Zarncke’s acute article was a praiseworthy attempt to construct a working hypothesis of the growth of the cycle. But it is full of grave misconceptions, as was, perhaps, inevitable in a hasty survey of such an immense body of literature. The versions are “placed” most incorrectly. The argumentation is frequently marred by a priori reasoning, such as that Chrestien, the acknowledged leading poet of the day, could not have copied Kiot, and by untenable assertions, such as that Bran, in the Mabinogi of Branwen, the daughter of Llyr, is perhaps a distant echo of Hebron in Robert de Borron’s poem. He had, however, the great merit of clearing the ground for his pupil, A. Birch-Hirschfeld, and urging him to undertake what still remains the most searching and exhaustive survey of the whole cycle: “Die Sage vom Gral,” etc. As Birch-Hirschfeld’s analysis is at present the only basis for sound criticism, I shall give his views fully:—The Grand St. Graal, as the fullest of the versions dealing with the Early History of the Grail, is the best starting-point for investigation. From its pronounced religious tone monkish authorship may be inferred. Its treatment of the subject is not original as is shown by (1) the repetition ad nauseam of the same motive (e.g., that of the lance wound four times), (2) the pedigrees, (3) the allusions to adventures not dealt with in the book, and in especial to the Promised Knight. The testimony of Helinand (see supra, [p. 52]), which is of first-rate importance, does not allow of a later date for the Grand St. Graal than 1204. On turning to the Queste it is remarkable that though sometimes found in the MSS. in conjunction with the Grand St. Graal it is also found with the Lancelot, and, when the hero’s parentage is considered, it seems more likely that it was written to supplement the latter than the former work. This supposition is adverse to any claim it may lay to being held the earliest treatment of the subject, as it is highly improbable that the Grail legend occupied at the outset such an important place in the Arthurian romance as is thus accorded to it. Such a claim is further negatived by the fact that the Queste has three heroes, the second of whom is obviously the original one of an older version. In estimating the relationship between the Grand St. Graal and the Queste it should be borne in mind that the latter, in so far as it deals with the Early History, mentions only Joseph, Josephe, Evelach (Mordrain) and Seraphe (Nascien), from whom descends Galahad; that it brings Joseph to England, and that it does not give any explanation of the nature of the Grail itself. It omits Brons, Alain, the explanation of the name “rich fisherman,” the name of Moys, although his story is found in substantially the same shape as in the Grand St. Graal, and is silent as to the origin of the bleeding lance. If it were younger than and derived from the Grand St. Graal alone, these points, all more important for the Early History than the Mordrain episodes would surely have been dwelt upon. But then if the Grand St. Graal is the younger work, whence does it derive Brons, Alain, and Petrus, all of whom are introduced in such a casual way? There was obviously a previous Early History which knew nothing of Josephe or of Mordrain and his group, the invention of the author of the Queste, whence they passed into the Grand St. Graal, and were fused in with the older form of the legend. There is, moreover, a positive reference on the part of the Grand St. Graal to the Queste (vol. ii., p. 225). The author of the Queste introduced his new personages for the following reasons: He had already substituted Galahad for the original hero, and to enhance his importance gives him a fictitious descent from a companion of Joseph. From his model he learnt of Joseph’s wanderings in the East, hence the Eastern origin of the Mordrain group. In the older form the Grail had passed into the keeping of Joseph’s nephew, in the Queste the Promised Knight descends from the nephew of Mordrain; Brons, as the ancestor of the original Quest hero necessarily disappears in the Queste, and his place is in large measure taken by Josephe. The priority of the Queste over the Grand St. Graal, and the use of the former by the latter may thus be looked upon as certain. But if Mordrain is the invention of the Queste, what is the meaning of his illness, of his waiting for the Promised Knight, of the bleeding lance, and of the lame king whom it heals? These seem to have no real connection with the Grail, and are apparently derived from an older work, namely, Chrestien’s Conte du Graal.
