“We learned more from a 3-minute record baby than we’ve ever learned in school”
(M2 student Aïlis, citing B. Springsteen in her informal learning blog)
1As Internet technologies have become commonplace for French university students, learners have many informal opportunities to listen to English. Various European studies point to such informal activities as a key element in second language development (European Commission 2012; Safar et al. 2011; Sockett & Toffoli 2012).
2The study of the effect of exposure to language outside the classroom has in the past focused both on deliberate attempts to learn (out of class language learning [OCLL]), as defined by Benson (2001) and on the effects of incidental exposure to language through online leisure activities (Online Informal Learning of English [OILE]), as defined by Sockett & Toffoli 2012).
3A survey conducted at the University of Strasbourg in 2009 (Toffoli & Sockett 2010) sought to identify the OILE practices of EFL students majoring in subjects other than English (henceforth “MSOE students”). Both the initial study and a replica study from 2012 found that the most prevalent OILE activities of these students are viewing American television series (with or without English or French subtitles) and listening to on-demand music (often with accompanying lyrics). Through an in-depth exploration of the data from these studies, the present article looks more closely at MSOE student practices when listening to music in English and provides some insight into one type of learning that may be taking place.
4Listening to pop songs in English has long been a pastime of students of all disciplines, probably stretching back to the early years of radio and vinyl records or further, but new technologies, in particular the “digital prostheses” which now accompany students wherever they go, have introduced new affordances in terms of access to and manipulation of music and lyrics. Music that was once just heard on the radio is now actively chosen, played and replayed by learners who also use online resources to understand the propositional contents of the songs. How do students actually use these resources? Do they do so with a view to language learning? Are specific learning strategies employed? Does such use lead to recognition and perhaps uptake of lexical items, as has been found with the language of television series (Kusyk & Sockett 2012)? Can the hours spent listening to English songs actually help students learn the language?
5The following review of the literature provides details on OILE research to date, on previous studies concerning music and language learning and on the new affordances Internet technology provides for music and language learning (albeit incidental) in this context.
6The specificities of OILE, as related to out-of-class language learning (Benson 2001; Pitkänen et al. 2011) are twofold. Firstly, OILE focuses specifically on the online components of OCLL and seeks to identify language development through online activities such as social networking, streaming or downloading television series or films, e-mailing, blogging, chatting, conversing via VOIP (voice over Internet protocol), participating in on-line forums, listening to music on demand and web browsing in English. A second distinction concerns the largely incidental rather than deliberate nature of OILE, as the aim of the activities is primarily for leisure, seeking information or purposes other than language acquisition (Toffoli & Sockett 2010). OILE has been described as a process driven by the intention to communicate, with language learning being only a by-product of this communication (Sockett 2011), while OCLL focuses on deliberate attempts to improve one’s English (Benson 2006: 26).
7Studies of online informal learning have shown that it is extremely widespread among French University students, often involves many hours of activities per week, involves a wide variety of activities and often includes production as well as comprehension skills (Toffoli & Sockett 2010; Sockett 2013).
8Some studies (Kusyk & Sockett 2012; Milton 2008; Rieder 2003; Sockett 2011) are beginning to provide insight into exactly which aspects of the language are being influenced and how this is occurring. For example, Kusyk & Sockett (2012) have been able to ascertain that frequent viewers of American series correctly recognise and understand more of the most frequent 4-grams (blocs of 4 words) appearing in these series than do occasional viewers. Having determined the 50 most frequent 4-grams in a 500,000-word corpus of 5 of the students’ favourite series, these chunks were integrated into some simple sentences and recorded using an American native-speaker. Students listened to the sentences and were asked to rate them on a 4-point scale of recognition and understanding. In order to verify comprehension, translations were requested. Frequent viewers significantly outperformed infrequent viewers.
9Sockett (forthcoming) has also documented more frequent use of these same 4-grams in elicited writing of fan fictions.
