Structure de l'univers - quand l'observation guide la théorie... ou pas

The scientific method is often presented, e.g. to children, as a linear process, starting by a question and ending by the elaboration of a theory, with a few experiments in-between. The reality of the building of science is much more complex, with back-and-forth motions between theories and observations, with some intervention of technology and randomness. This complex process is not always correctly understood and assimilated, even amongst scientists. The hero cult, mixed with some revisionism, still exists despite in-depth historical studies. In this context, it may be useful to comparatively examine the reaction to crucial observations, their interpretation and their impact on the contemporaneous theory development. Four examples are presented here, all linked to the question of the 'construction of the heavens' but at different epochs.
invited contribution, in French, published in Azimuth, vol II, nr. 4, p15-39 (2014), special issue "ratio mundi: possible cosmologies between narrative and logic"