Chrestien’s work, which ended at v. 10,601, may be dated as having been begun not later than 1189 (vide supra, [p. 4]). Its unfinished state accounts for its having so little positive information about the Grail, as Chrestien evidently meant to reserve this information for the end of the story. But this very freedom with which the subject is handled is a proof that he had before him a work whence he could extract and adapt as he saw fit; moreover we have (Prologue, v. 475, etc.) his own words to that effect. With Chrestien’s account of the Grail—a bowl bejewelled, of wondrous properties, borne by a maiden, preceded by a bleeding lance, accompanied by a silver plate, guarded by a king wounded through both ankles (whose only solace is fishing, whence his surname), ministering to the king’s father, sought for by Perceval, nephew to the fisher king, its fate bound up with a question which the seeker must put concerning it—may be compared that of the Queste, in which nothing is known of a question by which the Grail kingship may be obtained (although it relates the same incident of Lancelot), which knows not of one wounded king, centre of the action, but of two, both of secondary importance (though possibly Chrestien’s Fisher King’s father may have given the hint for Mordrain), in which the lance is of minor importance instead of being on the same level as the Grail. Is it not evident that the Queste took over these features from Chrestien, compelled thereto by the celebrity of the latter’s presentment? The Queste thus presupposes the following works: a Lancelot, an Early History, a Quest other than that of Chrestien’s, and finally Chrestien as the lame king and lance features show. It thus falls between 1189 (Chrestien begun) and 1204 (Grand St. Graal ended).
With respect to the three continuators of Chrestien it would seem that Gautier de Doulens’ account of the Grail, as found in the Montpellier MS., knowing as it does only of Joseph, and making the Fisher King and Perceval descendants of his, belongs to an older stage of development than that of Manessier and Gerbert, both of whom are familiar with the Mordrain group, and follows that of the original version upon which both the Queste and the Grand St. Graal are based. There is nothing to show that Gautier knew of the Queste, whilst from Gautier the Queste may have possibly have taken Perceval’s sister and the broken sword. Gautier would thus seem to have written immediately after Chrestien, and before the Queste, i.e., about 1195. As for the date of the other two continuators, the fact of their having used the Queste is only one proof of the lateness of their composition (as to the date of which see supra, [p. 4]). It must be noted that whilst in their account of the Grail Chrestien’s continuators are in substantial accord with the Queste versions, and yet do not contradict Chrestien himself, they add considerably to his account of the lance. This is readily explained by the fact that as Chrestien gave no information respecting the origin of either of the relics, they, the continuators, had to seek such information elsewhere; they found all they could wish respecting the Grail, but nothing as to the lance, the latter having been first introduced by Chrestien, and the Queste versions knowing nothing respecting it beyond what he told. Thus, thrown upon their own resources, they hit upon the device of identifying the lance with the spear with which Jesus was pierced as He hung on the Cross. This idea, a most natural one, may possibly have been in Chrestien’s intent, and may have been suggested to him by the story of the discovery of the Holy Lance in Antioch half a century before. It must, however, be admitted that the connection of the lance with the Grail legend in its earliest form is very doubtful, and that Celtic legends may possibly have furnished it to Chrestien, and indicated the use to which he intended putting it. The analysis, so far, of the romances has resulted in the presupposition of an earlier form; this earlier form, the source or basis of all the later versions of the legend, exists in the so-called Petit St. Graal of Robert de Borron. Of this work, found in two forms, a prose and a poetic one, the poetic form, pace Hucher, is obviously the older, Hucher’s proofs of lateness going merely to show that the sole existing MS. is a recent one, and has admitted new speech-forms;[62] moreover the prose versions derive evidently from one original. The greater simplicity of the poem as compared with the Grand St. Graal proves its anteriority in that case; Paulin Paris’ hypothesis that the poem in its present state is a second draft, composed after the author had made acquaintance with the Grand St. Graal, is untenable, the poem’s reference (v. 929 etc.) to the “grant livre” and to the “grant estoire dou Graal,” written by “nul home qui fust mortal” (v. 3,495-6) not being to the Grand St. Graal, but having, on the contrary, probably suggested to the writer of the