Cognitive Instability of Mindreading: Why Simulation Theory Cannot Support Process Reliabilism
Alvin Goldman's simulation theory offers a compelling description of how human beings read other minds, yet, insofar as it is advanced as a cognitive practice that yields knowledge, it must answer to epistemological standards, most pointedly to Goldman's own process reliabilism, on which a true belief counts as knowledge only if it is produced by a cognitive process that reliably leads to truth. This paper argues that the simulation mechanism lacks any independent error-correction procedure and therefore cannot screen out contamination arising from the subject's own states, such as biases and emotions. This structural deficit makes the accuracy of simulation heavily dependent on the contingent factor of how similar the target is to the subject, so that its truth-production rate fluctuates sharply across contexts and fails reliabilism's demand for cross-contextual stability, leaving it trapped in what Pritchard calls veritic luck. We do not deny the descriptive value of simulation theory; we contend only that, until simulation is finely subdivided and its operative environments delimited, the assertion that simulation produces knowledge cannot satisfy the reliabilist standard. The argument is strictly normative in its target: the paper does not dispute that simulation plausibly describes how mindreading is in fact carried out, but only whether, given the structural deficit identified below, the outputs of simulation are yet entitled to the status of knowledge.
Authors
- Jingnan Lin
Institutions
- Huaqiao University (CN)
Publication Details
- Journal
- Communications in Humanities Research
- Published
- 2026-11-17
- DOI
- https://doi.org/10.54254/2753-7064/2026.39029
- Primary Topic
- Epistemology, Ethics, and Metaphysics
- Type
- article
- Field-Weighted Citation Impact
- 0.00