… is from page 235 of the final volume (2016) – Bourgeois Equality – of Deirdre N. McCloskey’s ingenious trilogy on the essence and role of bourgeois values in modern life (footnote deleted; link added):
The best question one can ask of a scientific proposition – or for that matter of an ethical or aesthetic one – is How Do You Know? (Milton Friedman used to ask it regularly in the Money Workshop at the University of Chicago during the 1970s, striking terror into the hearts of students and colleagues.) Clive James, though a literary chap, admires it in the form of “the scientist’s unsleeping attention to the question of what constitutes evidence.”
The tired scientist, too drowsy to do good work – and the poor scientist, wide-awake and well-caffieneated but trained, circus-seal like, only to follow by rote a tired recipe of fine-sounding steps for gaining knowledge – ignore the complexities of epistemology. In the natural sciences, such lazy or mechanistic approaches to gaining information, knowledge, and understanding are more readily than in the social sciences exposed as flawed. The bridge collapses if its design is poor.
Too few such clear and unmistakable real-world tests are available as constraints on the work of social scientists. The enormous complexity and sheer number of human social interactions, and the unobservability of many of the data that are essential for a more-full ‘picture’ of today’s economy (such as Sam’s subjective preference for risk and Suzy’s subjective estimation of success for her new business venture), make simple recipes for describing and understanding economies and economic processes practically useless (save as means for impressing those people who ponder with insufficient seriousness the nature of knowledge and the incomprehensible complexities of modern economies). In the doing of theory and history (including very recent history) in the social sciences, bad work is much easier than in the natural sciences to pass off as good work, and good work is much easier than in the natural sciences to dismiss as being ‘simplistic.’