The value of meaning for autonomous robots

![CDATA[This paper examines the related problems of meaning and symbol grounding with respect to epigenetic robotics. While symbol grounding aims to give artificial systems “intrinsic meaning”, most existing approaches to symbol grounding address meaning as a problem of categorical perception, i.e. a theory of reference for maintaining correspondence between internal representations and external entities at a linguistic level. We argue that reference is only one aspect of meaning, and that nonlinguistic and pre-verbal creatures are also meaning users. As such, we argue meaning plays an important role in an agent’s value system, providing intrinsic motivation and reinforcement for life-long development and learning. Lastly, we explore how models of meaning can help shift the intellectual burden of grounding from the programmer to the program by designing robots capable of grounding themselves.]]