Expert lock-in and interpretation updates: the transfer of the Prisoner’s Dilemma to mathematical oncology

Source abstract

Abstract Source domain knowledge can hinder successful model transfer when it fuses with modeler’s interpretation of a formalism. This paper diagnoses this failure—called expert lock-in—through the case of the Prisoner’s Dilemma in mathematical oncology. In doing so, this paper also identifies a gap shared by two major template accounts: Humphreys’ formal template account reduces success to mapping and leaves the user’s vocabulary out of the analysis. Knuuttila and Loettgers’ model template account locates interpretive content inside the template rather than in the user’s competence to update it. Since neither can explain expert lock-in, this paper offers two contributions that together fill this gap. The companion thesis argues that the vocabulary recurring with a transferred formalism travels alongside it, not inside it: its interpretation is re-established in each use, not carried across, which is what makes updating both possible and necessary. The competence account argues that successful transfer requires three abilities: perception, verification, and articulation. Expert lock-in is a failure mode: it occurs when a modeler stops updating and treats the resulting interpretation as a property of the formalism rather than a disciplined choice. The cases of Axelrod and Maynard Smith show lock-in in a specific, recoverable form: each performed genuine updates and then treated their last update as the formalism’s boundary; mathematical oncologists instead show what continued updating looks like.