[86] Cf. Erec, l. 1526; list of knights, l. 1691 et seq.
[87] Cf. Erec, l. 1699; Hartmann, Erec, l. 1635; Diu Krône, l. 1379 (Adventure of the Cup); Lanzelet, ll. 7353-64.
[88] If Malduz, or Malduc, were a well-known enchanter, and connected with the Arthurian story, as he appears to have been, how did he vanish from it? Was it the greater popularity of Merlin which displaced him? What is the origin of his name? It sounds as if it might be Celtic, or can he be in any way connected with Maugis, the resourceful cousin of 'Les quatre fils Aginon'?
[89] So far as the Perceval story is concerned, there is certainly evidence of varying forms, e.g., Whence did the continuators of Chrétien, notably Gerbert, draw their versions? And what of the Perceval embodied in the Dutch Lancelot, which appears to be independent, so far as the working out of the adventures suggested by the Grail messenger are concerned, of any known version?
[90] Professor Foerster's attempt to base an argument on the source of Cligés cannot for a moment be accepted, cf. Introduction, Charrette, cxxxviii. We only know that the source was a book; but what that book contained, no one can say. We can never argue from the unknown to the known. We do not know much of Chrétien's sources for the other poems, but the grounds for an investigation do exist in the above instances, they do not in Cligés. We must find out how Chrétien dealt with Erec, Yvain, and Perceval before we are in a position to offer the slightest hypothesis as to his treatment of Cligés. The fact that Mark of Rome gives a short summary of the story is interesting, but so brief a résumé is of little critical value. It is certainly not a book, therefore cannot possibly be identical with Chrétien's source.
[91] On this subject, cf. any scientific collection of folk-tales, e.g., The Science of Fairy Tales, by Mr. E. S. Hartland, or in the same author's Legend of Perseus, the tabulated variants of the Dragon story in vol. iii. These would help the reader to realise the number of motifs often combined in a single story. The lais of Lanval, Graalent, and Guingamor, comparatively short though they be, yet combine at least three distinct story-motifs, i.e. what we may call the Joseph and Potiphar's wife, Tannhäuser, and Lohengrin themes. Any one of these lais would be capable of considerable expansion.
[92] I have studied the Yvain versions carefully, and have read those of Erec, but not compared them critically; but I should not be surprised if it were ultimately found that in The Lady of the Fountain we have the story at a stage anterior to Chrétien, and probably that at which it came into his hands, redacted by the Welsh scribe under the influence of Chrétien's poem; while in Geraint we have the process reversed, i.e. a rendering of Chrétien's poem modified by the earlier version. In the statement, 'Gwiffert Petit he is called by the Franks, but the Cymry call him the Little King,' we have, I think, a hint of this. The writer must have been too good a French scholar to think the one term a translation of the other; it rather implies that the Welsh knew the character only by a sobriquet borrowed from his diminutive size, which is exactly what we should expect, the earlier stages of story-telling being anonymous. So far as the correspondence in word and dialogue is concerned, the conclusion to be drawn depends entirely on the nature of the parallel passages; if they be merely such ordinary dialogue (question and response) as would naturally spring from the incidents of the story, both may well be reminiscences of the oral version. Analytic, self-communing passages would, of course, point to a later stage in evolution; but the Welsh version dialogue is of the simplest description.
[93] Professor Foerster recognises this argument in a measure, but does not appear to realise its full bearing.
[94] I should myself be inclined to limit Chrétien's share in the work to the rearrangement of existing combinations. I do not think he ever made any new combination, unless it were in the case of Cligés, and that is only a 'perhaps.'
[95] Cf. Lays of Graalent and Lanval, p. 175.