What is the basic structure of belief systems? Clear answers to this fundamental question are not forthcoming. This is because we typically measure a belief system by averaging across several relevant items and forming a scale (e.g., of conservatism). This approach cannot assess a system of beliefs because it cannot assess the network of connections between the beliefs that make up the system; it collapses across them and the interrelationships are lost. In this talk, I will present new work where I conceptualize and analyze attitudes and identities as interactive nodes in a belief system network. With this approach, representative survey data, and simulation studies, I examine two key questions in research on political belief systems: (Q1) What is central to belief systems and behavior (A1: identities, preprint)? (Q2) What belief system structure leads people to become ideologues (i.e. people with consistent and stable belief systems) (A2: dense networks with consistently positive connections)? These questions (and preliminary answers) are an initial step towards taking seriously the idea that belief systems are in fact systems.