We propose new articulation of the differences between the natural sciences and the social sciences. Drawing on science and technology studies (STS) we reconstruct the organizational and cognitive mechanisms of a certain type of natural sciences, one which is referred as laboratory science or highconsensus, rapid discovery sciences. The key features of those sciences crucial for their cognitive and engineering success include: • experimental reproduction of the studied phenomena in the laboratory; • laboratory interventions and modifications of the phenomena thus evoked and broadly understood scientific ‘tinkering’; • attempts to transfer the artificial arrangements developed in laboratory to non-laboratory settings. The STS perspective not only helps us to explain the differences in status and effectiveness between the social sciences and the natural sciences. It also allows us to formulate certain general recommendations for the development of the social sciences. We attempt to show that sociologists are able to implement engineering projects in certain domains of social reality, projects involving the creation of closed sociotechnical systems—analogous to the ones which are generated by natural laboratory sciences. We refer to those systems as ‘social machines’ and the proposed research methodology is called ‘syntheticmethodology