[In the following essay, Bell discusses the relationship of the L'ai d'Haveloc to Gaimar's account of the story, particularly concerning the battle between Haveloc and Odulf.]

The suggestion has been made in a recent number of this Review¹ that the account of the meeting of Canute and Edmund Ironside at Olney, given by Henry of Huntingdon and others, is not due primarily to a simple misunderstanding of the phrase 'comon togædere' of the A.S. Chr. s.a. 1016; that a tradition of an earlier and equally decisive single combat was a predisposing factor in the choice of the hostile rather than the friendly sense of the phrase; and that this tradition is to be sought amongst those which had gathered round the historical and romantic figure of Anlaf-Haveloc. Though the evidence there (l.c. pp. 119 ff.) adduced from a consideration of the battles of Brunanburh and Vinheith renders the existence pf such a tradition possible, it is on a passage in the Lai d 'Haveloc—and apparently on that alone—that the conclusion is reached: 'there is no good reason to doubt that the single combat formed part of the original story' (l. c. p. 118). When, however, this is based on the statement that 'the earliest version of the Haveloc story which has come down to us appears to be that of the French Lai d 'Aveloc, which probably belongs to the first half of the twelfth century' (l. c. p. 116), there seems to be some confusion between the extant version of the Lai and the earlier one from which it and Gaimar's account have been supposed to derive.

In the course of my work on the Estoire des Engleis I have been led to review the whole question of the relations between Gaimar and the Lai, and, as a result of a detailed investigation which I hope soon to publish, I very much doubt whether the passage in the Lai d'Haveloc cited by Miss Ashdown has quite the evidential value she ascribes to it.

In the first place it is not correct to say that the existing Lai d 'Haveloc is the earliest version of the story, for that honour belongs to the one of which Gaimar is the author. The exceptional regularity of the language and the absence of dialectical features make it extremely difficult to date the Lai on linguistic grounds alone, but, so far as this evidence goes, it points to a period later than Gaimar, i.e. in the second half of the twelfth century. This result was arrived at long ago by Kupferschmidt², though the phenomenon on which he chiefly relied—the use of -eit in imperfects of the first conjugation—is shown by a critical study of the text to have been unknown to the author. The date is supported, and defined more closely, by other considerations: the rule of the couplet is no longer strictly observed, and not only is the technique of the 'lai' adopted, but there has been some measure of direct influence by those of Marie de France; and the nature of the local allusions suggests the period of the revival of the Scandinavian trade in Lincolnshire and the consequent rivalry of the seaports of that county—-say c. 1200—as the date of composition.

Secondly, as the single combat is related neither in Gaimar nor in the English Havelok but only in the Lai d'Haveloc, which is not the earliest version, it becomes essential to determine the position of the latter in the Haveloc tradition. If—which is the generally accepted view—it and Gaimar are both derived independently from an earlier French poem of the first half of the twelfth century, then it is conceivable that the former retained and the latter omitted the account of the combat, and there is thus some justification for the assumption that such a combat originally formed part of the Haveloc story. If, on the other hand (which is the conclusion I have arrived at), the Lai is derived, entirely or in the main, from the Haveloc episode in Gaimar, then this passage must be carefully scrutinised before it is taken as evidence that the incident belongs to the Haveloc story.

Premising that my results are based on a study of both MSS. of the Lai d'Haveloc, whereas the printed editions follow one only, and that the later of the two, my grounds for asserting the dependence of the Lai on Gaimar are, briefly stated, as follows:

