Literary Change. German Poetry between Realism and Early Modernism and Its Relation to Literary, Cultural and Social Developments

该项目的目的是在其他文学体裁、公共话语以及社会和文化历史发展的背景下，模拟19世纪之交诗歌从现实主义到早期现代主义的发展，从而能够回顾相关的文学史假设，并将这些局部假设整合为一个综合模式。它建立在项目第一阶段的结果基础上，该阶段对比了现实主义和早期现代主义的抒情文本。这个新项目旨在捕捉随着时间的推移而发生的文学变化。根据我们的指导性论断，早期现代主义的发展与文学领域地位的多元化和扩张是相关联的。在文学史研究成果的基础上，将通过三个参数来显示变化：性别角色，包括关系模式、技术以及达尔文和尼采分别作为科学人类学和哲学人类学代表的影响。对于每个参数，它将检查它是否以及在多大程度上影响文本的内容、情感和风格的变化。审查诗歌和其他数据中这三个文本维度的每一个参数都需要将九个变化指标付诸实施，主要是通过调整和应用神经模型。每个维度的起点都是CLS中已经建立的方法。这些都需要针对现有的缺点加以改进。主题建模为报纸文章等产生了良好的内容表示，但当应用于文学文本时，主题就会变得不清晰。我们测试这一发现是否也适用于LDA以外的新方法。风格相似性是用Delta来衡量的。将测量转换为神经模型的尝试应该允许应用于短文本并检测文体特征。情感识别将被扩展为基于方面的组件，以便可以特定地提取与主题或实体相关联的情感。诗歌将是研究的重点，但也将探索其与其他文学体裁和公共话语的关系，并将辅之以社会和文化历史发展的数据。将这些论文的一部分汇集到一个综合模型中，并在分析的语料库的基础上对其进行测试，重点是上述三个参数，这是该项目的总体目标。这也是第一次尝试将对文学现象的孤立观察作为文学-历史变化的综合正式模式的基础，其中许多CLS项目仍然存在。

The aim of the project is to model the development of poetry from realism to early modernism at the turn of the 19th Century in the context of other literary genres, public discourses, and social and cultural historical developments, thus enabling a review of relevant literary historical hypotheses and an integration of these partial theses into an integrative model. It builds on the results of the first phase of the project, which contrasted lyric texts of realism and early modernism. The new project aims to capture literary change over time. According to our guiding thesis, the development towards early modernism is associated with a pluralization and expansion of positions in the literary field. Building on the findings of literary historical research, change will be made visible through three parameters: gender roles including models of relationships, technology, and the impact of Darwin and Nietzsche as representatives of a scientific and philosophical anthropology, respectively. For each of the parameters, it will be examined whether and to what extent it affects changes in the content, emotion, and style of the texts. Examining each of the parameters along these three textual dimensions in poetry and the other data requires operationalizing nine indicators of change mostly through the adaptation and application of neural models. The starting point for each dimension are methods already established in the CLS. These are to be improved with respect to existing shortcomings. Topic Modeling produces good content representations for e.g. newspaper articles, but when applied to literary texts, topics lose clarity. We test whether this finding also holds for newer methods beyond LDA. Stylistic similarity is measured with delta. The attempt to transform the measure into a neural model should allow application to short texts and to detect stylistic features. Emotion recognition will be extended into an aspect-based component so that emotions associated with topics or entities can be specifically extracted. Poetry will be the focus of the study, but its relationships to other literary genres and public discourses will be explored as well, which will also be supplemented with data on social and cultural historical developments. To bring together a part of these theses into an integrative model and to test it on the basis of the analyzed corpora with a focus on the aforementioned three parameters is the overarching goal of the project. This is also a first attempt to take the isolated observation of literary phenomena, of which many CLS projects still consist, as a basis for an integrated formal model of literary-historical change.