On the thread of discourse : Homogeneity, trends, and rhythms in texts

Content analysts share with literary critics a vested interest in the text, which is their common world. Both further share the presence of an observer (the critic or analyst). Apart from that, content analysis and literary critic differ on one basic issue. Between the content analyst and the text, one finds a dictionary i.e., a thesaurus that is semantic in nature. On the other hand, between the literary critic and the text, there is no such objective tool, such as the content analysis dictionary. These introductory remarks are useful on two grounds. First, they help to settle the place of content analysis in the world of textual analysis. Secondly, they identify the concepts that content analysis needs to expand when speaking about itself —i.e., procedures, dictionaries, and texts. Before moving any further, a remark is in order. Once processed by content analysis (through the dictionary), the text is really no more the same as before. In one sense, the text is less than before because, by its very nature, the dictionary orders content along one dimension at a time; but, in another sense, the text is also more than before because the dictionary adds knowledge, thought limited, to the text. This results in a redescription that is both repeatable and controlable. Note that the text analyzed by qualitative textual analysis (literary critic or political analysis) is also redescribed, gaining a new reading, yet a reading of the text as it was before. In other words, content analysis and qualitative textual analysis both produce a redescription of the text, but content analysis further produces an instrumental transformation of the text. It is the purpose of the present paper to demonstrate the power of a fine-grained procedure of content analysis (as opposed to an analysis by gross aggregates). In particular, we want to bring the demonstration to bear on issues related to the theory of language performance, to the measurement model involved, to the statistical requirements, and, to the quality and quantity of the resulting information usable by the social sciences. Texts used to demonstrate the efficacy of the procedure are of two sorts, political pamphlets, written expressedly to justify political violence, and other texts, mostly short stories, used as contrasts to the political pamphlets. All mentioned texts are scanned through general dictionaries, appropriate to the nature of the texts, that order their content along several dimensions.