The Lexical Coverage of Movies PDF

Preview

Summary

Applied Linguistics Advance Access published March 26, 2009 Applied Linguistics: 1–21 ß Oxford University Press 2009 doi:10.1093/applin/amp010 The Lexical Coverage of Movies STUART WEBB1 and MICHAEL P. H. RODGERS2 1 Victoria University of Wellington and 2Fukuoka University The scripts of 318 movies ...

Description

Accelerat ing t he world's research.

The Lexical Coverage of Movies Stuart Webb, Michael Rodgers Applied linguistics

Cite this paper

Downloaded from Academia.edu 

Get the citation in MLA, APA, or Chicago styles

Related papers

Download a PDF Pack of t he best relat ed papers 

A corpus driven st udy of t he pot ent ial for vocabulary learning t hrough wat ching movies St uart Webb

2010 webb IJCL t he pot ent ial for vocabulary learning t hrough wat ching movies.pdf St uart Webb Vocabulary Demands of Television Programs St uart Webb, Michael Rodgers

Applied Linguistics Advance Access published March 26, 2009 Applied Linguistics: 1–21 doi:10.1093/applin/amp010

ß Oxford University Press 2009

The Lexical Coverage of Movies STUART WEBB1 and MICHAEL P. H. RODGERS2 1

Victoria University of Wellington and 2Fukuoka University The scripts of 318 movies were analyzed in this study to determine the vocabulary size necessary to understand 95% and 98% of the words in movies. The movies consisted of 2,841,887 running words and had a total running time of 601 hours and 33 minutes. The movies were classified as either American or British, and then put into the following genres: action, animation, comedy, suspense/crime, drama, horror, romance, science fiction, war, western, and classic. The results showed that knowledge of the most frequent 3,000 word families plus proper nouns and marginal words provided 95.76% coverage, and knowledge of the most frequent 6,000 word families plus proper nouns and marginal words provided 98.15% coverage of movies. Both American and British movies reached 95% coverage at the 3,000 word level. However, American movies reached 98% coverage at the 6,000 word level while British movies reached 98% coverage at the 7,000 word level. The vocabulary size necessary to reach 95% coverage of the different genres ranged from 3,000 to 4,000 word families plus proper nouns and marginal words, and 5,000 to 10,000 word families plus proper nouns and marginal words to reach 98% coverage. The implications for teaching and learning with movies are discussed in detail.

INTRODUCTION There has been very little research on second language (L2) learning with movies. This is surprising because learners are motivated to learn language through watching movies (Chapple and Curtis 2000; King 2002; Colwell and Ipince Braschi 2006), and movies may be a valuable source of L2 aural input in the English as a foreign language (EFL) context, where there may be relatively few opportunities to improve L2 listening skills. L2 movies with English language soundtracks are often shown on television and in theatres and are available for rental or purchase in many countries where English is not the first language (L1). Although there does not appear to be any data on the amount of time that L2 learners watch movies, Gieve and Clark (2005) found that both European and Chinese learners were more likely to try to learn independently through watching films than through extensive reading, and Ding (2007, p. 275) found that among the participants in his study, all eight of the prize winners in national level speech contests in China ‘regarded watching English movies and television series as one of the most effective ways of improving their English’. The reason for the lack of research on movies might be because the vocabulary in movies is too difficult for some learners

