John Maynard Keynes, dans son Treatise on Probability (1921), distingue clairement l’incertitude et le risque, le risque étant de l’incertitude complètement structurée par des probabilités objectives. En baptisant « principe d’incertitude » le fameux principe qu’il a mis au cœur de la mécanique quantique, Heisenberg a ignoré la distinction keynésienne. La mécanique quantique en effet, tout en écartant la causalité déterministe, caractérise tous les phénomènes physiques de son ressort en termes de distributions complètes de probabilités objectives. En revanche, des sciences comme l’écologie, la climatologie, l’océanographie, et, à bien des égards, la science médicale, sont au moins partiellement des sciences incertaines au sens de Keynes. Cela ne les rend pas inaptes à éclairer des décisions collectives dont les conséquences peuvent être lourdes, dans la mesure où des méthodes rigoureuses ont été développées à cet effet – et expérimentées – dans un passé récent. Mais cela les rend vulnérables à des manipulations de l’incertitude qu’elles comportent, manipulations qui visent à les discréditer dans leur capacité à éclairer les décisions à prendre.
Scientific uncertainty and fabricated uncertaintyEconomics is erratic, politics is manipulated, but at least science is objective and certain, hence its predictions are reliable. This image of science is inherited from the creators of modern science and the scientific method, a few centuries ago, and runs deep in the minds of most people when they consider scientific issues. Without calling into question the method itself, Werner Heisenberg’s most famous contribution to quantum mechanics, the Uncertainty principle that bears his name, seems to bar predictability. However quantum mechanics provides perfect statistical predictability, and this is what one perceives when using electronic devices, lasers or magnetic resonance imaging. With sciences like ecology, climatology, oceanography, and even medical sciences, the situation is significantly different; to various extents they are genuinely uncertain, embodying the distinction between risk and uncertainty made by John Maynard Keynes in his Treatise on Probability. Despite significant methodological difficulties, for the last twenty years scientists, including economists, have developed structured, rigorous and operational approaches to evaluating and deciding under conditions of genuine uncertainty. These achievements however do not appear to carry enough weight in the controversies surrounding these «uncertain» sciences; while scientists work on reducing or taming uncertainty, many powerful and wellorganized actors in the economy, in politics and in the media, are busy denying scientific results and fabricating more uncertainty than actually exists, in order to undermine policies that hurt their particular interests and ideological prejudices. Such controversies -rather different from productive controversies in scientific research- are particularly virulent in the USA, but are also developing in China, Europe and India, with the rather serious effect of blocking efforts towards implementing more sustainable forms of development.Classification JEL : D81, Q54
