In pursuit of formaldehyde: causally explanatory models and falsification

Falsification no longer is the cornerstone of philosophy of science; but it still looms widely that scientists ought to drop an explanatory hypothesis in view of negative results. We shall argue that, to the contrary, negative empirical results are unable to disqualify causally explanatory hypotheses—not because of the shielding effect of auxiliary assumptions but because of the fact that the causal irrelevance of a factor cannot empirically be established. This perspective is elaborated at a case study taken from the history of plant physiology: the formaldehyde model of photosynthesis, which for about sixty years (1870s to 1930s) dominated the field—despite the fact that in these sixty years all the attempts to conclusively demonstrate even the presence of formaldehyde in plants failed.