The purpose of the current chapter was to introduce the foundations of classical cognitive science—the “flavour” of cognitive science that first emerged in the late 1950s—and the school of thought that still dominates modern cognitive science. The central claim of classical cognitive science is that “cognition is computation.” This short slogan has been unpacked in this chapter to reveal a number of philosophical assumptions, which guide a variety of methodological practices.
The claim that cognition is computation, put in its modern form, is identical to the claim that cognition is information processing. Furthermore, classical cognitive science views such information processing in a particular way: it is processing that is identical to that carried out by a physical symbol system, a device like a modern digital computer. As a result, classical cognitive science adopts the representational theory of mind. It assumes that the mind contains internal representations (i.e., symbolic expressions) that are in turn manipulated by rules or processes that are part of a mental logic or a (programming) language of thought. Further to this, a control mechanism must be proposed to explain how the cognitive system chooses what operation to carry out at any given time.
The classical view of cognition can be described as the merging of two distinct traditions. First, many of its core ideas—appeals to rationalism, computation, innateness—are rooted in Cartesian philosophy. Second, it rejects Cartesian dualism by attempting to provide materialist explanations of representational processing. The merging of rationality and materialism is exemplified by the physical symbol system hypothesis. A consequence of this is that the theories of classical cognitive science are frequently presented in the form of working computer simulations.
In Chapter 2, we saw that the basic properties of information processing systems required that they be explained at multiple levels. Not surprisingly, classical cognitive scientists conduct their business at multiple levels of analysis, using formal methods to answer computational questions, using simulation and behavioural methods to answer algorithmic questions, and using a variety of behavioural and biological methods to answer questions about architecture and implementation.
The multidisciplinary nature of classical cognitive science is revealed in its most typical methodology, a version of reverse engineering called functional analysis. We have seen that the different stages of this type of analysis are strongly related to the multiple levels of investigations that were discussed in Chapter 2. The same relationship to these levels is revealed in the comparative nature of classical cognitive science as it attempts to establish the strong equivalence between a model and a modelled agent.
The success of classical cognitive science is revealed by its development of successful, powerful theories and models that have been applied to an incredibly broad range of phenomena, from language to problem solving to perception. This chapter has emphasized some of the foundational ideas of classical cognitive science at the expense of detailing its many empirical successes. Fortunately, a variety of excellent surveys exist to provide a more balanced account of classical cognitive science’s practical success (Bechtel, Graham, & Balota, 1998; Bermúdez, 2010; Boden, 2006; Gleitman & Liberman, 1995; Green, 1996; Kosslyn & Osherson, 1995; Lepore&Pylyshyn, 1999; Posner, 1991; Smith&Osherson, 1995; Stillings, 1995; Stillings et al., 1987; Thagard, 1996; Wilson & Keil, 1999).
Nevertheless, classical cognitive science is but one perspective, and it is not without its criticisms and alternatives. Some cognitive scientists have reacted against its avoidance of the implementational (because of multiple realization), its reliance on the structure/process distinction, its hypothesis that cognitive information processing is analogous to that of a digital computer, its requirement of internal representations, and its dependence on the sense-think-act cycle. Chapter 4 turns to the foundations of a different “flavour” of cognitive science that is a reaction against the classical approach: connectionist cognitive science.