 28 5,014 may use fewer word families than unrelated programs but may have a greater number of encounters with members of those word families. The findings support earlier research on narrow reading (Sutarsyah et al., 1994), which found that related texts had fewer types and families than unrelated texts. The difference between the number of word types and families in the related and unrelated television programs was not nearly as large as in Sutarsyah et al.’s (1994) study (over twice as many word families and word types in the unrelated academic texts) or as large as in Schmitt and Carter’s (2000) study (7.5% more word types in the unrelated articles) of newspaper stories. This could be a function of the type of discourse—spoken vs. written—or it could also indicate that the vocabulary in television dramas does not vary as much as it does in written text. A summary of the number of tokens, word types, word families, and the differences between related and unrelated texts in the three studies is given in Table 7. The results provide evidence that it may be easier for learners to understand the vocabulary in related programs than in unrelated programs, because there are likely to be fewer word families encountered in related programs, and the word families that do occur are more likely to reoccur, increasing the potential for vocabulary learning. This provides some direction on how television may be more effectively used for language learning. If learners do not have high comprehension of television programs, it may be most effective to watch different episodes of a single program rather than single episodes of different programs. Watching different, or ideally successive, episodes of one program may also be more effective, because it enables viewers to make cumulative gains in background knowledge, which may increase their comprehension of subsequent episodes. For example, in the first episode of a program learners will need to become familiar with characters’ names and other key proper nouns, the relationships between the characters, as well as the story line. In subsequent episodes, viewers will have greater understanding of those variables. Background knowledge has consistently been found to have an influence on comprehension of written text (Stahl, Hare, Sinatra, & Gregory, 1991; Stahl & Jacobson, 1986; Stahl, Jacobson, Davis, & Davis, 1989). In answer to the second research question, the results indicate that there is greater potential for vocabulary learning through watching related television programs than unrelated television programs. There was a higher percentage of low-frequency (4,000- to 14,000-level) word families encountered 10 or more times, and a smaller percentage of word families encountered once in all six of the seasons than in the six sets of random television programs. It is useful to look at the total number of tokens and the word families that were encountered 10 or more times in the 4,000- to 14,000-word lists to see where differences VOCABULARY IN RELATED TELEVISION PROGRAMS 705 706 TABLE 4 Number and Percentage of Encounters With 4,000- to 14,000-Level Word Families in a Season of 24, Alias, Crossing Jordan, and Random Matching Programs Number of encounters 1 2 3–4 5–7 8–9 10+ Mean encounters for 10+ Total word families Total tokens RTP matching 24 24 Amount 280 136 61 43 8 27 32 % 50 25 11 8 1 5 555 1,926 TESOL QUARTERLY Note. RTP 5 random television programs. Amount 877 264 158 46 15 34 16 1,394 2,861 % 63 19 11 3 1 2 Alias Amount 664 198 149 50 21 40 16 1,122 2,654 % 59 18 13 4 2 4 RTP matching Alias Crossing Jordan Amount Amount 857 255 160 40 14 38 17 1,364 2,902 % 63 19 12 3 1 3 985 304 241 92 19 57 25 1,698 4,539 % 58 18 14 5 1 3 RTP matching Crossing Jordan Amount 967 279 211 85 24 34 20 1,600 3,569 % 60 17 13 5 2 2 VOCABULARY IN RELATED TELEVISION PROGRAMS TABLE 5 Number and Percentage of Encounters With 4,000- to 14,000-Level Word Families in a Season of CSI, Grey’s Anatomy, House, and Random Matching Programs Number of encounters 1 2 3–4 5–7 8–9 10+ Mean encounters for 10+ Total word families Total tokens RTP matching CSI CSI Amount 910 282 220 115 32 63 17 % Amount 56 17 14 7 2 4 945 268 157 66 16 32 17 % 64 18 11 4 1 2 Grey’s Anatomy Amount 738 233 201 95 33 81 17 RTP matching Grey’s Anatomy % Amount % 53 17 15 7 2 6 1,049 369 236 128 23 45 18 57 20 13 7 1 2 RTP matching House House Amount 932 357 244 133 35 113 22 % Amount % 51 20 13 7 2 6 1,058 348 231 108 20 43 16 59 19 13 6 1 2 1,622 1,484 1,381 1,850 1,814 1,808 4,215 3,086 4,109 4,345 5,960 3,978 Note. RTP 5 random television programs. 