Social Memory: A Game Theoretic Approach

社会记忆：博弈论方法

• 批准号: 0617789
• 负责人: Roger Lagunoff
• 金额: --
• 依托单位: Georgetown University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2006
• 资助国家: 美国
• 起止时间: 2006-07-01 至 2008-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0617789&HistoricalAwards=false
• 关键词: Social; Memory; Game; Theoretic; Approach

This project examines the creation and consequences of social memory. Social memory refers to beliefs about past behavior created by communication across generations in an ongoing society. Since individuals do not live forever, and since written and oral accounts of the past can be manipulated, social memory need not be accurate or credible. Yet, accountability of people and their governments requires some knowledge of the past.This project examines the nature of social memory, how it is created, and what are the consequences. It builds on a model of dynastic games developed by the researchers. Dynastic games are infinite horizon games with periodic, generational replacement. Each dynasty is a succession of short-lived individuals, each of whom cares about the well being of his successors, and each has no direct knowledge of the past before his entry. Past behavior is therefore communicated by members of one generation to the next.The older generation's incentives to accurately communicate the past, and the younger generation's beliefs about what is communicated are the central objects of concern. The project demonstrates how various anomalies can arise: (a) it is very possible for equilibrium beliefs--hence social memory--about the past history of events to be wrong; (b) accurate social memory may paradoxically entail dubious beliefs; (c) social memory is sometimes deliberately erased, the results of which can beneficial.Results so far establish that social memory surpasses full or direct memory in terms of outcomes that can be sustained. In fact, some efficient outcomes that could not be sustained under full memory are sustainable under social memory. Two broad areas of inquiry are therefore proposed. First, the project investigates the formal relation between social memory and beliefs about unanticipated messages (i.e., "mistakes"). Social memory emerging from a "trusting" society appears quite different than that from a "skeptical" one. The consequences of each are examined. Second, the deliberate erasing of memory and its consequences are examined.Dynastic games are natural models of ongoing societies--they are an important complement to models with infinitely lived agents in terms of both theory and relevance. Results indicate that the two types of models behave quite differently. However, little is known about social memory or about dynastic games generally. The project fills a critical gap in the literature since standard models concern whether individuals learn about present fundamentals, whereas this project concerns whether social memory encodes the past. The project makes explicit the connection between equilibrium beliefs and social memory, and between social memory and equilibrium outcomes. Finally, the research team has expertise with game theoretic models of communication, as well as substantial breadth in areas of social norms, legal systems and political institutions.Broader impact. Social memory has enormous implications for conflicts around the world. In some conflicts those who control access to information manipulate social memory. In other conflicts, recipients of truthful information doubt its credibility. Those most vulnerable to these problems are usually underrepresented groups excluded from social and economic elites. The project enhances our understanding of these problems by showing how they arise as equilibrium phenomena.

这个项目探讨了社会记忆的产生和后果。社会记忆是指在一个持续的社会中，通过代际交流而形成的对过去行为的信念。由于个人不可能永生，由于对过去的书面和口头记录可以被操纵，社会记忆不需要准确或可信。然而，对人民及其政府的问责需要对过去有所了解。这个项目考察了社会记忆的本质，它是如何产生的，以及后果是什么。它建立在研究人员开发的动态游戏模型的基础上。王朝游戏是具有周期性代际更替的无限视界游戏。每个朝代都是由一些短命的个体继承而来，每个人都关心他的继任者的福祉，每个人在进入朝代之前都对过去没有直接的了解。因此，过去的行为由一代人传递给下一代。老一辈人想要准确地传达过去的动机，以及年轻一代对传达内容的信念，是关注的核心对象。该项目展示了各种异常现象是如何产生的：(a)关于事件过去历史的平衡信念（即社会记忆）很可能是错误的；(b)准确的社会记忆可能矛盾地导致可疑的信念；(c)社会记忆有时被故意抹去，其结果可能是有益的。迄今为止的研究结果表明，社会记忆在结果可持续性方面优于完整记忆或直接记忆。事实上，一些在完全记忆下无法持续的有效结果在社会记忆下是可持续的。因此，提出了两个广泛的调查领域。首先，该项目调查了社会记忆和对意外信息（即“错误”）的信念之间的形式关系。来自“信任”社会的社会记忆与来自“怀疑”社会的社会记忆截然不同。对每一种情况的后果都进行了研究。其次，研究了蓄意清除记忆及其后果。动态游戏是正在进行的社会的自然模型——就理论和相关性而言，它们是具有无限生存主体的模型的重要补充。结果表明，这两种模型的行为完全不同。然而，我们对社会记忆或王朝游戏知之甚少。该项目填补了文献中的一个关键空白，因为标准模型关注的是个人是否学习了当前的基本原理，而这个项目关注的是社会记忆是否编码了过去。该项目明确了平衡信念与社会记忆之间以及社会记忆与平衡结果之间的联系。最后，研究团队在传播的博弈论模型方面具有专业知识，并且在社会规范，法律制度和政治制度领域具有相当的广度。更广泛的影响。社会记忆对世界各地的冲突有着巨大的影响。在一些冲突中，控制信息获取的人操纵社会记忆。在其他冲突中，真实信息的接受者怀疑其可信度。最容易受到这些问题影响的群体通常是被排除在社会和经济精英之外的代表性不足的群体。该项目通过展示它们如何作为平衡现象出现，增强了我们对这些问题的理解。