how difficult it is for some scholars to retreat from a position they have once taken up.
An unexpected contribution has been made to the Peredur question by a young German scholar, Dr. Paul Hagen, writing in Germania. Readers of my Grail legend may recollect that I claimed this Welsh tale as representing, in part, a purer version of one of the motifs worked into the Conte del Graal of Crestien de Troies, but contaminated with incidents and passages derived from that poem. For this I was taken to task by Dr. Golther, who asserted the entire dependence of the Welsh tale upon the French poem. I may fairly claim to have disproved this assertion,[1] which is indeed absolutely untenable. Dr. Hagen brings forward fresh arguments in disproof of Dr. Golther's theory, and is indeed quite at one with me respecting the anteriority of the Welsh tale. But according to him it is the homogeneous adaptation of a pre-Crestien French work based upon Breton lais and prose tales, which also served as the main source both of Crestien and of the lost French original of Wolfram von Eschenbach. The difficulty lies, it will be seen, in the fact that Welsh and French texts have features in common which point to a definite literary connection. I explain these features as due to the influence of the French poem upon an already existing Welsh tale; Dr. Hagen, as due to derivation from a common original. I fully see the difficulties of my explanation, and I grant that Dr. Hagen has criticised it acutely and vigorously. But destructive criticism is nowhere easier than in dealing with this inextricably tangled literature. The difficulty is to construct a theory that will fairly fit the facts. Has Dr. Hagen fully reasoned out his theory? I doubt it. We both agree that the Welsh tale must belong to an earlier stage than the French poem, because it gives in orderly and coherent sequence incidents of which a shadowy jumble is all that exists in French.
- ↑ In the already cited article, Revue Celtique, April 1891; Folk-Lore, June 1891.
