ABSTRACT
The intuition that the world can be ordered into a hierarchy of levels is nearly ubiquitous in cognitive neuroscience and the special sciences generally. However, scientists and philosophers use the term “level” to describe a wide variety of relations, and considerable confusion arises from a failure to keep them distinct. For this reason, it is best not to speak of “levels” simpliciter, but rather to speak of “levels of R,” where R is some explicit specication of a relation between levels or of the relata at different levels. Here I develop a eld guide to different senses of level in the special sciences. My ultimate goal is to focus on levels of mechanisms, which play a central explanatory role in sciences such as experimental biology and neuroscience. I extract some features of this variety of levels by looking at a well-understood exemplar, which I call the levels of learning and memory. These levels (sketched in the third section) are described as aspects of one of the most successful, if controversial, explanations in contemporary cognitive neuroscience. I develop a taxonomy of different senses of level, and I show that each is either inappropriate or incomplete as a description of the levels of learning and memory. I then argue that levels of mechanisms capture most clearly the explanatory sense of levels ubiquitous in contemporary cognitive neuroscience. I close by contrasting levels of mechanisms with levels of realization, the target of much recent debate in the metaphysics of mind.
