arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2601.00433physics.soc-ph
收藏

文化进化建模

Modelling cultural evolution

Mälardalen University(马尔默大学);Stockholm University(斯德哥尔摩大学);Institute for Futures Studies(未来研究所)

Fredrik Jansson

AI总结:

本文提出了一种文化进化建模的通用模板,通过动态循环连接系统状态、认知过程、行为和宏观结果,以理解文化变化的多元但一致的机制。

摘要:

形式化建模提供了一套理解文化动态的工具包,从个体决策到反复出现的变化模式。本章解释了模型是什么以及为什么它们重要。通过精确且共享的语言,它们通过将模糊的假设转化为清晰、可比较和可测试的主张来帮助思考和沟通。本章描述了建模过程,权衡解释清晰度与预测特异性。调查了四种模型家族,从微观层面的优化代理到宏观层面的启发式或甚至隐式代理,涵盖推理(贝叶斯推理、博弈论)、适应性更新(强化学习、进化博弈)、均值场方法( compartmental 模型、种群动态)以及复杂系统(基于代理的模型、社会网络)。在这些基础上,概述了一个通用的文化进化建模模板,将系统状态、认知过程、行为和宏观结果连接成动态循环,连接个体、群体、机构及其环境。总体而言,这些工具支持一种多元但一致的文化变化理解。

英文摘要:

Formal modelling provides a toolkit for understanding cultural dynamics, from individual decisions to recurring patterns of change. This chapter explains what models are and why they matter. Using a precise, shared language, they aid thinking and communication by turning fuzzy assumptions into clear, comparable, testable claims. The chapter describes the modelling process, trading explanatory clarity against predictive specificity. Four families of models are surveyed, from the micro-level with optimising agents to macro-level dynamics with heuristic or even implicit agents, covering reasoning (Bayesian inference, game theory), adaptive updating (reinforcement learning, evolutionary games), mean-field approaches (compartmental models, population dynamics), and complex systems (agent-based models, social networks). Building on these, a general template for modelling cultural evolution is outlined that connects system states, cognitive processes, behaviour, and macro-level outcomes in dynamic loops, linking individuals, groups, institutions, and their environments. Taken together, these tools support a pluralist but coherent understanding of cultural change.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