707 708 TABLE 6 Difference Between Total Tokens and Word Families in Encounters With 4,000- to 14,000-Level Word Families in a Season of All Six Programs and Random Matching Programs 24 Total tokens Difference for tokens Difference for tokens (%) Total word families Difference for families Difference for families (%) 1,926 2935 RTP matching 24 2,861 248.5 555 Alias 2,654 2248 RTP matching Alias 2,902 1,122 4,539 +970 3,569 +21.4 29.3 1,394 Crossing Jordan RTP matching Crossing Jordan 1,364 1,698 CSI 4,215 +1,129 RTP matching CSI 3,086 +26.8 1,600 1,622 Grey’s Anatomy 4,109 2236 4,345 1,484 1,381 2242 +98 +138 2427 2151.2 221.6 +5.8 +8.5 234.0 House 5,960 +1615 RTP matching House 3,978 +33.3 25.7 2839 TESOL QUARTERLY Note. RTP 5 random television programs. RTP matching Grey’s Anatomy 1,850 1,814 236 +0.3 1,808 occurred between the individual programs and the sets of random television programs. For example, there were 5,960 low-frequency running words in House, and 2,486 of these were from word families which were encountered 10 or more times (113 6 22). The random television programs set matched with House had only 3,978 lowfrequency running words, and 688 of these were from word families encountered 10 or more times (43 6 16). Although it should be expected that there would be fewer running words from word families encountered 10 or more times in the random set because there were fewer low-frequency running words, the difference between the two (1,798 tokens) is surprising because it accounts for 91% of the 1982 (5,960–3,978) excess tokens in the 4,000- to14,000-word lists of House. Thus, 91% of the excess low-frequency words in House were members of word families which were encountered 10 or more times. The results were similar for Crossing Jordan, which also had more low-frequency words than the random set; 77% of the excess words in Crossing Jordan were from word families which were encountered 10 or more times. Results were not quite as high for the other program, which had more low-frequency tokens than its matching set. Only 47% of the excess words in CSI were from word families which were encountered 10 or more times. However, there were also higher percentages of word families encountered 5–7 times and 8–9 times in CSI. Together, the results indicate that, if there were more low-frequency words in an individual program, there was greater potential to learn those words. TABLE 7 Comparison Between the Present Study and Earlier Research on Narrow Reading Including Number of Tokens, Word Types, Word Families, and the Percent Difference Between Related and Unrelated Texts Present study Tokens Types Difference for types Difference for types (%) Families Difference for families Difference for families (%) Sutarsyah et al. (1994) Related texts Unrelated texts Related texts Unrelated texts 665,134 19,256 2956 665,134 20,212 295,294 9,469 211,930 311,768 21,399 25 9,401 21,110 211 256 10,511 5,438 27,306 Schmitt & Carter (2000) Related texts 7,843 2,068 2156 Unrelated texts 7,843 2,224 27.5 12,744 a a 257 Note. aNot reported. VOCABULARY IN RELATED TELEVISION PROGRAMS 709 In answer to the third research question, the results indicate that there is potential for significant incidental vocabulary learning through watching complete seasons of television programs in terms of both breadth (how many words are known) and depth (how well words are known) of vocabulary knowledge (Read, 2004). If we count the word families which were encountered 10 or more times as potentially learned items, then from 27 to 113 words may have been learned through watching the episodes from the six programs. However, this may be a conservative estimate. It is likely that viewers would make gains in partial knowledge for most of the words encountered five or more times in the programs. These gains from incidental vocabulary learning at different encounter frequencies would serve to build up the viewers’ breadth and depth of vocabulary knowledge. Moreover, they would also likely make gains (strengthening certain aspects of vocabulary knowledge, such as spoken form and collocation) for words, such as those in the 1,000- to 3,000-word lists, which they may have been able to demonstrate knowledge of prior to viewing. Encountering known words in original contexts is likely to increase knowledge of different aspects of knowledge for those items. The results also indicate that there may be substantial incidental vocabulary learning through watching unrelated