Review: Kemppinen and "Lady Isabel": A Review Article
by Samuel P. Bayard
Western Folklore, Vol. 14, No. 2 (Apr., 1955), pp. 114-120
Kemppinen and "Lady Isabel": A Review Article
SAMUEL P. BAYARD
The Ballad of Lady Isabel and the False Knight. By IIVAR KEMPPINEN. (Helsinki: Suomen Kulttuurirahasto [The Finnish Cultural Fund], 1954. Pp. 301. $5.85)
THE INVESTIGATION OF AN INTERNATIONAL FOLK BALLAD requires much courage and learning. Both qualities, besides method, are shown in Iivar Kemppinen's study of Child No. 4, "Lady Isabel and the Elf Knight," a monograph purporting to solve the problems of the ballad's time and place of origin.
Obviously it takes especial effort to tackle a folk poem so widespread as Child 4; and Kemppinen has tried to review all versions. If he has not quite succeeded in this aim, he has still scrutinized 1,865 variant items, as well as a number of related and derivative songs in the traditions of many nationalities. Especially noteworthy in this survey is Kemppinen's revelation of the true Finnish and Estonian forms of the ballad (see pp. 128, n. and 143, n.), hitherto unrecognized.
Two especially valuable features are the setting forth of all variants (in groupings according to close relations) to show the presence of defined diagnostic traits, and the survey of various scholars' theories concerning the age and origin of Child 4. Both are material for, and introduction to, the author's own conclusions. In any rich study like this one, there is bound to be much of importance aside from the immediate ends and concerns of the work. Kemppinen, while stating methods and explaining circumstances, constantly becomes involved with problems that make his book quite provocative, since they inevitably must exert some influence on efforts to evaluate his conclusions. Kemppinen displays recognition of the difficulties faced by a student who wants to determine the area and period of origin for a traditional poem. His critical principles appear to be sound, theoretically. Unfortunately, the difficulties persist; the unequalled wealth of data seems yet to be not enough to banish them; and this reviewer is thus placed in the curious position of inclining toward Kemppinen's conclusions, yet never being wholly satisfied either with the nature of the evidence or the line of reasoning involved.
Briefly, the conclusions are these: That the ballad was not a retelling of the Judith and Holofernes story, but that its "false knight" is a supernatural being. That the poem thus consists of ideas and materials out of the European folkloristic background, and its false knight was the ancient demonic king of the "Wild Hunt" or "Wild Horde," often called Herlewin-of which the Halewijn of the Flemish version is a clear derivation. That the poem was composed in a territory of Germanic language and Celtic mythic background; this area, by reason of the completeness and clarity of the old Halewijn form, must have been along the lower Rhine. That the time of composition is shown by the 11th-century dating of the name Halevinus, the pre-12th-century vigor of the Herlewin tradition, the fact that round dance and accompanying ballad song had reached the Northern Countries by the i2th century, and the nature of the texts. Furthermore, that the ballad was composed between 1100 and 1200 and "was originally sung primarily as the song accompanying the round dance at the castles of the nobility, and by this time several different forms of the ballad had already appeared through the modifications of the story introduced by the minstrels (p. 265)." Kemppinen's evidence is used scientifically. The difficulties about the above listed conclusions are caused mainly by the failure to discuss certain perfectly entertainable possibilities about which we shall probably never know enough to decide on the real succession of facts; otherwise by insufficiency of evidence presented. Now to discuss these difficulties.
One of the most salient occurs to the reader's mind when Kemppinen enunciates his doctrine (which he calls "the present view") that in a folk poem the original author is traceable by his work (see p. 221). Certainly Kemppinen does not attempt the restoration of a single original text-that method which undid the work of Doncieux, as demonstrated by Coirault-but his mind is set firmly on the goal of finding what material the individual folk poet used. Here, of course, we run head-on into the circumstances of traditional poetic diction. What if the "original poet" or "author" of this ballad had used (entirely or mainly) formulized diction already conventional and in recognized, long-time use? The evidence of "textual analysis" would then become (as it has with Kemppinen) only a succession of events and recognizable narrative details; the first poet's work would be, perhaps, indistinguishable from that of the second or third; for the stylistic traits, being communal, could be of no use in disclosing the identity of individual folk poets, nor even (in the case of so long-lived and cross-cultural a ballad as this) in fixing the social background which must have been the original poet's.