10The relationship between music and language has been studied from many angles, including the similarities of perception and cognitive processing involved in their comprehension, their syntactical similarities, the effects of both on brain plasticity, or the primacy of emotional involvement in their learning and performance (Arbib 2013; Miras 2013).
11Regarding language learning, Lowe (1998) documents multiple sources indicating that
- 1 All the ellipses in this citation concern Lowe’s sources, which are too numerous to be cited exten (...)
certain commonalities exist between the processing, structure, and properties of language and music [...]1. Moreover, since auditory training is fundamental in second-language learning [...] and that music, like language, is initially acquired through the aural sense [...], researchers in second-language education [...] suggest that the incorporation of music into the second-language is a viable teaching strategy and should be considered (1998: 34).
12This quote points to the importance of listening in second language acquisition and suggests that the commonalities between music and language appear to be sufficient indication for using music in language teaching. This would render the predominance of listening activities reported in OILE studies all the more relevant, despite the important differences that can be cited between song and spoken language, for example deformations of both syntax and phonology in songs, the choice of particular types of vocabulary, or the potential interference of specifically musical information: melody, rhythm, instrumentation, voice quality and so on.
13As a possible aid to language learning, it has been suggested that music can help in several areas, from the development of listening comprehension, as indicated above, to pronunciation or vocabulary acquisition. Beasley & Chuang (2008) cite several authors who invoke music as a tool to enhance cultural literacy. Milovanov et al. (2010) found better pronunciation of English in young Finnish adults with higher musical aptitude, based on their comparative results on a pronunciation test, a phonemic listening discrimination task, and a test of musical aptitude.
14Boulton (1999) studied words that students associated with music and found that these opened up whole new lexical areas, varying from genres to examples of musicians. He also asked students to translate the names of well-known pop groups and found that the meanings (though sometimes approximate) of many of this sort of common noun, while often low-frequency words (such as pumpkins, settlers or worms), were often clear to students, producing no particular difficulties of translation. Other words were more problematic: although familiar to the students, they seemed to function more as proper nouns than as semantic items referring to something other than the pop group in question. Boulton’s study is helpful in pointing out the role of such informal activities in the development of vocabulary.
15Li & Brand (2009) carried out a comparative study on three groups of university ESL learners in Hong Kong, using a pretest-posttest-delayed-posttest experimental design, with one group exposed exclusively to learning English through music, another group exposed to no music for their English classes and a final group having half their lesson time taught using music and half without. Their findings showed that varying the degree of use of songs produced differential English language achievement, with the all-music group attaining the highest scores on both vocabulary and language use achievement posttests and delayed posttests. Deliberate classroom use of music would seem to show that language learning is taking place.
16Beasley & Chuang (2008) studied the effects of listening repetition, song likeability and song understandability on students’ perceptions of progress in English and on a vocabulary test. They found no significant correlations, but esteem that this might be due to the task design, where repetitions were generally limited to a single session (although often numerous within that session) and not over an extended period of several weeks. As Beasley & Chuang’s task design seems to affect precisely those aspects of listening which informal listeners control themselves, the transposition of such questions to the field of incidental acquisition would appear all the more pertinent.
17Given the promising nature of these studies, as well as those indicating language uptake by watching television series (see above), we wondered to what extent students picked up frequently occurring lyrics from popular hits and if so whether they delved also into the meanings of such lyrics. This is explored below in the results section.
18Recently, use of music services such as Spotify or Deezer, online music sales services such as iTunes and a range of practices by which listeners obtain MP3 copies of music from sites such as Youtube, mean that non-native speakers may be more actively engaged with their selection of music than young people of a previous generation, especially if we consider those for whom radio was the primary source. In comparison to the listening habits of previous generations, listeners today can be more active in their choice of music, have more control over how it is played and have more resources available to help them understand the lyrics.
