In the corresponding passage, Geoffrey of Monmouth gives as his authorities 'Breton' tradition and the clerk Walter of Oxford (cf. note to above passage). Layamon in his account is even more severe towards the guilty pair:
In the passage corresponding to that quoted above from Wace, Layamon adds the detail, that none knew the manner of the queen's death, whether she had drowned herself:
From these passages it is abundantly clear that Guinevere was no victim of treachery, but a willing sinner; and that the tradition of her infidelity to her husband existed prior to the formation of the Arthurian romantic cycle.
Granting, then, that the feature formed part of the early Arthurian legend, are we to consider that the version given by the chronicles faithfully represents the original tradition, and that it was Mordred who was Guinevere's original lover? I think not. It is an extremely curious feature of the problem, that though in each of the pseudo-historic versions Guinevere, as we have seen, is genuinely in love with Mordred, and is roundly condemned by the chroniclers for her conduct, in no single one of the Arthurian romances is there any trace of the slightest affection existing between them. Mordred, save as traitor in the final scenes, plays no rôle in the story; he is never represented as a persona grata at court; in one important version, as we shall see, the queen dislikes him because she suspects his true relation to Arthur. Guinevere's moral character is held to be untarnished, even by her liaison with Lancelot.
I suspect that we have here to deal with a lapse of tradition. Mordred is not the original lover, but he represents him; and between that original lover and Lancelot there intervenes a period in which Guinevere's lapse from virtue was smoothed over, and partially forgotten. It is certainly remarkable that in each of the three great prose branches, the Merlin,[120] the Tristan, and the Lancelot, Guinevere's moral character is apparently unaffected by her conduct with Lancelot. The compilers all agree in extolling her as the noblest of queens and best of women. Even so aggressively virtuous and clerical a romance as the prose Perceval li Gallois, though quite aware of the connection, regards Guinevere in a favourable light—indeed, as morally superior to Arthur! Nor can we quote the Queste as representing the opposite view; true, Lancelot is blamed for his relations with the queen, but Guinevere, when she appears upon the scene, is treated with marked respect, and the reader has an uncomfortable suspicion that the writer objected to her rather as woman than as wife,—he objects to the sex as a whole, only forgiving Perceval's sister on account of her virginity. It seems clear that if the character of Guinevere has, among the Welsh, been handed down to posterity under the unfavourable light in which Professor Rhys tells us she is popularly regarded, this must be due either to a tradition emanating from an earlier and healthier state of society, when conjugal infidelity was not regarded with complacency, or to a later and more enlightened verdict on her relations with Lancelot, but in no case can it be due to the influence of those who told the story of these relations.
The second cause will, I think, account for the nineteenth-century presentment of Guinevere's character; we judge her on the grounds of her relations with Lancelot, which we regard as blameworthy, though not undeserving of sympathy—in fact, we do but emphasise Malory's verdict.
But this does not account for the Welsh tradition, which, as I have before pointed out, knows practically nothing of Lancelot; that must rest upon other grounds, and I believe it rests upon the tradition preserved to us in the Mordred story.
What this original tradition was, we can now only surmise; it belonged to a period of which but few and fragmentary traces survive, but I think that most probably the primitive story ascribed the rôle of lover to Gawain. I made this suggestion some four years ago,[121] and subsequent study has shown me nothing to induce me to alter my opinion, though it has suggested sundry important modifications.
I think now that Gawain and Mordred really represent the two sides of one original personality; and that a personality very closely connected with early Celtic tradition.
What the exact nature of the relation between Gawain and early Irish mythic tradition may be we cannot yet say: that such a relation exists is practically beyond doubt.[122]
Among the characteristic features of the early Irish heroes with whom Gawain is connected, we find the following: Adventurous hero and nephew on the female side to royal centre of cycle (Cuchulinn and Diarmid[123]); son to that uncle (Cuchulinn); lover of uncle's wife, eloping with her (Diarmid); deadly combat between father and son (Cuchulinn and Conlaoch). This latter incident I believe to be of greater importance in heroic-mythic tradition than has yet been realised. As I interpret it, the father and son combat in heroic tradition really represents the 'slayer who shall himself be slain,' the prehistoric combat of the 'Golden Bough' (to which I have referred in chap. v.) influenced by the doctrine of re-birth, as set forth by Mr. Nutt in vol. ii. of the Voyage of Bran, i.e. it is a conflict of the god with his re-born and re-juvenated self, and as such has a very real place in Celtic tradition.
As we see above, we do not at present possess a version in which all these characteristics are united in one hero, but they might very well be so united. I think that the earlier Gawain was at once Arthur's nephew and son by his sister,[124] adventurous hero of the court, lover of the queen, and eventually slayer of his father-uncle.[125]
Very probably in the original story there was some such device as the beauty-spot of Diarmid, which aroused involuntary passion in every woman who beheld him; or the love-potion of the Tristan story; a device whereby the earlier tellers of these tales secured sympathy for the lovers, without lowering the character of the husband, so that Gawain, no less than Diarmid and Tristan, would be regarded as a gallant and sympathetic figure.
But the peculiar line of evolution followed by the Arthurian story, the strongly ethical and Christian character which it early assumed (due probably to the heathen belief of the historic Arthur's genuine antagonists, the Saxons), made a change necessary, if Gawain was to preserve his position as leading hero of the legend, and I now think it most probable that that change was effected by divesting Gawain of the characteristics incompatible with his later position, and bestowing them on another personality, created for the purpose, since they could not altogether be dropped out of the story. It is significant that, as I remarked above, the earliest tradition gives Gawain no brother save Mordred, and Layamon remarks emphatically, 'he never had any other.'
Further, I suspect, that exactly the same process took place with regard to Guinevere, and that we have a survival of it in the person of that mysterious lady, the false Guinevere.
I would therefore modify my original views on the subject, by saying that I now think that though Gawain was Guinevere's original lover, Lancelot did not succeed him in that rôle, in fact that Lancelot does not represent the original lover at all, that that tradition is now represented by the Mordred story, and that there was a period in the evolution of the legend, preceding the introduction of Lancelot into the cycle, during which the tradition of Guinevere's voluntary betrayal of her husband was dropped, and she was regarded in an altogether favourable light.