For this is another of Kemppinen's preoccupations (see p. 222); he says that the ballad "illustrates the social background of the poet; [the scholar] can place the cultural milieu introduced by the ballad in its historical setting." Now a good deal of this social material cited by Kemppinen is psychological, consisting of traditional ideas and beliefs-e.g. the myth of the wild hunt and the belief in the seducing demoniac being, both of which were flourishing in popular tradition for long centuries both before and after the i2th century. How, then, can elements such as these be restrictive as to "historical setting" when the latter phrase connotes the approximate time of this ballad's composition? Other similar considerations to which this observation applies are the popularity of the harp and the belief in the compulsive power of fairy music; as well as the "old ritualistic circumstances" of headhunting, throwing bodies into water, and hanging them on trees (see p. 217).
Like others who have studied popular ballads, Kemppinen believes that a movement of a ballad from place of composition (center) to periphery of its area of currency is marked by increasing text-and-content differentiation (see p. 182). With regard to "Lady Isabel and the Elf Knight," however, this theory will hold only for a part of the evidence; the more westerly texts, especially, do not support it. Actually there seems to be here a clash of the principle just enunciated with another which is likewise widely demonstrable-namely, that in the products of oral folk art, the traits of peripheral versions are likely to be more antique than those of versions nearer the geographic center of distribution. The possibilities and implications of these two opposing tendencies seem not to have been thoroughly thrashed out in Kemppinen's study; but this is no place to set them forth. Rather, may it not be a fair question to ask: "In the diffusion of a piece like the present ballad, will not transmission among peoples speaking closely related forms of the same linguistic family help to preserve old forms-and contrariwise, will not the movement of the ballad into areas of different culture and more widely varying or structurally different languages assist differentiation to the point of producing new entities?" In any case, until a balancing decision can be reached concerning the effects of these two contrary tendencies, there can hardly be a satisfying use of the evidence toward determining either the original material or the place of origin of the present ballad.
Involved in the foregoing considerations there is yet another difficulty. When a study of this sort is read, the question of the extent of traditional theme substitution cannot but come to mind. If a ballad is carried from an area of stanzaic verse with rhyme or assonance (e.g. the Low Countries or Germany) to one of unrhymed, unstanzaic verse, in which the songs are in lines of about even length (e.g. Finland or Russia), the chances for the suppression of one favored traditional theme for another-hence for progressive alteration of a ballad's content-are greatly increased, since theme, formula, and traditional metre are all inseparably bound up with language. An accumulation of varying versions with marked thematic changes (such as we have in "Lady Isabel," between the Teutonic, Romance, and Finno-Esthonian forms) will increase the investigator's troubles immeasurably when he works, as here, with the aim of determining the homeland of a ballad, and also (to some extent) the routes and stages of its migrations.
This difficulty is the one seemingly encountered by Kemppinen in his analysis of the Finnish versions' evidence. It emerges in his conjecture that "the Herlewin tradition as a whole, on which the ballad is based, is... associated with the ancient Manala myth" (p. 264). Now in this conclusion about the folklore background of the piece, there are at least two assumptions whose implications can be recognized when they are named: (1) The false knight could not have become associated traditionally with Herlewin, but he must originally have been Herlewin himself; and (2) the relations with the Manala myth shown by the Finnish versions could not have been accretions due to the absorption of local thematic material; or else could not have been local substitution of a more familiar for a less familiar oral-narrative theme about a denizen of the other world (of faery, of the dead, etc.). In fact, throughout this study, the full possibilities inherent in the ousting of one oral theme by another in different versions of the "Lady Isabel" ballad seem to have been disregarded. The fact is recognized adequately, but its implications for the total history of a ballad everywhere subjected to traditional processes are not.
