The Child Ballad in Canada: A Survey

Canadian Journal for Traditional Music (1981)
The Child Ballad in Canada: A Survey
Laurel Doucette and Colin Quigley

While many folklorists might find the tabulation of Child ballad statistics a tedious and somewhat anachronistic exercise, few would deny the historical value of describing in detail the results of one phase of Canadian folksong research. It is our contention, however, that such an exercise serves deeper scholarly concerns. As an examination of a regional classical ballad repertoire, this survey provides a base for comparison with other regions. Given the broad geographic distribution and time depth represented by this genre, the Child ballads are well suited for such comparative study. Moreover, as the oldest level of the Anglo-Canadian song tradition, the national Child ballad repertoire forms an important base for folk-song study within this country. An examination of maintenance and loss of story types within the Child corpus, as well as analysis of transformations undergone by individual ballads, could reveal culturally-specific concerns which may be reflected throughout the entire Canadian folksong repertoire.

In addressing this topic, we have very little to go on other than the recorded corpus of material, for most song collections were made before the concept of "context" was introduced to folklore field research. The corpus, however, is substantial, and bears close examination in terms of its composition, distribution, and narrative style. Throughout this paper, when we use the term "ballad," we mean "Child ballad"; and we use the titles adopted by Child, followed by the number assigned by him.2

Ballad Research in Canada
The earliest record of a British traditional ballad in Canada dates to 1898 when Phillips Barry published a text of "The Gypsy Laddie" (Ch 200) collected in Nova Scotia.3 In 1910 and 1912, W. Roy Mackenzie published a total of 5 ballad texts from Nova Scotia.4 During this early period, F.W. Waugh and W.J. Wintemberg each published one ballad text from Ontario,5 and Marius Barbeau collected one from Quebec.6

It was not until the late 1920s that major collections containing substantial numbers of Child ballads began to appear. In 1928, Mackenzie published Ballads and Sea Songs of Nova Scotia (Boston: Harvard University Press), with 16 ballad types, some in several variants (including the texts earlier published). In the following year, Barry and his collaborators published 30 separate ballad types collected from Canadian informants, although the title of their publication, British Ballads from Maine (New Haven: Yale University Press) gives no hint that 20 per cent of the texts are from Canadian sources. Arthur Huff Fauset included 3 ballad texts in his Folklore from Nova Scotia (New York: American Folklore Society) in 1931. Within this period of field research, Helen Creighton published her first song collection;7 it contained 11 ballad types. In 1933, Ballads and Sea Songs of Newfoundland (Cambridge: Harvard University Press) appeared containing material collected by Elisabeth Bristol Greenleaf and Grace Yarrow Mansfield. Out of the 185 songs published, 19 are Child types. During trips made to Newfoundland in 1929 and 1930, Maud Karpeles collected 24 Child ballad types (Folk Songs from Newfoundland, London: Faber and Faber, 1971).

This early concentration of collecting activity in the Atlantic Provinces continued, although publishing lagged somewhat from the early thirties until the 1960s. Creighton did extensive work in every part of her home province, Nova Scotia, and in addition collected a large amount of material in southern New Brunswick, and somewhat less in Prince Edward Island. She produced several books and articles which contain a total of 45 ballad types.8 Her massive taped field collection, containing other Child variants, is deposited at the Canadian Centre for Folk Culture Studies. The Centre also holds tapes which Laura Boulton made during a collecting trip which produced 12 ballad types recorded in Nova Scotia. Another Nova Scotia collection from this period, Carrie Grover's . Heritage of Songs (privately printed by the Gould Academy, Bethel, ME), circulated as an undated mimeographed manuscript until it was reprinted by Norwood in 1973. Carrie Grover was an expatriate Nova Scotian whose childhood memories of song included 5 ballad types. An unpublished manuscript originating in Missouri, "New Versions and Variations of the Ballad of Barbara Allen Discovered 1959-1961," by Sherman Lee Pompey, contains 11 Canadian texts, two of which are from Nova Scotia.

Edward D. Ives published one ballad text from Prince Edward Island in 1963,9 and in 1968 Louise Manny and James Reginald ilson published Songs of Miramichi (Fredericton: Brunswick Press), with 5 Child ballads.

In Newfoundland, two major publications appeared during this decade. Kenneth Peacock published Songs of the Newfoundland Outports in 1965 (Ottawa: National Museum of Canada), based on fieldwork conducted between 1951 and 1961. Of 1020 published songs, 25 are Child ballad songs. In the same year, MacEdward Leach published Folk Ballads and Songs of the Lower Labrador Coast (Ottawa: National Museum of Man), including 8 Child types. Leach also conducted fieldwork on the Avalon Peninsula of Newfoundland during 1950 and 1951. In his collection deposited at the Memorial University of Newfoundland Folklore and Language Archive are 8 Child ballad types. Still more recent collections are also deposited at this archive. Those indexed to date contain 20 Child ballad types.

