Social mechanisms and mechanismic explanation

Researchers in all fields dealing with reality wish to explain facts once they have described them. The prevailing account of explanation in the philosophical literature is the "covering law model". According to this opinion, to explain a fact is to deduce the proposition(s) describing it from a theory together with the appropriate data, such as initial conditions. This is not how explanation is conceived of in the advanced factual sciences. In these, to explain a fact consists in unveiling the mechanism that makes the thing in question "tick". In turn, a mechanism consists in the process(es) characteristic of the given thing. For example, metabolism is the central mechanism of living beings, learning is that of schools, and trade that of markets. I call mechanismic this kind of explanation. Therefore, anyone wishing to control a concrete system had better start by finding out its typical mechanisms, so as to maintain or alter them. This is what social control mechanisms, from cooperation and charity to legislation and violence, are supposed to accomplish. Typically, social mechanisms involve at least two levels, micro (individual) and macro (institutional). Consequently they can be neither understood nor designed on the basis of either individualism or holism. I argue that only systemism, the view that every thing is either a system or a component of one, can satisfactorily account for the centrality of mechanisms. Warning: Merton, Giddens and others often called "structure" what I call "mechanism". Following mathematical usage, I define the structure of a system as the set of all the relations (in particular cohesive bonds) among the system component (endostructure), as well as the relations between the system components and things in the environment.