The invention of the Lancelot love-story, which I think we must regard as in its origin an invention, was probably brought about by two causes, the growth of Minne-dienst, and the popularity of the Tristan story.
To be absolutely accurate, I think we ought to consider it as invented to satisfy the demands of the first, and developed under the influence of the second. That it is, as some writers have held, a mere imitation of the Tristan story, I do not think, rather it is marked by certain complex characteristics which cannot be explained on the hypothesis of other than a dual source. Thus it is impossible not to feel that the relations of the lovers are dictated by the rules of a conventional etiquette rather than by the impulse of an overmastering passion. Even in the scene in which Lancelot first reveals his love to the queen, there is no touch of genuine passion or self-abandonment; the confession has been foreseen and expected, and you feel that Guinevere has carefully regulated her conduct in accordance with the etiquette prescribed for such an occasion.
In the Charrette, this artificial character is strongly marked; Lancelot's bearing becomes absolutely grovelling in its humility. The fact that he has been guilty of a momentary hesitation before mounting the cart is regarded by his capricious lady as a deadly offence against the rules of love, and resented accordingly, while Lancelot is so overcome by the assumed indifference of the queen that he promptly attempts suicide. Compare this with the story of Gawain and Orgeluse in the Parzival. Gawain is heartily in love with the lady, who treats him, not merely with indifference, but with absolute insolence—insolence to which Gawain opposes the most serene and unruffled courtesy, till the lady comes to her senses, when he reads her a well-deserved lecture on the correct behaviour due to a knight from a well-bred lady. Gawain is quite as well aware of the rules of the game as Lancelot, but understands how to play it with becoming dignity, and remain master of the situation.
There are moments in the Lancelot-Guinevere story when one wonders whether the whole business be not as platonic and artificial as the love-rhapsodies of the would-be poets of mediæval Italy, or of certain of the troubadours; but the night interview in the Charrette, the story of Lancelot's relations with King Pelles's daughter, and Guinevere's frantic jealousy, together with the final scene of discovery, forbid this charitable assumption.
Again, as I remarked above, the problem is complicated by the high character ascribed to Guinevere, and the absolute lack of any real condemnation of her relations with Lancelot. This is carried so far that even after the final discovery the kingdom of Britain is threatened by the Pope with an interdict unless Arthur will consent to take back his faithless wife; while throughout the war with Lancelot the sympathies of the reader are asked for the knight, not for the king. Nothing could well be lower than the morality of the Lancelot story as it now stands: the cynical indifference of what we may call the 'secular' sections, on the one hand, coupled with the false and wholly sickly pseudo-morality of the Grail sections on the other, cannot but be utterly distasteful to any healthy mind. For my own part, I must needs think the immense popularity of the Lancelot-Grail romances wholly undeserved.
Another point which is often overlooked is the discrepancy of age between Lancelot and the queen; the hero's birth takes place some considerable time after the marriage of Arthur and Guinevere. In the final war with Arthur we are told that Lancelot is twenty-one years Gawain's junior, this latter being seventy,[126] while Arthur is ninety years old! It is quite clear that we have here no tale of the genuine spontaneous love of youth and maiden such as we find in Tristan and Iseult, but rather the account of the liaison between a young knight and a lady, his superior in years and station.
All these discrepancies and difficulties in the Lancelot story can, I believe, be best explained on the lines above suggested. The original story of Guinevere's infidelity had been dropped out of the legend, a reminiscence only surviving in the account of Mordred's treachery. Shortly after the middle of the twelfth century the tone given to courtly society by certain influential princesses, among them Eleanor of Aquitaine and England, and her daughter, Marie de Champagne, demanded the introduction into the popular Arthurian story of a love element, conceived after the conventions of the day. Doubtless the popularity of the older Tristan story was an element in the matter, but we must, I think, guard carefully against regarding the one as an imitation of the other; in colouring and characteristics the tales are, as I said above, diametrically opposed.[127]
Why Lancelot was selected as the queen's lover is a question which it is extremely difficult to answer with any certainty. When I treated the subject in my Legend of Sir Gawain, I suggested that he simply took the place of Gawain here, as elsewhere. That may have been the case, but the fact that, as I now think, we have distinct evidence of an intervening period, or rather of intervening stages, between the stories, somewhat militates against this idea. The choice may have been determined by quite simple considerations. It is noticeable that in each of the poems in which Chrétien mentions Lancelot previous to the Charrette he places him third in the list of Arthur's knights; in Erec the two first are Gawain and Erec; in Cligés they are Gawain and Perceval. None of the three here named would be available: Gawain from his relationship alike to Arthur and to Mordred, besides the fact that the character he early acquired as 'the Maidens' Knight' rather militated against the exclusive fidelity requisite for the post; Erec was already provided with a lady-love; Perceval was impracticable, not so much from the ascetic character ascribed to him, which was probably[128] a later accretion, as from his essentially uncourtly manners, and very slight connection with Arthur's household. It may very well be that at the 'psychological moment' Lancelot, by his new-won position in the cycle, was the one hero who approved himself fitted for the rôle, and thus reached in the character of the queen's lover his final evolution as an Arthurian knight.
Again, as I suggested in discussing the Lanzelet, it may be that some peculiarity in his relations with his mysterious protectress gave the required suggestion. With the knowledge at our disposal the question cannot be definitely answered.
But the central idea once conceived, the process of evolution proceeded merrily: doubts, hesitation, despondency, on the part of the hero, gracious advances on that of the queen; advances on the part of other ladies, jealousy on the part of Guinevere; despair and madness of Lancelot; reconciliation, suspicion, detection, danger, deliverance, all the well-known formulæ of such a love-tale are employed, well interspersed with the knightly adventures of Lancelot and other companions of the Round Table. Such a story could be expanded, ad infinitum, and there is no doubt that it was expanded to an inordinate length, as we shall find when the day comes for a critical edition of the various redactions of the prose Lancelot.
