Laboratory Models of Cultural Evolution: Fitting Theoretical Models to Experimental Data

人类与其他动物的不同之处在于他们相互学习的巨大能力。近年来，文化进化理论为社会科学中的许多重要问题提供了一些见解，如文化变异、合作、社会学习和不确定性下的决策。 不同的社会科学学科目前正在用不同的词汇和方法来解决这些问题。 文化进化模型和进化理论一般有潜力弥合学科分歧。 不幸的是，文化进化理论在实地或实验研究中几乎没有得到实证检验。 本项目将进行一系列微观社会实验，以测试文化进化理论模型的行为含义。微型社会是一群相互作用的人类参与者，他们根据他们在多个时间段内在实验任务中的表现获得不同数量的金钱。实验室微观社会是一种有吸引力的方法，因为随着时间的推移，个人和群体水平的行为可以进化并产生有趣的涌现模式。研究人员将通过控制实验任务的结构和参与者可用的信息种类来模拟进化模型的关键特征。他们将开发一系列传统和计算机介导的实验室实验，以及基于互联网的实验，这些实验可以利用来自不同背景和文化的全球受试者，以测试人们使用他人的回报来指导自己的决策的社会学习策略。关于人们如何利用他人的成功来获得自己的行为的理论在社会科学中很常见，但对这些理论的实证检验和改进非常罕见。因此，这项工作将具有普遍的重要性，许多领域关注的细节，人们如何在社会环境中作出决定。该项目将使用最大似然技术来估计参与者在不同实验条件下对个人与社会学习的依赖程度。研究的更广泛影响涉及到有关合作模式，机构以及文化规范和符号演变的新知识的价值。 这项研究将有助于揭示这些过程背后的基本行为机制，因此将有助于政策制定者关注对人类福祉有深远影响的大规模社会变革。

Humans differ from other animals in their great capacity to learn from each other. Recently, theories of cultural evolution have come to provide some insights to many important questions in social sciences, such as cultural variation, cooperation, social learning, and decision-making under uncertainty. Different social science disciplines are currently addressing these questions with divergent vocabularies and methodologies. Cultural evolution models and evolutionary theories in general have the potential to bridge disciplinary divides. Unfortunately, theories of cultural evolution have received little empirical testing in either field or experimental studies. This project will conduct a series of micro-society experiments to test the behavioral implications of theoretical models of cultural evolution. Micro-societies are groups of interacting human participants who receive variable amounts of money based on how they perform over multiple time periods in an experimental task. Laboratory micro-societies are an attractive method because individual and population-level behavior can evolve and produce interesting emergent patterns over time. The researchers will simulate critical features of evolutionary models by controlling the structure of the experimental task and the kinds of information available to participants. They will develop a series of traditional and computer-mediated laboratory experiments, as well as Internet-based experiments that can tap into a global pool of subjects from diverse backgrounds and cultures, to test success-biased social learning strategies where people use the payoffs of others to guide their own decision-making. Theories of how people use the success of others to acquire their own behavior are common in the social sciences, but empirical tests and refinement of these theories is very uncommon. Thus this work will be of general importance to many fields concerned with the details of how people make decisions in social environments. The project will use maximum likelihood techniques to estimate how much participants rely on individual versus social learning under different experimental conditions.The broader impact of the research relates to the value of new knowledge about the evolution of patterns of cooperation, institutions, and cultural norms and symbols. The research will help reveal the basic behavioral mechanisms that underlie these processes, and therefore will be useful to policy makers concerned with large-scale social changes that have profound implications for human welfare.