programs, if television is watched regularly. Counting word families which were encountered 10 or more times as potentially learned items, then from 32 to 45 words may have been learned through watching one of the six random sets of programs. In one comparison between 24 and the random matching television programs, there were more word families encountered 10 or more times in the random television programs. Although this was due to a greater number of low-frequency running words in the random programs, television programs are not viewed based on the number or frequency of running words. These results indicate that, in some cases, it may be just as effective to learn vocabulary through watching unrelated programs, and if learners watch L2 television programs regularly, regardless of the content, they are likely to increase their vocabulary. Overall, if vocabulary can be incidentally learned through watching L2 television programs in a similar fashion demonstrated in reading (Horst et al., 1998), the results of this study point to the potential efficacy of narrow viewing. In answer to the fourth research question, the results indicate that a vocabulary size of 3,000 word families plus knowledge of proper nouns and marginal words is sufficient to reach 95% coverage of television programs. The 1.3 million word corpus created for this study was more than 5 times the size of the television corpus analyzed in Webb and Rodgers’ (2009b) study. The cumulative coverage at the 3,000-word level was 95.83% in this study, which was slightly higher than in the earlier 710 TESOL QUARTERLY study (95.45%). It is also important to note that the results of the present study also indicate that coverage is likely to vary considerably between different programs. Although the mean coverage of 3,000 word families plus proper nouns and marginal words for all of the programs in the study is 95.83%, one of the six individual programs examined reached 95% coverage at the 2,000 word level, whereas two did not reach 95% coverage until the 4,000-word level. Coverage of 24 at the 2,000-word level was 96.64%. In contrast, coverage of House and CSI at the 4,000word level was 95.03% and 96.00%, respectively. This finding suggests that it may be of considerable value to analyze the vocabulary in programs to provide some guidance to teachers and learners on their respective vocabulary loads. These results are significant because they support the earlier findings of Webb and Rodgers (2009b) that knowledge of 3,000 word families plus proper nouns and marginal words will result in 95% lexical coverage of television programs. This coverage level might be sufficient for adequate comprehension and for incidental learning to occur. This finding supports the potentiality of television, especially in a narrow viewing context, as a source of aural input for language learners reaching or approaching this vocabulary level. Language learners and their teachers need to be made aware of the implications the results of this study have for learners seeking to improve their vocabulary and learn from authentic aural input. The higher rate of occurrence of low-frequency words and fewer word families in related television episodes means that a learner has more opportunities to gain knowledge of these words. Language teachers and learners also need to be made aware of the potential for increased comprehension from watching related television programs. The lower vocabulary load in a season of television versus unrelated episodes of different programs is something that a language learner should consider when choosing to learn from television. Add this to the gains in background knowledge a viewer gets as he or she becomes more familiar with the setting, character relationships, and the series’ main plot, and learners can expect to have the potential to comprehend authentic television that they might otherwise find overwhelming. Learners and teachers have two ways of implementing a narrow viewing approach. The first is through self-study, where television would be used in the same way books are used in extensive reading programs. Language learners would have access to full seasons of television in a range of titles and genres. This wide range of titles would allow for various tastes and interests. This study’s findings can also be implemented in the classroom as part of a teacher-led course. In this situation, viewing of episodes of a single television program would take place in the classroom accompanied by a VOCABULARY IN RELATED