This chapter explores the ways irony unfolds in music. Turner and DiBernardo examine representative pop songs, both original compositions and cover versions, to suggest several ways that irony is created and perhaps detected by listeners. As they argue, “Musical irony requires an interpretive ear for hearing contradictory or disjunctive sounds (and lyrics) within a musical context.” But inferring irony from music involves a special challenge given that music lacks it own semantic or representational signification. Lyrics are clearly a driving force in expressing ironic intent, but instrumental sounds often interact with the spoken words to convey richer ironic complexes, including both rhetorical and situational ironies. Listeners may be especially attentive to the tension, or the discrepancy, between the musical form, style, or genre of a song (e.g., the upbeat, lyrical form in Randy Newman’s song “Political Science”) and its lyrical content (e.g., the use of weapons of mass destruction). Many musical ironies may be “post-modern” because of their self-referential style (e.g., not just criticizing others, but ourselves as well). This chapter offers a compelling, beautifully detailed, argument that “music is a largely underexplored wellspring of ironic activity.”

The model fitting and estimation approach is laid out using two simple linear models, one for a continuous biological predictor variable and one for a categorical predictor. These two models are the familiar simple linear regression and the single-factor ANOVA. We show how these two models are variations on a theme and describe how to fit them to data. The model fitting is treated in detail, laying the foundation for more complex models in the following chapters. We emphasize what the model parameters mean, how to estimate them, calculate standard errors and confidence intervals, and test hypotheses about them. For categorical predictors, we introduce and recommend planned comparisons (contrasts) to examine patterns across categories. Checking assumptions and identifying unusual and influential data is detailed, as is the use of power analysis to determine necessary sample sizes.