Decisions are central to cognition, but many decisions, and much of life, consist of larger problems that can be decomposed into smaller ones. How then can simple, individual decisions, such as those regarding the direction of motion of a patch of moving dots, be organized sequentially and/or hierarchically to give rise to complex thought? The circuit structure of a modern computer certainly allows for such spatio-temporal organization, but it is likely that some features of computer architecture will not be found in the brain (e.g., a central system clock that updates most circuit elements synchronously). Even if such an organizational structure held for the brain, however, are drift-diffusion models that unify simple conditioning, decision making and timing processes up to the job of complex sequencing? I have focused much of my theoretical work on this question, with the goal of unifying models of simple behaviors with production system models – or more precisely, cognitive architectures – capable of simulating human problem solving, reasoning and planning (e.g., ACT-R or Soar). Preliminary attempts at this unification – my own and those of others – have shown promise in explaining sequential, hierarchical organization by asynchronous neural systems, and in explaining the cognitive, problem solving deficits that should be expected under conditions of, e.g., prefrontal brain damage and Parkinson’s disease (Simen & Polk, 2010). Complete unification has been held up by problems like the classic “binding problem”, and I am currently investigating theoretical models that have the potential to solve them (Simen, Van Vugt, Balci, Freestone & Polk, 2010).