TELEVISION PROGRAMS 711 variety of explicit instructional activities related to the individual episodes and the series as a whole. There are a number of strategies and techniques that can be used to improve comprehension and vocabulary acquisition prior to viewing, while viewing, and postviewing. In the self-study situation, teachers should take care to recommend that, upon finding a television program that piques their interest, students should start at the beginning of preferably the first season and work through it successively. This instruction needs to come prior to students beginning the viewing process, so as to guide their choices. The learners also need to be given the opportunity to try watching a program and to choose a different one if they don’t like it. They should be encouraged to watch the same episode multiple times if they enjoyed it, especially early on in a season or series, to solidify knowledge of the background of the story and the relationships between the characters. Short background summaries could be developed to guide learners to programs in which they might be interested. The summaries of the television programs in simplified English or, if the resources are intended for a homogeneous population, in the L1, would increase the chance of a learner finding a program he or she likes more quickly. Information concerning the relative lexical load of the programs would also aid students in deciding on a suitable television series. Activities in the classroom prior to viewing should begin with a rationale for using class time to watch television. Because there is a perception that watching videos is a recreational exercise (Bird, 2005), students need to be instructed about its potential benefits so they take television viewing as seriously as they would any other classroom activity. The students should be made aware that through watching successive episodes of one program they can expect to learn the spoken form of English, improve their L2 listening skills, learn about foreign culture, and enjoy the process at the same time. The learners should be warned that they may not understand everything at first, but that comprehension will improve in successive episodes. Prior to viewing, learners should be taught about the background story of the series. Included in this explanation can be details about the setting, the context of the story, and how the story is culturally relevant. The main characters should also be introduced before viewing starts, and should include brief descriptions of their relationships to one another. Before each episode, if appropriate, new characters can also be introduced. Preteaching low-frequency words from the television program is another classroom-based activity that can aid comprehension. Webb (2010a) found that knowledge of an episode’s 10 most frequently occurring topic-related words from the 4,000- to14,000-word BNC lists represented an increase in lexical coverage that may facilitate 712 TESOL QUARTERLY comprehension. Because of the nature of television, teaching the spoken form of the target words and their contextualized meanings would have the greatest value. Because learners need to be able to recognize the aural form and link it to meaning, language-focused instruction should include opportunities for the learners to hear the target words in a variety of contexts. The television series itself can be used to preteach the target vocabulary. Short scenes from the episodes that include examples of the target words can be used in a variety of ways. One approach would be for the learners to view a scene and identify which target words were used in the scene. This helps the learners to recognize the spoken form of the words. Another approach would be a cloze activity for which the learners are provided with transcripts of a scene with the target vocabulary removed, and they fill in the missing words as they listen to the dialogue. A dictation activity could also teach the spoken form of the target words. Learners would listen to isolated sentences from the episodes, transcribe them, and finally highlight occurrences of target words. Before the teacher provides the answers, it might take repeated listens for the learners to fully transcribe the sentences. Viewing the scenes in conjunction with these activities would also provide more context for the introduction of the television series’ characters and background story. Another approach that can promote comprehension is the use of glossaries made up of the most frequently occurring