What We (Do and Don’t) Mean When We Talk About Computational Implementation
MetaScientia: Journal of the History and Philosophy of Science · 2026
DOI: 10.68469/xbyz2h18 ↗Kurz gefasst
Fachartikel in der Zeitschrift MetaScientia: Journal of the History and Philosophy of Science, 2026. Verfasst von Danielle J. Williams. Die Arbeit nennt 61 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
The simplest of all mapping views says that what is required for some physical system to implement a computational structure is that the state transitions of the physical system mirror the state transitions described by the computational description. This view, of course, is considered overly permissive as it allows systems such as pails of water and rocks baking in the sun to satisfy the implementation relation. Anderson and Piccinini’s (2024) robust mapping account modifies the simple view by specifying a robust mapping requirement. If a robust mapping is made, then the physical system bears the physical signature of computation. However, their account departs from standard mapping views in several ways, including its characterization of the relata at either end of the implementation relation. In this paper, I evaluate the robust mapping account and argue that the constraints it imposes on the descriptions of the respective relata prevent it from generalizing across contexts. Using physical descriptions in neuroscience, I demonstrate that the formulation of the physical description of the physical system in the robust mapping account leads to two problems: What I call the multiple descriptions problem and the granularity problem. Then, focusing on the computational description, I argue that the authors’ mistaken Platonist assumption about standard mapping views leads to an additional problem: What I call the hybridization problem. Although this paper engages directly with the robust mapping account, it serves as a general overview of the scope and aims of standard computational implementation views by demonstrating how the metaphysical project of computational implementation relates to computational methodologies in cognitive science. This paper will be helpful for those who are interested in how a metaphysical project can contribute to explanation in the mind sciences and for understanding how and why standard mapping views are formulated the way that they are.
Zitieren
APA
Williams, D. J. (2026). What We (Do and Don’t) Mean When We Talk About Computational Implementation. MetaScientia: Journal of the History and Philosophy of Science. https://doi.org/10.68469/xbyz2h18
Vancouver
Williams DJ. What We (Do and Don’t) Mean When We Talk About Computational Implementation. MetaScientia: Journal of the History and Philosophy of Science. 2026. doi:10.68469/xbyz2h18
BibTeX
@article{Williams2026,
author = {Williams, Danielle J.},
title = {What We (Do and Don’t) Mean When We Talk About Computational Implementation},
journal = {MetaScientia: Journal of the History and Philosophy of Science},
year = {2026},
doi = {10.68469/xbyz2h18}
}
EUVIMED