| Abstrakt: |
Statistics are impossible without concepts. As the sociologist of invention Colum Gilfillan put it in 1952, ‘counting begins with definition of the thing to be counted’. This article is concerned with statistics on science, technology and innovation (STI). It documents how official statisticians have, over time, defined the concepts used for measurement. Debates on definitions of the concepts measured started at the very beginning of STI measurement in the first half of the twentieth century. Then, from the early 1960s onward, methodological manuals were developed to conventionalize the definitions. This article claims that the manuals did not have the expected result. They did not stabilize the definitions and the statistics based thereon. [ABSTRACT FROM AUTHOR] |
