Full Program »
Information Processing ArtifactsArtifacts are distinguished from other physical objects by their usability. Whether created, evolved, or appropriated for a particular purpose, artifacts possess identifiable functional capacities that are harnessed by agents through execution of prescribed use plans (Vermaas and Houkes 2006, Hughes 2009). Digital computing devices—the signature technical artifacts of our time—naturally admit such a characterization: their computational capacities are deliberately “engineered in” and are user accessible by design. These devices are unambiguously “information processing artifacts” (IPAs)—technical artifacts that have computational capacities and are used by agents as means to their information-processing ends.
Much less obvious are general criteria that would have to be satisfied to qualify a physical system as an IPA. Clear specifications of computational function and usability face nontrivial challenges. These challenges come not only from pancomputationalism, with its claims of arbitrariness and subjectivity in the very notion of computational function. (See Piccinini (2015, 51-73) for a discussion). They also come from the frontiers of contemporary research on nanoscale, unconventional, and natural computing substrates that, by nature or design, transform physical states in ways that rule out any simple isomorphism between physical state mappings and deterministic computational state mappings. The characterization of usability presents its own challenges in computational contexts, requiring clear distinction between use of a system for computation and the otherwise identical physical transformation of that same system when it is not so used.
In this work I provide an account of IPAs that addresses these challenges. Computational function ascription, use-plan requirements, malfunction, and artifact efficacy are all addressed in this account, which aims to clarify what it means for an agent to use a physical system for deterministic computation even in the face of physical indeterminism and state indistinguishability. Focusing for clarity on IPAs that implement logical transformations, and using Hughes’ (2009) characterizations of instrumental function ascription and artifact efficacy as a guide, I characterize artifacts that are based on the ideal classical L-machines of Ladyman (2009) and their noisy, quantum generalizations by Anderson (2010). I identify the essential ingredients of these logical IPAs and distinguish them from systems that have identical computational capacities but do not qualify as artifacts. Finally, I show how functions defined by deterministic computational state mappings can be ascribed to IPAs that do not support similarly structured physical state mappings—and without risk of descent into pancomputationalism.
Anderson, Neal G. 2010. “On the physical implementation of logical transformations: Generalized L-machines.” Theoretical Computer Science 411:4179-99.
Hughes, Jesse. 2009. “An artifact is to use: an introduction to instrumental functions.” Synthese 168:179-99.
Ladyman, James. 2009. “What does it mean to say that a physical system implements a computation?” Theoretical Computer Science 410: 376-83.
Piccinini, Gualtiero. 2015. Physical Computation: A Mechanistic Account. Oxford: Oxford University Press.
Vermaas, Pieter E., and Wybo Houkes. 2006. “Technical functions: A drawbridge between the intentional and structural natures of technical artefacts.” Studies in History and Philosophy of Science A 37: 5-18.
Department of Electrical and Computer Engineering
University of Massachusetts Amherst