topic-related words from the 4,000- to14,000-word lists (Webb, 2010b). The information presented in the glossaries can include L1 definitions, L2 definitions, pictures, usage examples, and phonetic transcriptions. However, not all words in the glossary need all this information included. Words such as cortex from House (13 encounters in the season analyzed in this study), directorate from Alias (48 encounters), or spatter from CSI (10 encounters) may have limited use to language learners outside those in specialized fields. Knowledge of these words can still help comprehension, so they should be included in a glossary but with shorter, easily understood definitions and notation explaining that they are words important to the program but perhaps not so valuable in most English language settings. Although still potentially useful in a classroom environment, the use of glossaries lends itself better to the self-study situation. This is because to use the glossary to its fullest it would be beneficial to pause the video to consult the glossary when the need arises, something which would be inconvenient in a classroom setting. Comprehension questions based on the content of the individual episodes can also be used to focus the students’ attention on key plot ideas or details. These comprehension questions can be completed while viewing, postviewing, or a combination of both. Regular breaks in the viewing would give the students time to answer more VOCABULARY IN RELATED TELEVISION PROGRAMS 713 detail-orientated questions that they may forget by the end of the episode. Questions that deal with the broader ideas of the episode could be completed postviewing when the entirety of the episode can be considered. These comprehension questions work equally well in a classroom setting as in a self-study situation. Postviewing activities could also include exercises in which students summarize the episode they just watched, or make predictions about the next episode or important story arcs. These activities would serve to heighten a learner’s involvement with the series. In teacher-led instruction not all viewing need be done in class, and for homework the students could watch the successive episode or finish watching episodes started in class. CONCLUSION Although it may be most effective for learners to watch different episodes of the same television program rather than watching different programs, it would be unreasonable to expect all learners to be motivated to only watch episodes of one program. If television is to be used regularly, following the approach taken in extensive reading, it may be most effective for learners to select the programs which they watch. The large number of complete seasons of television programs available on DVD makes it relatively simple to provide access to a wide variety of series. Allowing learners to find programs which they are interested in watching from a range of programs, and guiding learners to watch all of the available episodes of that program before moving on to a different program, may be an effective approach to using television for language learning. This may be most useful for learners who have some difficulty understanding L2 programs. If learners are at a level where they can easily understand L2 television programs, then the primary aim should be for them to encounter as much L2 aural input as possible rather than specific input. ACKNOWLEDGMENT The authors would like to thank the anonymous reviewers for their helpful comments. THE AUTHORS Michael P. H. Rodgers is an instructor at the Fukuoka University Language Education and Research Center in Fukuoka, Japan. His research interests are vocabulary acquisition, media corpus studies, and language testing. Stuart Webb is a senior lecturer and the Master of Arts programme director in the School of Linguistics and Applied Language Studies at Victoria University of 714 TESOL QUARTERLY Wellington, New Zealand. His research interests include vocabulary studies and extensive reading and listening. REFERENCES Allan, R. (2009). Can a graded reader corpus provide ‘authentic’ input? ELT Journal, 63, 23–32. doi: 10.1093/elt/ccn011. Bauer, L., & Nation, P. (1993). Word families. International Journal of Lexicography, 6, 253–279. doi: 10.1093/ijl/6.4.253. Bird, S. A. (2005). Language learning edutainment: Mixing motives in digital resources. RELC Journal, 36, 311–339. doi: 10.1177/0033688205060053. Bonk, W. (2000). Second language lexical knowledge and listening comprehension. International Journal of Listening, 