During the l950s, Newbell Niles Puckett and Edith Fowke were collecting in Ontario. The former recorded only 2 ballad types in the Kawartha Lakes region between 1956 and 1963.10 The latter, who ranged more widely over the province and into the neighbouring English-speaking areas of western Quebec, collected 21 ballad types, some in several versions, in Ontario alone. Most of these have been published in various articles and in the book, Traditional Singers and Songs from Ontario (Hatboro, PA: Folklore Associates, 1965). While collecting for the National Museum of Man in 1977, Laurel Doucette recorded 1 Child ballad in Gatineau County, Western Quebec, and in the following year, Sheldon Posen noted several in neighbouring Pontiac County.

While it might appear that central Canada has been well covered, at least in recent decades, in truth research has concentrated on just a few Ontario and Quebec counties, chiefly the former lumbering areas where a strong tradition of song was known to exist.

If research has been incomplete in the central provinces, it has been all but non-existent in most of the western ones. Maud Karpeles in 1929 noted 1 Child ballad in Saskatchewan.'1 The previously mentioned Pompey manuscript contained 9 versions of "Bonny Barbara Allan" (Ch 84) collected by correspondence from the 4 western provinces. A pair of researchers identified only as Pickering and Peterson, while on a research contract for the National Museum in 1967, recorded 1 ballad in British Columbia. In 1975, Edith Fowke, Barbara Cass-Beggs, and P.J. Thomas each published the text of, or noted the occurrence of 1 ballad from Manitoba, Saskatchewan, and British Columbia, respectively.'2 Two other Child pieces are included in the Thomas collection of British Columbia songs.

From this brief survey, it would appear that the collecting of song in general has been intensive and relatively early in Newfoundland and Nova Scotia, and in parts of New Brunswick and Prince Edward Island; selective and relatively late in central Canada; and very recent and totally haphazard in most of the West. The collected corpus of Child balladry in Canada represents only the minimum of what must have existed even as recently as the mid-point in this century. Indications are, however, that, given an equally intensive and continuous pattern of research, all of the provinces east of the Prairies would have proven as rich a field for British balladry as Nova Scotia; and that the western provinces, where British settlers formed a smaller percentage of the original colonists, would have provided a smaller but still substantial number of texts. The popularity of Child balladry across the country is indicated by the fact that the "Old Favourites" column of the Family Herald, a weekly farm newspaper published in Montreal, printed a total of 22 ballad types in 85 printings between 1895 and 1968, partly in response to requests from readers.13

The Ballad Corpus
The corpus of ballad texts to be discussed has been gathered from publications of field collections, from archival sources, and from personal communications with collectors and archivists. Every recorded incidence of a Child ballad type or variant was noted, whether or not a text was printed or recorded. Thus, if a collector published 1 text of a ballad story, noting that 3 other informants knew the same ballad, 4 variants were counted. Similarly, if a collector published a composite text from 3 informants, 3 variants were counted. This method of tabulation provides a clearer picture of the distribution of specific ballad types than would a mere computation of published texts. No distinction is made between complete texts and fragmentary ones. Apart from the fact that the distinction is difficult in many cases (what is a complete text to a Canadian singer may appear deficient in contrast to a Child text), the presence of even a brief fragment indicates that the song once existed in oral circulation in this country. Nevertheless, the presence of a text does not necessarily indicate that the ballad has had wide circulation here over a long period, or indeed that it has had any circulation at all, for some ballads were collected from immigrant informants, who perhaps had never sung their songs in Canada.

The corpus of Child ballads actually recorded or observed in circulation in Canada consists of 77 ballad types in 586 versions. Among the most popular titles, there are few surprises, for the national pattern conforms to that established by other scholars for the United States.'4 "Bonny Barbara Allan" (Ch 84) ranks first with 61 occurrences. "Lady Isabel and the Elf-Knight" (Ch 4) is in second place, having been recorded 31 times. Third place is held by "Little Musgrave and Lady Barnard" (Ch 81) with 27 occurrences. "Lord Randal" (Ch 12), "The Cruel Mother" (Ch 20), and "The Gypsy Laddie" (Ch 200) share fourth place with 26 occurrences each. However, only 8 recordings of "Lord Randal" were in English. The other 18 were recorded in Acadian areas of New Brunswick and Quebec, and, according to Robert Paquin, represent a direct translation from English to French, rather than an importation of a European variant from France.15 Two ballad types were reported 25 times: "Young Beichan" (Ch 53), and "The Sweet Trinity" (Ch 286). "Sweet William's Ghost" (Ch 77) and "The Farmer's Curst Wife" (Ch 278) were each noted 21 times. However, all 21 versions of the former are from Newfoundland, where it is the most popular of all the Child pieces, even outranking "Bonny Barbara Allan."

