Ballad Source Study: Child Ballad No. 4 as Exemplar
by Holger Olof Nygard
The Journal of American Folklore, Vol. 68, No. 268 (Apr. - Jun., 1955), pp. 141-152
BALLAD SOURCE STUDY: CHILD BALLAD NO. 4 AS EXEMPLAR [1]
BY HOLGER OLOF NYGARD
TRADITIONAL ballads have interested people in many different ways, and this interest has variously obscured, as well as enriched, our understanding of ballads. One activity, certainly, the hunting of sources, has been relatively fruitless. We are all led by natural curiosity to wonder how this or that song originated, and in response to this natural interest, vast amounts of energy and argument have been expended in efforts to trace the different ballads to their root ideas, with varying success.
The question of origin may indeed not always be difficult to answer. The general impression is that historical ballads are, after all, history; they are songs that identify time, place, and person. The inference is therefore that these songs originated after the events they describe. By such reasoning the ballad of The Death of Queen Jane (Child, No. 170) originated not long after Jane Seymour gave birth to Prince Edward in October 1537; Thomas Cromwell (No. 171) arose from events of the summer of 1540; Musselburgh Field (No. 172) celebrated the victory of the English over the Scots in the battle of 1547. With the well-known Mary Hamilton (No. 173) we are on less firm ground, for despite the air of historicity about the song, there is no certainty as to the identity of the four Maries or the truth of the incident. Attempts have been made to underwrite the historical origin of the piece, notably by C. K. Sharpe and J. W. Courthope. But is the historical event necessarily the genesis of a song so traditional in outline and so classic in content? Is it not as conceivable that the historic event has merely reinvested a song of earlier date with new vigor through an adventitious historical applicability? A notorious event is able to give new life, another habitation and name, to the narrative which has passed current in tradition- much as anecdotes gravitate to the famous. Certainly some ballads are so lacking in circumstantial detail, so fraught with commonplace event and phrasing, that one must recognize the possibility that the single ballad often no better describes one historical event than another. The Scottish bride-stealing ballads may be taken as cases in point.
Robin Hood, who figures large in English balladry, has been fair game for those who have seen history between the lines. After it had become unpopular to identify him as the Earl of Nottingham or as Robin Fitzooth, a more subtle historical approach was called for. The Gomme school chose to interpret ballads as history obscured, even as anthropologists like Andrew Lang prized the ballads for their obscured anthropological record. It may be illuminating and rewarding for us to think that the Robin Hood songs reflect social and cultural tensions between the lowly and aristocratic, between the English community and the Norman master, but the idea remains no more than an hypothesis. Robin Hood, insecure in the annals of history (despite the longish entry in the Dictionary of National Biography, and despite the recent effort to identify our ballad hero with the Robert Hood in the Wakefield Manor Court Rolls),[2] can be interpreted as no more than a comment on historical event; where he came from, what he was originally, and what gave the tradition such vitality remain as puzzles for ballad scholars.
If the attributed historical origins of a great many historical ballads are open to question, then certainly the beginnings of the non-historical material in folksong are shrouded in far greater mystery. Yet it is noteworthy that the ballads that have received the greatest attention as regards source are not the historical or quasi-historical songs, but precisely those that have been recognized by Child, Grundtvig, and others as old and international, those which Child honored with an early place in his collection. In dealing with the most ancient of ballads and those that least reflect time and place, source studies have the greatest room for speculation, and for error. Source studies deserve scrutiny, for they reveal patterns of approach and thought that might well be noted for their inherent virtues or dangers. These patterns are in themselves a chapter in the history of ballad scholarship. By way of contribution to that chapter I propose to review the various attempts that have been made to explain away, to lay the ghosts of Lady Isabel and the Elf-knight (Child, No. 4), the ballad which more than any other has led scholars to exercise their learning and ingenuity in this direction.
During the past century many assertions and theories of origin for Lady Isabel and the Elf-knight and its continental analogues have been promulgated. Some have been highly provincial, innocently narrow in view, rather falling into the classification of local antiquarian misinformation, and not deserving of the term "theory" at all; others have been rational and informed, even super-rational and symbolic in essence, partaking of a rationale beyond reason itself. The contributions have for the most part been made by men prominent in ballad scholarship who have all been motivated by the same desire: to demonstrate the ballad's antecedents, to reduce the ballad to its source or germinal idea.
