THE CHARACTER OF GUNNLAUG SERPENT-TONGUE

discussing Gunnlaugs saga ormstmgu, this essay will deliberately avoid the questions of date, sources, literary borrowings, freeprose vs. bookprose, authorship of the verses, etc. not because these things are not important, but because on one level the saga can be profitably discussed without reference to them. In fact, there is need for more discussion of this and all the Icelandic family sagas on what might be called, for want of a better term, the purely "artistic" or "literary" level. Perhaps too much time, has been spent sorting out sources and dividing what is borrowed from what is invented, and too little time, in proportion, talking about the finished product. To anyone familiar with the sophisticated criticism of other literary monuments ancient, medieval, and modern what has been written about the Icelandic sagas as works of art comes inevitably as a disappointment. The reason for the meagre harvest is that critics have not disentangled themselves sufficiently from the scholarly problems and from false conceptions about the nature of the sagas into which their concern for these problems has led them. Research which should have aided judgement has instead hindered it. This is not to argue for an irresponsible application of the "new" criticism to the sagas, but it is a plea for a responsible historical criticism which is also responsive to the artistic integrity of the text. Nothing will be lost in the attempt, at any rate, if we look at the family sagas now and then in the same way we look at most other works of art, as if they reflected a more or less conscious intention and order and were more than the awkward products of several centuries of oral and written tradition as set down by a thirteenth century writer or scribe.1 The first part of this essay will show how artistic readings have been