1. Scientific explanation and prediction, plus levels of explanation.
2. The mind/brain as a computing (information manipulator) machine (input-output device).
2a. What kind of a machine? In traditional terms, it must be one that is more powerful than a finite-state device, but more constrained than a universal Turing machine.
2b. What is the language of thought (i.e., mental code or mental programming language) of such a machine?
2bi. a symbolic code, rich and abstracted form the external world with units, patterns, and rules expressly specified? That is, a ready-made knower?
2bii. a shallow network of associations, more closely reflecting the external world, with units, patterns and rules emergent out of the overall operation of the code? That is, a ready-made learner?
2c. What kind of machine is the brain? Depends on the level of explanation: a network of neurochemical associators (i.e., good learners?) organized in assemblies of dedicated processors (i.e., ready-made knowers?).
3. The overall mental architecture: i.e., the full configuration of the mind-machine: modularity and its discontents. Some dichotomies (or gradients?): modularity/interactionism, symbolic/associationist, serial/parallel. At the cognitive level, is the architecture modular, symbolic, and parallel, but implemented as a modular, associationist, and parallel system?