Research into solving the problems that the recent rapid evolution of hardware and software technology and increasing consumer requirements has yielded, produced several Autonomic Computing initiatives, spearheaded by IBM. This paper contends that whilst this research is inspired by the Autonomic Nervous System (ANS), current implementations have treated the ANS as a single system, rather than as two disparate, competing systems: the sympathetic and parasympathetic nervous systems. In this paper, the authors show how the current limitations in state-of-the-art autonomic system implementations that hinder the successful representation of such sub systems and their conflicts can be overcome through the specification, design and development of a runtime introspective programming language, Neptune. An introduction to the Neptune specific CA-SPA (Concept Aided Situation-Prediction-Action) methodology for policy definition is given, as is the Cloud framework that hosts Neptune objects in an autonomic grid framework An evaluation of the language is given using illustrative examples based on sympathetic and parasympathetic scenarios, concluding with a summary of the further aims of the research.