Explaining the Computational Mind, by Marcin Milkowski.
Erik Myin, Karim Zahidi · Mind · 2015
The idea that the mind is ‘computational’, and so should be explained in a computational way, is ubiquitous in both philosophical and scientific thinking. But behind the semblance of unity created by the use of this word, one observes fundamental disagreements about the concept of computation. Some philosophers have famously held that computation remains the only hope for a science of thought and behaviour. Others have argued that cognition is not in fact computational at all. In the sciences of the mind, a wide variety of different approaches and types of models for mental and cognitive processes, be they symbolic, connectionist, or dynamical, have all been labelled as ‘computational’. Given that the notion of computation is both foundational and yet controversial, it should come as no surprise that a distinctive area of philosophy dealing specifically with the nature of computation has emerged. One of the most elaborated accounts of computation to have arisen out of this branch of philosophy is Gualtieri Piccinini’s laudably precise and clear mechanistic account of computation (see his ‘Computation without Representation’, Philosophical Studies , 137 (2008), pp. 205–41 for an early version, and Gualtieri Piccinini and Andrea Scarantino, ‘Information Processing, Computation, and Cognition’, Journal of Biological Physics, 37 (2011), pp. 1–38 for some further elaborations). According to this view, abstract computation is a process in which medium-independent vehicles are transformed according to general rules. A computational process then is a causal process that implements the general rules of an abstractly conceived computation. This notion of computation subsumes digital computation, analogue computation, and neural computation, and thus, according to Piccinini, accommodates various notions of computation used by scientists. It is a non-semantic account of computation, since the general rules do not operate on the basis of the contents of the vehicles, but rather on differences between certain portions of the vehicles. Such a non-semantic account is heralded to be a philosophical bonus, given that in semantic accounts the fate of one’s account of computation becomes dependent upon the fate of one’s theory of content.