Maxi Viola Mirabel Seidl
LMU Munich University
Within the last years, a variety of strategies broadly aimed at vindicating the value-free ideal (VFI) have been put forth. One strand of contemporary vindication attempts to argue for exactly that, namely the realizability of the VFI. Such an attempt has recently been made by Sikorski (2024). With reference to the debate between Betz (2013) and John (2015), surrounding the IPCC’s methodology, Sikorski attempts to vindicate the realizability of the VFI by rebutting John’s criticism of Betz’s solution to the trouble with the so-called methodological critique. I will begin by summarizing first Betz’s attempt at vindicating the VFI through methodological adjustments such as carried out by the IPCC, followed by the critical response by John, which seeks to undercut Betz’s position. Next, I will give an overview of Skorski’s account. Sikorski orients himself on Levi (1960), who, in analyzing the original discussion involving Rudner (1953), Jeffrey (1956) and Churchman (1956), rejects Rudner’s and Churchman’s position that scientists ought to respond to the seriousness of error in their assessment of a hypothesis. Instead, he argues, they can adopt conventions. Consequently, Sikorski turns to show how conventionally fixed statistical thresholds can allegedly help realize the VFI. With reference to Elliott (2011), I then want to draw up a distinction regarding the methodological critique, by differentiating between a ‘problem of context-sensitivity’ and a ‘problem of translation’, which come apart and can be applied individually within an analysis of the discussion surrounding VFI and its realizability. With regard to John’s paper, I consider the view that the conventions might reflect genuinely epistemic values due to the nature of knowledge and its relation to security, against which I will seek to illustrate the irreducible contingency of any set, conventionally established threshold. Any raw data, so to speak, must be translated and its translator must necessarily make choices which cannot come from internal benchmarks alone. The problem can be delegated to a past communal decision, yet it cannot be overcome by the means that Levi and Sikorski present. The ‘problem of translation’, as I argue, has not been met by the responses I discuss.
Chair: Caterina Lupieri
Time: 17:30 – 18:00
Location: SR 1.004
SOPhiA 2026 