Reviews of literature often summarise research within a given field to provide succinct information about state of the art as well as gaps for future research to address. In educational research, such information includes different positions within a field; for example, theoretical positions underpinning empirical studies. Such theoretical positions may influence interpretations of research results and remain a challenge to map and present. Here, we show how to use thematic analysis and network analysis to construct maps, analyse, and synthesise theoretical positions within educational research. We use early childhood inquiry-based science education (ECIBSE) literature as a case because of its historical roots in Dewey’s educational philosophy. This allows us to discuss findings in light of a well-known theoretical framework. Using our methodology, we analyse four theoretical positions for teaching and learning: (1) science should be learned through inquiry, (2) teaching should model scientific practices, (3) children should develop science-related competencies (4) child’s experience should take precedence. After discussing these positions, we turn to methodological possibilities and challenges of using thematic network analysis for literature reviews. R-code, data, and instructions for reproducing our results are available at: https://github.com/jbruun/thematicDiscourseNetworkAnalysis