In the course of our investigation we found it necessary to devote especial attention to the work of Chrétien de Troyes, endeavouring to ascertain the exact position which, in the evolution of the Arthurian romantic cycle, should be ascribed to this famous poet. It became clear that a very considerable portion of the matter with which he dealt belonged by its nature to the domain of what we call folk-lore; and by reason of that nature could not have been invented by the poet, but must have ante-dated, in some instances by many centuries, any possible literary rendering. Judged by the rules laid down by scientific authorities on comparative religion, and story-transmission, Chrétien could not have been an inventor, but only a brilliantly successful re-teller of stories long known and popular. Instead of standing at the source of Arthurian romantic tradition, he was swept into the current at a comparatively late period of its evolution. To solve the complex problems of Arthurian romance we must go behind Chrétien: it is the period preceding, not following, his work in which the solution of our puzzles must be sought.
To this Chrétien himself bears witness. The position claimed for him by certain modern scholars is not that which he claimed for himself; he never professed to be telling a story no one had ever heard before, though he may have flattered himself, not without reason, that he was telling it better than it had ever previously been told. He was dealing with heroes and adventures already well known to his public. The manner in which he introduces, or refers to, incidental characters makes it abundantly clear that he expected his readers to understand his allusions. Especially is this noticeable in the case of Perceval, who has been claimed, with more zeal than discretion, as one of his most famous creations. He alludes to the hero in a manner that makes it quite evident that this story was well known, and the name familiar, to the public, some decades before Chrétien himself undertook to tell it.
As practical results arising from these studies I would claim:
a That we, in future, place the evolution of the Perceval story at a much earlier date than we have hitherto been willing to assign to it.
b That we admit the possibility of very important variations in the tale, some of them being anterior to Chrétien's version.
c That we recognise that this story of Perceval was of capital importance in the general evolution of the Arthurian cycle.
d That in the mutual relations between the Perceval-Grail and Lancelot stories we have the key to the final shaping of the entire cycle.
These principles admitted, and I think the evidence adduced goes far to prove their soundness, it is obvious that in order to establish and appraise the above relations at their full value, we must have complete and critical editions of all the principal texts. As matters stand at present, the only texts which can be said to have been in any sense critically treated are the Didot Perceval, and the Parzival of Wolfram von Eschenbach for the older story, and the Charrette for the younger. We have been waiting for years for a critical edition of the Conte del Graal, and when we get it will the editor have taken into consideration the various additions to Chrétien's text, and the version of the Dutch compiler, or will it be Chrétien's portion of the poem alone? In that case it will not help us very far. We need sorely a critical edition of the curious Perceval li Gallois, with its blending of wild, folk-lore features with late proselytising and allegorising tendency, its baffling parallels to the German Parzival.
And if we are at a loss for material to adequately criticise the earlier story, what of the later? Considering the highly mythic, prehistoric character of so much of the Arthurian tradition, the disappearance of so many of the intermediate stages, and the consequent difficulty in fixing the earliest form of any characteristic feature, it would seem that our best plan would be to start from the final form assumed by the cycle and work gradually backward, since for a certain period, at least, we might hope to find solid ground beneath our feet. But the most important text for this final form of the Arthurian cycle, the prose Lancelot, remains unedited. And indeed it might well seem to be a work beyond the powers of any one scholar; the number alike of MSS. and of printed editions is so large; they are so scattered, no important library but can show one or more Lancelot texts, and we cannot afford to leave even one of all this mass unexamined. The great discrepancy between the printed texts which the foregoing comparison has shown us; the pregnant hints as to earlier redactions, which the passages I have quoted from M. Paulin Paris and Professor Heinzel assure us may be found in the MSS., are all indications of the vast extent of the task which confronts us.
Yet this much is certain, until it is boldly grappled with, and scholars are in possession of a complete critical edition of the Lancelot in which all the varying adventures shall be carefully chronicled, and all the traces of earlier redactions duly noted, any studies such as these in the preceding pages, be they the work of scholars of the very first rank, will always be liable to the necessity of revision, or the risk of subversion, by the accidental discovery of some hitherto unknown factor.[212]
This appears to me to be the great and pressing question which confronts Arthurian scholars; we desire our work to have a permanent value, yet we are leaving undone that which, to all appearance, is the surest means of securing such permanence.
A work of such magnitude can, I think, only be grappled with by a body of scholars, a chief editor, assisted by a group of sub-editors. The great extent and diffusion of the material (the Lancelot MSS. are, as I said before, practically scattered all over Europe), render it impossible for any one man to hope to complete the task within a reasonable term of years. I do not know what may be the principles regarding the choice of publications by the Sociétié des anciens textes Français, whether their aim be the introduction to the public of MSS. of which unique copies alone exist, rather than to publish critical editions of more easily accessible texts; but if the latter should lie within their province, I cannot imagine any publication that would be more warmly welcomed by Arthurian scholars, or that would be of greater interest and more enduring benefit to the students of mediæval literature, than a full and complete edition of the prose Lancelot.