An examination of temporal agency in courtship narratives

The reported study investigated temporal agency (i.e., the assignment of cause for temporal shift) in newlyweds’ courtship narratives. Transcripts of courtship narratives generated by each partner of 23 recently married couples (approximately 3 months) participating in the PAIR project (Huston, McHale, & Crouter, 1986) were analyzed for the presence of different linguistic strategies for encoding temporal shift. Statements were coded as “human agency assignments” when they assigned the cause of temporal shift to humans (e.g., we started seeing each other in June); statements that assigned temporal shift to abstract entities such as the events themselves (e.g., the summer started out well for us) or to the relationship (e.g., the relationship started to slow down) were coded as “abstract agency assignments.” The frequency with which narrators mentioned positiveand negative emotions was also coded to explore the possibility that emotional valence mediated agency assignments. The frequency of different agency assignments and emotion words were considered in the context of portions of the courtship accounts that narrators designated as describing “upturns” (episodes that increased the likelihood of marriage) or “downturns” (episodes that decreased marriage likelihood). Results indicated that the frequency of human agency assignments and positive emotion mention were higher in upturn than downturn narrative segments; in contrast, abstract agency assignments and negative emotion mention were more frequent in downturn than upturn segments. Subsequent analyses indicated that positive word mention partially mediated human agency assignments in upturns and that negative word mention partially mediated abstract agency assignments in downturns. These findings are consistent with previous research demonstrating an association between the emotional valence of an event and temporal agency assignment: In general, people assign temporal agency to themselves when describing positive events, but prefer abstract agency assignments for negative events (McGlone & Pfiester, 2009).