Despite several decades of criticism, dichotomous thinking about behavioral development (the view that the behavioral phenotype can be partitioned into inherited and acquired components) remains widespread and influential. This is particularly true in study of birdsong development, where it has become increasingly common to diagnose songs, elements of songs, or precursors of songs (song templates) as either innate or learned on the basis of isolation-rearing experiments. The theory of sensory templates has encouraged both the dichotomous approach (by providing a role for genetic blueprints to guide song learning) and an emphasis on structural rather than functional aspects of song development. As a result, potentially important lines of investigation have been overlooked and the interpretation of existing data is often flawed. Evidence for a genetic origin of behavioral differences is frequently interpreted as evidence for the genetic determination of behavioral characters. The technique of isolation rearing remains the methodology of choice for many investigators, despite the fact that it offers only a rather crude analysis of the contribution of experience to song development and provides no information at all about genetic contributions to development. The latter could in principle be elucidated by the application of developmental-genetic techniques, but it is unlikely that these can easily be applied to the study of birdsong. Because developmental questions are so often posed in terms of the learned–innate dichotomy, “experience” is taken to be synonymous with “learning” and the possible role of nonobvious contributions to song development has largely been ignored. An alternative approach, based on Daniel Lehrman's interactionist theory of development, permits a more thorough appreciation of the problems that have yet to be addressed, and provides a more secure conceptual foundation for theories of song development.