A Spin-based Measure of the Coherence of Belief Systems

In social science and humanities, coherence is an important but vaguely defined concept used for understanding the structure of belief systems and their physical representations like political parties. In our contribution, we provide a specified definition and measure of coherence based on superparamagnetic clustering. The measure captures the intuition, that coherence includes both a static component (diversity of sub-groups in the system) and a dynamic component (stability of the system under stress). We apply our measure of coherence to data representing the political beliefs of candidates in the Swiss national elections of 2003 and 2007 and explain the split-up of the Green party in 2004 as a result of lacking coherence both in its static and dynamic component.