2 STUART WEBB AND MICHAEL P. H. RODGERS

to understand, and that until recently movie scripts have been relatively difficult to obtain and analyze. Whether or not film is used for language learning may depend on whether or not learners have the receptive vocabulary size sufficient to comprehend the vocabulary in L2 movies. Teachers are unlikely to use L2 movies in the classroom and encourage learners to watch L2 movies outside of the classroom if they are not sure whether learners will be able to understand movies. Studies which discuss approaches to teaching with movies report that it is important to use movies that are at the appropriate level for learners (King 2002; Colwell and Ipince Braschi 2006); however, there is currently no research which indicates for which level of learners movies are appropriate. Moreover, different movies may have different lexical demands with movies from some genres being more difficult to understand than movies from other genres. Establishing the vocabulary size necessary to watch different types of movies would help teachers to determine which students should be able to understand movies and which students may not be ready to learn with movies. This would allow teachers to use film effectively in the classroom and may encourage learners to start using movies outside of the classroom for language learning. The aim of this study is to determine the lexical coverage of movies. Coverage refers to the percentage of known words in discourse and is a valuable measure because it may indicate the vocabulary size necessary to understand a text as well as to incidentally learn words in the text. It is particularly important to know the vocabulary needed for comprehension because learners are unlikely to watch L2 movies if they cannot understand them and watching movies which are too demanding may discourage learners from using movies for language learning. Determining the vocabulary size necessary to understand movies may provide both teachers and learners with a vocabulary learning goal, which when reached, would allow movies to become a valuable source of L2 aural input.

How many words do you need to know to understand a movie? L2 research investigating comprehension has sought to determine the coverage necessary to understand written text. Although there are many factors involved in comprehension, coverage may have the greatest effect on whether or not discourse is understood (Laufer and Sim 1985). It is a particularly useful measurement because certain coverage points may indicate the vocabulary size necessary to use or understand discourse types. The coverage needed for comprehension may vary depending on the discourse type and the degree of comprehension required. For example, Laufer (1989) suggests that 95% coverage is necessary for reasonable comprehension of a text, Webb and Rodgers (in press) suggest that 95% coverage is necessary to have adequate understanding of television programs, Hirsh and Nation (1992) suggest that 98% coverage is needed for learners to read for pleasure, Hu and Nation (2000) report that 98% coverage is necessary for adequate unassisted reading comprehension,

THE LEXICAL COVERAGE OF MOVIES

3

and Nation (2006) suggests that 98% is the ideal coverage for comprehension of written text. Nation (2006) also suggests that coverage greater than 98% may be needed for spoken text, and Donkaewbua (2007) reports that learners may need 99% coverage for adequate L2 aural comprehension of graded readers. It is important to note that a particular coverage does not ensure that learners will have the same percentage of comprehension. Knowing the words does not ensure that the discourse will be understood. Coverage may be the most important factor in determining comprehension, but it is one of many factors which are involved in comprehension. Once coverage is determined, researchers may then analyze discourse to find how many words learners need to know to reach the coverage point. For example, Nation (2001) reported that knowledge of the most frequent 2,000 word families provides coverage of 90% of the words in spoken discourse, Adolphs and Schmitt (2003) found that the most frequent 3,000 word families accounted for 95.91% of spoken discourse, and Webb and Rodgers (in press) found that the most frequent 3,000 word families plus proper nouns and marginal words provided coverage of 95.45% of the dialog in television programs. Nation’s (2006) study of the vocabulary needed to understand several different discourse types is the only research to examine the coverage of a movie. He analyzed the script of a children’s movie and found that a vocabulary of the most frequent 2,000 words provided 87.91% coverage, a vocabulary of the most frequent 4,000 word families plus proper nouns and marginal words provided 96.74% coverage, and a vocabulary of the most frequent 7,000 word families plus proper nouns and marginal words provided 98.08% coverage. There have been four studies which have measured the coverage necessary for comprehension. Laufer (1989) found that 90% coverage led to unsatisfactory comprehension more often than satisfactory comprehension and that 95% coverage could ensure reasonable comprehension of a general academic L2 text. Hu and Nation (2000) investigated L2 learners’ comprehension of a relatively easy fiction text with differing amounts of text coverage. They found that a small number of learners gained adequate comprehension at 90% coverage, more learners gained adequate comprehension at 95% coverage, but they were still a minority, and at 100% coverage, most learners were able to understand the text. Hu and Nation used a regression analysis to show that 98% coverage was needed for adequate comprehension without use of a dictionary or glossary. They also suggest that learners may need greater coverage of texts from different genres such as newspapers and academic texts. Hu and Nation’s study is supported by research by Carver (1994) who found that 98 to 99% coverage provided adequate L1 comprehension of a text with coverage being dependent on the difficulty of the text. However, the results of Bonk’s (2000) study examining the effects of coverage on listening comprehension indicated that lower coverage may provide adequate listening comprehension. He investigated comprehension of four short passages of increasing lexical difficulty and measured comprehension with a written recall test and a dictation test.

