The most convincing contemporary approach to explanation in social
research comes from Boudon (1984).With regard to social change, Boudon
distinguishes relatively closed and relatively open situations and social
systems. Open situations show a relatively large degree of ‘disorder’. He
sees two categories of generalizations, matching these types of situations.
In the first place, there are laws and theories ‘stricto sensu’. Laws and theories
belonging to this category, which might be falsified according to the
requirements formulated by Popper, are rare in the social sciences, and presumably
they will remain rare, because the subject matter of the social sciences
shows a lot of disorder. Boudon’s second category contains formal
theories. These theories he sees as not in themselves directly applicable to
reality, but offering a mode of discussion or describing ideal examples
which may be useful in the analysis of certain process.