INTRODUCTION

One of the earliest issues in cognitive ethology concerned the meaning of animal signals. In the 1970s and 1980s this debate was most active with respect to the question of whether animal alarm calls convey information about the emotional states of animals or whether they “refer” directly to predators in the environment (Seyfarth et al. [1980]; see Radick [2007] for a historical account), but other areas, such as vocalizations about food and social contact, were also widely discussed. In the 1990s, ethologists largely came to a consensus that such calls were “functionally referential” (Evans and Marler [1995]) even if they did not satisfy all the semantic requirements imposed by philosophers of language. More recently, though, it has been argued that ethologists should eschew the concept of reference and return to a focus on the affective aspects of animal communication (Rendall and Owren [2002]). We propose to take a new look at this debate in the light of recent developments in the philosophy of language under the heading of “neo-expressivism” (Bar-On [2004]). This view provides two different senses in which an utterance satisfies an expressive function. We intend to use neo-expressivism to provide a philosophical framework for understanding the relationship between the affective and referential aspects of animal signals by seeing them as both acts that express some motivational state of the animal and products that express propositions with truth-evaluable content.