An Oral Canon for the Child Ballads: Construction and Application
by J. Barre Toelken
Journal of the Folklore Institute, Vol. 4, No. 1 (Jun., 1967), pp. 75-101
An Oral Canon for the Child Ballads: Construction and Application
J. BARRE TOELKEN
Ever since Francis P. Magoun first enunciated his application to Anglo-Saxon poetry of the oral-formulaic theories developed by Milman Parry, literary scholars - especially medievalists - have taken part in a lively controversy over the existence, nature, and function of the formula in oral prosody.[1] At this late date there seems to be little, if any, scholarly objection to the concept itself; rather, the discussants have centered their attention on the identification of formulas and on their significance to
the poems in which they are observed to occur. It is not the aim of this paper to augment the theoretical aspects of formula study in any way except to point out and discuss a rather well-defined problem area which requires, but has not received, adequate folkloristic examination.
The leading formulists have in general drawn their data from the characteristics of oral epic poetry observed and recorded among illiterate Yugoslavian singers; their subsequent theories have been applied to such diverse materials in print as Greek epic and Anglo-Saxon narrative poetry. Thus, the extant oral poetry of one language and culture in modern times has been used as a lens through which to study the prosodic peculiarities in the presumed oral poetry - now extant only in print - of
other times, other places, other languages.[2] More lately, an attempt has been made to apply the theories of Parry and Lord to the British traditional ballad, a genre whose oral status is at least ambiguous.[3]
In all of these cases the conclusions drawn by a variety of talented and perceptive scholars may or may not be valid, but certainly the evidence adduced militates against the incautious acceptance of many of their unqualified assertions. We can do nothing about that gap in time which separates the modern critic from the primitive poet, or about that gap in attitude which divides the literate scholar from the illiterate (more often in modern times, aliterate) singer; but evidence and criteria can be found within the culture and language of the poetry under discussion, be it Greek or English.It is reasonable to suggest, on these grounds, that any generalizations about oral poetry in the English language must take into account, among other things, the English traditional ballad, for it is the one sizeable body of oral poetic narrative in our language. Yet studies of the ballad as
oral poetry run the risk of being beset by the same limitations found in the recent examinations of Anglo-Saxon poetry, those attendant to the extrinsic imposition of formulaic criteria upon the genre. In the case of Old English studies (echoed by Jones in "Commonplace and Memorization ..."), the argument runs somewhat as follows: (a) Yugoslavian epic poetry is oral; (b) it is made up chiefly of formulas and themes; (c) all oral poetry is therefore characterized by use of formula and theme; (d) Beowulf (the ballad, etc.) contains many formulas and themes; (e) therefore, Beowulf (the ballad) is oral poetry; (f) and corollary to this, Beowulf (the ballad) must have been composed in the same manner as was the Yugoslavian epic. Space does not here permit me to comment on each of the logical fallacies in this sequence, and perhaps to begin with I have misrepresented the argument somewhat by oversimplifying it.
Suffice it to say that Magoun's unqualified pronouncement that "Oral
poetry ... is composed entirely of formulas, while lettered poetry is never
formulaic,"[4] has been followed by a plethora of articles showing principally
by a formula count that this or that early poem must have been
oral. Since no one sings Beowulf any more, perhaps the scholar has no
recourse but to count formulas, but the solidity of the theory is oftentimes
upset, as it is in the case of the Anglo-Saxon poem, Phoenix,
probably the richest in formulas of any Old English poem, in which the
poet declares that he is composing with pen in hand. fn short, the theory,
as it pertains to oral poetry in English, lacks a certain amount of validity,
for it assumes, rather than demonstrates, that formula and theme play
the same role in English oral poetry as they do in the Yugoslavian epic.
