As Bernard Lahire notes, the social sciences are increasingly marked by hyper-specialization, driven both by disciplinary compartmentalization—between fields such as sociology, law, or economics—and by sub-disciplinary fragmentation within sociology itself, for instance between the sociology of work, urban sociology, or the sociology of education. Family sociology is no exception to this trend. While it has helped make visible and legitimate a research area long relegated to the “private” sphere, it also constrains cross-fertilization between research domains within sociology. This is compounded by an individual tendency to concentrate on a narrow set of topics within these sub-disciplines, adding yet another layer of hyper-specialization. Several factors contribute to this evolution: the growing complexity of contemporary societies, the deterioration of academic working conditions that leave little time for research, and the race for excellence measured through quantitative indicators such as the h-index—pressures that often encourage scholars to consolidate their reputation within a well-defined “niche” rather than venture into new (sub-)fields where legitimacy must be (re)built. In this presentation, I will take the opposite approach. My aim is to show why breaking down (sub)disciplinary boundaries matters—why we should step outside our comfort zones, dare to explore new questions and draw inspiration from alternative perspectives. To illustrate this, I will retrace the path that led me to temporarily move away from the study of transnational families to investigate the socialization of children whose parents are separated or divorced and who grow up in equal shared custody arrangements. I will discuss the conceptual frameworks I mobilized for the first time in this work, the original methods developed to engage with children on this topic, and the challenges encountered along the way—before concluding with how this venture beyond my comfort zone has, in turn, enriched my perspective on transnational families.
Merla et al. (2025) studied this question.