"Hind Horn" (Ch 17) was reported 17 times, "Captain Wedderburn's Courtship" (Ch 46) 16 times, and "Sir James the Rose" (Ch 213) 15 times. Thirteen occurrences of "Katharine Jaffray" (Ch 221) and "Our Goodman" (Ch 274) were noted, 12 of "Willie o' Winsbury" (Ch 100), and 11 of "Lamkin" (Ch 93). Out of the total 77 ballad types located in Canada, the above 17 were the only ones reported over 10 times. Of these titles, 6 do not appear on the Abrahams and Foss list of Child types commonly found in North American traditional repertoires:16 "Sweet William's Ghost," "Hind Horn," "Captain Wedderburn's Courtship," "Sir James the Rose," "Katharine Jaffray," and "Willie o' Winsbury." Moreover, these authors list 4 types which are not found in Canadian collections at all: "Riddles Wisely Expounded" (Ch 1), "Edward" (Ch 13), "Sir Lionel" (Ch 18), and "The Maid Freed From the Gallows" (Ch 95).

Other types found in Canadian collections are as follows:

Reported 9 times — "Lord Thomas and Fair Annet" (Ch 73). Reported 8 times — "False Knight Upon the Road" (Ch3; one of these recordings is in French); "The Three Ravens" (Ch 26); "The Bailiff's Daughter of Islington" (Ch 105); "The Grey Cock, or, Saw You My Father?" (Ch 248). Reported 7 times — "Fair Margaret and Sweet William" (Ch 74); "The Baffled Knight" (Ch 112); "Mary Hamilton" (Ch 173); "Geordie" (Ch 209); "Henry Martyn" (Ch 250 or 167). Reported 6 times — "Babylon" (Ch 14); "The Twa Brothers" (Ch 49); "The Cherry Tree Carol" (Ch 54); "The Unquiet Grave" (Ch 78); "The Mermaid" (Ch 289). Reported 5 times — "Johnie Scot" (Ch 99). Reported 4 times — "The Elfin Knight" (Ch 2); "Earl Brand" (Ch 7); "Young Hunting" (Ch 68); "The Knight and the Shepherd's Daughter" (Ch 110); "The Bold Pedlar and Robin Hood" (Ch 132); "James Harris" (The Daemon Lover) (Ch 243); "Get Up and Bar the Door" (Ch 275). Reported 3 times — "Lord Lovel" (Ch 75); "Robin Hood's Progress to Nottingham" (Ch 139); "The Bonnie House o Airlie" (Ch 199); "The Braes o Yarrow" (Ch 214); "The Suffolk Miracle" (Ch 272); "The Wife Wrapt in Wether's Skin" (Ch 277); "John of Hazelgreen" (Ch 293). Reported twice — "The Twa Sisters" (Ch 10); "Lady Alice" (Ch 85); "Young Johnstone" (Ch 88); "Robin Hood and Little John" (Ch 125); "Sir Hugh, or, The Jew's Daughter" (Ch 155); "Bonnie George Campbell" (Ch 210); "Andrew Lammie" (Ch 233); "The Laird o Drum" (Ch 236); "The Jolly Beggar" (Ch 279); "Captain Ward and the Rainbow" (Ch 287). Reported once — "King John and the Bishop" (Ch 45); "Sir Patrick Spens" (Ch 58); "Fair Annie" (Ch 62); "The Lass of Roch Royal" (Ch 76); "The Wife of Usher's Well" (Ch 79); "Child Maurice" (Ch 83); "The Famous Flower of Serving

Men" (Ch 106); "Johnie Cock" (Ch 114); "Robin Hood and the Prince of Aragon" (Ch 129); "Robin Hood Rescuing Three Squires" (Ch 140); "Willie Macintosh" (Ch 183); "Rare Willie Drowned in Yarrow" (Ch 215); "The Broom of Cowdenknows" (Ch 217); "Rob Roy" (Ch 225); "Lizie Lindsay" (Ch 226); "The Rantin Laddie" (Ch 240); "King Edward the Fourth and a Tanner of Tamworth" (Ch 273); "The Keach i the Creel" (Ch 281); "The Brown Girl" (Ch 295); "The Trooper and the Maid" (Ch 299).

