Investigating song-based language teaching and its effect on lexical learning

The present thesis addresses the following question: Can popular songs as they are currently used in second and foreign language classrooms benefit lexical learning? Lexical learning is defined as the acquisition of new vocabulary as well as the consolidation and further elaboration of familiar words and phrases. To answer this research question, three methodologically distinct studies are reported.  In the first study an international questionnaire explored teacher cognitions as well as actual teaching practices involving songs. The responses of 568 informants in 41 countries indicate that a majority of respondents believe in the usefulness of songs for language learning and that many respondents utilize songs in class for clearly defined pedagogical purposes, including vocabulary learning. The questionnaire also elicited information from the respondents about the way they incorporate songs in lessons, including details about how often a song is played and what types of form- and meaning-focused activities are used to engage learners with the lyrics of a song.  The second study investigated the lexical characteristics of teacher-selected songs and the vocabulary learning opportunities they afford. For this purpose, a corpus of 635 songs used for ESL/EFL purposes, comprising 177,384 tokens, was compiled and analysed. Results indicate that teacher-selected songs are short, repetitive and relatively undemanding as far as lexis is concerned compared to other authentic text genres. Knowledge of the 4000 most frequent word families of English provides 98% coverage of the running words in this song corpus. Little difference was found in terms of the overall vocabulary demands between songs intended for use with beginners, intermediate and advanced learners.  The third study investigates whether participating in a song-based lesson results in higher verbatim text retention compared to a lesson based on a poem or a prose text. For the sake of ecological validity, the procedures and the materials used in the classroom intervention study were informed by the findings of the teacher questionnaire (study one) and the song corpus analysis (study two). Results indicate that a song-based language lesson but also a poem-based lesson result in significantly higher recognition and cued recall of verbatim text than a lesson based on a prose text.  In response to the overall question, this thesis provides evidence that songs as they currently tend to be used by language teachers around the world indeed benefit certain aspects of lexical learning, perhaps in particular the entrenchment in memory of already (half-)familiar words in association with their phraseological patterning. It is argued that, while certain structural characteristics of songs (and poems) have the potential of rendering text (and the lexis therein) memorable, it is the way that songs tend to be exploited in the classroom that capitalizes on this mnemonic potential.