The narrowly provincial interpretations of the events described in the ballad need hardly concern us here (save for the amusement they afford in their quaintness); for they are nothing more than local custom foisted upon the song or, perhaps more properly, legends that the song has foisted upon the locality. Robert Chambers reported how in Ayrshire the country people pointed to "a tall rocky eminence called Gamesloup, overhanging the sea," where it is said the false Sir John "was in the habit of drowning his wives, and where he was finally drowned himself." These people who look upon the ballad as a representation of fact "further affirm that May Collean was a daughter of the family of Kennedy of Colzean."[3] The Scots of Ayrshire, not unlike ballad singers elsewhere, made the ballad vividly their own by framing the narrative as if the event did indeed happen "in these very parts."
Svend Grundtvig, in his informative Headnote to Kvindemorderen, the Danish form of the ballad (Danmarks gamle Folkeviser, No. 183), gave little attention to the question of origin, for he found the matter beyond demonstration; but he tentatively suggested that to his mind some obscured elf song ("fordunklet Elvevise") lay behind the narrative.[4] His suggestion was in great part predetermined by his belief, also held with reservations, that the ballad may have originated in the Scandinavian North, for he would read the names of the villain of the ballad as deviations from a Danish name. Grundtvig would have been hard put, however, to explain why this elf destroyed his victims, for elves in Scandinavian lore are not predatory and murderous; and Grundtvig's critics have not missed the opportunity of raising this question.[5] The villain in the Scandinavian variants is nowhere identified as an elf; he is called an elf only in British variant A.
As a possible elf story suggested itself to the Dane Grundtvig, so a possible merman story suggested itself to the German Franz Bihme,[6] for mermen are as characteristically North German as elves are Scandinavian. Rivers and seas are a constant feature in the ballad, and the villain of the piece drowns his victims in the variants of France, England, and Poland. Child, in his first collection of ballads, following J. H. Dixon, had espoused this theory: "The Merman or Nix may be easily recognized." [7] But he quite dropped the idea for the 1882 edition after having been soundly twitted by Grundtvig for suggesting that the heroine escapes from the merman by pushing him into the sea and so bringing about his death through drowning. This would make the ballad a variant of the "Wise Men of Gotham" who tried to drown the eel.
There is insufficient reason for giving credence to either the obscured and distorted elf-tale or the submerged merman-tale. They are both shots in the dark, and we remain, and probably will remain, ignorant of how far off the mark they in fact are.
The association of Lady Isabel a nd the Elf-knight with the Bluebeard tale leads us to apparently firmer ground, for here we may consult the evidence and reach conclusions not entirely built on faith. Goethe w as the first to make the association, and a number of scholars have followed his example, including Ludwig Uhland, Franz Bohme, and Lutz Mackensen. It is not unusual to find references in German scholarship to the Blaubartlied, suggesting that the association is sound and proven. But, in point of fact, the association is all too readily made; the tale and ballad in all their multiple forms are distinct and separate in their histories, as far as the records dating from as far back as the mid-sixteenth century will permit us to judge. Ballad and tale do share a narrative idea: a man or demon murders a number of women and is in turn destroyed by one too clever for him. The startling thing is not, however, the similar narrative idea (similar only if sketched in such a summary way), but the complete difference of treatment and interpretation of the narrative. There appear to be no details in the total body of variants of either ballad or tale that would render possible a demonstration of relationship. Each appears sufficient unto itself and will reveal no affinities with the other. If the doctrine of polygenesis of a narrative idea commands respect, its adherents might well cite this ballad and tale as instances.
Both Franz Bohme[8] and Paul Kretschmer[9] have considered the possible relationship. The variance of their opinions is good demonstration of the difficulties involved. Bohme thought that the ballad was originally an elf-song (here he disconcertingly follows Grundtvig), into which had been stuffed the Bluebeard story of multiple murders, a story with a historical basis. Kretschmer thought that the tale was originally about an otherworld demon, a tale into which was stuffed the ballad narrative of multiple murders, a ballad with a historical basis.
