Over the past decade, sociologists have become increasingly interested in the formal study of semantic relations within text. Most contemporary studies focus either on mapping concept co-occurrences or on measuring semantic associations via word embeddings. Although conducive to many research goals, these approaches share an important limitation: they abstract away what one can call the event structure of texts, that is, the narrative action that takes place in them. I aim to overcome this limitation by introducing a new framework for extracting semantically rich relations from text that involves three components. First, a semantic grammar structured around textual entities that distinguishes six motif classes: actions of an entity, treatments of an entity, agents acting upon an entity, patients acted upon by an entity, characterizations of an entity, and possessions of an entity; second, a comprehensive set of mapping rules, which make it possible to recover motifs from predictions of dependency parsers; third, an R package that allows researchers to extract motifs from their own texts. The framework is demonstrated in empirical analyses on gendered interaction in novels and constructions of collective identity by U.S. presidential candidates.

中文翻译：

---

谁对谁做什么？使文本解析器适用于社会学调查

在过去的十年中，社会学家对文本中语义关系的正式研究越来越感兴趣。大多数当代研究要么关注映射概念共现，要么关注通过词嵌入测量语义关联。尽管有利于许多研究目标，但这些方法有一个重要的局限性：它们抽象出人们可以称之为文本的事件结构，即发生在其中的叙事动作。我的目标是通过引入一个从涉及三个组件的文本中提取语义丰富的关系的新框架来克服这一限制。首先，围绕文本实体构建的语义语法区分了六个主题类：实体的动作，实体的治疗，作用于实体的代理，实体作用的患者，实体的特征和实体的财产；第二，一套全面的映射规则，这使得从依赖解析器的预测中恢复主题成为可能；第三，一个允许研究人员从他们自己的文本中提取主题的 R 包。该框架在小说中的性别互动和美国总统候选人的集体身份建构的实证分析中得到证明。