Meanwhile, what of the romance which had given the initial impulse to the formation of the Lancelot story, the Tristan? As a matter of fact the Tristan was in the unenviable position of a Frankenstein. It had created, or rather helped to create, a monster which was its eventual destruction. So far as incidents go, the Lancelot has borrowed but little from the Tristan; the episode of the blood-drops, which betray the nocturnal meeting of Guinevere and Lancelot in the Charrette, is generally admitted to be borrowed from the similar episode in the Tristan poems, while the version given by Hartmann von Aue of the abduction of Guinevere shows points of contact with that of Iseult by Gandîn, but the incidental parallels between the stories are in reality very slight. Turn, however, to the prose Tristan, and you find the influence of the Lancelot absolutely dominant. Following the example of Lancelot, Tristan believes himself to have lost the favour of his adored queen, flies to the woodland, where he goes mad; attempts suicide; Iseult pours out her woes in letters to Guinevere, who is regarded as the noblest of queens, and a recognised authority on love! Guinevere invites the lovers to Arthur's court; Lancelot places his castle of Joyous Garde at their disposal. The details of the beautiful old love poem, the poignant tragedy of Tristan and Iseult, are lost sight of. In a fragmentary form they still exist, but are buried out of sight underneath the great mass of Arthurian accretion. It is no longer the love of Tristan for Iseult which is the central interest of the story, but the rivalry between Tristan and Lancelot, which of the two shall be reckoned 'the best knight in the world.'
Dr. Wechssler, in his study on the various redactions of the Lancelot-Grail cycle, points out the manner in which the two versions of the Tristan have been worked over and modified so as to bring them more into harmony with the Lancelot.[129] But how thoroughgoing was this modification, and how disastrous to the older story, can only be understood by a first-hand study of the texts. An interesting point for future criticism to determine will be whether there was ever an earlier, and independent, prose Tristan, or whether the prose versions of this tale are not all posterior to and dependent upon the Lancelot. I do not think that any question can here arise as to the priority of the poetical relative to the prose form.
To sum up the conclusions arrived at in these pages, I would suggest that the order of Guinevere's lovers, so far as can be determined from the surviving Arthurian tradition, was as follows:
1. Gawain.—This being indicated by Gawain's close connection with kindred Celtic legends; traces of the relation surviving in the accounts given in the Merlin of Gawain as the 'queen's knight,' and in passages of Chrétien's Perceval, Wolfram's Parzival, and early English romances.[130]
2. Mordred.—Representing a period when such a relationship was held incompatible with Gawain's character as chivalrous hero, and the more unamiable features of the primitive conception were transferred to another character who was regarded as Gawain's only brother. The later stages of this period are preserved in the Chronicles.
3. Intervening period wherein Guinevere undergoes same process as Gawain, and false Guinevere is evolved. The queen's character is regarded as irreproachable and Mordred as an unwelcome suitor. Strong traces of this period remain, both in the earlier metrical and prose romances, and complicate the subsequent presentment.
4. Lancelot.—His introduction in this character being due (a) to social conditions in courtly circles, (b) to desire to create within the Arthurian cycle a love-tale which should rival in popularity the well-known and independent Tristan story. Mordred, however, remains in the story, and he, rather than Lancelot, should be considered as representing the original 'infidelity-motif.'[131]
THE PROSE LANCELOT—LANCELOT AND THE GRAIL
We now approach the most difficult and complicated part of an exceptionally difficult and complicated question; rather, to be more accurate, we are now confronted with the union of two questions, each of them, in a high degree, intricate and obscure. We have not yet succeeded in solving the problems connected with the evolution of the Grail romances; we can scarcely be said to have begun the examination of the Lancelot legend; the union of the two might well appear to present such insuperable difficulties that the critic might shrink from grappling at close quarters with so formidable a task. And yet it may well be that this union of the two legends, which at the first glance appears so seriously to increase our difficulties, is precisely that factor which shall play the most important part in their final solution; that inasmuch as the Lancelot legend was the dominant factor in the later cyclic development of Arthurian romances, the disentangling of this particular thread will be the clue which sets free the other members of the cycle, and enables them to fall once more into their original and relative positions.
The elements composing the Grail problem are so well known that here I need do no more than briefly recapitulate them. The Grail romances are practically divided into two families: that dealing with the history of the relic—the Early History romances as they are very generally called; and that dealing with the search for the relic, the Queste, which latter family is again sharply divided into two sections differentiated from each other by the personality of the hero—the Perceval and Galahad Questes. I am not sure whether we ought not to go a step further and recognise a third clearly defined family, that of the Gawain Queste. Mr. Nutt in his Studies on the Legend of the Holy Grail[132] partly recognises this, but does not, I think, attribute sufficient importance to the matter, regarding Gawain as an understudy of Perceval. I incline to think that before the question is finally solved we shall require to study very closely the variants which regard Gawain as Grail hero, and compare them with the Perceval versions. I am not sure that we shall find the result quite what we expect!
So far criticism has confined itself to the question of the relation existing between the Early History and Queste versions, and that between the two main families of the Queste. In this latter case the general consensus of opinion is to regard Perceval, whose story is marked by certain definite and widely spread folk-lore features, as an earlier Grail hero than Galahad, whose Queste is strongly allegorising and mystical in character.
It is this latter Queste which here mainly concerns us, but we shall find that before we are in a position to examine it closely we must deal with certain features both of the Gawain and Perceval variants.
The Gawain versions will not detain us long. There is, correctly speaking, no definite Gawain-Grail romance, but we find records of Gawain's visits to the Grail castle scattered throughout the latter part of the Conte del Graal, Diu Krône (where he is really the Grail hero), prose Lancelot, and Dutch Lancelot (this latter, as we shall see, differing in very important particulars from the prose Lancelot). In each case these adventures are marked by peculiarly wild and fantastic features, sometimes apparently borrowed from the hero's feats at the Château Merveil, as recorded by Chrétien and Wolfram, sometimes entirely independent of those feats, but strongly reminiscent of Perceval's experiences in the Grail castle. In the distinctively Lancelot romances, where Gawain, Lancelot, and Bohort all attempt the adventures of Corbenic, Gawain is the first to do so, and his experiences are repeated, with a more fortunate result, in the case of the other two. The Grand S. Graal, which gives an account of the founding of Corbenic, and the establishment of its marvels, states that none are to escape with their lives till Gawain shall come, and he shall receive shame and dishonour.[133] This same romance makes Gawain a descendant of Joseph of Arimathea.