This corpus appears to be a rather miscellaneous selection of Child ballad types, a fact which makes it difficult to sort these pieces into meaningful categories in order to understand their appeal. If we divide these types into the general categories suggested by Buchan,17 we find 42 titles which could be classified as Romantic and Tragic, 16 Historical and Semi-historical (including those of legendary characters), and 7 Magical and Marvellous (including those primarily of the supernatural world). In addition there are 7 comic ballads (practically Child's entire stock), 4 riddling ballads, and 1 religious ballad.

However, Buchan's categorizing of individual ballad types may be somewhat less than appropriate within the North American context. We might question, for example, to what extent Canadians, given a limited knowledge of British history, would recognize "Mary Hamilton" as an Historical ballad. Rather might they not consider all such tales of long ago to be equally historical, perceiving no sharp distinction between stories recounting events which actually happened in history (and which are therefore documented), and stories recounting the kinds of events which regularly do happen in human life. The ballads of an historical or semi-historical nature collected in Newfoundland, "Henry Martyn," "The Sweet Trinity," and "Captain Ward and the Rainbow," would seem to illustrate the same point. All of theseballads concern sea-going heroes. Since the stories have no historical significance in Newfoundland, they would seem to appeal as tales of adventure.

Similarly, we might query the use of the category "Magical and Marvellous" in relation to any traditional song corpus, since a community which has preserved old ballads is likely to have also preserved an acceptance of supernatural occurrence as an ordinary part of everyday life. Hence a story of the return of a dead lover or child might be viewed as romantic and/or tragic, but not as magical or marvellous. This category presents a second problem as well, because these elements can be found in a number of story types. The revenant motif in particular is common in Newfoundland ballads, as is the dream motif in which contact is made with the "other world" in less corporeal form. The riddling ballads as recorded in that province also reflect the widespread presence of supernatural motifs.

An additional problem is encountered in attempting to make any categorization of Canadian versions of Child ballads. Many of the versions collected here are considerably shorter than the Child texts. In some cases they have lost the very elements by which they might have been classified: supernatural motifs, or specific historical details. In the light of these considerations, it might prove more useful to view the Canadian corpus in terms of the reasons for its appeal, rather than to attempt to classify individual ballad types. On this basis, it would seem that romance, adventure, and tragedy are the chief reasons for the appeal of a majority of ballad stories.

Among the minor categories, the comic and religious ballads have an appeal of their own, while the riddling ballads represent a romantic encounter (as with "The Elfin Knight" and "Captain Wedderburn's Courtship") or a humorous deception ("King John and the Bishop"). Another of the riddling ballads, "The False Knight Upon the Road," presents a phenomenon which must be considered when attempting to understand the popularity of any specific ballad type in Canada. The earliest version collected by Helen Creighton features the usual question and answer format involving a little child and the false knight, or devil. But the melody is jig-like, and the singer dances to a refrain of mouth music.'8 In this version at least, the ballad has moved far away in tone from Child's encounter between child and devil, and has come closer in spirit to a humorous devil-human encounter such as that featured in "The Farmer's Curst Wife."

Any definitive categorizing of the Canadian corpus would involve a major project of comparing and correlating variants, with a consideration of melody as well as text. For the present, we can conclude that romantic and tragic elements in a ballad story do much to ensure its popularity, as does sheer entertainment appeal; in this light, the ballads are no different from the general Canadian repertoire.

Provincial Distribution
While the irregular pattern of song collecting makes it very difficult to come to any conclusions about provincial distribution, a few trends are noticeable. Of the 77 ballad types which have been traced in Canada, 47 have been noted in Nova Scotia, a not surprising fact when one considers that collecting has been more or less continuous there since W. Roy Mackenzie first began fieldwork in the early years of the century. Newfoundland, as might be expected, comes second in terms of total ballad types with 44. New Brunswick ranks third with 38, many of which date to the fieldwork of Phillips Barry in the 1920s. However, Ontario, a province where intensive collecting began as late as the 1950s and even then only in select areas, has produced a total of 25 ballad types, 4 of which have never been reported from the other provinces. In addition, the adjacent, once predominantly English-speaking Quebec counties of Gatineau and Pontiac, which are low in population density although vast in area, have produced a total of 9 types. From figures such as these and from the ratio of ballad types to total versions reported, it would appear that the old stereotype of the Atlantic region as the best collecting ground for Child ballads should be reconsidered. For example, it is possible that in Nova Scotia, with an aggregate of 47 types in a total of 235 reported versions, the provincial repertoire of Child ballads has been almost completely gleaned. In other provinces, where the ratio of types to incidence is much closer (38 to 80 in New Brunswick, 9 to 17 in Quebec, 25 to 46 in Ontario), we can presume that a body of balladry comparable to that of Nova Scotia has gone unrecorded. Other details of provincial distribution can be seen in the chart on page 10.

