The thesis investigates the scientific citation and its various functions in the scientific community and develops it as a tool for research in theory of science, scientometrics and science studies. Through empirical and theoretical studies that utilize both quali-tative and quantitative methods the citation’s history, method, as well as its use in research policy is examined. Through a historical study the thesis shows three stages of development as re-quired for the construction of the citation as an indicator of intrinsic aspects of sci-ence. These consisted of a) citations as technology, within the citation index, b) the citation as a research method as theorised within Mertonian sociology of science, and c) the citation as a research subject. Following this the “citation debate” in Science and Technology Studies (STS) is described and analysed, which questions the use of gen-eralized quantitative methods. Inspired by an STS approach a performative model of “the mangle of the citation practice” is developed. This aims to understand the cita-tion existing in a context where researchers, articles and the citation index are mutu-ally creating and recreating each other. The thesis uses the HistCite scientometrics tool to develop a novel methodology that highlights local dynamics of citation prac-tices between scientific authors and texts using a visual approach of identifying pat-terns of citations in graphic representations of articles and their citation patterns. For this a “citation typology” is created to identify specific patterns and phenomenon in HistCite graphic representations. The last empirical study is of the introduction of quantitatively based performance-based models for funding of research in Norwegian and Swedish research policy 2003-2010 which problematizes the part played by the citation in a research policy setting as “unobtrusive” indicators of scientific practice. The thesis demonstrates the significance of the citation in research through its de-sign as a reflection of the scientific reference, and result of it being constructed – and used – as an indicator of scientific quality. Furthermore, it shows an emerging awareness in the scientific community that quantifiable indicators of scientific achievement – of which the citation is perhaps the main element – has gained a prominent role in both internal and external domains of scientific practice