Genre analyses of the rhetorical structure of research articles (RAs) have been extensively conducted. Many studies have sought to identify communicative purposes or moves and their constitutive actions or steps in the different sections. However, studies examining move structure of RAs have often neglected why moves are sequenced in specific manners. Despite this, previous studies investigating other genres have reported cases where moves are achieved in a relational manner. Taboada and Lavid’s (2003) analysis of scheduling dialogues revealed that many speakers used background/foreground relations when they proposed a plan to the hearer, whereas they used a “cause and effect” type of relation when hearers rejected a proposal. Kong (1998) examined relational structure between moves in business letters by Chinese and English-speaking writers. He found that Chinese-speaking writers tended to present a request move after a premise move indicating the legitimacy of the request as background information, while English-speaking writers tended to present a request move followed by a move more explicitly justifying the request. Gruber and Muntigl’s (2005) analysis of students’ essays showed that the four moves of the essays – Orientation, Consequential Explanation, Factorial Explanation, Discussion and Summary – can be realised in a relational manner. For example, textual parts that realise Consequential and Factorial Explanations, and Discussion were shown to be related in the way that the former provides background information to the latter. Although very few, there are also studies that investigated relational structure of moves in RAs. Lim’s (2012) analysis of the introduction section of management science RAs demonstrated that writers often used concession relation between a move describing contexts of the research problem and a move providing justification of the research. By contrast, Kawase’s (2022) analysis of applied linguistics RA introductions revealed that when writers sought to justify their studies by indicating how they were to extend previous studies, they had a greater tendency to do this after describing contexts of the research as background information, instead of using a concession relation.