Scholarlink

<p>I propose the Dual-Constraint Model (DCM), a formal framework that locates perceptual content at the intersection of (i) internal generative inference and (ii) explicit world-imposed invariance constraints. Percepts are cast as solutions of an optimization that combines sensory likelihood, endogenous priors, and a penalty for hypotheses that violate ecological or physical invariances. DCM preserves the positivist insight that the world imposes non-arbitrary constraints while preserving constructivist insight that internal models select among admissible interpretations under uncertainty. The formulation yields clear, falsifiable dissociations: manipulations that degrade invariance structure produce perceptual and early-sensory effects that cannot be mimicked merely by shifting priors. I present explicit operational forms for the invariance penalty, outline empirical designs and analysis procedures, and provide concrete criteria under which DCM is empirically distinguishable from pure-prior or pure-constraint accounts. The argument is presented as a theoretical synthesis rather than a definitive resolution, recognizing that empirical and philosophical questions remain open.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Perception Under Constraint: A Dual-Constraint Model Reconciling Positivism and Constructivism

  • 期刊论文

  • 发表日期：

    2026年2月6日

查看全文
  • Levente Orbán

摘要

I propose the Dual-Constraint Model (DCM), a formal framework that locates perceptual content at the intersection of (i) internal generative inference and (ii) explicit world-imposed invariance constraints. Percepts are cast as solutions of an optimization that combines sensory likelihood, endogenous priors, and a penalty for hypotheses that violate ecological or physical invariances. DCM preserves the positivist insight that the world imposes non-arbitrary constraints while preserving constructivist insight that internal models select among admissible interpretations under uncertainty. The formulation yields clear, falsifiable dissociations: manipulations that degrade invariance structure produce perceptual and early-sensory effects that cannot be mimicked merely by shifting priors. I present explicit operational forms for the invariance penalty, outline empirical designs and analysis procedures, and provide concrete criteria under which DCM is empirically distinguishable from pure-prior or pure-constraint accounts. The argument is presented as a theoretical synthesis rather than a definitive resolution, recognizing that empirical and philosophical questions remain open.