The understanding of agency that animates Onuf’s Constructivism is complex and worthy of more space and more depth than it received here. I think of this essay mostly as an invitation for Nick to clarify what I may have muddied. My principal argument is that Constructivism can benefit from divorcing agency from actors, intentions, and texts, and marrying it to power-laden negotiations. Deeds are necessary for agency to be attributed, but agency is not inherent in the deed, even if that deed is the speech act proclaiming a new rule. Agency is a quality attributed to particular actors at particular moments; it is social, negotiable, and changeable. While locating agency within the modern individual may be useful for exploring social dynamics in Western, late-modern societies, beginning with assumptions about who can be an agent carries both ontological and methodological problems for Constructivists. The co-constitution dynamics of each society (broadly defined) include deciding the rules for who should be considered an agent, so the definition of agency should be an empirical question that can only be understood by examining the discourses of the construction process in that particular society
Conceiving of agency as an effect of discourse improves Constructivist explanations of social phenomena. First, it eliminates IR’s Level of Analysis Problem. Agents are whatever entities society designates through discourse - corporations, non-human animals, gods, ancestral spirits, viruses, familial networks, nations, genders, castes, clans, or ethnicities. It also allows for all kinds of non-corporeal agents that (Nick’s qualms notwithstanding) many social arrangements currently treat as agents. In addition, thinking of agency in this way helps reframe how Constructivists explain change by directing attention away from acts (all of which produce change) to whether those acts are interpreted as change or stability. Distinguishing authored change from the agency of participation and recognition means that change becomes a normal backdrop for analysis rather than something that has to be explained. Finally, it would also mean that I have put to good use the creative spark that I found in Onuf’s Constructivism.