Clearly, our observations on oral poetry would be considerably more
accurate and solid if we could determine first which poems were oral
(in composition or in transmission, or both) and then study those poems
to see what their poetic devices were and how they were applied. As I
have admitted, this is out of the question with Beowulf, but it is not out
of the question in the case of the ballads. Further, since the ballads are
in a language descended from Anglo-Saxon, it is distinctly within tlhe
realms of possibility that observations on the ballad prosody may afford
some valuable approaches to the older poetic traditions.
In the particular instance of the ballads, however, folklorists and literary
critics alike have generally assumed that "the ballad" is a poetic
phenomenon which can be treated in a single sweep. Most accept the
unsupported notion that all the ballads in the Child collection are (or
were) traditional, i.e., orally transmitted, and most subscribe to the
theory that the ballads passed orally from one singer to another, intact,
by means of a process which utilizes memorization. Jones, in the article
referred to above, has brought this latter point into question, but only
to substitute its direct opposite: total formulaicity. Bruce A. Beatie
has more recently suggested the possibility of two kinds of "oral traditional"
poetry, on the grounds that one can observe formula in action
4 Magoun, "Oral-Formulaic Character," p. 449.
77
J. BARRE TOELKEN
here, memorization in action there.5 Obviously, in the case of extant
oral poetry in English, we have both kinds, for children's jump-rope
rhymes, for example, and adult obscene songs and ballads have remained
very stable in transmission while the "Child" ballad has always tended
toward textual fluidity.6 There are, then, complicated aspects of ballad
study for which formula counting (or any other exclusive and unqualified
approach) offers little elucidation. Instead, it seems to me, the oral
formulist and the folklorist must begin with a collection of ballads which
are, or which have been, in oral tradition and work from there. The
development of such a corpus is a particularly fitting (and necessary)
task for the folklorist, for, as Arthur K. Moore has pointed out, most
people in the literary world who deal with the ballad in classroom and
in journal are untrained in folklore and rather have gravitated to the
belief that traditional ballads are those most commonly anthologized,7
even though these in not a few instances have undergone patent literary
"adjustments." It is therefore to the task of constructing a basic oral
canon from the Francis James Child collection of English and Scottish
ballads that this paper addresses itself.
Assuming that Child construed the term "popular" to mean something
akin to Grundtvig's folkevise ("folk-ballad"), scholars have generally
spoken of his type, as did Gerould, as the "Ballad of Tradition"; that is
to say, in spite of what D. K. Wilgus has aptly called "The Ballad War,"
there has been rather general agreement that the Child ballads fall more
appropriately into the province of folklore than into the area, of say,
eighteenth-century literature. However, even though Thelma James and
others have at various times called one aspect or another of the Child
collection into question, there has never been, it seems to me, adequate
examination of the traditional status of each item presented by Child.
Of course, Child himself implied doubt of the 'traditional authenticity of
some variants, listing them by lower case letters instead of capitals; in
other instances, he admitted using a broadside text on the strength that it
might have been the only remnant of an older traditional ballad. As the
present study will attempt to show, however, the number of unreliable
texts stretches far beyond Child's admissions or suspicions: fewer than
Op. cit.
6 For this observation on the stability of texts in the oral poetry of children and in
the obscene songs of adults, I am much indebted to Ed Cray.
7 Arthur K. Moore, "The Literary Status of the English Popular Ballad," Comparative
Literature, X (1958), 1-20.
