The presentation's focus is the Swedish science policy debates on bibliometrics used for the national funding allocation system for higher education and how it is torn between two qualitatively different systems of research impact measures. At present, the national Swedish bibliometric model recognizes and measure research quality as equivalent of publication and of ”being cited”– how well researchers make themselves cite-able in citation based metrics. However, a public inquiry last year proposed a new model similar to the “Norwegian model” of impact based metrics where high quality is seen as equivalent to the prestige of the publication channels, i.e. the status of the scientific journals and publishers. This model is proposed to be implemented from 2014, barely five years after the introduction of the current model. What are the consequences of such a constant reorganization of funding practices and how does it influence the publishing and research practices of scientists? To discuss this, Swedish humanities research will be used as a case, as it during the last years appears - erroneously it will be argued - to have made a great increase in publication rates in Web of Science. In addition I will discuss the lack of transparency and comparability between the various quantitative models of performance indicators such as the bibliometrics models used at faculty and national levels. All in all, the aim is to provide grounds for discussing how citations as well as other bibliometric variables not just reflect but actually effect the quality of research – how bibliometrics is performative.