"In general, when modeling phenomena in science and engineering, we begin with simplified, incomplete models. As we examine things in greater detail, these simple models become inadequate and must be replaced by more refined models."

Clearly stated, from Structure and Interpretation of Computer Programs. Interesting how best process for learning mirrors the process for discovery. (Although that nice, neat story of discovery is also a useful but inadequate model, that people like Kuhn challenge and refine.)