However, a lot has changed in the field since him, even if his critique still has validity.  Heckman, for example, has re-focused much of what is considered economic inquiry precisely to question of cause and effect based on observables.  However, the primary focus of social science remains on what one can deduce about unobservables -- justice, well-being, inherent traits -- from observable data.  This differs from much work in natural science in one way very little: even if a phenomenon is unobservable to human senses, it's presence can be deduced from theory and observation of what can be observed.  However, the key difference in social science are that the unobservable components are dependent upon social contexts, which are not the case in most questions of natural science.

Anyway, Migeru has warned me to not discuss this anymore in this thread.  Good discussion.