78
AN ORAL CANON FOR THE CHILD BALLADS
half of his collected ballads may be in fact treated as clearly traditional.8
Most teachers of folklore require as a very basic necessity and a minimum
essential some notation by their students concerning source and
informant, not because a tale or a ballad must be proven "folk" before
it can be enjoyed, but because it must be proven traditional before it
can be fruitfully studied as folklore. To state this in terms of any other
academic field would border on the ridiculous: "The medievalist must
first determine that a given poem is indeed medieval and indeed a poem
before he can soundly treat it as a medieval poem." But scores of studies
done on "the Child ballad" assume tacitly that all 305 are ballads, and
that they are all traditional; moreover, they apparently accept comments
by Child which would be sure to fail a student in Folklore I. Such notes
as "from the recitation of an old woman," "procured by David Webster,
Bookseller, from tradition," "taken down from recitation in the North
of Scotland," "copied from the mouth of a milkmaid," and "sung by
an aged nurse, once resident in the north," which Child provides, mostly
reprinted from his printed sources, are so vague and hence inadequate
for the folklorist that they invite not only scholarly distrust, but suspicion,
especially since many other ballads in his collection appear with complete
and verifiable references. One feels not that Child was trying to defraud
his readers at all; rather, the lack of scholarly concern for these important
details during Child's era simply renders his notes comparatively valueless
for scholars in ours. Thus, for scholarly purposes at least, it is not
possible, or desirable, to assume in the lack of contradictory evidence
that the songs were all in fact in oral tradition, especially when many
of them came to print in an age when a premium was placed on collections
of "rude" poetry, and when ballads copied from the mouths of nurses
in the North were in vogue among buyers of books in the South. Moreover,
"evidence" of any sort concerning ballad authenticity is exceedingly
rare, and what might confidently be called contradictory evidence -
proof that a ballad was NOT in oral tradition - would be impossible to
come by. Thus, while we cannot point to any particular ballad and say
with certainty that it was not in oral circulation, positive evidence for
some ballads can be found and studied.
8 It should be made clear here that "unreliable"is not meant to imply "esthetically
bad," but only that - for better or for worse - some ballad texts in Child's collection
were or are in oral tradition and may thus be dealt with as folklore. The remainder
are simply without firm credentials.
79
J. BARRE TOELKEN
Since oral transmission, in the case of the British ballads, has usually
assured that an item will be anonymous and variable, those texts may
be eliminated from consideration here which retain a definite connection
with a specific personality and which exist only in one extant printed
version. Reliable collection from oral tradition will of course be of
paramount importance, although dependence on recently collected
materials risks the development of a canon of those ballads which chance
to have survived into modern times. This is a necessary risk, for at least
it allows for the establishment of a group of texts considerably more
reliable than that which is now available. Further, the stress on modern
collection may be tempered by reference to the printed works and studies
of those sholars in the past whose seriousness and rigor have matched
their enthusiasm for publishing. Most importantly, over-all, since Child's
notes are inadequate to the aims of this paper, the principal testimony
on behalf of anyballad must come from sources outside the Child
collection itself.
I propose, then, that oral status, or traditional reliability if you will,
be accorded any Child ballad meeting one or more of the following
requirements:
1. Existence of two or more English language variants collected from
oral tradition outside England by reputable folklorists who give the
details (e.g., name of informant, place, date, etc.) of collection. To
simplify the matter of finding reliable texts in North America I have consulted
as the principal work for this requirement Coffin's The British
Traditional Ballad in North America,9 for in it is combined all previous
scholarship on collected ballad materials in the New World. Coffin lists
all Child ballads which have been collected reliably in North America,
plus those which have been reported only, and those which have been
found only in fragments, discussing each entry critically and fully
enough so that in each case it is quite clear whether the ballad is or
was in actual oral circulation. Thus, in those cases where a ballad is
listed which was collected from a person who learned it from a book
or a broadside, the present study has been able to eliminate it from consideration
as authentically oral at the present time.10 It is interesting
9 Philadelphia, 1950; revised and enlarged, 1963.
,0 Such a distinction does not mean that a given ballad in broadside circulation will
not eventually enter the stream of oral transmission and be collected some years from
now from an oral informant. Neither does this elimination suggest that such a ballad
80
AN ORAL CANON FOR THE CHILD BALLADS
and significant to note that in Coffin's 1963 revised edition there are few
additions to his 1950 listing, though in some cases the editor's hesitance
about a particular fragment has been dispelled by the discovery of an
independent oral text. In the main it would seem that those Child ballads
which did come over from England via oral transmission have all been
found; new versions are continually making their appearance, as one
might expect, but it is probably safe to conclude that Coffin's list will not
be appreciably enlarged by future collectors. After applying the above
strictures on printed texts to Coffin's entries, there are ninety-five Child
ballads listed by him which may be seen as reliably observed in oral
tradition in North America.11
II. Existence of two or more demonstrably traditional tunes which are
connected with the ballad in question. Obviously there is a problem in
showing, especially from printed texts, the demonstrably traditional
nature of any tune. Such investigation requires extreme musical competency,
and as far as I can determine, only Professor Bertrand H.
