Social Norms, Social Boundaries and Inequality

社会规范，那些在日常生活中指导我们行为的非正式理解，很难科学地研究，因为它们既难以定义，也难以衡量。然而，它们规范了大量的社会行为，也有助于形成社会不平等。 该项目将通过开发一种新的方法来促进我们对不平等的科学理解，该方法可以识别描述正确行为的书面文本中的社会规范。该项目将归纳出文本中规定和禁止的行为与社会阶层等更可衡量的不平等维度之间的关系。该项目还将揭示这些书面规范与造成不平等的社会界限和排斥之间的联系。 该项目将有助于了解社会不平等的普遍性，这种不平等以前一直阻碍科学研究，并将提供有关社会规范如何随时间变化的证据。该项目将通过分析由描述正确行为的文本组成的原始数字语料库来研究社会规范。该语料库包括过去95年中产生的700多万字和9万页文本。本研究采用了语料库语言学的主题建模、词嵌入、词网络映射、命名实体识别、情感分析等分析技术。这些战略将确定社会规范，象征性的边界，规范遵守的执行技术，以及将道德价值观附加到社会规范的正当性叙述。机器学习将以来自同一语料库的文本的传统手工内容分析为指导和补充。该项目还将利用档案研究，将研究结果置于其社会和历史背景中。通过这些方法，研究将产生一个新的方法来研究社会规范和不平等。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Social norms, those informal understandings that guide our behavior in everyday life, are difficult to study scientifically because they are both challenging to define and to measure. Yet they regulate a large body of social behavior and also help to pattern social inequality. This project will advance our scientific understanding of inequality by developing a new approach that identifies social norms in written texts describing correct behaviors. The project will inductively derive the relationships between behaviors prescribed and proscribed in the texts and more measurable dimensions of inequality, such as social class. The project will also uncover the connections between those written norms and the social boundaries and exclusions that create inequality. The project will inform understanding regarding the pervasiveness of social inequality that has previously been resistant to scientific study and will provide evidence regarding how social norms change over time. The project will study social norms through analysis of an original digital corpus consisting of texts describing correct behaviors. The corpus includes more than 7 million words and 90,000 pages of text produced during the last 95 years. The research uses multiple methods including corpus linguistics analytic techniques of topic modeling, word embedding, word network mapping, named entity recognition, and sentiment analysis. These strategies will identify social norms, symbolic boundaries, techniques for the enforcement of norm compliance, and narratives of justification attaching moral values to social norms. Machine learning will be guided by and supplement traditional by-hand content analysis of texts from the same corpus. The project will also use archival research to place the research findings in their social and historical contexts. Through these methods, the research will produce a new approach to studying social norms and inequality. The results will fuel discussion of the cultural foundations of social inequality as well as shed light on how social norms generally both operate and evolve historically.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.