latter his fiction of Christ’s being the real author of his work. The Grand St. Graal used the poem conjointly with the Queste, piecing out the one version by help of the other, and thereby entirely missing the sequence of ideas in the poem, which is as follows: Sin, the cause of want among the people; the separation of the pure from the impure by means of the fish (symbol of Christ) caught by Brons, which fish does not feed the people, but, in conjunction with the Grail, severs the true from the false disciples; punishment of the self-willed false disciple; reward of Brons by charge of the Grail. In the Grand St. Graal, on the contrary, the fish is no symbol, but actual food, a variation which must be laid to the account of the Queste. In a similar way the two Alains in the Grand St. Graal may be accounted for, the one as derived from the poem, the second from the Queste. As far as conception is concerned, the later work is no advance upon the earlier one. To return to Borron’s work, which consists of three sections; there is no reason to doubt his authorship of the second, Merlin, or of the third, Perceval, although one MS. only of the former mentions the fact, and it is, moreover, frequently found in connection with other romances, in especial with the Lancelot; as for Perceval, the silence of the unique MS. as to Borron is no argument, as it is equally silent in the Joseph of Arimathea section. All outward circumstances go to show that Borron divided his work into three parts, Joseph, Merlin, Perceval. But, if so, the last part must correspond in a fair measure to the first one; recollect, however, that we are dealing with a poet of but little invention or power of giving unity to discordant themes, and must not expect to find a clearly traced plan carried out in every detail. Thus the author’s promise in Joseph to speak later of Moses and Petrus seems not to be fulfilled, but this is due to Borron’s timidity in the invention of new details. What is said of Moses does not disagree with the Joseph, whereas a later writer would probably follow the Grand St. Graal account; as for Petrus he is to be recognised in the hermit Perceval’s uncle. There may be some inconsistency here, but Borron can be inconsistent, as is shown by his treatment of Alain, who at first vows to remain virgin, and afterwards marries. But a graver argument remains to be met; the lance occurs in Perceval—now ex hypothesi the first introduction of the lance is due to Chrestien. The lance, however, only occurs in two passages, both obviously interpolated. The identity of authorship is evident when the style and phraseology of the two works are compared; in both the Grail is always li graaux or else li veissel, not as with the later versions, li saint graaux; both speak of la grace dou graal; in both the Grail is bailli to its keeper, who has it en guarde; the empty seat is li liu vit, not the siège perilleux. The central conception, too, is the same—the Trinity of Grail-keepers symbolising the Divine Trinity. The secret words given by Christ with the Grail to Joseph in prison, by him handed on to Brons, are confided at the end of the Perceval by Brons to the hero—and there is no trace of the Galahad form of the Quest, as would inevitably have been the case had the Perceval been posterior in date to the Queste. As the Perceval is connected with the Joseph, so it is equally with the Merlin; it is remarkable that neither Merlin nor Blaise play a prominent part in the Queste versions, but in Borron’s poem Merlin is the necessary binding link between the Apostolic and Arthurian ages. Again the whole character of the Perceval speaks for its being one of the earliest works of the cycle; either it must have used Chrestien and Gautier or they it; if the former, is it credible that just those adventures which were necessary to supply the ending to the Joseph could have been picked out? But it is easy to follow the way in which Chrestien used the Perceval; having the three-part poem before him he took the third only for his canvas, left out all that in it related to the first two parts, all, moreover, that related to the origin and early history of the Grail; the story of the childhood is half indicated in the Perceval, and Chrestien may have had Breton lays with which to help himself out; all relating to the empty seat is left out as reaching back into the Early History; the visit to Gurnemanz is introduced to supply a motive for the hero’s conduct at the Grail Castle; the wound of the Fisher King is again only an attempt of Chrestien’s to supply a more telling motive; as for the sword Chrestien invented it; as he also did the Grail-messenger, whose portrait he copied from that of Rosette la Blonde. The order of the last episodes is altered by Chrestien sensibly for the better, as, with him, Perceval’s doubt comes first, then the Good Friday reproof, then the confession to and absolution by the hermit; whereas in the Perceval the hero after doubt, reproof, and absolution rides off again a-tourneying, and requires a second reproof at Merlin’s hands. It is easy to see here which is the original, which the copy. Chrestien thus took with clear insight just what he wanted in the Perceval to fit out his two heroes with adventures.