Bronson has been able to combine musical and literary talents in the
production of reliable data on the ballad tunes for the Child collection
to the general satisfaction of ballad scholars. Even though the ability
to detect clearly authentic folk materials in musical texts (especially
printed ones) is at best a hybrid process combining musical skill, necromancy
and extrasensory perception, Bronson's The Traditional Tunes
of the Child Ballads is the only work I have consulted for data on this
requirement, for in it he has brought together all previously published
tunes for the ballads and all available materials from private collections.
Further, his musical competence is such that in most cases he is able to
identify the mode and style of a given tune so clearly that any commercial
or "sophisticated" material is almost immediately recognizable. Since
is not now actually in oral circulation, waiting in some corner of Buncombe County
to be collected. It assures us only that on the basis of texts now available we are discriminating
between ballads learned and passed orally and ballads learned and passed
in printed sources.
11 I have eliminated from consideration a few ballads in Coffin's list which, though
apparently collected from an oral informant, exist only in a single fragmentary text.
For example, Miss Creighton, in Songs and Ballads from Nova Scotia (Toronto, 1933),
p. 15, gives a fragmentary version of Child 139, "Robin Hood's Progress to Nottingham,"
as sung by a Mr. Ben Henneberry. Without derogation to Miss Creighton as
eithers cholaro r collector, I feel that until more is known of Mr. Henneberry'sp osition
in an oral train of transmission, or until other supporting evidence can be found,
we are justified in exercising great caution in calling items like this authentic.
81
J. BARRE TOELKEN
Bronson studies all available ballad tunes, his items often coincide with
the textual findings of Coffin (who lists by text, the music being commonly
lacking in printed collections) and of Alexander Keith (below). In these
cases, testimony provided by Bronson will only serve to reinforce that
drawn from these other sources. However, since Bronson has had
access to numerous musical texts not generally available to scholars, it
seems advantageous to list in the table below all his items which are
relevant to the present study.
Some strictures on Bronson's listings are in order, however. First of
all, it should be noted that for the same reason that one text is not enough
to establish the fact of oral tradition, one tune, in the absence of other
evidence, cannot be enough. For a similar reason, tunes collected from
only one singer cannot, without further support, admit a ballad to the
oral canon. Tunes must further be either certifiably traditional (by
structure and style) or must have been collected from oral tradition
directly by someone whose competence is not in question. That is to
say, the mere existence of a ballad tune in print is no assurance that it
must have been traditionally sung, for it might have been composed by
a trained musician to accompany a printed text from broadside. Lastly,
many early collectors (and a few modern ones) have ignored tunes altogether
in the publishing of their findings. For this reason, while the
presence of a tune at least may mean the ballad was sung, the absence
of a tune in print does not indicate the ballad was not sung. Thus tunes
can lend only positive evidence, except in those cases where a tune with
definite connections with a ballad is clearly commercial.
In his first two volumes, Bronson lists ninety-six items, ninety-three
of which are given as Child ballads, the rest as illuminating appendices.
Of these ninety-three entries, four are given without tunes, and have thus
been eliminated from consideration under requirement [I.12 Nineteen
other entries are eliminated, as follows: #6 (connection of tune with
text is dubious; Mrs. Brown of Falkland is only source), 16 (tune not
clearly traditional), #22 (Bronson: "... until other traditional variants
of the ballad are disclosed, we may reserve judgment as to the authenticity
of this one."), #27 (not clearly traditional), #31 (variant given is an
analogue; there is no evidence of actual connection with the ballad, or
of development from it through variation), #32 (Mrs. Brown is only
source), #33 (data inconclusive), #34 (Mrs. Brown is only source;
12 70, 71, 87, 91.