Both points of view are sheer speculation. Just as there are no discernibly valid reasons for arguing the precedence of one form over the other, or for admitting the relationship itself, for that matter, so there is no reason for believing either tale or ballad to be historical in its inception. Kretschmer thinks the ballad may have derived from some possible broadside account of actual murders such as were perpetrated by the Frenchman Comorre in the early thirteenth century, or by the more notorious Breton Gilles de Laval in the early fifteenth century. The point is not very well taken when one considers by how much both of these gentlemen antedated the printing of fliegende Blotter. But more significantly, Kretschmer failed to study the ballad's history closely and so recognize that it was not "einfach ein Lustmord" (simply a sex murder); its motifs point back to an earlier supernatural content. This oversight would render Kretschmer's position less secure than Bohme's. Bohme, who would give precedence to the ballad, saw the tale as a tradition possibly relating to Gilles de Laval, a supposition that Kretschmer flatly denied. It is to my mind reasonable to consider this historical attribution of source made by both Bohme and Kretschmer as akin to the naivete of the Scottish country people with their Gamesloup.
We move next to a study made about the time when the appearance of the copious Handworterbuch des deutschen Aberglaubens focused attention upon superstitions. In I929 Friedrich Holz[10] presented the theory that the germinal idea of the ballad was the medieval belief that maiden's blood cured leprosy. Holz cites tales which illustrate the superstition (Hartmann von Aue's Der arme Heinrich is the best known example), as well as cases from medical history that prove the wide currency of the belief and the respect with which it had on occasion been held by medical practitioners. But when we turn to the variants of the ballad the idea does not inspire us with confidence. There are streams of blood and springs of blood in Scandinavian and Dutch variants as well as the more usual streams of water. But no variant from these areas gives the slightest hint of what the villain does with the maids, aside from hanging or burying them. No statement is made of his using their blood. And no indication is to be found of ill health on his part. It is perfectly true that the villain seems motiveless in the older tradition (gold and robbery are a latterday tradition in the ballad), but that is no reason for attributing to him the motive that Holz suggests. Only one single variant in the entire tradition of the song furnishes him support for his idea, a variant from Switzerland clearly removed from the center of the tradition. In this variant the knight promises to teach the maid the "Baderliedli" if she will come away with him. Holz interprets this song as a reference to a particular healing bath: "gemeint ist hier das Schongauerbad am Linden-berge."[11] His case is bolstered by the fact that this ballad from Aargau in Switzerland has a parallel folktale from the same district, a tale in which leprosy is cured by baths in the blood of maidens. Linking the entire European tradition of Lady Isabel and the Elf-knight with this superstition on the basis of the single folktale and variant is questionable method. The chain of causation moves in the opposite direction:
a well known and much sung ballad, having reached the district of Aargau, has probably submitted to the influence of a local tradition. The song has been maintained in tradition over all Europe for centuries without the awareness on the part of singers of such a reasonable motive for what is, after all, an other than human villain. Holz' demonstration of origin is a forced argument.
The solar mythologists have had fair game with Lady Isabel and the Elf-knight, with Leon Pineau[12] as spokesman. By a process of fabulous analogizing he identified the hero as the spirit of shadows, death, night, and winter, finally overcome by the warmth of summer. The murdered maids thus represent months of the year. Refrains from Scotland and Denmark are cited as final proof of the ur-meaning of the song: Scottish variant A has the refrain, "Aye as the gowans grow gay, / The first morning in May;" and Danish variant A, "Men linden groer." Such refrains are, of course, the most common of common places in balladry a nd have less to do with the narratives they accompany than with the tunes. Andreas Heusler has fittingly pointed out that Pineau went so far as to date the Scandinavian ballads of magic and the supernatural much earlier than the Eddic poetry.[13] The solar mythologists of the school of Max Miiller have long since been laughed out of court with their poetic subjugationt o their simple formula of all imaginativec reationst hat seem to stand outside history. I have no interest in belaboring a dead issue; I have touched upon Pineau's interpretation of the ballad only because it forcibly illustrates a common denominator among the source attributions, the tendency to write the ballad's early history not on its own terms but on terms dictated by something else.
Another theory of inception, not unlike Pineau's adventures of a soul among the masterpieces of folk poetry, is Paul de Keyser's[14] psychoanalytic interpretation of the morphology of the ballad. Unmindful of the fact of change in ballads as they pass through tradition, and oblivious of the national variants outside his native Holland, de Keyser points out how in the usual Dutch ballad, the brother of the maid (in a passage of incremental repetition-itself a commonplace passage) gives his sister permission to follow Halewijn, the villain, provided she remain chaste. According to de Keyser, Halewijn is, in the singer's subconscious mind, the brother of the maid. The beheading of Halewijn by the maid is then interpreted as punishment by castration arising from the suppressed desires of the singer. The ballad is an expression of unconscious drives of sibling incest on the part of those who have given the song its narrative shape.
