On the diversity of stationary cosmologies in the first half of the twentieth century

Before the establishment of the hot Big Bang scenario as the modern paradigm of cosmology, it faced several early competitors : the so-called stationary cosmological models. There were truly plural and independent approaches incorporating cosmic expansion but without time evolution of the total cosmological density, thanks to the inclusion of some processes of continuous matter creation. We distinguish here three different independent motivations leading to a stationary vision of the universe. First, Einstein’s concerns on the asymptotic behaviour of gravitation led him to consider continuous matter creation in a recently discovered unpublished work dated of 1931. Second, there is the quest by Dirac and Jordan for a scientific explanation of numerical coincidences in the values of fundamental constants, leading both to a time variation of Newton’s constant and spontaneous matter creation. The third one appears in the steady-state theory of Bondi and Gold as the postulate of the Perfect Cosmological Principle according to which the properties of the universe do not depend in any way of the location and the epoch of the observer. Hoyle developed a mathematical model of spontaneous matter creation from a modification of general relativity that should be considered as a direct legacy of Einstein’s and Dirac’s approaches. Hoyle’s model allows obtaining a “ wide cosmological principle ” as an end-product, echoing Bondi and Gold’s Perfect Cosmological Principle. Somewhat ironically, many modern key questions in hot Big Bang cosmology, like dark energy and inflation, can therefore be directly related to the physical motivations of the early stationary cosmologies.