19Guichon, in his research into language learning and teaching technologies, lists controls such as pause and rewind buttons as aids to listening comprehension (2011: 124). Roussel et al. (2008) filmed and analysed language students’ use of controls during listening comprehension exercises and determined four different types of listening strategies. More competent listeners tended to use a more global listening strategy, at least the first time around, pauses and rewinds being used during second and third listenings in order to verify what was heard or its meaning. The simple fact of being able to exercise control over the sound input was determined to improve overall comprehension. The existence of these affordances within OILE practices, whether they be music or video, would therefore suggest that they are likely to facilitate comprehension.
20Guichon (2011) also includes L2 texts (transcriptions) as important aids to listening comprehension. Hamon’s (2007) inventory of help functions in on-line learning environments documents an overwhelming use of text provided as support for various types of language learning activities, probably as a result of facility for the authors/editors. She deplores the lack of multimodal help functions (for example images to explain audio extracts) which would exploit the full range of ICT. Chanquoy, Tricot, & Sweller (2007) explain how text aids to listening lessen cognitive load through the “modality effect”: the use of two different channels of perception (in this case auditory and visual) allow superior processing of the (complementary) information provided because of an increase in short term memory (the information is not all processed in the same “place” at the same time). However, this sort of multiplication of the same information via different modalities can have negative consequences, especially for “experts”, as it will use up cognitive resources (split attention) to no added value. This is known as the “redundancy effect” (Sweller et al. 2011). Sockett & Toffoli (2012) found substantial evidence of concurrent listening to songs and use of lyric websites in their longitudinal diary study of five informal MSOE students, who indicated many instances of such use during the 2-month duration of the study.
21We therefore hypothesize that the combined influence of the three elements mentioned above, choice, control and supplementary resources, is likely to favour the active engagement of OILE practitioners with their music, including an interest in its propositional content. Our present investigations explore these three aspects of listening to music as part of an ongoing study of online informal learning.
22In order to address the questions concerning both the breadth of these phenomena and students’ perceptions of their impact on learning, this article compiles results from two data sources. It first exploits a pre-test designed to identify the English-language music listened to by students whose only formal contact with English was a two hour per week language requirement over the 24 weeks of a French academic year. The ensuing survey, carried out in November 2012, asked Humanities and Arts students more detailed questions about the music they listen to in English, the type of listening practices they engage in and elicited translations of four lyrics from songs by artists determined to be popular with respondents.
23227 copies of the questionnaire, distributed to MSOE students during their two-hour once-a-week required English class, were collected. Of these, 207 were returned with responses completed for the questions on music. Two thirds of this panel were women (67 %) and the vast majority (90 %) of French nationality. The age spread was from 17 to 34, the average being 20.
24Basic identifying information (age, gender, nationality, level in English) was included in a first part of the questionnaire on OILE in general. Questions relating to use of music appeared on a second page. Initial questions on this second page target information concerning how and where music is listened to (duration, mobile devices, at home or outside, music sources), quantity and proportion of music in English. (See Appendix 1 for a complete version of the questionnaire.)
25The extent to which respondents actively chose the music they listened to was investigated through a range of questions relating to favourite artists and the means by which audio files were obtained. The use of controls (such as pause and rewind buttons) and resources (such as lyrics websites) were also addressed. A final question attempted to elicit some information relating to content uptake, through the identification and translation of a few short lyrics of widely downloaded songs.
26Since one of the purposes of this study was to examine ways in which learners interact with their music, the extent to which music was actively chosen was an important element of the research. Identifying whether students mostly all listened to the same music or had highly individualised tastes is an important preliminary step in characterising this choice, while information about which English-language singers were listened to most frequently is essential if the choice of lyrics for the identification and translation exercise is to be pertinent.
- 2 The same type of students attending the same learning centre, but not necessarily the same individ (...)
27In a preparatory study, 122 learners2 had been asked to list the five English language groups or singers they listened to most, in order to determine whether a questionnaire on lyrics of songs listened to by a majority of students might be a worthwhile methodology. Previous studies of television series (Kusyk & Sockett 2012) revealed a small number of series viewed by a large number of learners as well as a second much larger number of series watched by only a small number of viewers.
