Is computation an intrinsic property of physical systems, or is there a distinction between a computer and other objects in the universe? Computer science as a theoretical discipline, traditionally dealing only with issues of abstract computation, has tended to ignore this question. Amongst more philosophical approaches to computing, the first view, pan-computationalism, has been argued for in various guises. Assertions such as “the universe is a computer” are indeed superficially appealing, given the great success of modern computing theory and technology. Yet they lose their apparent content as we look at them more closely. If everything computes merely by virtue of existence, then what more do we say about an object when we call it a computer? How are novel and unconventional computing devices to be characterised, if computing occurs universally and intrinsically in physical objects?
We present here a more discriminating view, in which the use of a physical system to carry out an abstract process (a computation) depends on a number of specific properties that both the physical device (the computer) and physical process (it computing) must have. Not all physical processes constitute computing: a key element in physical computation is the use of a physical system to manipulate the representation of abstractly-encoded data in specific processes. In presenting our framework, Abstraction/Representation (AR) theory, we show in that physically carrying out computation and doing science are closely related activities. Both involve representational activity. By looking at how computers are developed from both fundamental science and then engineering and technology, we show the crucial physical nature of computing.
@inproceedings(Horsman-2018-art, author = "Dominic Horsman and Viv Kendon and Susan Stepney", title = "Abstraction/Representation Theory and the Natural Science of Computation", pages = "127-149", crossref = "PPC-2018" ) @proceedings(PPC-2018, editor = "Michael E. Cuffaro and Samuel C. Fletcher", booktitle = "Physical Perspectives on Computation, Computational Perspectives on Physics", publisher = "Cambridge University Press", year = 2018 )