De Keyser's interpretation is not without its interest, and the reinterpretation of the events of the human scene by sexual symbol may not be without value, but does it in fact tell us what we wish to know about the origin of Lady Isabel and the Elfknight? We must be prepared to rewrite the psychological correlatives with every major change that the ballad has undergone in its transmission. We cannot, for instance, be certain that the brother played a part in the ballad in its earliest form.
Apart from the Dutch variants, the brother enters the ballad only in a minor tradition in Germany, and then quite differently, as the protector of the maid at the close of the narrative. One is also forced to observe that a psychological interpretation such as de Keyser's does not distinguish the song under consideration from a number of other songs that might be reduced to the same drives. The Cruel Brother (No. 11) is undoubtedly read by many as an exhibition of sibling love; psychological explication makes good sense of the ballad story. But the psychological analogizing of Halewijn by de Keyser does not enrich and support our understanding in the same way.
It must be admitted, of course, that de Keyser is not seeking the same kind of root that the literalists are attempting to find. De Keyser cannot help us determine an early from a late form of the same narrative, but that is precisely what a source study must do. In defense of the literalist it must be remembered that whoever first fashioned the ballad, despite his troubles and joys, confused Halewijn no more with the brother than with winter. To say that he did so unwittingly is to undertake a proof hardly feasible within the limitations of the pragmatic foundations of the psychoanalytic science. These then are a goodly number of attempts to disclose the possible beginnings of one ballad. Their very number and the fundamental disagreements among them bring home to us the difficulty of the task. There is clearly room for more hypotheses, for none of the half-dozen so far reviewed has been widely accepted. But where so many have failed, there seems as little hope of success as in the turning of the sands of Egypt for the tomb of Sanakht.
We have not in fact finished the review of source hypotheses for Lady Isabel and the Elf-knight; one remains, which will occupy our attention at greater length if only because it has gained credence among scholars by its deceptive plausibility. I call Sophus Bugge's[15] source theory deceptively plausible, for it makes its appeal to one's reason, is supported by a show of linguistic techniques, and is the address of a scholar to his problem. What is more, Bugge's conclusion has been accepted without questioning by George Doncieux, Knut Liest0l, W. J. Entwistle, and Marius Barbeau, to name a few.
Bugge argued as long ago as 1879 that Lady Isabel and the Elf-knight is derived from the Book of Judith in the Apocrypha, that heroine and villain are Judith and Holofernes. Child presents Bugge's argument very fairly in his Headnote to the ballad, without comment, although he did have misgivings. The forcefulness of Bugge's argument resides in the wealth of minute points he offers for the reader's consideration. In our bid for scepticism we shall consider the nature of the minutiae, the premises of his argument, and its mainstays.
Phrased in a general way the two narratives have a likeness: a man who stands in enmity to a woman is killed by her, and her method of killing him is by decapitation. But there are some major differences too: Holofernes does not intend to deprive Judith of her life, nor has he indulged in a series of murders. This multiple murder motif is the very heart of the ballad narrative, for it appears in all variants from all countries. It is difficult to conceive of the series of murders as absent from the ballad at its beginning. And Bugge would be hard put to explain how the supernatural element entered the ballad if the original was indeed a redaction of the Apocryphal tale.
Analysis of Bugge's argument reveals that it rests upon a number of assumptions that rather beg the question. He assumed, first, that any small detail in any of the many wide spread variants from Iceland to Italy was "original" if it was in any way suggestive of a parallel with the Judith story.[16] He assumed, second, that an early German poem about Judith, another source of the ballad maker, is corrupt in its phrasing so that it differs from the ballad in the one significant instance in which the ballad shows a supposed direct borrowing from the poem. He assumed, third, that the ballad maker, in writing his ballad after the Vulgate account of Judith and Holofernes, misread and misinterpreted the Latin in such ways as account for particular details to be found in certain scattered variants of the ballad. He assumed, fourth, that the earliest form of the ballad, the original, differed greatly from the extant forms. This assumption looks innocent enough; we take the statement to be a truism. And yet a comparison of the national forms of the song during the last 400 years reveals an overall correspondence and consistency that do not permit us to think that prior to the mid-sixteenth century vast changes w ere being worked in the narrative outline. Fifth, he stretched the concept of parallels and echoes beyond the point permitted a careful student and one less enamored of the argument. And sixth and finally, he assumed that the name Holofernes is the root name from which the various names of the villain have arisen.
