Beyond the Google ngram viewer

Abstract

In the last years the Google Ngram Viewer has become a popular tool for quick and dirty analyses of how certain concepts emerge and develop over time. So far most scholars using this tool do not reflect on its methodological problems and on whether the corpus presented by Google is actually appropriate for their specific research problem. Therefore, digital humanities experts tend to be both fascinated and sceptical about the Ngram Viewer. In this paper we do not take up the diverse lines of critique, but rather focus on what we deem to be the most important innovation coming with the Ngram Viewer: the very simple idea of tracing relative frequencies of words and phrases over time. We propose to take this basic idea one step further, particularly in regard to scientific and meta-scientific concepts. We show how bibliographic databases and journal archives can be used to analyze how often certain concepts appear in scientific publications. As these databases have not been established with any idea of linguistic analysis in mind, but rather as tools to find and access scientific publications, what we propose can be characterized as “off-label” use of bibliographic databases and journal archives.