Department of Theoretical Philosophy
Prediction Error Minimization:
Metaphysics, Methodology, and the Bounds of Cognition
In “The Extended Mind” (1998), Andy Clark and David Chalmers argue that, if certain conditions are met, human cognition extends beyond the brain boundary into the external world. This is the thesis of Extended Cognition (EC). The EC thesis has resulted in a fierce debate with authors who maintain the traditional view that cognition is bounded by the brain. In recent years, the debate on extended cognition has reoccurred in all its glory within Clark and Jakob Hohwy’s respective discussions of the Prediction Error Minimization (PEM) framework. PEM is a neurocomputational theory that states that the brain is essentially a hierarchical prediction machine. Within the PEM framework, cognition comes down to minimizing the error value of predictions that the brain makes about the world. Whereas Clark argues PEM is compatible with the EC thesis, Hohwy believes PEM and EC are incompatible.
The re occurrence of the EC debate within the PEM framework seems to be a rather puzzling phenomenon because both Clark and Hohwy reach their opposite conclusions from within the fine theoretical strictures set by the PEM framework. Whereas the received view considers the EC debate to reoccur within the PEM framework as the result of Clark and Hohwy’s entertaining two different interpretations of PEM, this paper proposes that the debate’s reoccurrence is best understood as the result of two different interpretations of the EC thesis. Whereas Clark argues for the compatibility between PEM and a methodological EC thesis, Hohwy argues against the compatibility between PEM and a metaphysical EC thesis. This proposal is advantageous over the received view because it enables us to explain the disagreement between Clark and Hohwy while still doing justice to their mutual commitments to the PEM framework.
|Last modified:||31 August 2017 12.22 p.m.|