|
|
| Program > Themes in the Philosophy of Science |
| Programme > Thèmes en philosophie des sciences |
| Gualtiero
Piccinini (University of
Pittsburgh) Computing Mechanism |
|
I offer a novel explication of what it means to be a computing mechanism. The main tool I employ is functional analysis understood as in engineering and physiology, where a functional analysis of a mechanism is a partition of a mechanism into component parts and an assignment of functions to the system’s parts. The present account analyzes calculators and computers in terms of their components parts (processor, memory, input and output devices) and their functions. Those components are also analyzed in terms of their component parts (registers and circuits) and their functions. Those, in turn, are analyzed in terms of primitive computing components (logic gates) and their functions. Primitive computing components can be further analyzed but their analysis does not illuminate the notion of computing mechanism. In contrast to extant philosophical explications of computing mechanisms, the present account applies very naturally to stored-program digital computers and matches the language and practices of computer designers. It also satisfies four desiderata better than the competition: (1) it is observer-independent, i.e. it does not allow different observers with different goals to attribute different computations to the same mechanism; (2) it explains why calculators and computers compute while planetary systems and hurricanes don’t; (3) it explains the difference between computers properly so called (i.e. all-purpose, stored-program computers) and less powerful computing mechanisms; and (4) it explains what it means for a computing mechanism to make a mistake in a computation. This makes the present account ideal for grounding discussions of computational theories of mind and brain.
IDepartment
of History and Philosophy of Science
|
|
|
