are sometimes clearly cognisable in the compound sagas for which they supply the genuine material.
The next phase is one in which several such small sagas are combined by an artistic narrator into a whole complex large saga.
Specimens of this sort are the Eyrbyggia Saga, Gudmund's Saga, Laxdaela and Nial's Saga, where in different degrees, and with very different skill, many small biographies are melted down into an often elaborate and highly- wrought history of the district for several generations.
Stories might pass through both these stages and still be left pure from foreign matter, save that with which tradition itself had garnished them, while they were being told round bivouac fires or on a night-watch at sea by the comrades or contemporaries of the hero to a younger generation, or to persons from a different part of the country. Thus a folktale might adhere to a genuine character or a genuine anecdote, and it may be sometimes difficult to distinguish between this early natural adhesion, and a late and deliberate addition. But stories that have reached the second phase have usually been helped into it by book-learned scribes.
For these phases (the first of which certainly preceded the use of writing for recording these stories) were not the only forms that the original traditionary material was forced to assume in the course of time.
When it became interesting to writers to gather traditions, and necessary to put them into accepted saga form for audiences and patrons, the merest fragments of a forgotten story would suffice to the professional restorers and imitators. They would build up a large and perfectly worthless (worthless both as regards art and history) structure upon a nickname, a bit of genealogy, the name of a homestead, or a single tradition or scrap of verse. So the saga of Gunnlaug Snakestongue is almost entirely unhistorical, untraditional, and the work of a scribe deliberately "making bricks without straw".
