Colloquium by Dinçer Çevik: "The Curious Case of Explanation Models: 'The Explanation Paradox'"

In this presentation, I revisit Julien Reiss's "Explanation Paradox" (EP, 2012) and its sequel, "Explanation Paradox Redux" (2013). My aim is to critically assess the structure of Reiss's argument and to propose an alternative way of interpreting the paradox.

Reiss takes the Hotelling model (1929) as his case study, claiming that economic models are typically built on false, idealized assumptions. He therefore argues that such models are themselves false in an "elliptical" sense-that is, as a manner of speaking (2012, p. 49). This raises the central question: how can an economic model, whose assumptions are known to be false or irrelevant to real-world phenomena, still provide genuine explanations?

Drawing on insights from Elgin & Sober (2002), 1 argue that-despite reasonable criticisms of Reiss's (2012) position -the overall structure of EP is best understood as analogous to Cartwright's (1983) critique of physical laws. This reframing shifts the puzzle: the issue is not how false models can explain, but rather how models that lack applicability can nonetheless be explanatory.

Read in this way, the paradox is not about the explanatory power of false models, but about the paradox of how non-applicable models can still explain.