The foregoing remarks will perhaps explain why this reviewer is not convinced that "textual analysis leads us to approximately the 12th century." Actually, the only philological evidence adduced is that about the Herlewin-Halewijn relations. All the other textual analysis is not linguistic, but centered about the persistence and/or variation of four main diagnostic incidents, and is therefore not historic, but identificatory; hence dateless. In fact, when the alteration of this ballad's texts is considered-e.g. such a translation as that between western and Finnish traditional meters-the fact that the round dance and its accompanying ballad song had arrived in northern countries by the 12th century is not so important for dating as one might assume. I certainly agree with Kemppinen that this ballad is medieval; but the fact that it survives in western European tradition in a form claimed to have been evolved in the dance does not preclude the possibility either of an origin later than the 12th century, or of the refashioning of an earlier non-dancing narrative poem-in which case, the identifiable, basic narrative as we know it could be considerably older than the 12th century. The folklore material forming the spiritual world in which "Lady Isabel" is set does not militate against interpretation or speculation toward either increasing or lessening the age of the composition.
In a study that identifies a ballad and tries to fix its time and place of origin, there may or may not be a musical problem involved. In the case of "Lady Isabel," the obvious variety of the tunes set to versions renders the musical aspects of more or less local concern, and robs them of the power to throw light on origin or stages of migration from country to country. But the very diversity of music joined to texts of this piece should have been in itself an object lesson to Kemppinen. He considers only one tune-the wellknown Credo-derived melody of the Halewijn texts-and his discussion is exceedingly superficial. In short, he uses the tune to bolster his theory of the time and place of composition, without taking any of the truly musical factors of folk song into consideration. His dictum is that "the Catholic Mass was already known by the time the ballad originated; and it was used as a charm against evil spirits." This statement may well be correct; but what it seeks to demonstrate is an assumption pure and simple. A student who knows anything at all about the ways in which traditional tunes and texts are associated will appreciate that: (1) the relationship of this Halewijn melody to the Credo of the Mass may or may not have been recognized by the singing people of the Low Countries when the air was joined to the ballad; (2) the practice of folk singers is to set texts to well-known tunes to which they can easily be sung; and (3) a folk tune and text-version are not necessarily joined permanently, or even always consciously. It is extremely likely that this particular melody came to be associated with Halewijn simply and solely because it was a favorite folk air, and for no other reason. Also, its association may have been long subsequent to the original composition of the ballad, since Kemppinen notes that it apparently retained its popularity in the Dutch-Flemish area up to the present day (see p. 262).
Kemppinen is indulging in more assumption when he says (p. 265) that this ballad "was originally sung primarily as the song accompanying the round dance at the castles of the nobility, and by this time several different forms of the ballad had already appeared through the modifications of the story introduced by the minstrels." One can appreciate that a newly introduced dance-and-song custom may well have gotten to the north of Europe travelling from chateau to chateau; but are we compelled to believe that this was invariably its means of diffusion? Could it have spread among no other classes and in no other ways? We must recall that round dances are of prehistoric antiquity in the Mediterranean area, and not class-bound. Also, are we forced to assume that "Lady Isabel" was composed instantaneously, just as soon as the dancing-song fashion made its first appearance in the Low Countries? Could it not have been composed after this fashion had become familiar-hence not necessarily in association with either castles or nobility? Furthermore, if it were originally sung to the round dance at some nobleman's castle, how could several forms of it have "already" appeared "by this time"?