28Responses to this preparatory study were remarkable in their diversity. The 122 learners surveyed listed 299 different singers and groups including 199 performers mentioned only once. Only 5 performers were mentioned by more than ten respondents (table 1).
Table 1. Performers mentioned by more than ten respondents
29The fact that no more than 20 respondents (16 % of those questioned) claimed to listen to the same music, and that many respondents listened to music listed by no-one else in the sample, both hinted at the choice and control levels students have over the music they listen to and made it apparent that asking learners to identify particular lyrics may be a problematic methodology. It was nevertheless determined that the translation exercise would be included in the questionnaire in order to see whether any more general effect could be attributed to frequent listening in spite of the great diversity of individual listening choices.
30The results from the main questionnaire are summarised below, with particular attention paid to the content translation questions.
31The survey responses allow some qualification and quantification of the changing practices students are engaging in (what music, where and how it is obtained, how often and how long it is listened to, using static or mobile equipment, and so on). The analysis of the translations of song excerpts hints at language uptake that may be taking place.
32The reported levels in English of the respondents, as determined by placement tests and confirmed by instructors at the beginning of the academic year, situate the population as follows: A1 – 15 %; A2 – 39 %; B1 – 41 %; B2 – 5 %.
33Listening to English language music either at home or on the move is an activity which involves 100 % of students surveyed, with some 89 % of the 207 respondents answering positively for home-based listening and 79 % for mobile listening.
34On average, students claim to listen to 1-3 hours of music daily, although 30 % of respondents quantify their listening as being in excess of 3 hours per day. These numbers are important in that they describe an activity which may take up some 700 hours annually for an average student. Furthermore, over 75 % of students report listening to a majority of English language music, with a third of those listening almost exclusively to songs in English. Taken together these initial results suggest that university students in France listen to several hundred hours of English language songs every year, while their exposure to English in formal classes is probably less than 50 hours over the same period of time.
35In order to investigate the extent to which learners interact both cognitively and physically with their music, respondents were initially asked whether understanding lyrics interested them. English music has long been a background presence in France, particularly since the end of the state monopoly of radio in 1981. However song lyrics are not always easy to understand even for native speakers, and obtaining transcriptions of lyrics in the past would often require purchasing the album which may or may not include a sheet of lyrics. The affordances of Web 2.0, particularly multi-user interfaces, mean that informally transcribed song lyrics can easily be found online, often with suggested translations into a speaker’s native language. Other recent changes relate to recording quality, with digital recordings and modern speakers and headphones allowing music to be heard more clearly than it might have been in the past. These changes could have an impact on whether students consider English music primarily as background noise or as meaningful content. It has also become possible with portable digital music players to carry around a very large quantity of listening material and to pause and replay songs or parts of songs or to choose to listen to the same song repeatedly. Over 80 % of participants in this study indicate that the propositional content of songs is something they are interested in. These listeners also demonstrate a belief that they understand the words of songs, situating their level of comprehension at around 50 % of what they listen to.
36Questionnaire responses situate the proportion of students making occasional or frequent use of lyrics websites at 46 %, while 43 % of those questioned make occasional or frequent use of controls (such as pausing) while listening to music.
37From the websites given as examples of sources for lyrics, two main modes of operation become apparent. Respondents either list the website they most frequently use to search for lyrics, or they indicate that they type partial lyrics into Google and select either the first website listed, or a website with which they are familiar from the search results. Hence searching for lyrics can either be deductive, based on the title and artist, or inductive, based on learner perceptions of part of the lyrics. The most frequently consulted website for lyrics is www.lacoccinelle.net (cited 34 times), followed by www.google.com (11 citations), and trailed by 15 others (with 1 to 4 citations each).
38In this survey, students were also invited to indicate the means by which they obtained the media files they listened to when on the move. Although Youtube is by far the dominant mode of obtaining music files cited, using conversion software to make MP3 files, a wide range of other practices can be observed, from buying music from iTunes and making digital copies of their own or borrowed CDs to using specific peer to peer or other downloading services (although this only concerned some 10 % of students questioned) to various combinations of the above. What is important to retain from these results is the high degree of variety and personalization in learner behaviour.