82
AN ORAL CANON FOR THE CHILD BALLADS
connection of tune with text is uncertain), #38 (Bronson doubts tune
was ever sung traditionally), #42 (Mrs. Brown is only source; connection
of tune with text uncertain),l3 #57 (no definite connection with the
Child ballad; Buchan's MS the only authority for tune), #66 (connection
of tune with any known text of the ballad is unlikely), #72 (no certain
connection of tune with text), #86 (uncertain connection between tune
given and this ballad), #92 (another ballad is given, close only in topic),
#94 (no proof that it was traditionally sung, Bronson admits), #101
(Mrs. Brown's tune not metrically suited to her text, and many of her
words are themselves "artificial, or modern." Christie's tune is treated as
unreliable because he leaves unclear whether he himself mated tune to
text; he admits "arranging" the tune.), #102 (unreliable testimony;
American version was collected from a traditional singer, but one who
had access to Buchan's text, probably through Kittredge's edition of
Child), #103 (the only authentic part of this version actually belongs
to "Leesome Brand," #15. Tunes reproduced by Bronson have greater
kinship with other ballads).
Bronson, then, provides positive data which may be used to authenticate
only seventy of the 113 ballads in the first two volumes of Child,
or roughly sixty-two per cent. Of these, twenty are not found in America,14
and thus the ninety-five ballads certified by Coffin are expanded by that
number. In publishing these first two adjuncts to Child's volumes
(I and II), Bronson estimates he has used up fifty to sixty per cent of his
total materials associated with volumes I through V; in short, he has
used more than half of the available musical evidence to authenticate
slightly over twenty-two per cent of the total Child canon. This leads
me to suspect that when his entire study is finished he will have supplied us
with data to certify musically not more than 150 of Child's 305 ballads;
the figure may actually be much smaller, considering that a goodly
portion of his remaining materials will naturally be connected with
several widely popular ballads, such as "Sir Hugh," "Mary Hamilton,"
13 A subtle but significant distinction militates for the exclusion of those ballads
for which Mrs. Brown is the only source. If she had been illiterate, her versions -
even though unique - would not be brought into question, for she would perforce have
gotten them from oral tradition. But she was quite literate, and often sang, as Bronson
tells us, with the text (sometimes a corrected one at that) in hand. Thus she was not in
the strictest sense a traditional singer, and in the lack of other evidence, we are justified
in not leaning too heavily on her versions as reliable documents of oral tradition.
1" 5, 9, 19, 24, 25, 39, 41, 44, 47, 52, 55, 61, 64, 67, 69, 89, 96, 97, 98, 113.
83
J. BARRE TOELKEN
"The Gypsy Laddie," "James Harris," "Our Goodman," and "The
Farmer's Curst Wife," all of which we already know have numerous
versions in tradition.
III. Existence of ballad versions in England - even when not found
elsewhere in the English-speaking world - if present in variation and
collected by reputable scholars from oral tradition. In establishing the
reliability of British texts, however, a serious problem arises: as Wilgus
has noted,15 the ballad collections made in England are not generally
trustworthy for use by folklorists. Heavy-handed collation of texts,
bowdlerization, extension, emendation, etc., by such scholars as the
Reverend Sabine Baring-Gould, Cecil Sharp and others make unreliable
the use of such texts in any study which is dependent on soundly
traditional materials. The fact that a Child ballad may appear in their
pages is thus of primary value to our curiosity, for when we find in the
same collection an instance where an entire set of new words has been supplied
for some "charming" traditional air, we are entitled to professional
skepticism concerning the whole lot.16 British collections are furthermore
sparse and most of them are dedicated - as was Cecil Sharp himself-
to the study and preservation of tunes more than texts. Of all
British collections, Wilgus singles out but one which is scholarly and
reasonably accurate: Alexander Keith's edition of Gavin Greig's Aberdeenshire
ballads, Last Leaves of Traditional Ballads and Ballad Airs
(Aberdeen, 1925). Even here there is more collation than the folklorist
would wish, but we are more certain with this collection that the notes are
scrupulous and complete. Thus while many of the versions printed by
Keith do not lend themselves to a close textual analysis as documents of
oral tradition, most of his items were taken originally from tradition and
can therefore help to certify that a given ballad was in oral circulation.