The Ballad Singers
It was W. Roy Mackenzie's belief that ballad singers were "of an entirely different race from other singers."19 Perhaps this was true in his area of fieldwork during the first two decades of the century, but Mackenzie's limited experience of performance within a community context invites scepticism. The facts of the various collections suggest just the opposite: that the ballads were sung by very typical singers of the communities where collecting was done. Many of the ballad singers had substantial repertoires of traditional song, of which the Child ballads formed only a small segment. Ben Henneberry of Devil's Island, Nova Scotia, one of Helen Creighton's most prolific informants, supplied her with 90 songs, only 9 of which were Child ballads.20 Another, Angelo Dornan of Elgin, New Brunswick, recalled 135 songs, only 3 of which were Child types.2' Edith Fowke considered O.J. Abbot her best informant, but of his 124-song repertoire, only 2 were Child ballads.22 Two or three would appear to be the average number of Child ballads in a typical repertoire.

While Ives has stated that "most of the Child ballads have been collected from women in the Northeast,"23 this is not true of Canadian collections, although it might pertain to those of New England. Ballad singing appears to be more or less evenly distributed between the sexes, with men rather than women having a slight edge, possibly because men may have been more accessible as informants in some collecting situations.

The ethnicity of ballad singers has received attention from some collectors. Mackenzie was the first to note that those of Scottish ancestry had ceased to sing the old ballads (by which he meant both Child and broadside ballads) by the early years of this century.24 He attributes the change primarily to the religious fanaticism of the latter decades of the nineteenth century, but also to changing social conditions in Nova Scotian communities. When French-speaking Swiss Hugenots adopted the language and songs of their Scots neighbours (often learned while serving as hired help in a Scottish household), the Scots were presented with an additional reason for dropping the old songs, that of maintaining their superior social status.

Edith Fowke has mentioned a similar phenomenon in Ontario in her discussion of the predominating Irish quality of the song tradition there:

Most of the songs come from the Irish Catholics rather than the Ulstermen who have also played a prominent part in Ontario's history. Just as Dr. Mackenzie noted that in Nova Scotia there came a time when the Scottish Protestants put aside their old ballads as fit only for the ungodly, so the straitlaced English and Scotch Protestants, who set their imprint on Ontario's pioneer communities, preferred hymns to worldly songs. Fortunately the Catholics had no such prejudices, so we have the somewhat anomalous situation that Canada's most Protestant province, long the stronghold of the United Empire Loyalists, the Daughters of the Empire, and the Orangemen, is also a treasure house of Irish ballads.25
While much research remains to be done on the cultural consequences of the nineteenth-century religious reforms which changed the shape of Canadian Protestantism in general and Canadian Presbyterianism in particular, it is interesting to note that while English-speaking Lowland Scots were discarding their "worldly" songs and ballads, the Catholic Irish were picking them up as a replacement for the Gaelic culture they had dropped shortly after their arrival in Canada. To understand the full implications of such cultural changes and exchanges, the folklorist must await full descriptions of migration and assimilation patterns by social historians. Meanwhile we can assume that the religious convictions of the Protestant Lowland Scots in much of Canada account in large measure for the short supply of Scottish ballad variants ma country which has seen massive Scottish immigration for the past two hundred years. This fact is particularly telling when we consider how heavily Child himself relied on Scottish sources for his collection. Even in Nova Scotia, the most Scottish of Canadian provinces, English names vastly outnumber Scottish ones on informant lists.

Ballad Content and Style
Two scholars have already devoted considerable attention to variation over time in the traditional British ballad. Buchan, in outlining the change in northeastern Scotland from an oral culture to a literate one, has presented a model for the study of any ballad text of the "verbal" rather than "oral" period.26 Buchan views the diverse transformations in ballad content, structure, and style as being derived from the social and cultural change undergone by a specific society at a specific point in history, and from the creative response of the singers to such changes. Coffin's explanation for change is based on the transfer of materials from the Old World to the New, and the concomitant passage of time.27 His description of variation in the ballad emphasizes the kinds of devolutionary change which occur after a society has entered the literate or verbal phase, in which singers depend upon memory rather than recomposition in their performance.

Within the Canadian corpus, we can find ample evidence to support the statements of both Buchan and Coffin concerning transformation in later ballad versions. We can also find remarkable evidence of preservation of elements of the older oral style. And in most texts, the old and the new are mixed.