14, 14–31. Chapple, L., & Curtis, A. (2000). Content-based instruction in Hong Kong: Student responses to film. System, 28, 419–433. doi: 10.1016/S0346-251X(00)00021-X. Colwell, H., & Ipince Braschi, M. (2006, September). Using films with mixed level ESL classes. Paper presented at the English Australia Conference 2006, Perth, Western Australia Retrieved from http://www.englishaustralia.com.au/ea_ conference2006/proceedings/pdf/Colwell.pdf Crawford, J. (2002). The role of materials in the language classroom: Finding the balance. In J. C. Richards & W. A. Renandya (Eds.), Methodology in language teaching: An anthology of current practice (pp. 80––91). Cambridge, England: Cambridge University Press. Donkaewbua, S. (2007). The effects of previous partial word knowledge on vocabulary learning through listening. Unpublished manuscript, School of Linguistics and Applied Language Studies, Victoria University of Wellington, New Zealand. d’Ydewalle, G., & Pavakanun, U. (1995). Acquisition of a second/foreign language by viewing a television program. In P. Winterhoff-Spurk (Ed.), Psychology of media in Europe: The state of the art—perspectives for the future (pp. 51––64). Opladen, Germany: Westdeutscher Verlag. d’Ydewalle, G., & Pavakanun, U. (1997). Could enjoying a movie lead to language acquisition? In P. Winterhoff-Spurk & T. Van der Voort (Eds.), New horizons in media psychology (pp. 145–155). Opladen, Germany: Westdeutscher Verlag. d’Ydewalle, G., & Van de Poel, M. (1999). Incidental foreign-language acquisition by children watching subtitled television programs. Journal of Psycholinguistic Research, 28, 227–244. doi: 10.1023/A:1023202130625. Gardner, D. (2004). Vocabulary input through extensive reading: A comparison of words found in children’s narrative and expository reading materials. Applied Linguistics, 25, 1–37. doi: 10.1093/applin/25.1.1. Hanley, J. E. B., Herron, C., & Cole, S. (1995). Using video as an advance organizer to a written passage in the FLES classroom. Modern Language Journal, 79, 57–66. doi: 10.2307/329393. Harmer, J. (2007). The practice of English language teaching. New York, NY: Pearson Longman. Hirsh, D., & Nation, P. (1992). What vocabulary size is needed to read unsimplified texts for pleasure? Reading in a Foreign Language, 8, 689–696. Holmes, G. (2008, November 24). Americans can’t get enough of their screen time [Press Release]. New York, NY: The Nielsen Company. Retrieved from http://en-us. nielsen.com/content/nielsen/en_us/news/news_releases/2008/november/ americans_cannot_get.html VOCABULARY IN RELATED TELEVISION PROGRAMS 715 Horst, M. (2005). Learning L2 vocabulary through extensive reading: A measurement study. Canadian Modern Language Review, 61, 355–382. doi: 10.3138/ cmlr.61.3.355. Horst, M., Cobb, T., & Meara, P. (1998). Beyond A Clockwork Orange: Acquiring second language vocabulary through reading. Reading in a Foreign Language, 11, 207–223. Hu, M., & Nation, P. (2000). Unknown vocabulary density and reading comprehension. Reading in a Foreign Language, 13, 403–430. Hwang, K., & Nation, P. (1989). Reducing the vocabulary load and encouraging vocabulary learning through reading newspapers. Reading in a Foreign Language, 6, 323–335. Jenkins, J. R., Stein, M. L., & Wysocki, K. (1984). Learning vocabulary through reading. American Educational Research Journal, 21, 767–787. Kamiya, T. (1994). Tuttle new dictionary of loanwords: A user’s guide to gairaigo. Tokyo, Japan: Charles E. Tuttle Publishing Company. King, J. (2002). Using DVD feature films in the EFL classroom. Computer Assisted Language Learning, 15, 509–523. doi: 10.1076/call.15.5.509.13468. Koolstra, C. M., & Beentjes, J. W. J. (1999). Children’s vocabulary acquisition in a foreign language through watching subtitled television programs at home. Educational Technology Research and Development, 47, 51–60. doi: 10.1007/ BF02299476. Laufer, B. (1989). What percentage of text-lexis is essential for comprehension? In C. Lauren & M. Nordman (Eds.), Special language: From humans thinking to thinking machines (pp. 316–323). Clevedon, England: Multilingual Matters. Mueller, G. A. (1980). Visual contextual cues and listening comprehension: An experiment. Modern Language Journal, 64, 335–340. doi: 10.2307/324500. Nation, I. S. P. (2004). A study of the most frequent word families in the British National Corpus. In P. Bogaards & B. Laufer (Eds.), Vocabulary in a second language: Selection, acquisition, and testing (pp. 3–13). Amsterdam, The Netherlands: John Benjamins. Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening? Canadian Modern Language Review, 63, 59–82. doi: 10.3138/cmlr.63.1.59. Nation, I. S. P. (2011). Web site: http://www.victoria.ac.nz/lals/staff/paul-nation/ nation.aspx Nation, I. S. P., & Heatley, A. (2002). Range: A program for the analysis of vocabulary in texts [computer software]. Wellington, New Zealand: Victoria University of Wellington. Retrieved from http://www.victoria.ac.nz/lals/staff/paul-nation/ nation.aspx Neuman, S. B., & Koskinen, P. (1992). Captioned television as comprehensible input: Effects of incidental word learning from context for language minority students. Reading Research Quarterly, 27, 95–106. Organization for Economic Cooperation and Development (OECD). (2007). Communications outlook 2007. Paris, France: OECD. Retrieved from http://213. 253.134.43/oecd/pdfs/browseit/9307021E.PDF Oetting, J. B., Rice, M. L., & Swank, L. K. (1995). Quick incidental learning (QUIL) of words by school-age children with and without SLI. Journal of Speech, Language and Hearing Research, 38, 434–445. Pavakanun, U., & d’Ydewalle, G. (1992). Watching foreign television programs and language learning. In F. L. Engel, D. G. Bouwhuis, T. B. Osser, &G. d’Ydewalle (Eds.), Cognitive modelling and interactive environments in language learning (pp. 193– 198). Berlin, Germany: Springer. Quaglio, P. (2009). Television dialogue: The sitcom Friends vs. natural conversation. Amsterdam, The Netherlands: John Benjamins. 716 TESOL QUARTERLY Read, J. (2004). Plumbing the depths: How should the construct of vocabulary knowledge be defined? In P. Bogaards & B. Laufer (Eds.), Vocabulary in a second language: Selection, acquisition, and testing (pp. 209–227). Amsterdam, The Netherlands: John Benjamins. Rice, M. L., & Woodsmall, L. (1988). Lessons from television: Children’s word learning when viewing. Child Development, 59, 420–429. doi: 10.2307/1130321. Rott, S. (1999). The effect of exposure frequency on intermediate language learners’ incidental vocabulary acquisition and retention through reading. Studies in Second Language Acquisition, 21, 589–619. doi: 10.1017/S0272263199004039. Saragi, T., Nation, I. S. P., & Meister, G. F. (1978). Vocabulary learning and reading. System, 6, 72–78. doi: 10.1016/0346-251X(78)90027-1. Schmitt, N. (2008). Review article: Instructed second language vocabulary learning. Language Teaching Research, 12, 329–363. doi: 10.1177/1362168808089921. Schmitt, N., & Carter, R. (2000). The lexical advantages of narrow reading for second language learners. TESOL Journal, 9, 4–9. Stahl, S. A., Hare, V. C., Sinatra, R., & Gregory, J. F. (1991). Defining the role of prior knowledge and vocabulary in reading comprehension: The retiring of number 41. Journal of Reading Behavior, 23, 487–508. Stahl, S. A., & Jacobson, M. G. (1986). Vocabulary difficulty, prior knowledge, and text comprehension. Journal of Reading Behavior, 18, 309–323. Stahl, S. A., Jacobson, M. G., Davis, C. E., & Davis, R. L. (1989). Prior knowledge and difficult vocabulary in the comprehension of unfamiliar text. Reading Research Quarterly, 24, 27–43. doi: 10.2307/748009. Sutarsyah, C., Nation, P., & Kennedy, G. (1994). How useful is EAP vocabulary for ESP? A corpus based case study. RELC Journal, 25, 34–50. doi: 10.1177/ 003368829402500203. Vandergrift, L. (2007). Recent developments in second language and foreign language listening comprehension research. Language Teaching, 40, 191–210. doi: 10.1017/S0261444807004338. Waring, R., & Nation, P. (2004). Second language reading and incidental vocabulary learning. Angles on the English Speaking World, 4, 96–110. Waring, R., & Takaki, M. (2003). At what rate do learners learn and retain new vocabulary from reading a graded reader. Reading in a Foreign Language, 15, 130– 163. Webb, S. (2007). The effects of repetition on vocabulary knowledge. Applied Linguistics, 28, 45–65. doi: 10.1093/applin/aml048. Webb, S. (2010a). Pre-learning low-frequency vocabulary in second language television programmes. Language Teaching Research, 14, 501–515. doi: 10.1177/ 1362168810375371. Webb, S. (2010b). Using glossaries to increase the lexical coverage of television programs. Reading in a Foreign Language, 22, 201–221. Webb, S., & Rodgers, M. P. H. (2009a). The lexical coverage of movies. Applied Linguistics, 30, 407–427. doi: 10.1093/applin/amp010. Webb, S., & Rodgers, M. P. H. (2009b). Vocabulary demands of television programs. Language Learning, 59, 335–366. doi: 10.1111/j.1467-9922.2009.00509.x. VOCABULARY IN RELATED TELEVISION PROGRAMS 717