Let us look more closely at the arguments themselves. The High German poem from c.200 on which follows the Vulgate in its outline and major details, has a pair of lines that read: du zahiz wiblichi / uf slabranihichi.[17] The lines are part of the angel's advice to Judith about dispatching Holofernes; they have been variously emended by editors, for they do not make good sense. Bugge translates these words as, "You draw it in a womanly fashion and strike. . . ."[18] In one Danish variant we meet the lines: Saa qwindelig hun det suerd uddrog, / saa mandelig hun til hannem hug. These lines from Danish D are a commonplacein Danish balladry, a fact that Bugge does admit. But nevertheless he thinks a relationship with the old German poem is sufficiently clear to permit h is making an editorial change in the text of the poem, thereby bringing it into conformity with the Scandinavian commonplace. He alters the poem to read, on the basis of what is in the Danish ballad: sil zohiz wiblichi/ undi sluoc mannllchi. [19] His procedure is a clear begging of the question; he has not demonstrateda similarity between poem and ballad, but has instead made one.
Bugge disarmingly suggests that some of the parallels he presents may be accidental agreements, but only after he has completed a lengthy barrage of citations that do not convincet he reader. He does not recognize ballad commonplaces as such, but cites them as illustrations of how the ballad and the Apocryphal tale (also not devoid of traditional commonplaces) are related. In some Low German and Danish variants the principals ride for three days and three nights. The statement occurs in a great many ballads as a standard folksong description of a long ride. What is more, Grundtvig points out that the statement in the Danish variants has entered the ballad from Den farlige Jomfru (DgF, No. I84); the nature of commonplaces is that they do wander. The figure three bears no singular significance in balladry, for most multiple things tend to three. Yet Bugge finds the three day ride revealing, for did not Judith kill Holofernes on her fourth day at his camp? [20]
Bugge disregards the changes incurred by traditional dissemination in the ballad. He draws his catalogue of parallels from any variant that suits his purpose. With him all motifs in the ballad are original which seem to echo the Judith story, and all other motifs are later interpolations and changes. In certain English variants the knight comes from the North; so must we infer that Holofernes did (Book of Judith, 16:5)? [21] The trait in the ballad is so distinctly confined to England (in a tradition that came from France, without mention of "North") that one can only assume that the North of the Outlandish Knight and the Assyria of Holofernes have nothing to do with one another. The same thing must be said of May Collin's (variant D) leaving her home at night, and Judith's doing the same thing.[22] Holofernes' armies are to be seen as metamorphosed in to the brothers, sisters, and followers of the villain found in certain variants.[23] Surely such an interpretation takes advantage of the most fortuitous details. The Scandinavian picture of a wonderland to which the false knight will take the maiden and his offer in the English ballad of castles over which she shall rule as lady are cited as correspondences for Holofernes' promise to Judith that she shall be great in the house of Nebuchadnezzar and her name shall be known in all the land (1 :21).[24] But such offers of attractive bait are to be met with in any story in which a seduction or artful vanquishment of a woman by a man is attempted. In the Danish ballad the heroine binds the murderer's hands and feet; Bugge points out that in the German poem Judith's handmaid stands by to restrain Holofernes if he should awake while she is preparing to take his life.[25] The passages are hardly correspondences. The closing stanza of the Icelandic ballad, in which the heroine is described as retiring to a nunnery, is a Scandinavian commonplace, not original with this ballad; Bugge believes the stanza to be an echo of Judith's living out the rest of her days as a chaste widow.[26] One lone Swedish variant gives the villain sisters who cry out when they find him dead; this suggests to Bugge the lamentation of Holofernes' lieutenants.[27] In one English variant( D), the maid and her parents journey to the seashore to behold the body of the murderer; Judith and the Jews journey to find the body of Holofernes after their victory at arms.[28] Clearly such "parallels" do not convince. Bugge, searching about among the vast number of variants of the ballad that were available to him, has made a case by pointing out every fortuitous resemblance that a rich ballad tradition through the vagaries of oral transmission could provide. His method was not to sort out in the first place those things that could clearly be argued as original or old in the ballad's tradition. To him, that which seemed closest to the Apocrypha was original.