I think it is quite clear that the Grail castle as depicted in the later romances is really a combination of the features of two originally distinct accounts, the Grail castle of the earlier Perceval story, and the Château Merveil of Gawain legend. The marvellous features which the Galahad-Lancelot Queste emphasises have clearly been borrowed from the Gawain romances, and are therefore to be considered as younger than these.
Dr. Wechssler's study, Über die verschiedenen Redaktionen des Robert von Borron zugeschriebenen Graal-Lancelot-Cyklus, to which I have previously referred, is of value in helping us to the next stage of our investigation. The writer points out that the redactors of the prose romances we possess were familiar with two compilations, practically covering the entire ground of Arthurian romance, one of which, the earlier, was ascribed to Robert de Borron, the other, the later, to Walter Map; or rather, as the author is careful to write throughout, pseudo-Borron and pseudo-Map.[134] The original cycle, which the writer designates A., consisted of Livre del Graal,[135] Merlin, Suite Merlin, Lancelot, Queste, and Mort Artur, but only traces of the Borron cycle remain, the romances as we have them belonging to the pseudo-Map redaction.[136]
Further, Dr. Wechssler claims to have detected clear traces of two subsidiary cycles formed by selections from the original; redaction B. consisting of the Livre del Graal, the Merlin, and Suite Merlin, and the Queste and Mort Artur. The redaction B. he considers the earlier shortened version of the pseudo-Borron cycle.[137]
A still later and shortened redaction was composed of the Merlin and Suite Merlin, Queste and Mort Artur; this also being attributed to the pseudo-Borron.[138]
According to Dr. Wechssler the distinguishing mark which separates the pseudo-Borron from the pseudo-Map cycle is the introduction into the former of the personages of the Tristan legend absent from the Map cycle.
This is very clear, and very interesting, but let us wait a minute before we examine it, and see how, in the hands of its own author, the theory works out. The study to which I have just referred was published in 1895; in 1898 another study appeared from the same pen, this time dealing exclusively with the Grail romances,[139] in which Dr. Wechssler practically adopted the standpoint of Professor Birch-Hirschfeld, that the Grail is ab initio a Christian symbol, but at the same time endeavoured to harmonise this view with that which regards the Grail as originally a heathen talisman, while, in the same way, he claimed to discover a viâ media between the conflicting variants of the Queste, presenting us, as the result, with a curious composite hero, who was named Galahad, but whose story was the story of Perceval.
I do not know if the author was himself really satisfied with the result of his ingenuity; I am convinced no other student of the Grail romances was; but the interest of the study for us lies in this, how did a scholar who three years before had published a really sound, solid, and valuable piece of criticism, such as that on the Grail-Lancelot cycle, come to wander so far astray in the quagmire of pure hypothesis and unfounded assumption? Simply and solely, I believe, because it had never occurred to Dr. Wechssler that the Lancelot romances could be associated with any Queste other than the Galahad Queste. He saw, and saw rightly, that the Lancelot story played a very important rôle in the cyclic evolution of the Arthurian romance; he saw that it was closely connected with a Grail Queste, and never suspecting that the hero of that Queste could be other than Galahad, while at the same time he recognised the priority of certain elements of the Perceval story, he endeavoured, with a fatal result, to combine the two, and evolve such a Queste as would suit the earlier redaction of the Lancelot story.
And yet the key to the truth was in his hand all the time, had he but known it. He knew M. Paulin Paris's 'Romans de la Table Ronde'; on p. 87 of vol. iv. the writer quotes a passage from a MS. of the Bibliothèque Nationale, to which I have previously referred, but which is of such paramount importance for the question before us that I make no apology for repeating it here: 'Et le grant conte de Lancelot convient repairier en la fin à Perceval qui est chiés et la fin de tos les contes ès autres chevaliers. Et tout sont branches de lui (c'est-à-dire se rapportent à Perceval[140]) qu'il acheva li grant queste. Et li contes Perceval meismes est une branche del haut conte del Graal qui est chiés de tos les contes' (MS. 751, fol. 144-48).
To this quotation M. Paulin Paris added the remark, 'Mais dans la Quête du Saint Graal, Perceval n'est plus le héros qui découvre le Graal et accomplit les dernières aventures. Galaad, le chevalier vierge, fils naturel de Lancelot, est substitué au Perceval des dernières laisses de Lancelot. La manie des prolongements aura conduit à ces modifications des premières conceptions. Et c'est la difficulté de distinguer ces retouches successives qui a donné à la critique tant de fils à retordre.'
The position could scarcely be more clearly stated to-day; one can only regret that this luminous hint of the great French scholar should have remained so long unfruitful. When the passage first attracted my attention, which it did some years ago, I made a note of it as important for the theory of the early evolution of the Perceval story, but not till I had read Dr. Wechssler's study of the Grail-Lancelot cycles did its immense importance as evidence for the evolution of the Arthurian cycle, as a whole, dawn upon me. Yet here we have a piece of evidence of the very highest value, a direct and categorical statement that at one period, and that an advanced one (otherwise it would not be termed 'le grant conte'), of its evolution, the Lancelot legend was connected with and subordinate to the Perceval story, and that in its full and complete Grail-Queste form.
In other words, the distinction between the cycles respectively attributed to Borron and to Map is not only the presence or absence of the personages of the Tristan story (as Dr. Wechssler supposes), but the much more important and radical distinction that, in the first the Queste was originally a Perceval, in the second always a Galahad Queste. It is surprising that this distinction had not occurred to the original framer of the thesis, any one familiar with the genuine Borron romances must be aware that the Queste they presuppose is a Perceval Queste. Probably the disinclination, to which I have referred above, to connect Lancelot with any Grail hero save his own son had very much to do with the matter; further, I do not think that Dr. Wechssler had formed a clear idea of the process of evolution of the cycle he postulated, which he represents as progressing by contraction, i.e. the earliest form being the fullest, or why that cycle should have been connected with the name of Robert de Borron. In fact, he reserves the discussion of the questions concerning original formation for another study.
