Cognitive science is an intrinsically interdisciplinary field of study. Why is this so? In the current chapter, I argue that the interdisciplinary nature of cognitive science necessarily emerges because it assumes that cognition is information processing. The position I take is that explanations of information processors require working at four different levels of investigation, with each level involving a different vocabulary and being founded upon the methodologies of different disciplines.
The chapter begins with a historical treatment of logicism, the view that thinking is equivalent to performing mental logic, and shows how this view was converted into the logical analysis of relay circuits by Claude Shannon. Shannon’s work is then used to show that a variety of different arrangements of switches in a circuit can perform the same function, and that the same logical abilities can be constructed from different sets of core logical properties. Furthermore, any one of these sets of logical primitives can be brought to life in a variety of different physical realizations.
The consequence of this analysis is that information processors must be explained at four different levels of investigation. At the computational level, one asks what kinds of information processing problems can be solved by a system. At the algorithmic level, one asks what procedures are being used by a system to solve a particular problem of interest. At the architectural level, one asks what basic operations are used as the foundation for a specific algorithm. At the implementational level, one asks what physical mechanisms are responsible for bringing a particular architecture to life.
My goal in this chapter is to introduce these different levels of investigation. Later chapters reveal that different approaches within cognitive science have differing perspectives on the relative importance, and on the particular details, of each level.