4 STUART WEBB AND MICHAEL P. H. RODGERS

Bonk (2000) found that with adequate coping strategies, L2 learners could have adequate L2 aural comprehension of short texts at far below 95% coverage. There are two reasons why the studies examining the effects of coverage on comprehension might differ. First, there are many factors involved in comprehension. Research has also shown that factors such as background knowledge (Stahl and Jacobson 1986; Stahl et al. 1989, 1991), the relevance of unknown vocabulary in context (Stahl 1990), the amount of redundant information (Kameenui et al. 1982) and individual differences (Mezynski 1983; Stahl 1990) have been found to affect reading comprehension. Although coverage may have the greatest effect on whether learners are able to understand a television program, it is one of many factors which may affect comprehension and 100% coverage does not always ensure comprehension. Second, there are likely to be degrees of comprehension ranging from basic understanding of the main ideas to very precise knowledge of the text. It is unlikely that tests used in different studies will measure comprehension to the same degree unless they use identical texts and tests. Comprehension of movies may be easier than comprehension of written or aural text because the aural input in movies is supported by visual input. Chapple and Curtis (2000) and Rubin (1994) report that video may provide ample visual input which may assist with comprehension. However, Rubin notes that the correlation between the visual and the aural input may vary from scene to scene and program to program. Mueller (1980) found that visual input in the form of drawings facilitated listening comprehension for lower level learners but had no effect for advanced learners. Hanley et al. (1995) found that learners demonstrated higher comprehension on a test when they learned with video than when they learned with pictures. They suggest that video may improve comprehension because it may effectively link L2 form with L1 meaning. Taken together, the research on comprehension and coverage indicates that coverage ranging from 90% to 99% is likely to provide adequate comprehension of movies. In the present study, 95% and 98% were chosen to represent the possible lower and upper boundaries which may indicate adequate comprehension of movies. These coverage points were selected because Laufer (1989) suggests that 95% coverage may signify reasonable comprehension of written text, and Nation (2006) suggests that 98% coverage may signify ideal comprehension of written text.

Can people learn L2 vocabulary through watching movies? Research indicates that viewers can incidentally learn L1 and L2 vocabulary through watching television and short movies. Rice and Woodsmall (1988) compared scores for children who watched a 12-minute cartoon with a narration which included 20 unknown L1 target words and for children who watched the cartoon with a narration which did not include the target

THE LEXICAL COVERAGE OF MOVIES

5

words. All of the target items were encountered at least five times in the video. The results showed that the children who encountered the target words while watching the cartoon had higher scores on a picture recognition task, and that five-year olds learned more words than three-year olds. The findings indicated that children do incidentally learn L1 vocabulary through watching television and that accumulated linguistic knowledge and experience watching television and videos may improve the potential for vocabulary learning. Oetting et al. (1995) replicated Rice and Woodsmall’s (1988) study with normally developed, and specific language impaired 6–8-year olds. The findings supported the earlier results indicating that watching video is likely to result in L1 incidental vocabulary learning with the size of the gains related to age and learner aptitude. The results indicated that the learners who were likely to have known the most running words and therefore had the greatest coverage, learned the most target words. A series of studies by d’Ydewalle and colleagues found that watching television without subtitles is likely to lead to L2 incidental vocabulary learning (Pavakanun and d’Ydewalle 1992; d’Ydewalle and Van de Poel 1999), L2 incidental vocabulary learning is higher when watching television with subtitles than without subtitles (Pavakanun and d’Ydewalle 1992; d’Ydewalle and Pavakanun 1995), watching television with reversed subtitles (L1 in the soundtrack and L2 in the subtitles) leads to greater vocabulary learning than standard subtitles (L2 in the soundtrack and L1 in the subtitles) for adults (Pavakanun and d’Ydewalle 1992; d’Ydewalle and Pavakanun 1995, 1997), presenting the target L2 in the soundtrack may lead to larger gains than presenting it in the subtitles for children (d’Ydewalle and Van de Poel 1999), and that there may be considerable L2 vocabulary learning from watching a short subtitled movie (d’Ydewalle and Pavakanun 1997). d’Ydewalle and Pavakanun (1997) report that many children in Belgium can speak a considerable amount of English prior to receiving any formal English language instruction and suggest that watching subtitled television may be responsible for some of those gains. Neuman and Koskinen (1992) examined L2 incidental vocabulary learning through reading, watching television, and watching television with captions (L2 subtitles) with young L2 learners. They found that watching television with captions led to significantly higher scores on three different measures of vocabulary knowledge than reading the scripts of the programs. They suggested that encountering L2 input in multiple modalities may increase incidental learning because both the imagery and the printed words support the aural input. Koolstra and Beentjes (1999) compared L2 incidental vocabulary learning through watching a 15-minute English language documentary with L1 (Dutch) subtitles, the same documentary without subtitles, and the same documentary with a L1 soundtrack. The participants were children in primary school. They found that the participants who watched the documentary without subtitles had higher scores on a multiple-choice test which involved matching the L1 meanings with the L2 target vocabulary than the participants who watched the same video with the L1 soundtrack indicating that watching

