What does it take to implement a computer? Answers to this question have often focused on what it takes for a physical system to implement an abstract machine. As Joslin observes, this approach neglects cases of software implementation—cases where one machine implements another by running a program. These cases, Joslin argues, highlight serious problems for mapping accounts of computer implementation—accounts that require a mapping between elements of a physical system and elements of an abstract machine. The source of these problems (...) is the complexity introduced by common design features of ordinary computers, features that would be relevant to any real-world software implementation. While Joslin is focused on contemporary views, his discussion also suggests a counterexample to recent mapping accounts which hold that genuine implementation requires simple mappings. In this paper, I begin by clarifying the nature of software implementation and disentangling it from closely related phenomena like emulation and simulation. Next, I argue that Joslin overstates the degree of complexity involved in his target cases and that these cases may actually give us reasons to favor simplicity-based criteria over relevant alternatives. Finally, I propose a novel problem for simplicity-based criteria and suggest a tentative solution. (shrink)

Without a proper restriction on mappings, virtually any system could be seen as implementing any computation. That would not allow characterization of systems in terms of implemented computations and is not compatible with a computationalist philosophy of mind. Information-based criteria for independence of substates within structured states are proposed as a solution. Objections to the use of requirements for transitions in counterfactual states are addressed, in part using the partial-brain argument as a general counterargument to neural replacement arguments.

What is a computer? What distinguishes computers from other artificial or natural systems with alleged computational capacities? What does use of a physical system for computation entail, and what distinguishes such use from otherwise identical transformation of that same system when it is not so used? This paper addresses such questions through a theory of information processing artifacts, the class of technical artifacts with physical capacities that enable agents to use them as means to their computational ends. Function ascription, use (...) plan requirements, malfunction, and efficacy of IPAs are all addressed in this theory, with emphasis on artifacts that can be used—reliably or otherwise—for digital computation. By explicitly distinguishing physically grounded computational capacities from user-ascribed computational functions, and by recognizing the distinct roles of each for the implementation of computations in artifacts, this theory clearly distinguishes the use of physical systems for computation from the transformations of physical system states that enable such use. As such, it provides a rigorous basis for distinguishing “computers” from other artificial and natural systems—a distinction whose nature and legitimacy faces ever-evolving challenges from multiple disciplines. This theory, and the associated “instrumental” view of computation in artifacts, naturally accommodates the openminded but scrupulous consideration of radically unconventional physical systems as potential substrates for future computers. (shrink)

Without a proper restriction on mappings, virtually any system could be seen as implementing any computation. That would not allow characterization of systems in terms of implemented computations and is not compatible with a computationalist philosophy of mind. Information-based criteria for independence of substates within structured states are proposed as a solution. Objections to the use of requirements for transitions in counterfactual states are addressed, in part using the partial-brain argument as a general counterargument to neural replacement arguments.