ABSTRACT
Thus far we have seen that significant changes for the better can be wrought in quantitative sociology by the incorporation of generative thinking and revised closure strategies into the research process. This leaves for consideration the third of the broader prerequisites of the realist strategy for measurement, namely the network structure of explanation. In brief, I shall be asking in this chapter whether sociology can adopt this mode of explanation and testing which is based upon a network of scientific discourse in which connections are made between relatively speculative ideas and certain other concepts which are understood well enough to control and measure. To what extent can sociology develop ‘knot concepts’, which are commonly identified by many theories and many instruments and so perform a key systematizing role when it comes to the testing of new theories?