Moreover, traditional ballads seem to have been in lively circulation in
Scotland later than in England, and Bronson notes (II, 37) that for this
reason Child's best texts for more than a third of his collection came from
northern Scotland, particularly Aberdeenshire, the area in which Greig
did the bulk of his collecting.
In all, Keith lists 102 Child ballads from the collections of Greig. Of
15 Anglo-American Folksong Scholarship Since 1898 (New Brunswick, New Jersey,
1959), pp. 125-143.
16 See James Reeves' perceptive study of the Sharp MSS, The Idiom of the People
(New York, 1958).
84
AN ORAL CANON FOR THE CHILD BALLADS
these, fifteen are questionable,17 one because it is taken directly from a
chapbook (#237), some because they are given in indecisive fragmentary
form, others because they come from only one informant. The latter
observation is of particular importance to Keith's collection, for the
informant most mentioned as a single source is Miss Bell Robertson, who
did not sing at all, but recited her ballads. Some of them she claimed to
have learned from the singing of her mother, but others she learned from
young ladies of her acquaintance, which would have increased the
possibility that some of her materials came through "popular" or commercially
printed channels. Again, this cannot establish that these
ballads were not in oral tradition, but it eliminates the certainty that
Miss Robertson's knowledge of them was traditional, and hence places
her in question as a reliable informant in these cases. Of the eightyseven
reliable ballads presented by Keith, thirty-one are not found in
North America (according to Coffin's listing). Of these, eleven are dealt
with by Bronson. Thus the number of ballads certifiable as oral under
sections I and II (above) must be expanded by twenty to accommodate
these ballads which apparently did not leave England in traditional
circulation, and for which Bronson can offer no data.
Of these three sections, it seems to me that requirement one is the
most significant. To find a ballad circulating orally outside England
would seem to be the sharpest sort of proof that it has been in the
traditional domain for some time, for the settlers of this and other
English-speaking countries were primarily those among whom the folklore
processes were most likely to be in operation. This consideration
will not, of course, explain the existence of those broadsides which may
have been brought over on paper and which slipped later into tradition;
in those cases we may simply go further into the credentials of the singer
and collector before according any particular text traditional status. It
is certainly apparent that the other English-speaking areas of the world
must be as effectively canvassed as has the United States before we can
know the full story of ballad exports and survivals. It is my hope that
studies on the Child ballad in Australia, New Zealand, India, South
Africa, and so forth, will supplement the figures 1 have presented in this
prolegomenon.
On the basis of present information, then, a total of 135 Child ballads
can be seen as certifiably oral (ninety-five by Coffin, twenty more by
17 71, 103, 104, 134, 138, 191, 192, 196, 237, 241, 252, 267, 269, 288, 294.
85
J. BARRE TOELKEN
Bronson, twenty more by Greig/Keith). Hopefully this figure will be
expanded somewhat by future studies, although Bronson's calculations,
as noted above, indicate that we may as well be pessimistic about authenticating
more than roughly half of Child's items finally. My own persistent
search - by no means completed - through the journals and bulletins
published here and abroad has thus far failed to add a single ballad
to the above number.