Now I would submit that the rational progress of evolution is by expansion, not by contraction, and that the name of Robert de Borron became associated with a cycle representing the ensemble of Arthurian romance because there was a smaller cycle which was really the work of the genuine Robert de Borron, which smaller cycle formed the germ of the later and more extended body of romance.[141]
Scholars have long ago recognised that the three works attributed to Robert de Borron, and which, as we possess them, probably represent prose versions of that writer's original poems, are closely connected with each other, and have every appearance of having been intended to form one consecutive work. These three are the Joseph of Arimathea, Merlin, and Perceval, which latter is only represented by one MS. and is what we generally call the 'Didot' Perceval.[142]
Now if we examine the Didot Perceval, as printed by Mr. Hucher in vol. i. of Le Saint Graal, we shall find that the last twenty pages, succeeding Perceval's achievement of the Grail Quest, are devoted to Arthur's expedition to France, his conquest of Frollo and war with Rome, succeeded by Mordred's treachery, the final battle and Arthur's departure for Avalon—in fact, precisely the contents of the Mort Artur, which, as we know, generally follows the Queste, only related in a more concise and summary manner;[143] and one more in accordance with the Chronicles than is the case with the other prose romances.
I think it is quite clear that the Perceval, whether in the original form in which Borron wrote it or not, as we possess it, shows distinct traces of having formed the concluding portion of a cycle.
It is quite obvious that a genuine Borron cycle, such as suggested above, would contain the germ of later expansion. Thus the Joseph of Arimathea certainly appears to represent what we may perhaps call the first draft of the Grand S. Graal. Merlin was certainly expanded into the Merlin Vulgate and Suite. Perceval represents Queste and Mort Artur. Only the Lancelot is unrepresented, and with that I do not think the original 'Borron' cycle had anything to do.
The introduction of the Lancelot probably belongs, as Dr. Wechssler suggests, to a subsequent writer, who borrowed the more famous name, to the pseudo-Borron; and from the quotation given by M. Paulin Paris, I should think it likely that, at first, the juxtaposition of the Lancelot and Perceval-Grail stories was purely external, and that they did not affect each other by contamination. The Didot Perceval may well have been the Queste of the earliest pseudo-Borron, whether or not it represents the Queste of the genuine Borron cycle.[144]
But the growing popularity of the Lancelot story would render such a contamination inevitable, and I am strongly tempted to believe that in that perplexing romance, the prose Perceval li Gallois, we have the Queste of a later pseudo-Borron cyclic redaction. The perplexing features of this version are well known: the whole tone is highly ecclesiastic, there are numerous references to an earlier Perceval story, Lancelot plays an important rôle, yet Galahad is unknown, and there are certain mysterious folk-lore features not met with elsewhere. Hitherto no one has succeeded in satisfactorily placing this romance. I would suggest that it represents the Queste of a late pseudo-Borron Lancelot-Perceval-Grail cycle; and I am encouraged in this supposition by the fact that this romance knows the Questing-Beast, a mysterious creation only found in the Suite Merlin and the Tristan Palamedes romances. Now the Suite Merlin claims to be by Robert de Borron, and the introduction of the Tristan figures into the Arthurian story is, as we saw above, held by Dr. Wechssler to be the distinctive 'note' of the Borron-cycle.[145]
This conclusion is further strengthened when we examine the rôle assigned to Lancelot in these two romances. In each case he is one of the most distinguished knights at Arthur's court, but he is much less en évidence in the Didot Perceval than in the Perceval li Gallois. In the first-named romance he is represented as overthrowing all the knights of the Round Table, till the appearance of Perceval, by whom he is himself overthrown. He would thus appear to rank next to the hero of the tale and to be the superior of Gawain. So far as we can gather, the order of superiority runs thus: Perceval, Lancelot, Gawain, Yvain. But he is, apparently, not of those who start on the Grail quest; nor is there any indication of his liaison with Guinevere. But the author mentions among the knights 'le fiz à la fille à la femme de Malehot.'[146] We do not know the lady of Malehault save through the medium of the prose Lancelot.
In the Perceval li Gallois (Perlesvaus Professor Heinzel prefers to call it), Lancelot is one of the three best knights in the world, the other two being Perceval and Gawain; he takes part in the Grail quest, but on account of his sinful relations with Guinevere is not worthy to behold the sacred talisman, which does not appear, even in a veiled form, during his stay at the Fisher King's castle, whereas it appears clearly to Gawain. The position, so far as Lancelot is concerned, is thus nearer to the presentment of the Galahad Queste than is the Didot Perceval. This last-named, we have seen above, shows clear indications of betraying a cyclic redaction; these indications, though differing in form, are not less clear in the Perceval li Gallois. The concluding passage runs thus: 'Après iceste estoire commence li contes si comme Brians des Illes guerpi le roi Artus por Lancelot que il n'aimoit mie et comme il aséura le roi Claudas qui le roi Ban de Bénoic toli sa terre. Si parole cis contes comment il le conquist et par quel manière, et si com Galobrus de la Vermeille lande vint à la cort le roi Artus por aidier Lancelot, quas il estoit de son lignage cist contes est mout lons et mout aventreus et poisanz.'[147]
In quoting this passage, Professor Heinzel remarks: 'Auch der Perlesvaus ist einem grösseren Romanwerk einverleibt, aus dem die Handschrift von Mons den Perlesvaus ausgeschrieben hat. Was ihm folgte muss eine Art Lancelot gewesen sein.'[148]
There is a further and interesting possibility before us. The compilers may—in one instance, I think, we can show reason to believe that they did—have incorporated the Chrétien Perceval (or a version closely akin to it) into their cycles as representing the Queste.
