Myrdal's works are usually analysed with a dual and separated point of view : on the one hand the methodological papers concerning the value problem and based on a strong non neutrality thesis ; on the other part the theoretical analysis concerning monetary theory and policy, with a Wicksellian filiation. In fact both the dimensions are strongly connected by a common way : the application of the Hägerström's Swedish guillotine between is and ought, but also the construction of a bridge between economic science and political views on social engineering and economic policy. Myrdal wants to address this problem : how economic science can become politically relevant ? This paper analyses two stages of that unique project : the proposition of a "technology of economics" (1930), and the selection process for a "norm for monetary policy" (1939). It shows that Myrdal distorts an initial end and means scheme by proposing some intermediary concepts between positive and normative fields. From a theoretical and statistical framework and an explicit value judgment these concepts enable to elaborate an iterative tree of selection of a speci-c monetary policy. If the Myrdal's project encounters difficulties in conciliating a non-cognitivist thesis with economic prescriptions and in proposing a tractable method, it remains an important benchmark for the analysis of the links between positive and normative views concerning monetary policy.