But now for the ballad maker's misreading of the Vulgate text. At no point is the phrasing of the ballad like that of the Book of Judith. The words of the one approach the words of the other most closely in the instances where the words do not mean the same thing. In the Scandinavian ballad the villain lifts the maid upon his horse and they ride away. Bugge suggests that this lifting is an echo of the verb "elevaverunt" in the description of Holofernes' men carrying Judith into his tent (10:20).[29] He does not mention that lifting the maid upon the horse behind the rider is a commonplace in Scandinavian balladry. In the Dutch ballad the maid blew the horn like a man; Bugge compares this with "cum audissent viri vocem ejus" (13:14), believing that the ballad maker read "viri" as being in the genitive case.[30] Bugge believes that the couplet from a High German variant, "Der Ulinger hat eilf junkfrawen gehangen, / Die zwolft hat er gefangen," is a rendering of "Dixit se incessurum fines meos et juvenes meos occisurum gladio, infantes meos dare in praedam, et virgines in captivitatem" (16:6).[31] Bugge believes further (although he admits the idea is daring) that a High German form of the ballad in which the murderer is bound and himself hanged may stem from a misreading of the Vulgate "accessit ad columnam, quae erat ad caput lectuli ejus, et pugionem ejus, qui in ea ligatus pendebat, exsolvit" (13:8). He believes the misreading took place by a misplacement of modifier, as in "She loosened the knife, which belonged to him, that hung on supports."[32] It is inconceivable how such an echo of the Vulgate (if we were to admit it as one) could remain intact through oral transmission of the ballad into Southern Germany and reveal itself in one isolated variant that is not in the usual stream of the tradition. These are instances of the ingeniously drawn parallels
between ballad and source.
In the Apocryphal story an angel visits Judith to give her advice as to how to dispatch Holofernes. Bugge finds traces of the angel in two places in the tradition of the ballad: first, in the form of the white dove that warns the maid in the High German ballad that she is going to be killed, and second, in his reconstruction of the refrains of the Scandinavian variants. As others have pointed out, the talking dove is a commonplace in German ballad lore and folktales, and is, as best one can determine, a late entry into the tradition of the ballad. Bugge's reconstruction of the refrain is the epitome of ingenious argument. As he expends an entire page in the demonstration,[33] and as it is his closing point, we shall examine it here. Danish A has as its Indstev (refrain following the first narrative line of a stanza) the following:
"Men lienden groer," or, less frequently, "Men leinenn grodt." Danish B has the fuller "Se Vindelraad til edele Herre din!" Danish C has an Indstev much like Danish B, in which both the gentleman's and lady's names are mentioned: "Se Hollemen ind til Vendelraad!" Because Danish A has the same Efterstev (refrain following the second and last line of narrative text in the stanza) as has Danish C, Bugge believes that the Indstev of A should therefore be the same as that of C. "Men leinenn grodt" (the less usual form) is to his ear and understanding a corruption of the name of the maid as it is found in Danish B and C, Vindelraad, Vendelraad. He believes that "leinenn" in the Indstev represents the name of the maid, which originally must have been Lenel, Lennel, or Linnel. The "raad" (which "grodt" in A suggests to him) was in his view not part of the original name. Lenel or Linnel is, like Linnich of a Low German variant, the diminutive of the name Helena to be found in one German variant. And so Bugge reconstructs the Indstev for Danish A to read: "Se, Lennel, raad til Hollevern aedele Herre din!" The Indstev in this form is then interpreted by Bugge as an admonition on the part of the angel to the heroine that she should bear a weapon against the villain. The phrase "raad til" is not as specific as Bugge would have it; the import of Bugge's refrain might be, "Be advised regarding Hollevern your noble lord." In terms of other Scandinavian ballads it would be very strange if the refrain were as explicitly a part of the action of the ballad as Bugge would make it. The B and C refrains are commonplaces that suggest the dance motions that once accompanied the singing of the Danish ballads. The A refrain as it appears in the Karen Brahe MS of c.1550 is understandable as a commonplace statement about the natural scene, much like the Scottish "Aye as the gowans grow gay."
But the reader will wonder how the name Hollevern entered the reconstructed refrain. Bugge put it there. For he argues that the mainstay of his source attribution is that the villain's various names seem to have arisen from Hollevern, a name palpably suggestive of Holofernes.