[63] As for Borron’s guiding conception, his resolve to have nothing to do with the Early History made him neglect it entirely; he only cared to produce a knightly poem, and we find, in consequence, that he has materialised all the spiritual elements of his model. Gautier de Doulens’ method of proceeding was much simpler: he took over all those adventures that Chrestien purposely left out, and they may be found brought together (verses 22,390-27,390) with but few episodes (Perceval’s visit to Blanchefleur, etc.) entirely foreign to the model amongst them.[64] The Perceval cannot be later than Gautier, as otherwise it could not stand in such close relationship to the Joseph and Merlin; it must, therefore, be the source of the Conte du Graal, and a necessary part of Borron’s poem, which in its entirety is the first attempt to bring the Joseph of Arimathea legend into connection with the Arthur sage. The question as to the origin of the Grail would thus seem answered, the Christian legendary character of Borron’s conception being evident; but there still remains the possibility that that conception is but the Christianised form of an older folk-myth. Such a one has been sought for in Celtic tradition. The part played by Merlin in the trilogy might seem to lend colour to such an hypothesis, but his connection with the legend is a purely artificial one. Nor is the theory of a Celtic origin strengthened by reference to the Mabinogi of Peredur. This knows nought of Merlin, and is nearer to Chrestien than to the Didot-Perceval, and may, indeed, be looked upon as simply a clumsy retelling of the Conte du Graal with numerous additions. A knowledge of the Didot-Perceval on Chrestien’s part must be presupposed, as where could he have got the Fisher King and Grail Castle save from a poem which dealt with the Early History of the Grail, a thing the Mabinogi does not do. But, it may be said, Chrestien used the Mabinogi conjointly with Borron’s poem. That the Welsh tale is, on the contrary, only a copy is apparent from the following considerations:—It mixes up Gurnemanz and the Fisher King; it puts in the mouth of Peredur’s mother an exclamation about the knights, “Angels they are my son,” obviously misread from Perceval’s exclamation to the same effect in Chrestien’s poem; Perceval’s love-trance over the three blood drops in the snow is explained in Chrestien by the hero’s passion for Blanchefleur, but is quite inexplicable in the Mabinogi; again, in the Welsh tale, the lance and basin episode is quite a secondary one, a fact easily explained if it is looked upon as a vague reminiscence of Chrestien’s unfinished work; moreover the Mabinogi lays great stress upon the lance, which has already been shown to belong to a secondary stage in the development of the legend. Again the word Graal occurs frequently in old Welsh literature, and invariably in its French form, never translated by any equivalent Welsh term. As for the name Peredur, it is understandable that the Welsh storyteller should choose the name of a national hero, instead of the foreign name Perceval; the etymology Basin-Seeker is untenable. There is no real analogy between the Grail and the magic cauldron of Celtic fable, which is essentially one of renovation, whereas the Grail in the second stage only acquires miraculous feeding, and in the third stage healing powers. It is of course not impossible that such adventures in the Mabinogi, as cannot be referred directly to Chrestien, may belong to a genuine Peredur sage.
The question then arises—was Robert de Borron a simple copyist, or is the legend in its present form due to him, i.e., did he first join the Joseph of Arimathea and Grail legends, or had he a predecessor? Now the older Joseph legends know nothing of his wandering in company of a miraculous vessel, Zarncke having shown the lateness of the one commonly ascribed to William of Malmesbury. Nor is it likely Borron had before him a local French legend as Paulin Paris (Romania, vol. i.) had supposed; would he in that case have brought the Grail to England, and left Joseph’s fate in uncertainty? The bringing the Grail to England is simply the logical consequence of his conception of the three Grail-keepers (the third of British blood), symbolising the Trinity, and of the relation of the Arthurian group to this central conception; where the third Grail-keeper and the third of the three wondrous tables were, there the Grail must also be. What then led Borron to connect the sacramental vessel with the Joseph legend? In answering this question the later miraculous properties of the Grail must be forgotten, and it must be remembered that with Borron it is only a vessel of “grace;” this is shown in the history