1. At least one reading of the Lai seems incompatible with the existence of the common source, and two of the names in the Lai—'Achebrit' and 'Sigar l'Estal'—seem to derive from the text of Gaimar.
2. Of the numerous parallel passages in the two texts, a marked proportion are confined to two sections of the narrative which are peculiar to the French versions—Argentille's dream and the battle between Haveloc and Edelsi—and their differences of expression seem to be due to the author of the Lai rather than to Gaimar.
3. Since the later text is written in the form of a 'lai,' there must necessarily be some changes in the order of the narrative, and reasons of technique are sufficient to account for the varying explanations of Haveloc's presence at Edelsi's court, and the only difference which could be held to prove the independence of the Lai is its use of the 'strongest man' motive³, which is also found in the English version. Careful study of the text of Gaimar suggests that this motive was unknown to him and could not have been omitted by him, but that it is probably a later development of the story in local tradition.
4. One difference appears to be due directly to a misunderstanding of Gaimar's text. On two occasions in the French versions Haveloc makes use of an axe in self-defence. In Gaimar, after his arrival in Denmark and appeal for protection to the Danish lord, Sigar, he is attacked in his lodging by some of the latter's servants who abduct his wife; he seizes an axe which he finds hanging up in the house, rushes out into the street, rescues his wife and kills most of the assailants; later, when he is to be presented to an assembly in Sigar's hall, he is apprehensive of punishment and seizes an axe from one of the bystanders in order to defend himself if necessary. In the Lai the ambush takes place in the street and the axe is there seized from one of the assailants, but, in Sigar's hall, Haveloc passes undisturbed through the bystanders and, still unhindered, takes down an axe from the wall. I suggest that the author of the Lai, misunderstanding the phrase 'dans la ruelle' used by Gaimar with reference to the scene not of the abduction but of the subsequent rescue, imagined the whole affair as taking place in the street, adopted the second of the two methods of obtaining an axe as more suitable and, consequently, had to do the best he could with the other when he came to the scene in Sigar's hall.
5. There are four features peculiar to the French versions which, as they fit in with Gaimar's sources of knowledge and methods of composition, appear to have been introduced into the story by him. They are:
   1. Argentille's dream. This is made the turning point of the first part of the story, is quite different from the English account, seems reminiscent of Iseult's dream in the Forest of Morrois and has evidently been composed with the finish of the French versions in view. As there is some evidence from the Estoire des Engleis that Gaimar was acquainted with the Tristan story, the innovation may be due to him.
   2. The Capture by 'outlaws.' In the English Havelok Grim and his companions are driven by an unexpected storm to England; in the French he is a regular traveller between Denmark and England, and is attacked by 'outlaws.' As these do not appear to have been familiar to the author of the Lai, whereas there is ample evidence that Gaimar was well acquainted with their existence, it would seem that he is responsible for their introduction.
   3. The Geography of the poems. In Gaimar the two kingdoms concerned—of Edelsi and of Adelbrit—are very definitely in East Anglia, and the bounds of the former agree very closely with those of the Southumbrian realm subsequently described by him in his Estoire. In the Lai, though the author does not appear to have a very clear conception of the relations of the kingdoms with which he has to deal, to each other and to England as a whole, yet Edelsi's kingdom is described in the same detail as in Gaimar. It seems probable that this does not represent the original state of affairs and that Gaimar is responsible for their reduction in status from national to local sovereigns, though residence in Lincolnshire most likely accounts for the greater detail in describing Edelsi's realm, as opposed to Adelbrit's, noticeable in the Lai.
   4. The Chronology of the poems. Both in Gaimar and the Lai the events are ascribed to the period following the death of Arthur, and, though a general reference to that monarch might not be out of place in a 'lai,' actually he is referred to in the Lai in the same terms—historical rather than romantic—as in Gaimar, but with no obvious purpose. As there is in Gaimar a clear intention of linking up the events of the story with that period in order to provide a basis for the subsequent Danish claim to have reigned in England prior to the arrival of the English, and as his appeal to Gildas (v. 41) appears to be not entirely a mere literary device for securing credence, it is highly probable that he is responsible for attaching the story to this period.

If these features have been introduced by Gaimar—the arguments only have been outlined and no attempt has been made here to adduce evidence in their support—and if they are also found, as they in fact are, in the Lai, it follows that the latter must have derived them from the former. Consequently, in view of these and other points in which the Lai has been shown dependent on Gaimar and of the lack of proof to the contrary, it can no longer be regarded as representing an independent version of the Haveloc story, and the presence of an incident in the Lai cannot be accepted as proof of its occurrence in the original unless -other evidence is forthcoming in support.

Thirdly, a distinction must be made between the motive for the combat and the combat itself. Of the former Miss Ashdown remarks (1. c. p. 117): 'the humanitarian note is curious, and one might be inclined to see in it the refining tendency of French romance'; but, holding, as she does, that the Lai is older than Gaimar, she rejects this possibility and seems thereby to regard the motive as well as the combat as part of the original Haveloc story. There does not appear to be any compelling need to do this, for, in her own words (l.c. p. 124, n. 3), 'the fact that a certain motive is suggested in the version which has come down to us does not destroy the possibility that the original version implied a different motive'; and the mere fact that the story has been rewritten as a 'lai,' and has been influenced by their technique, renders it a priori probablethat the motive is derived from French romance rather than from Scandinavian tradition. Moreover, the concern for the common people attributed to Haveloc by the author of the Lai seems to me to have been introduced by him partly from the same literary considerations as the additional touches whereby he makes of Edelsi a model leader, who goes out on personal reconnaissance before calling on his followers to do battle, and to suggest that no adequate motive for the combat was offered by the form of the story from which he derived his account.

If we turn to Gaimar's description of the battle between Haveloc and Odulf, we are at once struck by the fact that there is no explicit mention of the latter's fate and that it is uncertain whether he was killed or pardoned, though Gaimar's language—'Li reis Odulf fud dune vencuz Kar Haveloc si se cuntint Il sul en ocist plus de vint' (vv. 742-4)—seems rather to imply the former. On the other hand, Gaimar lays considerable stress on Haveloc's elemency after the battle; he pardons two. enemy princes—apparently Gaimar's own invention—and 'del pais la menue gent Vindrent a merci ensement E Haveloc lur fist parduns Par le cunseil de ses baruns' (vv. 749-52). As he usually evinces some interest in the outcome of the battles he describes, even to the extent of turning an indecisive into a decisive engagement (e.g. vv. 1417 ff.), it seems reasonable to assume that a single combat between the two monarchs did not figure in the story as Gaimar knew it. Neither does it appear probable, in this case, that the author of the Lai developed the combat from the uncertain data at his disposal in Gaimar's text, though, if he knew in addition another form of the story in which such a combat figured, Haveloc's elemency in Gaimar would supply him with a motive for it.