This blending of elements can best be illustrated by a brief glance at a fairly random selection of ballads: the 11 texts in Creighton's first book, Songs and Ballads from Nova Scotia (pp. 1-24). In comparing these with the various Child versions, we can find a broad variety of features representing both the old and the new, or as some scholars would have it, "purity" and "contamination."

"The False Knight Upon the Road" has the same question-and-answer format as Child's two versions, but the questions here are those found in "Riddles Wisely Expounded," and a refrain of mouth music which is as long as the stanza has been added.

Only Child's E version of "The Cruel Mother" is longer than the Nova Scotia version, a very complete telling of the story. Contrary to Coffin's suggestion, not only is one marvellous element retained here, the return of dead children, but another is added in the form of a murder weapon which cannot be rubbed clean of blood and cannot be thrown away (stanzas 8 and 9, p. 4). The story is effectively told in two major acts, the crime and the retribution, with use of a firm structure of triads and balances, and a dramatic concentration on the mother-child relationship.

"Captain Wedderburn's Courtship," on the other hand, exhibits a lesser degree of preservation of the ballad form and style. While retaining a riddle format, it opens like a broadside ("As I rode out one May morning"), and features stanzas varying from 6 to 8 lines in length.

The Nova Scotia version of "Lord Thomas and Fair Annet" is a shorter 13-stanza variant of Child's D text, but it is just as effective in telling the story. All of the essential details have been retained, but the action here moves more swiftly. Dialogue is used to convey much of the plot, and such features of oral re-creation as annular framing and balanced structure are retained.

Only 3 1/2 stanzas of "Little Musgrave and Lady Barnard" could be recalled by the singer. As it stands, the fragment illustrates how memory loss can bring about retention of only the climax of a story (in this case, the confrontation between the wronged husband and his wife, with their subsequent deaths).

Creighton's variant of "The Bold Pedlar and Robin Hood" is very similar to Child's only text. However, the Nova Scotia version has dropped one stanza and picked up another from Child 142 ("Little John a Begging"). The added stanza has been altered to end the ballad with a moralizing message which has nothing to do with the previous action.

A second Robin Hood ballad, "Robin Hood's Progress to Nottingham," survives only as a 3-stanza fragment containing the climax of a confrontation between the hero and his enemies, the foresters and citizens of Nottingham. This version illustrates the same sort of decay seen above in connection with Child 81.

In "Sir Hugh, or, The Jew's Daughter," the magical element of the dead boy speaking to his mother from the well is lost, not because of rationalization, but because the singer forgot the second half of the ballad. We find an example here of a commonplace formula used in such a way that it makes no sense at all: "It rains in bower and ha'." (stanza 1, p. 16)

Creighton's 12-stanza version of "The Farmer's Curst Wife" is longer than either of Child's texts and contains details from both of them. It retains the whistling ballad burden of the Sussex variant, and the mouth-music refrain of the Scottish one, and is in every way as complete and amusing as the older versions.

"The Sweet Trinity (The Golden Vanity)" is fairly similar to Child's C version, with the loss of a few details, and the change from a Turkish to a Russian enemy. Both exhibit an irregular stanza pattern.

Finally, Creighton includes a version of "Katharine Jaffray," although she admits it is rather a secondary ballad similar to one which Child had collected from an Irishwoman in Massachusetts, and had presumed to have been derived from a broadside (p. 24). Its details are similar to those of the various Child versions, and the final moralizing stanza is adopted from his I text. In structure, however, the ballad is unsophisticated, with the story being recounted in linear fashion, with little use of framing structures and balanced stanzas.

One could pass through the whole repertoire of ballad texts collected in Canada and make similar observations. Some texts show signs of decay that would be obvious even to one not familiar with the original ballad plot: parts of the story are missing, the form is so irregular that rhythm is destroyed, concluding stanzas have nothing to do with the preceding action, individual lines defy comprehension, and so on. But there are other texts which exhibit many of the structural and stylistic details of the best of the British ballads, retaining all plot elements and perhaps even adding some. And there are also texts which tell their story in a linear, broadside-like fashion, but do it with such economy of language that the effect produced is one of great poetic strength and beauty.

It is obvious that a brief survey of the Canadian corpus of traditional balladry can produce no definitive conclusions concerning the nature of the repertoire, for the variety of style, structure, and content is great. However, a few general statements concerning the ballad corpus can be made with some certainty.

