Metaphors are indispensable figures of speech. By them the abstract is made concrete, the dull and dead is recreated with interest and vivacity, and the complex is cast in recollectable form. What would literature be without the elaborate metaphors of Dante's Commedia and Bunyan's Pilgrim's Progress, the enchanting allegories of George MacDonald and C. S. Lewis, and the evocative modern parables in Abbott's Flatland and Orwell's Animal Farm? Likewise metaphors, or ‘models’, are vital to the scientific enterprise. Without the model of ocean waves Huygens might not have formulated the wave theory of light; without the ‘billiard ball’ model of molecular collisions, the kinetic theory of gases would be difficult to explain; and apart from Bohr's planetary model of the atom, the development of atomic theory could have been greatly retarded.

Yet particular metaphors are always dispensable, if not in literature, whose themes are eternal, then in science and the philosophy of science, where discovery and understanding are often contingent on the model or metaphor employed. Thus Darwin acknowledged that the term ‘natural selection’, which he conceived by analogy with the ‘artificial selection’ of breeders, was less satisfactory than Herbert Spencer's phrase, ‘the survival of the fittest’, after Alfred Russel Wallace pointed out that this was ‘the plain expression of the fact’.