从元方法论视角理解计算模型认知效力问题——明知有缺陷的计算模型能供给我们怎样的知识?  

Understanding the Cognitive Effectiveness of Computational-Models from the Viewpoint of Meta-methodology:What Knowledge Can a Flawed Computational-model Provide Us

作　　者：彭文青 PENG Wentsing(Department of Philosophy,Sun Yat-sen University,Guangzhou,Guangdong,510006)

机构地区：[1]中山大学哲学系,广东广州510275

出　　处：《自然辩证法通讯》2020年第11期31-36,共6页Journal of Dialectics of Nature

基　　金：教育部哲学社会科学研究重大课题攻关项目“认知哲学研究”(项目编号:13JZD004)。

摘　　要：在工程实践中,构建计算模型的约束不仅是求真,可求解性也是重要约束之一,这使得我们建构计算模型所用的方法不能确保模型"正确"地表征物理实在,它们是有"缺陷"的,但何以有缺陷的计算模型却经常给我们提供有效用的知识?本文从科学元方法论视角切入,对此关涉计算模型认知效力的重要问题进行初步探讨:首先阐析构建计算模型过程中不可避免使用的若干重要元方法的"天然"缺陷性;进而,从知识的经验不完全决定性角度论证有缺陷的计算模型提供有效用知识的可能性与合理性;最后,给出一个具有强解释性的认知新框架——知识域"共生互荣"关系模型,作为对此问题更细致清晰的解释与回应。In engineering practice,the constraints of constructing computational-models are not only truth seeking,but also solvability,which makes the methods used in constructing computational models cannot ensure that the models represent physical reality correctly.They are defective.Nevertheless,Why do defective computational-models often provide us with effective knowledge?From the perspective of methodology of philosophy of science,this paper makes a preliminary discussion on this important issue concerning the epistemology and ontology of computational models.Firstly,the reasons why computational models inevitably have defects are expounded.Then,based on the philosophical theory of Experience-Underdetermination,the possibility and rationality of providing effective knowledge by defective computational models is demonstrated;Finally,a new cognitive framework with strong explanatory power,named"Symbiosis and Mutual Prosperity"Relationship Model for knowledge domain is given,as a more detailed and clear explanation and response to this problem.

关 键 词：计算模型 模型建构 抽象/理想化 模型虚构 模型验证 

分 类 号：N031[自然科学总论—科学技术哲学]