Too messy or too neat? Assessing the limits of computational models of the brain

Abstract

Mazviita Chirimuuta has written a timely book that reinvigorates a classic argument against mechanical minds: biological naturalism, the view that consciousness and general intelligence depend upon life. A central commitment of this book is a kind of Kantian humility about the prospects for our knowledge of the brain: it holds that because there is a fundamental difference in kind between silicon-based computational “maps” and the biological “territory” of the brain, no scientific model can provide us with knowledge of the “brain-in-itself.” In this commentary, we flesh out two arguments Chirimuuta offers for biological naturalism. We agree with her that these arguments provide support for biological naturalism, but only if she abandons her Kantian humility about our knowledge of the brain and embraces a more dialectical, Hegelian approach to the relationship between scientific models and their targets.