GILBERT Murray, a great liberal all his life, articulated the problem of knowledge's sovereignty in the twentieth century. He opposed ‘hugging the walls’ of conventional thinking. He had been associated with Jane Harrison, Francis Cornford and others who have been called the ‘ritualists’ who had intro-duced unconventional approaches to classical studies. He was also prominent in promoting political projects such as the League of Nations. In the 1940s he feared people would creep ‘back into their shells.’ Anthropology, he argued, showed that what he rather awkwardly called ‘inherited conglomerates’ had ‘practically no chance of being true or even sensible.’ Yet, he recognised, no society could live without ‘inherited conglomerates’ or ‘even submit to any drastic correction of them without social danger.’ The tension never ends between stabilising mental habits and what Clifford Geertz called escape from ‘stoppered fly-bottles.’

In Britain between 1900 and 1950 various societies formed, organised and also dissolved knowledge's sovereign legitimacy and authority. Their processes were both robust and charismatic but they were not mechanical classifications of mental materials. These processes recognised knowledge's instability and the difficulty of hedging it according to strict principles. The organisation of knowledge by learned people in learned societies consisted of a series of processes that were dynamic and complex. Organisations organise. Such is their heft and power. The most successful of these societies – the universities, the Royal Society and the British Academy, for example – found ways to renovate themselves as they adjusted their processes of organisation and dissolution. The University of London's capture of the Warburg Institute, to the institute's short-term advantage, is an example of such a renovation. It gave the Warburg reputation and, perhaps, the illusion, of permanence. Such an illusion, if that is what it is, might have advantages but also disadvantages.