That he was acquainted with the tradition in some other form than Gaimar—very possibly oral—is shown especially by his treatment of Sigar's recognition of Haveloc. In Gaimar, Sigar first sees Haveloc when besieged in the church tower, and his resemblance to the late king—his father—is so great that the Danish lord grants him a truce, takes him to his hall, learns his name and story, in consequence of which he has him watched in expectation of the mystic flame, and this convinces him of Haveloc's identity. In the Lai, the same events, in slightly different order, lead up to the same conclusion, but even more stress is laid on the physical resemblance. Yet, in spite of this being so great that it strikes Sigar in the conflict round the church tower, when, a short time before, Haveloc had sat as an honoured guest at his table, the resemblance passes unnoticed. In the English Havelok, the recognition depends entirely on the mystic flame, there is no question of resemblance, and consequently Haveloc attracts no special attention when at the Danish lord's hall prior to the attack on his lodging. These agreements show that the Lai is combining Gaimar's account with one derived from some other source; for, just as he sought to provide a motive for his introduction of Argentille's visit to the hermit, so he provides one for Haveloc's visit to Sigar in view of a version of the recognition which he does not adopt.

In favour of this suggestion, that the author of the Lai found mention of a single combat in his second source, it may be urged that, whilst he has throughout shown a decided tendency—under the influence, as I believe, of the Lai des deux amants of Marie de France—to make Argentille play a more important part in the story and to make Haveloc more than ever disinclined to act save at the instance of others, in this case he is made to show unwonted decision of character in proposing the single combat with Odulf entirely on his own initiative.

We have seen that this combat was probably unknown to the tradition used by Gaimar, but as probably known to that used by the author of the Lai to supplement the former's account, and the problem arises: was this combat an original feature of the story or is it an addition made in the later twelfth century? The evidence collected by Miss Ashdown seems, as far as I can judge, to render it likely that such a combat did figure in the Anlaf-Haveloc traditions; but, in seeking to link it up with that related in the Lai, is it not possible that she has overlooked one consideration? Assuming the correctness of her deductions from Brunanburh and Vinheith, we should expect the combat, which is to decide the fate of a kingdom and of which Anlaf-Haveloc is a protagonist,. to take place in England, but the one thing clear about the battle between Haveloc and Odulf is that it occurs in Denmark. In the English Havelok, however, though no single combat in the sense of this discussion takes place, yet, because 'Havelok saw his folk so brittene' (v. 2700), he makes for his opponent, Godrich, fights and captures him; the details are in full accord with the boisterous nature of this poem and its rough-and-tumble hero, but there is also the suggestion of a single combat and the battle takes place in England.

It is well known that the conclusion of the story in the French version is very different, involving as it does the account of the dead men set up on stakes to personate the living, but it has not, to my knowledge, been ascertained—I have been concerned only with the two French texts and not with the wider problems of the Haveloc tradition—which of the two versions represents more closely the original ending, though I cannot be sure, in my own mind, that the ruse of the dead men was not introduced into the story by Gaimar. There can be little doubt that, in addition to his interest in the Haveloc story for its own sake, he had in mind its importance for strengthening the Danish claim of prior possession of England put forward, in his account, by Canute at his celebrated meeting with Edmund Ironside; and that claim would be strengthened if Haveloc obtained Edelsi's kingdom by the latter's free gift rather than by right of conquest. Also, if the ending underlying that of the English Havelok be the original one, the outcome of the combat was probably fatal to Haveloc's opponent.

Therefore, if Gaimar knew the dead men ruse from another source—and he was not unacquainted with Danish traditions—it would, with his purpose in view, supply him with a better and more striking ending, and, to judge by his methods on other occasions, he would not have scrupled to adopt it instead of the original combat ending. Further, when both Gaimar and the second source conflict, the author of the Lai seems to prefer the former but likes to make use as well of any additional features from the latter. Hence, assuming that the single combat figured in his second source—here representing the original tradition—he would have a very striking incident at his disposal, after deciding to adopt the ruse ending from Gaimar, which he could use to good purpose in the, as yet, rather colourless Haveloc-Odulf incident.

Thus, though I have taken away from Miss Ashdown with the one hand in showing that this passage of the Lai cannot safely be used as direct proof of her contention, yet I have returned her somewhat with the other, and, should it be possible to substantiate the hypothesis of the preceding paragraph, it may be that she will consider herself the gainer, rather than the loser, by the exchange.

Notes

¹ M. Ashdown, 'The single combat in certain cycles of English and Scandinavian tradition and romance.' Mod. Lang. Rev. XVII, p. 113.

² M. Kupferschmidt, Die Haveloksage bei Gaimar. Rom. Studien, IV, p. 411.

³ Edelsi promises his dying brother-in-law to protect Argentille, then an infant, and, when she is of fit age, to marry her 'al plus fort home' in his kingdom; in order to deprive her of her inheritance, he adheres to the letter of his promise by giving her to his scullion, Haveloc, because of his prowess in wrestling and other feats of strength.