Cognitive Science: An Introduction/Computational theory of mind

Most cognitive scientists are computational functionalists. This means that a being is in a mental state not because of the physical properties of the being, but in terms of the computations implemented with the physical properties. So having a belief, or planning, is defined not by some set of neural processes, but the functional algorithm those neural processes are implementing. This idea originated with philosopher Hilary Putnam in the 1960s.[1]

Let's take some information processing machine as an example. We might be able to characterize its behavior in terms of a flow chart that describes how one state might lead to another state. At the same time, we can describe the machine in terms of its physical parts--it might be made of wood, or circuitry. The flow chart, in this example, might be the functional description of the system, and the description of the physical parts and how they interact would be the physical description of the system. This is how we tend to think of the difference between software and hardware in computer systems.

Likewise, as computational functionalists believe, you can describe human minds the same way. There is a functional description (at the cognitive level) of how the information interacts, and then there is a physical description of how the neurons interact. The neurons implement the functional system.

Crucial to this is that there can be multiple physical ways to implement the same function. You might do multi-column addition in your head, and I might too. We would both be implementing the same algorithm or function. But at some level of physical detail, your brain is doing it differently than my brain is. Taken to an extreme, any physical system implementing the right functional algorithm would have that mental state--even if it's a "computer" made of swiss cheese, if it's implementing the right algorithm, it would have beliefs, feel pain, or be conscious.

We use functional definitions outside of cognitive science. Let's take the concept of a bandage. A bandage might be defined as something that can be or is being used to stop bleeding or to hold a limb in place. Whether the bandage is made of leaves, linen, cotton, or some other material is irrelevant, as long as it can implement the function. Computational functionalists view mental processes the same way. What makes addition addition and consciousness consciousness is the algorithm being run, no matter how it's implemented.

Identity Theory edit

This is in contrast with identity theory, which attempts to describe mental states in terms of their physical makeup. The might say that pain involves c-fibres, for example. This means that any being without c-fibres cannot be in pain. This strikes many as implausible.

Behaviorism edit

Behaviorism defines mental states in terms of behavioral dispositions. Pain might be described as the tendency of a being to avoid the harmful stimulus, screaming in the presence of harmful stimuli, and so on. It has the benefit of being objective, but falls prey to other criticisms that most contemporary cognitive scientists see as damning.

  1. Putnam, H. (1967). The Mental Life of Some Machines. In Hector-NeriCastan ̃eda(ed.), Intentionality, Minds and Perception. Detroit: Wayne State University Press, pp. 177–200.