Since the establishment of information architecture (IA) as an area of expertise and research more than a decade ago, its community of scientists and practitioners has been seeking foundations to establish concepts, scope, relations with other disciplines. Some are motivated by the conceptual gap; others are also concerned about the lack of communication between theory and practice in the field. Attempting to find a scientific method to investigate questions arisen from information architecture, we suggest in this article the adoption of the Meta-Modeling Methodology (M³) by Van Gigch and Pipino (Future Comput Syst 1:71–97, 1986). We believe it can provide a comprehensive way to understand information architecture as a discipline, promote critical thinking and improve grounded discussions in the community.