In the work of preparing these studies I felt that I ought to leave no available version of the Lancelot unexplored, and therefore undertook to read carefully the immense compilation generally known as the Dutch Lancelot. Well was it for me that I did not shrink from the task! I had not read far before I began to suspect that the text represented by this translation was, in every respect, a fuller and a better text than that used by Dr. Sommer in his Malory collation; in the Queste section in particular was this the case. In the succeeding chapters I intend to go fully into what is, I believe, in the interests of Arthurian criticism, a very important discovery. Here I will only say that I eventually found that the text of the Dutch Lancelot, of the printed version of the prose Lancelot Lenoire, 1533 (which, as I have remarked before, Dr. Sommer does not chronicle), and Malory's Lancelot and Queste sections all stand together as representing a much fuller and more accurate text than that of the prose Lancelot 1513, or the Queste MSS. consulted by Dr. Furnivall for his edition of that romance. Whether we have not here an important part of the unshortened pseudo-Borron-Lancelot into which the Map Queste has been introduced is a matter for careful investigation. The point to which at the present moment I would draw attention is, that the Dutch Lancelot incorporates a very considerable section of a Perceval romance, which bears a very close resemblance to Chrétien's poem, with this curious difference, that it gives an account of the achieving of the adventures named by the Grail messenger, which, so far as I know, is found nowhere else. This section, which occupies over two thousand lines, demands a special study, but for us its significance lies in this that it seems to point to the conclusion that in the evolution of a Lancelot-Perceval cycle (the existence of which I think we may hold for proven) the compilers allowed themselves considerable latitude in the Queste section. There were several Perceval Questes to select from, and they took which they preferred, even pressing the original, manifestly independent, Perceval romances into their service. I suspect that this variation in the Perceval Queste helped towards its suppression in favour of the Galahad variant, which had the advantage of existing only in one form, though the cause mainly operating was an entirely different one.[149]
So far then we have traced the evolution of the Lancelot story, and found that at one period of its development, and that an advanced period, it was connected with a Grail story, which regarded Perceval as its hero and knew nothing of Lancelot's son, Galahad. How then did the latter appear upon the scene, and in what light are we to regard the romances dealing with him?
I have studied the Galahad Queste closely, and have compared versions gathered from widely different sources, French originals, and translations, and I am distinctly of the opinion that we possess the romance practically in its original form. It is a homogeneous composition, it is not a compilation from different sources and by different hands. There is no trace of an earlier and later redaction, save only in the directly edifying passages, which in some cases appear to have undergone amplification. The difference between the versions is not that of incident or sequence, scarcely even of detail, but rather of the superior clearness and coherence with which the incidents are related in some of the versions as compared with others. I am strongly inclined to think that there is no peculiarity in any of the Queste MSS. which cannot quite well be ascribed to the greater or less accuracy of the copyist, or his greater or less taste for discourses of edification.
Nor is the Queste by the same hand as was responsible for the final moulding of the Lancelot story; though so closely connected with, indeed dependent upon, that story, it yet in many points stands in flagrant contradiction with it, and there is little doubt that the Lancelot would gain greatly in coherence if the Queste were omitted, and the passages preparatory to it eliminated from the original romance. These remarks apply also to the Grand S. Graal in its present form, though, as we shall see, this last named romance does not stand on precisely the same footing as the Queste with which it is now closely connected.
The following facts seem to stand out clearly. Both these Grail romances, the Queste especially, depend entirely for their interest on Lancelot. They are the glorification of his race as that from which the Grail Winner is predestined to spring. The genealogies, however they may vary (as they do in the different versions), are all devoted to this object. They are most closely connected with, and practically presuppose each other; yet admitting, as I think we must admit, that they do not represent the original form of the Grail story, they do not produce the impression of romances which have been worked over with the view of substituting a new hero for the one in whose honour the tale was originally constructed.
Nevertheless in the case of the Grand S. Graal we must, I think, admit imitation; even as in the original Borron cycle the Joseph of Arimathea was designed as an introduction to the life and deeds of the Grail Winner, Perceval, so in this, the latest form of the cycle, the introduction to the Queste is built upon and expanded from the Joseph. The introduction is based upon and follows the lines of the old introduction, but the Queste is a new Queste.
Let us be quite clear on this point. Galahad may have in a measure supplanted Perceval, but he has neither dispossessed nor robbed him. He has taken over no one of his characteristics, no one of his feats. Such traces of the Perceval story as remain are found in connection with Perceval himself; he, too, achieves the Grail Queste. He has undergone a change, and a change for the worse, but that was quite as much due to the evolution of the Grail as a Christian talisman as to the invention of Galahad. The hero of the Didot Perceval and Perceval li Gallois is as inferior to the hero of Chrétien and Wolfram as is the Perceval of the Galahad Queste. The truth is that Perceval is still the Grail hero, but he shares that character with another whose invention is due to special and easily discernible causes.[150]
The point of view of the writer of the Queste is not that of the compilers of the Lancelot. As I remarked in the previous chapter, the view taken by the Lancelot of the relations between the hero and the queen is frankly unmoral. Neither is blamed for his or her action, neither is apparently conscious of wrong-doing. In the Queste Lancelot's conscience is sorely vexed, and his sin insisted upon. The compilers of the Lancelot have a very courtly respect for women—the author of the Queste despises them utterly. The interest of the Lancelot lies in the relation between the sexes—the respective duties of knight and lady—the theme which inspires the Queste is their abiding separation.
Again, compare the treatment of the various characters of the story in the two respective sections. Next to Galahad and Perceval, the hero of the Queste is Bohort (Bors). But for a single youthful lapse he yields in nothing to those doughty champions of celibacy: his purity, alike of body and soul, is emphatically insisted upon; his confession fills the priest who receives it with a fervour of admiration; yet it is precisely this saintly youth who, in the section preceding and following the Queste (the Lancelot and the Mort Artur), is the confidant and go-between of Lancelot and Guinevere. It is Bohort who seeks Lancelot at the secret bidding of the queen, Bohort who carries love-tokens between them, who arranges meetings. It is he and Lionel who consult the queen as to the delicate question of Lancelot's future relations with the lady who has cured him from the illness caused by drinking the poisoned spring; he who is the confidant of Guinevere's indignation at the supposed love-affair between Lancelot and the maiden of Escarloet; and if he tries to prevent the last fatal meeting between them it is with no view of hindering a wrong to his lord Arthur, but solely because he has reason to suspect the trap laid for the lovers. The two presentments not simply fail to agree, but stand in flat contradiction with each other.