Rhyme, or at least exact rhyme, is not used as extensively in Canadian texts as in those of Child, although this is not always immediately noticeable because the ballads are usually very rhythmic. Moreover, the rhythm is sometimes emphasized by the addition of a refrain of mouth music, which relieves the severity of the ballad story but also creates a dramatic contrast with it. We have seen this in the conversation between the devil and the little child in "The False Knight Upon the Road." It also appears in an Ontario version of "The Gypsy Laddie" collected by Fowke, where the heartbreak of the abandoned husband is both relieved and highlighted by the gay refrain:

Laddy fal the dinko dinko day, Laddy fal the dinko daisy.28
A second feature which occurs in a minority of texts but is nevertheless very striking is a highly dramatic opening stanza, brought about by the loss of the introductory section of the ballad. The audience is thus catapulted right into the action at a crucial point in the plot, and must surmise the emotional situation from the few details which are given. While the ballad may thus lose in completeness of detail, it gains in concentration of attention on emotional, rather than social or cultural, realities, and the impact so created can be stunning. A Maritime version of "Hind Horn" begins

"Where did you get it, by sea or by land,
Or did you get it off of a drownded man's hand?"29
Similarly, an Ontario version of "The Lass of Roch Royal" opens with dialogue:

"Oh, mother, I had a warning dream, Oh, mother, I had a dream.
I dreamed the bonny lass of Lochland Lane, She was tapping to get in."30
A Lunenburg County, Nova Scotia, variant of "Little Musgrave and Lady Barnard," which was complete to the singer although only 5 stanzas long, begins:

He took his bugle to his mouth
And blew it loud and strong,
He that's in bed with another man's wife
It's time that he was gone.3'

Many of the texts collected by Child exhibit similar opening stanzas, of course, but we tend to view these versions as incomplete, having lost an introductory scene or scenes. When such variants are perceived as separate entities, however, the emotional impact of the text becomes obvious, and we can imagine the effect produced upon a traditional audience which knows no other version of the story.

Another notable feature of Canadian texts is their language. To a person schooled in the texts of Child, they represent an impingement of everyday Canadian speech on the language of classical balladry. But to the tradition-oriented Canadian audience, they would sound just right, a proper blend of the poetic (the language of broadside and grammar school recitation, the diction of the "Old Favourites" page) with the ordinary. Thus the abandoned husband of Child 200 ("The Gypsy Laddie") is still aristocratic, but in giving orders to the servant sounds much more like an Ontario farmer than an English lord:

"Go saddle me my old gray steed;
The bay is not so speedy."32

Katharine Jaffray of Child 221 spurns her intended husband, a "rough spoken man," and rides off from the "wedding house" to Edinburgh Castle with her beloved farmer's son.33 This kind of change which serves to bring the ballad into the realm of the listener's experience is also noticeable in Canadian titles. "Hind Horn" is more regularly "The Old Beggar Man."34 "Lady Isabel and the Elf-Knight" becomes "Pretty Polly" in honour of the heroine; "The Dapherd Grey" in reference to the horse ridden by the villain; or "Doors of Ivory," alluding to the expensive cage promised to the pet bird as a reward for keeping secret the death of the false lover.35

Conclusion
These few comments on Canadian Child ballad texts should suggest what a ripe field this is for further investigation. A more intensive examination of various Canadian texts, coupled with a more detailed comparison with versions from elsewhere, would undoubtedly produce insights into the stylistic and structural features of this ballad corpus. Such analysis could also be expected to reveal some of the singers' cultural values and the underlying belief systems which support these values. As the oldest level of English language traditional song in Canada, the Child corpus could then be related to the broader repertoire of narrative song, with a view to investigating the role of classical British balladry in shaping the Canadian song tradition.

Memorial University of Newfoundland
St. John's, Nfld.

Footnotes
1 This paper, originally prepared for a course on classical British balladry directed by David Buchan, incorporates material treated more fully in "The Child Ballads as Found in Newfoundland: A Survey" by Colin Qaigley (Culture & Tradition, 5(1980), 16—32). We wish to express our appreciation to Dr. Buchan, Edith Fowke, Lucien Ouellet Sheldon Posen, Peter Narváez, Neil Rosenberg, P.J. Thomas, and to the Canadian Centre for Folk Culture Studies and the Memorial University of Newfoundland Folklore and Language Archive.

2 Francis James Child, The English and Scottish Popular Ballads (5 vols. Boston: Houghton, Mifflin, 1882-1898; rpl. New York: Dover, 1965).

3 Phillips Barry, "Traditional Ballads in New England 2," Journal of American Folklore. 18(1898), 191—224.

4 W. Roy Mackenzie, "Three Ballads from Nova Scotia," Journal of American Folklore, 23(1910), 371—380; "Ballads from Nova Scotia," Journal ofAmerican Folklore, 25(1912), 182—87.

