The aim of this article is to elucidate the notions of explanation and
mechanism, in particular of the social kind. A mechanism is defined a
s what makes a concrete system tick, and it is argued that to propose
an explanation proper is to exhibit a lawful mechanism. The so-called
covering law model is shown to exhibit only the logical aspect of expl
anation: it just subsumes particulars under universals. A full or mech
anismic explanation involves mechanismic law statements, not purely de
scriptive ones such as functional relations and rate equations. Many e
xamples from the natural, biosocial, and social sciences are examined.
In particular, macro-micro-micro-macro social relations are shown to
explain otherwise puzzling macro-macro links. The last part of the art
icle relates the author's progress, over half a century, toward unders
tanding mechanism and explanation.