Lionel, again, is throughout the Lancelot a valiant knight, warmly attached alike to his brother and to his cousin. Like Bohort he takes Lancelot's part on every occasion, with him he quits the court when the queen, in an access of jealousy, banishes Lancelot. When he is finally slain both Bohort and Lancelot are overcome with grief. But the Queste paints him in the most repulsive colours: violent, brutal, and unreasoning to a degree. He is so indignant with his brother for going to the rescue of a maiden rather than of himself (when both are equally in danger) that he does his best to kill him in revenge. He does kill an unoffending hermit, and a fellow knight of the Round Table who would intervene, and finally it needs a special interposition of Providence to part the two brothers before a fatal issue to the conflict forced on by Lionel has taken place.
Hector, Lancelot's half-brother, who in the later Lancelot story is one of the bravest and most distinguished knights of the court, is in the Queste held up to scorn and rebuke; while the author of this romance has no colours too black in which to paint the character of Gawain, who, though deposed from his position of chief hero, is, throughout the Lancelot proper, treated with the greatest respect. He is entirely loved and trusted by king and queen, and if his valour is in the long-run surpassed by that of Lancelot, the compiler is careful to preserve his honour intact by pointing out, first, that he never recovered from the severe wounds received in the war with Galehault, second, that he was over twenty years Lancelot's senior. The final conflict between them, the most deadly in which Lancelot was ever engaged, was fought when Gawain was seventy-two and Arthur ninety-two years of age; further, as we shall see presently, in some versions the conclusion is more of the character of a drawn battle than of a defeat for Gawain.[151]
It is, I think, quite clear that the Galahad-Grail romances are the work of another hand than that responsible for the main body of the Lancelot cycle; and the work of one who was at small pains to harmonise his story with the branches already existing. It is indeed doubtful whether the writer had any thorough acquaintance with the legend as a whole. It is noteworthy that the points of contact with what we may perhaps call the 'secular' section are all restricted to the later part of the story, that commencing with what M. Paulin Paris called the Agravain section. Between the Grand S. Graal, the Galahad Queste, and the later part of the Lancelot there are a number of what we may call cross-references, the precise value of which will be very difficult to determine. But they do not stray outside a certain limit—they are restricted to Lancelot, the Knight of the Round Table, the queen's lover, and father of the Grail Winner—they do not appear to know Lancelot the protégé of the Lady of the Lake. In this character the Grail romances ignore him, nor do they appear to know anything of his most famous adventure, the freeing of Guinevere from Meleagant.[152]
It is the later and not the earlier Lancelot story which is known to the writer of the Queste; and the more we study the romance the plainer this becomes. The Lancelot romance may really be divided into two great divisions, the Enfances, Charrette, and Galehault section, which is practically unaffected by the Grail tradition, though it shows evident signs of contact with the Perceval story; and the latter portion which (saving the Mort Artur, unaffected except by the addition of the concluding Queste paragraph, easily removed) has been redacted under the influence of the Galahad-Grail accretion.
Till the versions concerned have been critically examined we cannot determine the value or gauge the evidence of the matter common to the Lancelot, Grand S. Graal, and Queste. The most noticeable instances are the following: the keeping of the Grail at Castle Corbenic, the founding of which is related in the Grand S. Graal; the characters of King Pelles and his father, with regard to whom the evidence varies,—as a rule, the character of the Fisher King appears to be confined to the former, that of the Maimed King to the latter (the author of the Queste appears to have no idea that the two characters are one and the same);—the daughter of King Pelles, and his son Eliezer. This latter is, I think, peculiar to the Lancelot-Galahad story, the Perceval versions do not know him. The adventure of the broken sword borne by Eliezer, told both in Lancelot and Grand S. Graal, and achieved, though without satisfactory explanation, in Queste.[153] The Boiling Fountain and Bleeding Tomb adventures, also told in the two first, partly achieved in the Lancelot, and achievement summarily announced in Queste. The Perilous Cemetery, origin stated in Grand S. Graal, vainly attempted by Gawain and Hector in Lancelot, final achievement barely recorded in Queste.
In these last instances the story may well have been in the Lancelot, and taken over by the compiler of Grand S. Graal; the Queste makes very little of them; they only serve to keep up the connection between the 'secular' and 'religious' sections.
With regard to the Corbenic-Grail adventures, I am inclined, as I said before, to look upon them as due to the influence of the Gawain story, and as already existing, in a purely adventurous form, in the Lancelot, before it was formally united to the Grail Quest.
On the whole, I decidedly lean to the opinion that Grand S. Graal and Queste are by one and the same hand—the one based upon and expanded from an older poem, the other a practically new invention, the two being designed to replace the Joseph of Arimathea and Perceval of the earlier Grail cycle. As I said above, the author was very little concerned about the harmony of his work. So long as by a superficial rearrangement and interpolation of incidental adventures he could produce an appearance of harmony, he cared nothing at all about the more important questions of continuity of treatment, and preservation of tone and character. The result is that his work, which stands practically as he left it, is in flagrant contradiction with the story it is designed to complete.
But what was the motive which led to the setting aside of the earlier Perceval Queste, and what the causes which determined the particular form assumed by its successor?
I do not think they are difficult to detect.
During the later years of the twelfth and earlier years of the thirteenth century we see two stories in process of gradual evolution—the Perceval-Grail story and the Lancelot legend. One early took a decidedly mystical and ecclesiastical bent, the other became more and more worldly and secular. The two appear to have had an equal hold on popular imagination, they early came into touch with each other, but they never really blended. The Lancelot, as the younger, borrowed at the outset certain features from the Perceval, but it retained its own distinctive character; while the elder story slowly changed, the Grail, at first a subordinate element in the story, gradually but surely dominating the tale, which became more and more ecclesiastical, while the hero became more and more conventional.[154]
But at a certain point it became evident that these lines of tradition could no longer remain parallel, they must coalesce, or the one must yield to the other. The Grail quest had become the most popular adventure of Arthur's court, one after another the knights were being drawn into the mystic circle; how could the most popular and most valiant of the knights of the Round Table, for this Lancelot had now become, remain outside the chosen group? It was plain that Lancelot must take part in the Grail quest; it was equally plain that the first knight of the court could not be allowed to come out of the ordeal with any detriment to his prestige; yet the Grail demanded purity of life, and Lancelot was the queen's lover. More, the queen's lover he must remain or forfeit his hold on popular sympathy.