6 STUART WEBB AND MICHAEL P. H. RODGERS

television did lead to incidental vocabulary learning. Koolstra and Beentjes also suggest that video which links visual input with meaning, presents the language clearly and accurately, and kindles the interest of the viewers may lead to increased vocabulary learning. Taken together, the research indicates that materials which provide visual and aural input such as movies may be conducive to incidental vocabulary learning. The research also indicated that watching movies may be as effective in contributing to incidental vocabulary learning as reading. This suggests that movies may be a very useful resource for language learning. However, it is important to note that in most of the research, the viewed content was designed for children and represented conditions which were highly conducive for incidental vocabulary learning. Contexts were clear and rich with multiple encounters with target words. Research has shown that the more words are encountered in context, the more likely they will be learned (Jenkins et al. 1984; Horst et al. 1998; Rott 1999; Nation 2001; Waring and Takaki 2003; Webb 2007). The research also showed that older children learned more words than younger children (Rice and Woodsmall 1988; Neuman and Koskinen 1992; Oetting et al. 1995). This could be because older children may have a larger vocabulary and therefore higher coverage of the viewed programs. Unfortunately, the vocabulary coverage of the viewed materials was not reported in any of the studies. Examining the coverage of movies could determine whether or not incidental vocabulary learning is likely to occur for specific learners. Although the research indicates that the conditions present in movies (visual plus aural input) may contribute to incidental vocabulary learning, the research does not provide much information about the input apart from noting that target words were often encountered several times. While showing that watching L2 movies may lead to vocabulary learning is a very useful finding, it is not clear whether the language typical in movies is at an appropriate level for learners, whether they are likely to be understood by learners at different levels, and whether they represent adequate or ideal conditions for incidental learning. Perhaps the best way to analyze the potential for language learning through watching movies is to look at the vocabulary size necessary to reach certain coverage points. Research examining coverage suggests that most running words need to be known to correctly guess words in context. Coverage could provide a gauge to whether or not learners of a particular vocabulary size may be able to learn vocabulary incidentally through watching movies. Liu and Nation (1985) report that 95% coverage is needed to guess words from context when reading, and Nation (2001) suggests that 98% coverage is the ideal coverage for guessing words from context in written text. However, research indicates that the combination of visual and aural input may make it easier to learn words through watching movies than through reading. Reading research has indicated that illustrations may aid vocabulary learning for adult L2 learners reading a graded reader (Horst et al. 1998), and for children listening to stories (Elley 1989).

THE LEXICAL COVERAGE OF MOVIES

7

The aim of the present study was to determine the lexical demands of movies. Specifically this study may provide some indication of the vocabulary size needed to understand movies and to incidentally learn vocabulary through watching movies. Determining how many words learners need to understand movies may provide both teachers and learners with a target for vocabulary learning, which when reached, would allow movies to become a valuable source of L2 input. ...