认知科学中计算解释的自主性
Autonomy of Computational Explanation in Cognitive Science
高俊1
作者信息
- 1. 华东师范大学 哲学系,上海 200241
- 折叠
摘要
计算作为一种科学方法,在解释认知现象中扮演重要角色.最近,其解释地位受到部分机制论者的质疑.他们认为计算作为一种科学解释是不完全的,可以被整合到机制解释中,因此计算解释并不具有解释自主性.然而,机制论的质疑并不成立.认知科学中计算解释存在两种形态:How形式和Why形式.机制解释对How形式计算解释自主性质疑的根源在于其混淆了计算解释与计算实现.机制解释通过描述被解释系统的因果机制来完成解释任务,属于因果解释类.机制解释试图取代Why形式计算解释失败的原因在于,很多Why形式计算解释使用类似于数学解释中的基本原理,这使其无法被整合到任何因果或基于因果的解释理论中.
Abstract
Computing,as a scientific method,plays an important role in explaining cognitive phenomena.Re-cently,the explanatory status of computing has been questioned by some mechanists.They believe that computation as a scientific explanation is incomplete and can be integrated into mechanism explanation,so computational expla-nation does not have explanatory autonomy.However,mechanists'challenges are invalid.There are two forms of computational explanation in cognitive science:How-form and Why-form.The root cause for the questioning in the autonomy of How-form computational interpretation lies in its confusion between computational explanation and com-putational implementation.Mechanism explanation completes its explanatory task by describing the causal mecha-nism of the explained system,and it belongs to the category of causal explanation.The reason for failure that mech-anism explanation attempts to replace Why-formal calculation explanation is that many Why-form computational ex-planations use basic principles similar to mathematical explanations,which makes them unable to be integrated into any causal or causal-based explanation theory.
关键词
计算解释/自主性/机制解释/计算实现Key words
computational explanation/autonomy/mechanism explanation/computational implementation引用本文复制引用
基金项目
国家社会科学基金重点项目(23AZX020)
出版年
2024
NETL
NSTL
万方数据