The ‘I/we have heard/learned’ narrator by definition was not himself contributing the essential dramatic ingredient of direct witness. By alleging that he was the mouthpiece of tradition he was distancing himself from the events themselves by at least one remove and often many more. I call narrative with this inbuilt element of transmission ‘story’. Story, thus constituted, especially if it was dealing with happenings outside living memory, was in no position to enforce a hard-and-fast distinction between original ‘fact’ and later ‘fiction’. Beowulf himself, for instance, probably never existed. In all likelihood stories were attached to him which had been attached to others before: for example, as Campbell pointed out, Beowulf probably owed his youthful swimming contest with Breca (499–589) to taking over the adventures of another hero who competed with Breca in this way in a lay known to the Beowulf poet and who was perhaps a Finn, since the sea, for no apparent reason, is said to have carried Beowulf to Finland (579b–81a). Story's way of validating itself was to incorporate authenticating processes of witness as part and parcel of its own narrative. To carry conviction such testimony had to be true to collective experience, just as poetic language was. Story was symbolic in the observation it included as well as in the words it used: it reported representative witness in representative language.

Men when gathered in hall or other assembly were society's basic means of establishing corporate witness.