When I took my undergraduate courses in philosophy of science, the general dogma — laid down by mid-20th century philosophers and historians of science like Karl Popper, Imre Lakatos and Thomas Kuhn — was that logical empiricism was naïve and that the experimental sciences were thoroughly theory-laden. Post-Kuhnian science studies didn’t change this epistemological dogma; surely the Edinburgh school opened the black-box of scientific practice and actor network theorists eschewed epistemological issues altogether, but for the last fifty years or so nobody has ever really suggested that science might be an basically observation (data) driven enterprise.
But there are many signs that biomed/biotech practitioners are about to work out a new spontaneous philosophy of science. The whole jargon of systems biology (‘robust’ data, ‘high throughput analysis’) is based on the notion that hypotheses and theories can somehow be harvested from observational data. And in the editorial of the last issue (#2, 27 April 2007) of Lab Times, Craig Wenter’s recent large-scale collection of ocean bacterial genomic DNA is defended against critics who have suggested that such data hoovering is mindless because it doesn’t have the solution of any specific scientific problems in mind. In the editors’ words:
Science isn’t and has never been only hypothesis-based, as some of the high priesthoods of science would desparately have us believe. Science always starts with observaton and description, and only later proceeds by building and testing hypothesis based on the described.
Interesting! A CERN physicist would probably never reason like this. But Lab Time‘s editors are presumably tuned in with their peers. Are we witnessing a return to inductivism in the life sciences? And what does this aggressive language use (‘high priesthoods’, ‘desperately’) signal?