5 F.W. Waugh, "Canadian Folk-Lore from Ontario," Journal of American Folklore, 31(1918), 74; W.J. Wintemberg, "Lord Lovel and Lady Nancy: A Traditional Ballad," Transactions of the Royal Society of Canada, 13:2(1919), 19—36.

6 Marius Barbeau, Arthur Lismer, and Arthur Bourinot, Come A Singing! Canadian Folk Songs (Ottawa: National Museum of Canada, 1943), p. 19. This same publication contains a version of "The Elfin Knight" (Ch 2) collected in British Columbia in 1920 (p. 33).

7 Helen Creighton, Songs and Ballads from Nova Scotia (Toronto: J.M. Dent, 1932; rpt. New York: Dover, 1966).

8 Major song publications from this period are Helen Creighton, Folksongs from Southern New Brunswick (Ottawa: National Museum of Man, 1971), and Maritime Folk Songs (Toronto: Ryerson, 1962); and Helen Creighton and Doreen Senior, Traditional Songs from Nova Scotia (Toronlo: Ryerson, 1950).

9 Edward D. Ives, "Twenty-One Folksongs from Prince Edward Island," Northeast Folklore, 5(1963), 1-87.

10 See Laurel Doucette, "An Introduction to the Pucketl Collection of Ontario Folklore," Canadian Folk Music Journal, 3(1975), 22—29.

11 Maud Karpeles, "British Folk Songs from Canada," Journal of the Folk Song Society, 8(1930), 229.

12 Edith Fowke, "Songs of a Manitoba Family," Canadian Folk Music Journal, 3(1975), 35—46; Barbara Cass-Beggs, "Barbara Allen," Come All Ye. 4(1975), 205—7; P.J. Thomas, "British-Canadian Folk Music in B.C.," Come Alt Ye, 4(1975), 210—14.

13 Edith Fowke, "'Old Favourites': A Selective Index," Canadian Folk Music Journal, 7(1979), 29—56.

14 Tristram P. Coffin, The British Traditional Ballad in North America (rev. ed. Roger deV. Renwick; Austin: Univ. of Texas Press, 1977).

15 Robert Paquin, "Le Teslament du garcon empoisonné Un 'Lord Randal' francais en Acadie," Canadian Folk Music Journal, 7(1979), 3—17.

16 Roger D. Abrahams and George Foss, Anglo-A Pnerican Folksong Style (Eaglewood Cliffs, NJ: PrenliceHall, 1968), pp.92—93.

17 David Buchan, The Ballad and the Folk (London: Routledge and Kegsn Paul, 1972), p. 74.

18 Creighton, Songs and Ballads from Nova Scotia, pp. 1-2.

19 W. Roy Mackenzie, The Quest of the Ballad (Princeton, NJ: Princeton Univ. Press, 1919), p. 107.

20 Creighton, Songs and Ballads from Nova Scotia. p. xiv.

21 Creighton, Folksongs from Southern New Brunswick, p. 1.

22 Powke, Traditional Singers and Songsfrom Ontario, pp. 12—13.

23 Edward D. Ives, "Lumbercamp Singing and the Two Traditions," Canadian Folk Music Journal, 5(1977), 23.

24 W. Roy Mackenzie, "Ballad-Singing in Nova Scotia," Journal ofAmerican Folklore, 22(1909), 327—31.

25 Fowke, Traditional Singers and Songsfrom Ontario, pp. 2—3.

26 Buchan, pp. 227—43.

27 Coffin, pp. 1—19.

28 Fowke, Traditional Singers and Songs from Ontario, p. 19.

29 Creighton, Maritime Folk Songs, p. 5.

30 Fowke, Traditional Singers and Songsfcom Ontario, p. 106.

31 Helen Creighton, Folklore of Lunenburg County (Ottawa: National Museam of Man, 1950; cpt. Toronto: McGraw-Hill Ryerson, 1976), p. 79.

32 Fowke, Traditional Singers and Songsfrom Ontario, p. 19.

33 Creighton, Songs and Balladsfrom Nova Scotia, pp. 22—24.

34 Manay and Wilson, p.206; and Fowke, Traditional Singers and Songsfrom Ontario, p.80.

35 Fauset, p. 109; Fowke, Traditional Singers and Songs from Ontario, p. 102; Manny and Wilson, p. 202.

Résumé: Laurel Doucette et Colin Quigley ontfait un étude des anciennes ballades Anglaises et Ecossaises (Child) que l'on trouve au Canada et ont préparé un index démontrant lafréquenCe des différents items. Ils ont constate que les ballades dramatiques et romantiques sont les plus populaires, particulièrement celles qui mettent en relief les questions morales et ils suggérent que les etudes subséquentes puissent établir une relation entre le repertoire et sesfonctions sociales.