During the long state-making process after the end of Japanese colonization in Northeast China (1946-1952), the Chinese Communist Party (CCP) implemented various policies concerning colonial collaborators as part of its discourse politics. By invocating popular sentiments about the colonial past and blurring classification standards for different collaborationist behaviors, local cadres of the CCP favored militant anti-collaborator policies in rural areas between 1946 and 1948 and in cities from 1950 to 1952. In contrast, relatively pacificatory policies were launched in cities before 1950 through limited sentimental references and nuanced classificatory criteria. Why did the CCP adopt radical measures in its discourse politics on colonial collaboration only in the former cases but not in the latter ones? The current thesis contends that the extent of anti-collaborator policies was contingent upon the mediating role played by collaborators in effecting institutional change. By flexibly and evolvingly historicizing the boundary of colonial collaboration, the CCP aimed to distinguish positive institutions of past rule from negative ones and transform colonial institutions from exploitative origins to revolutionary assets. This study explores a potentially transportable theory of discourse politics and classification struggles in the context of post-colonial institutional transition.