To assemble the versions of "Lady Isabel" Kemppinen casts his nets wide. In addition to pieces which he recognizes as cognates, he also discusses many related and apparently derivative songs. No one can afford to minimize the perplexities attendant on sifting out, through multiple, cumulative variations, the variant forms of an ancient international folk song. The ambiguity of some of this textual material cannot but be seen occasionally despite the keenness and care with which Kemppinen has arranged the numerous items. At the edges, so to speak, of this vast family are songs whose relation to it must remain conjectural; and also there seem to be modern pieces using the same ancient plot, with rationalized and modernized details. For example, there is "The Sea Captain" (mistakenly called "an American ballad," p. 250), not assigned to any place in this version-comparison scheme. But the ballad had already been claimed as a remake of "Lady Isabel" by W. Roy Mackenzie (Ballads and Sea Songs from Nova Scotia, 1928, p. 74); though Phillips Barry (Bulletin of the Folk-Song Society of the Northeast, No. 7, 1934, pp. 12, 13) in his discussion takes in a number of Nordic and Romance ballads emphasizing the supernatural powers of the maid instead of the man. Nevertheless, this ballad is very much like Kemppinen's Italian forms listed as RI 25 ff., p. 77 if., his French forms listed as FR 30 if., p. 71, and the French (Breton) form BF 10, p. 66.
Among the ostensible forms of this song in America, Kemppinen lists one version of "The Cruel Ship's Carpenter," a comparatively modern British broadside piece (p. 174). It is here classified as a "literary variant"; actually it is the only variant cited of a quite widespread ballad. However, my object here is not to dispute Kemppinen's attitude toward literary forms, which is perfectly sound ("Literary modifications cannot be accepted as evidence," p. 8), but to question whether the Anglo-American piece in question can really be regarded as a literary variant. It has become traditional and multiform, and one may legitimately distinguish between a popular broadside utilization of old folk themes and the work of a cultured literary man or scholar reinterpreting such elements. Actually, these late broadsheet ballads are often the only records we have for the existence in English of very old traditional narratives and motifs, and as concerns the English material, they often show no more drastic revision than the peripheral Baltic versions that Kemppinen discusses. They are, in fact, very much on a par with broadsheet copies of the old Scottish form of "Lady Isabel." Among them a series of late Anglo-American songs displays a wider spread of "maid is killed" forms of the story than would otherwise be suspected of the farther western sets of versions.
In short, if "The Cruel Ship's Carpenter" be regarded as one of the spalds of this ancient ballad, so should "The Wexford Girl" (WG-"Oxford Girl," "Bloody Miller," etc.), "The Box Upon Her Head" (BH), "Rose Connolly" (RC), and the American pieces "Poor Omie" (PO) and "The Banks of Ohio" (BO-Banks of Pedee, etc.), as well as "The Sea Captain" (SC) already mentioned. In SC and BH the maid escapes; in the others she is killed. Other parallel motifs are: seduction and/or abduction (all); murderer volunteers to guide maid across country (BH); abduction on promise of marriage (WG, PO, BO); going to waterside (WG, PO, RC, BO, SC); stabbing or beating and throwing into water (WG, RC-this adds poisoning as well-PO, BO); cries of victim (WG, PO); killing of would-be murderer (BH); lulling molester to sleep (SC-here maid uses captain's own sword to row her skiff to shore); murderer's clothes bloody (WG-excuse given here is bleeding at nose); noise of killing heard in lane or at gate (BH); apprehension and punishment of murderer (WG, RC, PO); maid becomes bride of rescuer (BH). One may add that the singing in SC is, or was, regarded as magic: see P. W. Joyce, Old Irish Folk Music and Songs, 1909, p. 152, where the girl is also regarded as a mermaid. Incidentally, the language, attitudes and social situation implicit in these much-modified forms of the story are themselves no inconsiderable pieces of cumulative evidence that Halewijn and its close relatives are truly medieval in time of composition.
I have felt constrained to criticize this study closely because it seems to me that certain important factors in the solution of the origin-problem of this (or any) traditional ballad have not been given the needed consideration in these pages; hence, that the solution is based in insufficient evidence, or that equivocal data have been handled with too much assurance. But this monograph contains many more excellencies, which space prohibits my going into. I must, however, mention that the reader and student are helped by a good person-and-subject index and an extensive bibliography; and the many (often beautiful) illustrations by artists of various lands add an unusual and welcome interest to a work of this kind.