How was it possible to preserve intact at once Lancelot's superiority and the purity of the Christian talisman? Only in one way: by giving him a son who should achieve the quest and then vanish, leaving Lancelot still facile princeps among the knights of the Round Table, with the added glory of having been the father of the Grail Winner.
But this son could not be the child of Guinevere. The offspring of a guilty liaison could not be the winner of the sacrosanct talisman; yet Lancelot must be faithful to his queen—how solve this problem? The story in its primitive form gave the hint for the required development. Who more fitted to become the mother of the Grail Winner than the fair maiden who filled the office of Grail-bearer?[155] The obvious propriety of such a relationship was bound sooner or later to strike the imagination of some redactor. The Arthurian story already possessed the machinery by which Lancelot could become father of the elect child, while remaining Guinevere's lover; Brisane had but to do for Elaine what Merlin did for Uther, and the difficulty was overcome. Moreover, Helaine was, in the old story, the name of the Grail Winner's father, nothing more easy than to bestow the same name on the new hero's mother. All this was only a question of clever adjustment of already existing factors.
Perceval, of course, was in possession, but the later development of his story, which had converted him from a genuine, faulty, but loving and lovable human being, true man and faithful husband, into an aggressively proselytising and persecuting celibate, had made it possible for him still to retain a place in the romance; he could act as second to Galahad, and, like him, disappear, the quest once achieved. But having thus disposed in Lancelot's interest of the two who might have seriously challenged his fame as a knight, Perceval, the real, Galahad, the vicarious (for I think we can only regard him as his father's representative), achiever of the quest, it became necessary to add a third, who should bring back to court the tidings of their success. It is quite obvious, from the point of view of the Lancelot story, that Perceval and Galahad could not be permitted to return. The third was easily found in the person of Lancelot's nearest relative, the knight who, his shield unstained by the bar-sinister which marked that of Hector, had been gradually rising in popular favour; Bohort owes his position in the Queste to his position in the Lancelot proper.
The evolution of this character has not, I believe, attracted much attention hitherto, but it is one of the most remarkable features of the Lancelot story. In the earliest versions, represented by the Lanzelet, etc., he is not known at all.[156] When he first appears he plays but a small part, gradually his rôle becomes more and more prominent, till in the later portion of the prose Lancelot he has become a very efficient understudy to the hero, even surpassing in valour Gawain himself. Thus, on the return of the knights from one of their numerous quests in search of Lancelot, when they are called upon to rehearse their adventures, in order that a record of them may be made, it is decided that their rank, in order of merit, is Bohort, Gawain, Hector, Gaheriet, Lionel, and Baudemagus. Gawain and his brother, the representatives of the older stratum of Arthurian tradition, are the only two who can compete with the all-conquering race of Ban, and the bosom friend of that race, Baudemagus.
Finally he is represented as the father of a son who bids fair to rival his ancestors in valour. When a critical study of the Lancelot mss. is seriously undertaken, I think we shall find that the position occupied by Bohort in the story will afford a valuable indication of the relative age of the redaction.
I am quite prepared to find that among the objections which will doubtless be advanced against the theory here advocated one will be that it is too complete in detail, too 'cut and dried,' if I may use the term, to be free from suspicion. To this I would answer that I believe in examining the later stages of Arthurian romance we must follow a somewhat different process from that which we employ for the earlier. The Arthurian poems, being in a large measure independent, and never having formed part of a 'cyclic' whole, may well be studied separately, in, and for, themselves, though of course we would not leave out of sight variants of the same story. But the later prose romances, those which have avowedly formed parts of a cycle, must be studied, not separately, but in conjunction with the other romances with which they were connected. They are in the position of the parts of a dissected puzzle, the study of one part by itself will never really help us to understand the whole, it is only by studying collective sections, and trying continually new combinations, that we can hope to find the original disposition of the parts.
It is no use to study the Queste romance by itself. If we wish to know how it stands with regard to the Lancelot, we must study it with the Lancelot, and if we do this certain points become absolutely clear. The Queste pre-supposes a very advanced stage of the Lancelot story; one at which the family of the hero, quite as much as the hero himself, is the subject of glorification.[157]
The Galahad Queste is absolutely unthinkable without a previous knowledge of the Lancelot romances; as a matter of fact, it stands in closer relation to these than it does to any earlier Grail quest. The Lancelot romances, on the contrary, would be quite complete and far more coherent without the Queste. I have commented already on the striking discrepancies between the sections, but I have not so far dwelt at any length on the extraordinary lack of Grail references in the Mort Artur, the section immediately following the Queste. If we set on one side the introductory passage, which I have no shadow of doubt does not belong to the Mort Artur at all, but is the concluding passage of the Queste, there is no evidence of the influence of the latter throughout the whole of this last section of the cycle. Galahad is never mentioned; he was—and is not—as completely as if he had never been. Lancelot never thinks of, never refers to, his valiant son; his whole thought and care is for the queen, whom we were previously told he had renounced. I do not think it possible for any one to read the Mort Artur and believe that the Queste forms an integral part of the Lancelot story. On the other hand, cut out the Queste, suppress the few passages in the immediately preceding section of the Lancelot story which relate to it, and you have a tale as complete and coherent as is possible for any legend which has been the fruit of long growth and evolution, and has not possessed from the outset a clear and definite purpose and outline.
Admit, as I think we must needs admit, that the Lancelot and the Grail stories form two independent streams of tradition; recognise, as we must recognise, their diverse character,—one strongly secular, the other strongly ecclesiastical,—and I think we must own that if in their completed form they were to coalesce, that coalition could only be carried out under the conditions suggested above, which conditions we find fulfilled in the Galahad Queste. For me this romance is the last word of the Lancelot evolution, the final blending of two separate and important streams of tradition, the grant conte of Lancelot and the grant conte of Perceval and the Grail, the which is chiés et fin de tous les contes.