In early Scandinavian variants we meet such strikingly Germanic names as Oldemor, Olmor, Romor, Hollemen, Ulver, Alemarken, Rulleman, etc. The names in German variants are not unlike these: Ulinger, Ulrich, Adelger, Alleger, Helsinger, Halsemann, Olingen, Olbert, etc. Bugge thought that there was some obscured but ascertainable original form behind the series. He disagreed with Grundtvig's choice of Oldemor. He decided that the original name must have been Hollevern, Holevern, or Olevern. He explains that even before he had any thought about the source of the ballad's theme, he had arrived at the decision, from a comparison of the names, that the root form was (H)ol(l)evern, although, as he admits, it appears nowhere unaltered, "skjont dette ingensteds uforandret foreligger."[34] A counting of the frequencies of the successive phonemes that make up the twenty-one names Bugge had at his disposal gives us a choice among the following names: Olemor, Ulemor, Oleger, Uleger. Bugge suggests that an original v had given way to m, g, and b, an idea that is hard to accept without good reason. The v sound, which he argues for, appears in but three names, Ulver, Halewijn, and the English Elf (a questionable member in this company). None of the names end in double consonants as does his reconstructed name. The mainstay of his argument, as he calls the name derivation, is very insecure.
Bugge admits that the extant ballad is much different from his postulated original, that the ballad is but a wild shoot of the Judith story. But in making his case he has asked us to dismiss all our native caution and to stretch our credulity beyond reason. Andrew Lang's response was: "If so, the legend is diablement change en route."
In Hoffmann von Fallersleben's collection of Dutch ballads, which Bugge had to hand, appeared a nineteenth century broadside variant of Halewijn[35] that vor Fallersleben cites as an example of the depths to which the folk poetry of the Netherlands has fallen. A third line has been fashioned for every stanza, usually by merely varying the substance of the second line. Two of the stanzas read:
34. Ik heb van 't leven hem beroofd,
in mynen schoot heb ik zyn hoofd;
hy is als Holofernes, my gelooft.
37. Zy reed dan voord als Judith wys,
zoo recht nae haer vaders paleys,
daer zy wierd ingehaeld met eer en prys.
The vague similarity of the story of Halewijn and the Judith-Holofernes narrative had not escaped notice before Bugge. This mention of the two principals from the story in the Apocrypha lends no support to Bugge's theory (he strangely makes no mention of this particular variant), for the names are clearly a late addition to the ballad. The names are both introduced in unballad-like similes and as part of a line filling process. Is it beyond the realm of probability that Bugge took his cue from this
variant?
We have now come to the end of a lengthy list of source studies for Lady Isabel and the Elf-knight. This ballad has been most richly endowed with sources, none of which can safely be accepted as valid, even though each is heralded by its author as a certitude. We may learn from the experience of others that the certitude is not with the findings, but with the preconceptions and determination brought to the problem. It has become apparent in the course of this paper that all the attributions of source considered have been fathered by world views, Zeitgeist, climates of opinion, attitudes, partialities, rather than by the evidence that the subject itself, the ballad, affords. The difficulty is undoubtedly in the paucity of information that the ballad does afford. But the facts and inferences that we may read in the variants of the ballad do provide strong grounds for doubting each suggested source. It would be singularly wrong-headed to neglect what the ballad record can offer as a guide and restraint. We are left with a handful of improbable impossibilities as to the source of the ballad. And for these we may well be thankful, for their authors have trod the sands of surmise and have taught us how to avoid them, if we will but learn by example.
It is true that Bugge's argument, in its devious ingenuity in holding to the text, bears no outward mark of fashion, but nevertheless his position was much predetermined by other habits of thought. His addiction to Biblical parallels was noted by his Scandinavian colleagues. Grundtvig made a decidedly questioning appraisal of Bugge's theory in an addition to a letter to Child dated 29 January I880. Child's reply gives a much clearer view of his judgment of Bugge than does the Headnote to Lady Isabel, which was written two years later:[36]
The very important article of Sophus Bugge, I have not had time to consider, but should imagine that it was a corollary to his other theories which are now making such a stir. He will expect a Bible story in many a ballad, I dare say, and such suggestions being infectious, and not to be demonstrably refuted, many minds will be ready to follow him to any length.[37]
When Leon Pineau presented his theory of origin, he had this to say of Bugge:
"Nous ne croyons pas qu'il se trouve encore quelqu'un pour defend recetteth Corie."[38]
But his countryman, George Doncieux, defended the theory six years after Pineau's words appeared in print. A number of scholars have accepted Bugge's demonstration (possibly for lack of a better theory); the late W. J. Entwistle on six different occasions in his European Balladry identifies the source of Halewijn as the Book of Judith, and the identification has spread to other general accounts of balladry as well as to specific accounts of this ballad. A certain scepticism is a salutary thing in these matters, on the part of investigators and their readers alike.