Even so, the list of solidly oral Child ballads presented below can represent
only a reliable body of texts with which the folklorist and critic of
oral poetry may confidently work. It can in no way pretend to be a
complete listing of oral balladry in English, or even in Child, for many
of the excluded ballads do have Scandinavian and other European ballad
analogues. As Archer Taylor and Holger Nygard have shown so vividly,
some English ballads which seem to have started in England later rose
to prominence elsewhere, while others collected mostly in England are
directly traceable to other countries and languages. A ballad may have
died out after it came into English, or after it gave birth to foreign
progeny, but its oral nature must remain conjectural unless there is
separate testimony of at least one or two variants under 1, II, or III above.
In any case, it would be fruitless to study a Danish ballad for examples
of English folklore solely on the indication that it may once have been
orally transmitted in English. And the existence of a tale analogue
asserts only that the tale is known; it does not, as Bronson has observed,
allow an assumption that it was ever in ballad form.
Although it is not the aim of the present study, the foregoing comments
quite clearly bring into question those studies which have concluded
with various generalizations about the Child ballad as folklore. In
particular, one thinks of Professor Wimberly's examination of folklore
in the ballads, in which he deals without distinction with the folk motifs
and superstitions revealed in those ballads that are folklore and those
that are not, as well as those about which we can say neither. As valuable
and rich as his study is for the student of the ballad, his findings must
surely be modified or at least re-examined in light of the strictures suggested
above; at least we may suspect that items of folklore as they
appear in a questionable or palpably literary ballad have quite another
significance than those which appear in their natural habitat in an authentic
folk ballad. And Gerould's fine study: how accurately does it
actually represent the ballad of tradition?
86-
AN ORAL CANON FOR THE CHILD BALLADS
The very texts in Child's collection must also come under closer
scrutiny, for even though we may be able to authenticate a ballad, a
particular text in Child may be completely unreliable to stand as representative.
I am not at all sure how the high probability of editorial
meddling in Child's texts may be compensated for by the modern scholar,
except by leaning more heavily on variants recently recorded from oral
tradition than on the beautifully blurred but questionable texts in Child.
If we are English teachers looking for striking examples of narrative
poetry, naturally our favorites in the anthology will do handsomely; if
we are folklorists looking for traditional data, then we must be more
particular in where we look when using Child.
Those ballads for which Child presents several or many versions are
nearly all in the following list of oral ballads. In the non-oral category,18
most ballads appear in only one version in Child, some in as many as
three; two of them have five versions, one has six, one has seven, one has
eight. This observation is generally illustrative of the characteristic
mentioned earlier: that truly oral ballads are usually found in variation,
while materials which circulate chiefly in print tend to be copied quite
closely. That any single ballad receives small notice in Child is of little
consequence, however; "Bonny Barbara Allen" (#84), one of the most
widespread of all English ballads (Bronson alone gives 198 tunes),
appears in only three versions in Child. It is for this reason, as well as for
those mentioned and implied above, that the criteria for an oral canon
must be based on data independent of the Child collection.
A few further statistical observations on the data derived from the
above investigation are in order. Of the 170 non-oral ballads:
a) Nearly fifty can be classed as Border Ballads, the exact number
dependent on the way that class is construed.
b) Thirty-five are Robin Hood ballads, most of which were taken by
Child from broadside sources. With more careful studies of the Robin
Hood ballads this number will most likely be reduced by a few; in any
case, we are not obliged to believe, on the basis of reference in Piers
Plowman to "rymes of Robin Hood," that early forms must have been
either ballads or traditional.18 It cannot be stressed too strongly that the term "non-oral" is used here to denote
those ballads for which there is no conclusive evidence of oral circulation; it is meant to
be descriptive rather than restrictive, for the canon as presented is unavoidably
inclusive rather than exclusive. Neither should the term "non-oral" be construed as
a valuej udgmento n the literaryo r musicalc haracteristicso f any ballad.
87
J. BARRE TOELKEN
c) Eighteen exist only in a single Buchan manuscript or in collections
derivative of Buchan.
d) Nineteen exist only in a single Percy manuscript or in collections
derivative of Percy.
e) Of the five ballads for which Child's only source was Scott's
Minstrelsy, all are non-oral.19
f) The one ballad (#267) for which Percy and Buchan are the only
sources is non-oral; the one ballad (#195) for which Percy and Scott
are the only sources is non-oral; the one ballad (#86) for which Scott
and Buchan are the only sources is non-oral.
