(Journal of Philosophy, 1988, 85:303-318).

provides a superb and unusual opportunity to gain insight into the meaning of experiment as a method in science. The primary desideratum in all experiments is reduction of confusing variables: we bring all the buzzing and blooming confusion of the external world into our laboratories and, holding all else constant in our artificial simplicity, try to vary just one potential factor at a time. But many subject defy the use of such an experimental method—particularly most social phenomena—because importation into the (...) laboratory destroys the subject of the investigation, and then we must yearn for simplifying guides in nature. If the external world occasionally obliges by holding some crucial factors constant for us, we can only offer thanks for this natural boost to understanding. (shrink)