It may be gratifying for us to know that some ballads do have identifiable sources. That Lady Isabel is so unattached, as it were, should not disturb us, for the ballad is still with us in all its multifarious variety. We still have the adventure of Lady Isabel with an elf-knight as full of mystery as ever.
University of Kansas
Lawrence, Kansas
Footnotes:
1. This paper was read at the Sixty-fifth Annual Meeting of the American Folklore Society, in Tucson, December 1953.
2. J. W. Walker, The True History of Robin Hood (Wakefield, I952).
3 Cited in F. J. Child, English and Scottish Popular Ballads, I (Boston, 1882-98), 24.
4 Svend Grundtvig, Danmarks gamle Folkeviser, IV (Copenhagen, I853 f.), 29.
5 See Leon Pineau, Les Vieux Chants Populaires Scandinaves, I (Paris, 1898), 264.
6 Ludwig Erk and Franz Bohme, Deutscher Liederhort, I (Leipzig, I893-94), I50.
7 Child, English and Scottish Ballads, I, 195.
8 Erk and Bohme, Deutscher Liederhort, I, I48-I49.
9 Paul Kretschmer, 'Das Mirchen von Blaubart," Mittheilungen der Anthropologischen Gesellschaft in Wien, XXXI (1901), 62-70.
10 Friedrich Holz, Das Mddchenrduberballade (Heidelberg, I929), pp. 86-102.
11 Holz, Das Mddchenrduberballadpe,. 94.
12 Pineau, Les Vieux Chants, I.
13 Andreas Heusler, "Ober die Balladendichtung des Spitmittelalters namentlich im skandinavischen Norden," Germanisch-RomanischMe onatsschrift,X (1922), 19.
14 Paul de Keyser, "Het Lied van Halewijn. Een psycho-analytisch Onderzoek, "Nederlandsch Tijdschrift voor Volkskunde, XXVII (1922), 165-I74.
15 Sophus Bugge, "Bidrag til Nordiske balladedigtnings Historie," Det Philologisk-historiske Samfunds Mindeskrift, 1854-1879 (Copenhagen, I879), pp. 75-92.
16 In his own words: "Efter disse Bemaerkninger skal jeg gjennemgaa Balladen, som den foreligger i de mangfoldige skriftende Former, for at paavise Forbindelser med Fortaellingen om Judith og Holofernes" (p. 81).
17 Karl Miillenhoff and W. Scherer, Denkmaler deutscher Poesie und Prosa aus dem VIII-XII Jahrhundert (Berlin, I892), p. 141.
18 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 87.
19 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 87. Bugge's emendation does not agree with any suggested by various editors of the poem. See Miillenhoff and Scherer, Denkmaler deutscher Poesie und Prosa, p. 141.
20 Bugge, "Bidrag til Nordiske balladedigtnings Historie," pp. 84-85.
21 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 82.
22 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 84.
23 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 83.
24 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 83.
25 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 87.
26 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 89.
27 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 90.
28 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 89.
29 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 84.
30 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 88.
31 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 85.
32 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 88n.
33 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 92.
34 Bugge, "Bidrag til Nordiske balladedigtnings Historie," p. 79.
35 Hoffman von Fallersleben, Horae Belgicae II. Niederldndische Volkslieder, 2nd ed. (Hannover, 1856), pp. 43-46.
36 In his Headnote, he describes Bugge's argument as an "entirely novel and somewhat startling hypothesis"(C hild, E nglish a nd Scottish P opular B allads, I, 51).
37L etterd ated2 1 MarchI 880,c ited in S. B. Hustvedt, Ballad Books and Ballad Men (Cambridge, M ass., 1930), Appendix A , "The Grundtvig-Child Co rrespondencep,." 286.
88 Pineau, Les Vieux Chants, I, 263.