of (Moys) the false disciple, which obviously follows in its details the account of the Last Supper, and of the detection of Judas by means of the dish into which Jesus dips a sop, bidding the betrayer take and eat. Borron’s first table being an exact copy of the Last Supper one, his holy vessel has the property of that used by Christ. In so far Borron was led to his conception by the story as told in the canonical books; what help did he get from the Apocrypha? His mention of the Veronica legend and certain details in his presentment of Vespasian’s vengeance on the Jews (e.g., his selling thirty for a penny) show him to have known the Vindicta Salvatoris, in which Joseph of Arimathea appears telling of his former captivity from which Christ Himself had delivered him. Thus Borron knew of Joseph’s living when Vespasian came to Jerusalem. From the Gesta Pilati he had full information respecting the imprisonment of Joseph; he combined the accounts of these two apocryphal works, substituting a simple visit of Christ to Joseph for the deliverance as told in the Gesta Pilati, and making Vespasian the deliverer, whereto he may have been urged by Suetonius’ account of the freeing of Josephus by Vespasian (Vesp. ch. v.). But why should Joseph become the Grail-keeper? Because the fortunes of the vessel used by the Saviour symbolise those of the Saviour’s body; as that was present at the Last Supper, was brought to Pilate, handed over to Joseph, was buried, and after three days arose, so with the Grail. Compare, too, Christ’s words to Joseph (892, etc.) in which the symbolical connection of the laying in the grave and the mass is fully worked out. Thus Joseph who laid Christ’s body in the grave is the natural guardian of the symbol which commemorates that event, thus, too, the Grail is the natural centre point of all the symbolism of mass and sacrament, and thus the Grail found its place in the Joseph legend, ultimately becoming its most important feature. Need Perceval’s question detain us? May it not be explained by the fact that as Joseph had to apply twice for Christ’s body, so his representative, the Grail-seeker, had to apply twice for the symbol of Christ’s body, the Grail? But it is, perhaps, best to consider the question and the Fisher King’s weakness as inventions of Borron’s, possibly derived from Breton sources, the ease with which the hero fulfils a task explained to him beforehand favouring such a view. Borron, it must be noticed, had no great inventive power; in the Joseph he is all right so long as he has the legend to follow; in the Merlin and the Perceval he clings with equal helplessness to the Breton sagas, confining himself to weaving clumsily the adventures of the Grail into the regular Arthur legend.
The question as to the authorship of the Grand St. Graal and the Queste, the latter so confidently attributed to W. Map, may now profitably be investigated. Map, who we know flourished 1143-1210 (see supra, [p. 5]), took part in all the political and social movements of his time. If we believe the testimony of the MSS. which ascribe to him the authorship of the following romances: (1) the Lancelot, in three parts; (2) the Queste; (3) the Mort Artur; (4) the Grand St. Graal, he would seem to have shown a literary activity quite incompatible with his busy life, when it is remembered how slow literary composition was in those days. Nor can it be reconciled with the words of Giraldus Cambrensis,[65] although Paulin Paris (Rom. i. 472) has attempted such a reconciliation by the theory that the words dicere and verba dare referred to composition in the vernacular, and that Map was opposing not his oratorical to Gerald’s literary activity, but his French to Gerald’s Latin works. Against this initial improbability and Gerald’s positive testimony must be set, it is true, the witness of writers of the time and of the MSS. The most important is that of Hélie de Borron in his prologue to Guiron le Courtois.[66] After telling how Luces de Gast was the first to translate from the Latin book into French, and he did part of the story of Tristan, he goes on: “Apriés s’en entremist maistre Gautiers Map qui fu clers au roi Henry et devisa cil l’estoire de monseigneur Lancelot du Lac, que d’autre chose ne parla il mie gramment en son livre. Messiers Robers de Borron s’en entremist après. Je Helis de Borron, par la prière monseigneur de Borron, et pour ce que compaignon d’armes fusmes longement, en commençai mon livre du Bret.” Again in the epilogue to the Bret,[67] “Je croi bien touchier sor les livres que maistres Gautiers Maup fist, qui fit lou propre livre de monsoingnour Lancelot dou Lac; et des autres granz livres que messires Robert de Berron fit, voudrai-je prendre aucune flor de la matière ... en tel meniere que li livres de monsoingnour Luces de Gant et de maistre Gautier Maapp et ciz de monsoingnour Robert de Berron qui est mes amis et mes paranz charnex s’acourderont au miens livres—et je qui sui appelex Helyes de Berron