39In order to ascertain whether frequent listening to music had any effect on knowledge of the propositional content of songs, the questionnaire concluded with four song lyrics which the participants were invited to attempt to translate. As discussed above, this approach is inspired by Kusyk & Sockett’s (2012) study of vocabulary uptake from English language television series.
40The lyrics were chosen from the most popular artists listed by respondents to the preliminary survey, although, as noted before, the wide majority of participants might not have listened to these songs before, as these artists may not be part of their personal repertory. The phrases were considered to be pronounced clearly in each case (this was a criterion for selection).
41These sentences were:
1. “I’ve heard that you’ve settled down, that you’ve found a girl and you’re married now” (Adele: Someone like you)
2. “‘cause I didn’t mean to hurt him, could have been somebody’s son” (Rihanna: Man down)
3. “When she was just a girl she expected the world” (Coldplay: Paradise)
4. “Is our secret safe tonight and are we out of sight?” (Muse: Resistance)
42The attempted translations were assessed by two independent raters on a scale scoring a maximum of five points for sentences two and four, up to four points for sentence one and two points for sentence three. Inter-rater reliability was good with a standard deviation of 0.3 for the 511 attempted translations.
- 3 This listening could include sources other than music.
43Results were divided into two groups, those learners who claimed to listen to English informally more than once a week (the frequent group)3 and those who claimed to do so less than once a week (the infrequent group). This choice allowed the respondents to be divided into groups of relatively similar size, 132 for the frequent group and 94 for the infrequent group.
44For all four translation questions the frequent group outperformed the infrequent group in both attempts to translate and quality of translation as assessed by the independent raters. These results are presented in table 2.
Table 2. Results for the frequent / infrequent groups
45The four outcomes for translation quality, taken together, were then submitted to a two-tailed t-test. The overall difference in translation scores was found to be significant at the 0.05 level (t = -2.55), although falling slightly short of significance at the 0.01 level (critical value = 2.59)
46These results tend to support the perceptions of learners obtained from the earlier parts of the questionnaire which indicate an interest in and partial comprehension of the propositional content of songs.
47Ranking the translation scores also gave insight into other independent variables in the study. Since translating song lyrics also involves written skills, average total scores (out of 16 for the four translation questions) were established for learners claiming to read frequently on the Internet and for those claiming to read infrequently. The same calculation was made for those claiming to use lyrics websites frequently and infrequently. Since a question was asked about the use of controllers (such as pause buttons), average total scores for those who claimed to use such functions frequently and infrequently were also obtained. These totals are presented in tables 3a-d below.
Table 3a. Results for listening activities
Table 3b. Results for reading activities
Table 3c. Results for the use of lyric sites
Table 3d. Results for the use of controls
48From these results, it can be seen that frequent online reading activities are the best predictor of translation score, outperforming even frequent listening activities. Indeed in a two-tailed t-test the frequent reading group significantly outperforms the infrequent group at the 0.05 level (t = 2.14). While frequent use of lyrics sites corresponds to a small increase in average total translation score, use of controls cannot be shown to vary with translation score.
49The complexity apparent in the listening practices of students in this survey is illustrated by the great variety of performers listed in the preparatory survey. Whereas in the past one might have considered that popular music was dominated by a limited number of performers (the Beatles, the Rolling Stones, the Who, the Kinks, etc.), it is apparent from these results that the great diversity of music in the modern world and the range of modes of online access to it mean that learners can actively choose the music they listen to, and listen to unique repertoires of songs, exposing them to English in forms, ways and quantities which are specific to each learner.
