SOPhiA 2019

Salzburgiense Concilium Omnibus Philosophis Analyticis

SOPhiA ToolsDE-pageEN-page

Programme - Talk

The DEKI account of scientific representation. A solution to the problem of model-based representation?
(Philosophy of Science, English)

There is general agreement that scientific models play a central role in contemporary science. Models are used to represent, explain, predict and explore natural and social phenomena. Among these model functions representation plays a crucial role, because representational accuracy is a prerequisite for other model functions. Traditionally philosophical accounts of scientific representations made either similarity between the representatum and the represented phenomenon or the ability to draw accurate inferences from the representation, the central element of model-based representations. Recently, Roman Frigg and James Ngyuen suggested an alternative account of model-based representations and proposed to formulate the conditions for a scientific representation in terms of the following biconditional statement (Frigg, Ngyuen 2018). According to their account, a model (M) is a model-representation of a target-system (T) if and only if

(i) M denotes T (and in some cases parts of M denote parts of T).

(ii) M exemplifies Z-properties P1 , ..., Pn.

(iii) M comes with key K associating the set {P1, ..., Pn} with a set of properties {Q1, ..., Qm}: K({P1, ..., Pn}) = {Q1, ..., Qm}

(iv) M imputes at least one of the properties Q1, ..., Qm to T.

Although Frigg and Nguyen's account represents an important improvement over earlier theories of representation, I will argue that there remain some significant problems regarding their approach. I will show that certain elements of their account, like denotation and exemplification are redundant and that the notion of a key that associates model properties with other properties has to be carefully outlined to avoid the consequence that everything potentially represents everything. My talk will present an alternative version of the Frigg / Ngyuen account, which avoids some of these pitfalls.

Time: 15:20-15:50, 18 September 2019 (Wednesday)
Location: SR 1.005

Thomas Durlacher 
(University of Luxembourg, Luxembourg)

Thomas Durlacher is a PhD student at the University of Luxembourg. His PhD project is concerned with philosophical questions arising in the context of computer simulations that model the behavior of social agents. Before this, he worked at the Centre for the History of Science at the University of Graz in Austria. His philosophical interests lie in the area of the general philosophy of science, with particular focus on model-based reasoning, representation and scientific explanations.

Testability and Meaning deco