qui fui engendrez dou sanc des gentix paladins des Barres qui de tous tens ont été commendeour et soingnor d’Outres en Roménie qui ores est appelée France.” Now Hélie cannot possibly belong to the reign of Henry II (+ 1189) as asserted by Hucher (p. 59), as he speaks of Map in the past tense (fu clers), and Map outlived Henry, moreover the mention of Romenie proves the passage to have been written after the foundation of the Latin Empire in 1304. Hélie’s testimony is thus not that of an immediate contemporary, and it only shows that shortly after Map’s death the Lancelot was ascribed to him. It is, moreover, in so far tainted, that he speaks with equal assurance respecting the great Latin book which of course never existed; nor can we believe him when he says that he was the comrade of Robert de Borron, as this latter wrote before Chrestien, and must have been at least thirty years older than Hélie, who in the Guiron (written about 1220) calls himself a young man. How is it with the testimony of the MSS.? Those of the Lancelot have unfortunately lost their colophon, owing to the Queste being almost invariably added; those of the Queste show as a rule a colophon such as the one quoted by Paulin Paris from the Bibl. Nat., MS. 6,963 (MSS. Franç II., p. 361): “Maistre Gautiers Map les estrait pour son livre faire dou Saint-Graal, pour l’amor del roy Henri son seignor, qui fist l’estore translater dou latin en françois.” A similar statement occurs in a MS. of the Mort Artur (Bib. Nat. 6,782.). Both are equally credible. Now as the King can only be Henry II (+ 1189) and as the Queste preceded the Mort Artur it must be put about 1185, and Chrestien’s Conte du Graal about 1180, an improbably early date when it is recollected that the Conte du Graal is Chrestien’s last work. The form, too, of these colophons, expressed as they are in the third person, so different from the garrulous first person complacency with which Luces de Gast and Hélie de Borron announce their authorship, excites the suspicion that we have here not the author’s own statement, but that of a copyist following a traditional ascription. Whether or no Map wrote the Lancelot, it may safely be assumed that he did not write the Queste, or a fortiori the Grand St. Graal. The tradition as to his authorship of these romances may have originated in Geoffrey’s mention of the Gualterus archidiaconus Oxenfordensis, to whom he owed his MS. of the Historia Regum Britanniae. A similar instance of traditional ascription on the part of the copyist may be noted in the MSS. of the Grand St. Graal, the author of which is declared to be Robert de Borron. The ordinary formulæ (quoted supra, [p. 5]) should be compared with Borron’s own words in the Joseph (supra, [p. 5]) and the difference in form noted. What proves these passages to be interpolations is that the author of the Grand St. Graal especially declares in his prologue that his name must remain a secret. The colophons in question are simply to be looked upon as taken over from the genuine ascription of Borron’s poem, and there is no positive evidence as to the authorship of either the Queste or the Grand St. Graal; both works are probably French in origin, as is shown by the mention of Meaux in the Grand St. Graal. As for the date of Borron’s poem, a terminus ad quem is fixed by that of the Conte du Graal (1180); and as the poem is dedicated to Gautier of Montbeliard, who can hardly have been born before 1150, and who must have attained a certain age before he could become Robert’s patron, it must fall between the years 1170 and 1190.
The results of the investigation may be summed up as follows: the origin of the Grail romances must be sought for in a Christian legend based partly upon the canonical, partly upon the uncanonical, writings. This Christian legend was woven into the Breton sagas by the author of the oldest Grail romance; the theories of Provençal Spanish, or Celtic origin are equally untenable, nor is there any need to countenance the fable of a Latin original. Chronologically, the versions arrange themselves thus:—
(1) Between 1170 and 1190 (probably about 1183) Robert de Borron wrote his trilogy: Joseph of Arimathea—Merlin—Perceval. Sources: Christian legend (Acta, Pilati, Descensus Christi, Vindicta Salvatoris) and Breton sagas (Brut?). Here the Grail is simply a vessel of grace.
(2) About 1189 Chrestien began his Conte du Graal, the main source of which was the third part of Borron’s poem. Marvellous food properties attributed to the Grail; introduction of the bleeding lance, silver dish, and magic sword.
(3) Between 1190 and 1200 Gautier de Doulens continued Chrestien’s poem. Main sources, third part of (1) and first part of same for Early History—introduction of broken sword.
(4) Between 1190 and 1200 (but after Gautier?) the Queste du St. Graal written as continuation to the Lancelot. Sources (1) and (2) (for lance) and perhaps (3). New personages, Mordrain, Nascien, etc., introduced into Early History.