50These facts mean that the translation activity presented in this study provides only a limited insight into the impact of listening to songs on comprehension of lyrics. Where Kusyk & Sockett’s (2012) study was able to test learners on structures they had heard frequently in television series, it proved difficult in this research to demonstrate that the lyrics chosen for translation had indeed been frequently listened to by those taking part. It is however interesting that despite this methodological challenge, statistically significant general tendencies did emerge in the relationship between online informal reading and listening activities and scores in the translation test. This suggests that OILE activities in general may foster greater awareness of the propositional content of songs.
51A further limitation of our methodology is that self-report is a somewhat biased and not necessarily reliable indication of what is really going on in terms of learner practices and preferences. The type of tracking necessary to make more objective observations, comparable to what could be accomplished in a classroom study, was considered to be encroaching excessively into the learners’ personal habits.
52When seeking to assess the extent to which learning is taking place in OILE contexts, the quantitative data presented relating to the number of hours of exposure is a first significant element. It is a finding of this study that most students are exposed to several hundred hours of music with English lyrics annually, the use of mobile devices contributing significantly to this figure. It is likely that these hundreds of hours of listening also involve listening repeatedly to the same favourite songs, and mobile listening devices have features such as playlists and loop functions which facilitate these repetitions.
53Since 95 % of the learners surveyed were below B2 level according to the CEFRL their actual listening skills are likely to be limited; the descriptors for levels B1 and A2 emphasize the need for clarity which is often absent in songs.
A2: Can understand and extract the essential information from short, recorded passages dealing with predictable everyday matters which are delivered slowly and clearly.
B1: Can understand the main points of radio news bulletins and simpler recorded material about familiar subjects delivered relatively slowly and clearly.
B1: Can understand the information content of the majority of recorded or broadcast audio material on topics of personal interest delivered in clear standard speech
(Council of Europe 2000: 68, our emphasis).
54It can therefore be argued that without the high levels of repetition and interaction reported in this study, comprehension of lyrics would be extremely difficult for learners at this level.
55It can also be argued that listening to music plays a role of additional sensory stimulation (Wright et al. 2010) for all learners, contributing to the learning impact of other encounters with English. Finally, we can claim that many learners also actively engage with the music through their interest in understanding lyrics. This often takes the form of consulting transcripts, simultaneously to the listening. Referring back to the theory of cognitive load (Chanquoy et al. 2007) it would appear unsurprising that students at A2-B1 levels of English would make use of these affordances.
56Creating a unique musical universe is also an expression of identity, and the fact that second language content plays a role in this identity is significant, as Norton (2000) or Dörnyei & Ushioda (2009) have suggested. The change in listening experience, from collective to individual, underlines the concept of music as identity, as listening is less frequently a shared experience and more often an auditory barrier, provided by high quality headphones, separating the listener from his/her physical environment of bus or street and enabling Kuure (2011) and others to argue that place is performed through a series of choices rather than imposed by geography.
57The finding that music is chosen, and that learners often pause and rewind music and use lyrics sites indicates that far from being background noise, English-language music is an important part of learner identity and can potentially contribute to the incidental acquisition of English in interaction with other OILE activities.
58We have seen that the affordances of recent online technologies and mobile devices appear to be providing a new role for music in language learning through OILE. The results presented here indicate the widespread presence of English-language songs among MSOE students and also point to the fact that these students are interested in comprehending the meaning of the songs they listen to, perhaps even in affirmation of some particular aspect of their L2 identity. It would thus be desirable to obtain further information on individual differences as they relate to the use of music and informal language learning - for example how, why and when some students sing along with the music they listen to and their perceptions of the utility and efficiency of so doing. In a series of student blogs currently under study, some students have stated that listening to and singing along with music helps their pronunciation or even language proficiency in general. A research design to test these perceptions, perhaps for changes over time, would also complete the results presented here. Studying students’ own perceptions of these phenomena, through a close analysis of these blogs, or of some other type of reflexive learning documents would provide more in-depth and qualitative information about these music and language practices, rounding out the perspective that has been presented in this study and providing for a more robust all-round comprehension of the part informal listening to songs can play in long-term language acquisition.