Among the physical sciences, cosmology is distinctive: its domain is a unique and irreproducible system, the Universe; those who study it are inescapably immersed in the system; almost all of it is inaccessible to them; even on the accessible fragment there is almost no possibility of experiment.footnote1 At the same time, the employment of continually improving techniques produces a flood of observational results. These distinguishing characteristics are partly responsible for the present state of cosmology: the inadequacy of the theoretical foundations becomes increasingly apparent, the contradiction between theory and observations increasingly acute. To match current theory with observation, for example, one must believe that 90 per cent or more of the matter in the Universe is unseen and of unknown composition. The unsatisfactory and unstable state of the subject has given popularizers of cosmology great ideological scope. I shall discuss two examples below: Stephen Hawking’s A Brief History of Time, and Barrow and Tipler’s The Anthropic Cosmological Principle. First I shall establish a historical context in which the crisis can be understood.

The history of Western cosmology is in the first place the history of the displacement, and then the abolition, of a centre. Until Copernicus displaced it, the Earth was at the centre, and for some centuries after Copernicus, the Sun.footnote2 There were earlier advocates of the Sun as centre, for example Aristarchus, but their influence was slight.

Until early in the present century it was not clearly understood that the Sun is a star of rather common type, located within this Galaxy (the Milky Way).footnote3 Nor was it known until the 1920s where the boundaries of the Galaxy should be drawn, or indeed whether anything lay beyond it. The ‘spiral nebulae’ had been seen, but even with nineteenth-century telescopes they could not be resolved into stars, let alone located beyond the confines of the Galaxy.

Shortly after World War I, when the 2.5 metre Mount Wilson telescope became available, Harlow Shapley sketched a map of the Galaxy, and located the Sun in its outer parts. The Galaxy had a centre, and the Earth was far away from it. Soon afterwards Edwin Hubble showed that the ‘spiral nebulae’ were also galaxies—systems of stars similar to the Milky Way but much more remote. Without reason to prefer one galaxy above another, one could no longer appeal to astronomy to prescribe a centre for the Universe.

Subsequent observations have revealed billions of galaxies. Most galaxies are found not singly but in groups of a few dozen, or in clusters of hundreds or thousands. The Milky Way belongs to the Local Group, with about two dozen members. Most groups and clusters belong to superclusters. The Local Group belongs to the Local Supercluster, which has about one hundred member clusters.

With or without a centre, the Universe was presumed before the twentieth century to be a more or less static system. The planets, some comets and occasional novaefootnote4 had been seen since ancient times, while small relative motions of stars had been known since the early eighteenth century. The observation of Jupiter’s moons by Galileo in 1609 had been more relevant to the displacement of the centre than to any fresh recognition that there were moving objects in the heavens. Kant, and then Laplace, had advanced a nebular hypothesis to explain the formation of the Solar System, but it was generally accepted that on the whole the stars stayed where they were.

This led to a difficulty, known in the trade as Olbers’s paradox.footnote5 If the system of stars was limited in extent, and if Newton’s law of universal gravitation was indeed universal, then, as Newton himself had realized, gravitational attraction would have caused the system to collapse. This was contrary to observation; the visible Universe was not in a state of collapse. But if the stellar system, always supposed static, extended indefinitely in all directions then it was possible to showfootnote6 that, unless the Universe was of finite age, the whole sky should be at all times about as bright as the sun’s disk. This also was contrary to observation; the sky was dark at night.