In an age when there is much talk of 'canons' in the sense of 'authoritative lists of approved (rather than “genuine”) texts', it is curious that even the second edition of the Oxford English Dictionary (1989) does not recognize that meaning in English. But from classical scholarship that meaning has now spilled over into more general use. 1 The important point has been made, though, that the Greek word canon was never used by the Greeks or Romans in this sense. In fact, only in the Christian era does something approaching this meaning occur, and naturally it is used of what is perceived as the genuine body of Holy Scripture. It is never used to refer to any body of secular literature. 2
We need to go further than this simple observation, for we are not just dealing with semantics. Not only did the ancients not use 'canon' in this sense, but in fact they had no word at all to express what 'canon' has meant since the late eighteenth century in classical scholarship: that is, an authoritative select list of authors. 3 That very fact is significant. For the early Church, it was most important that some commonly accepted body of texts was regarded as authoritative: a religion based on books could hardly have it otherwise. But there was no equivalent need to have an agreement about the important texts for the study of pagan literature. And in fact there was no such agreement-hence, no word for 'authoritative select list'. For there were no such lists in antiquity: when speaking of the various 'canons' of ancient authors we are using a metaphor from later religion, and one that suggests a rigidity quite out of step with the way ancient literature was studied by the ancients themselves.
What certainly did exist in the ancient world were lists of authors regarded by their compilers as representing the best in a particular genre. This of course is something quite different from a 'canon'; it