Scholars of politics have been familiar with normative-empirical distinction. Yet this article reconsiders this divide through exploring another classification in terms of the “observable” and the “unobservable”. According to this new classification, we can assume two types of politics research. Firstly, there are researches which are based on the positivist epistemology and, therefore, deal with the observable. In these cases, the cooperation between positivist, empirical analyses and the normative political philosophy focusing on moral values such as justice, equality and freedom would be feasible. Secondly, there are researches which are based on the non-positivist epistemologies including both interpretivism and realism, and, therefore, deal with the unobservable in some senses. In these cases, the cooperation between non-positivist empirical analyses and the “politics-political” political theory focusing on topics about the nature and the role of politics and the political would be feasible. Consequently, this article contends that we can rethink the existing distinction between the normative and the empirical in political science; the distance either between normative political philosophy and “politics-political” political theory or between positivist empirical analysis and non-positivist one might be farther than what is ordinarily drawn between normative political theory and empirical analysis.