Advancing the inference of performance in birdsong

It is appealing to integrate different acoustic traits to infer differences in performance demands among birdsongs, and to use this as a tool for investigating which roles song performance plays in communication. But inferring performance from acoustic measurements introduces a degree of interpretation that can cause disagreement. Here I give an overview of approaches to assess song performance, associated methodological issues, and ways of addressing them. I note advantages and limitations of performance metrics derived from physiological principles or from acoustic trade-offs, discuss issues with the scaling of performance metrics, and with choosing and adapting metrics to different study species and research goals. Throughout I emphasize that these metrics provide tentative assessments of performance, and that empirical results should be interpreted by comparison to alternative hypotheses.