The following table lists the Child ballads which may safely and confidently
be treated as oral. In each case the basis of their certification
will be indicated by reference to sections I (North American versions as
established by Coffin), II (traditional tunes as established by Bronson,
through Vol. II of Child), Ila (traditional tunes collected by the author
from oral tradition, for Vols. III-V of Child), and III (English versions
as established by Greig/Keith). It is significant, I believe, that in most
cases more than one of these requirements will be met.
Child No. Title Basis
1 Riddles Wisely Expounded 1, II
2 The Elfin Knight 1, I,II
3 The False Knight Upon the Road I, 11
4 Lady Isabel and the Elf Knight I, 11, III
5 Gil Brenton II
6 Willie's Lady III
7 Earl Brand 1, 1, 111
9 The Fair Flower of Northumberland II, III
10 The Twa Sisters 1, II, III
11 The Cruel Brother I, 11
12 Lord Randall 1, 11, III
13 Edward 1, 11
14 Babylon 1, 1I, II1
15 Leesome Brand III
17 Hind Horn I,11, lII
18 Sir Lionel 1, 11
19 King Orfeo II
20 The Cruel Mother I, II, III
24 Bonnie Annie II, 11
25 Willie's Lyke-Wake II, III
19 186, 190, 202, 205, 206.
88
AN ORAL CANON FOR THE CHILD BALLADS
Child No. Title Basis
26 The Three Ravens I, II
33 Kempy Kay III
37 Thomas Rymer I, II
38 The Wee Wee Man I
39 Tam Lin II, III
40 The Queen of Elfland's Nourice 1, II
41 Hind Etin II, III
42 Clerk Colvill I
43 The Broomfield Hill , II, III
44 The Twa Magicians 11, III
45 King John and the Bishop I, II
46 Captain Wedderburn's Courtship 1, II, III
47 Proud Lady Margaret II, III
49 The Twa Brothers I, II
51 Lizie Wan I, II
52 The King's Dochter Lady Jane II, III
53 Young Beichan I, II, III
Volume If
54 The Cherry Tree Carol I, II, III
55 The Carnal and the Crane II
56 Dives and Lazarus I, II
58 Sir Patrick Spens 1, II
61 Sir Cawline II
62 Fair Annie I, II, III
63 Child Waters I, II, III
64 Fair Janet II
65 Lady Maisry , II
67 Glasgerion II
68 Young Hunting I, II
69 Clerk Saunders II
73 Lord Thomas and Fair Annet , II, III
74 Fair Margaret and Sweet William I, II
75 Lord Lovel I, II, III
76 The Lass of Roch Royal I, II, III
77 Sweet William's Ghost I, II
78 The Unquiet Grave I, II
79 The Wife of Usher's Well I, II
81 Little Musgrave and Lady Barnard I, II
83 Child Maurice 1, II, III
84 Bonny Barbara Allen I, II, III
85 Lady Alice I, II
87 Prince Robert I
89
J. BARRE TOELKEN
Child No. Title Basis
88 Young Johnstone I, If
89 Fause Foodrage I1
90 Jellon Grame I, II, lII
93 Lamkin 1, lI, III
95 The Maid Freed From the Gallows I, II
96 The Gay Goshawk II
97 Brown Robin II
98 Brown Adam 1I, III
99 Johnie Scot i, II, III
100 Willie o Winsbury I, II, III
105 The Bailiff's Daughter of Islington I, II, III
106 The Famous Flower of Serving-Men I, 1, III
110 The Knight and the Shepherd's Daughter I, 11, III
112 The Baffled Knight I, I, III
113 The Great Silkie of Sule Skerry II
Volume III
114 Johnie Cock I, Iil
132 The Bold Pedlar and Robin Hood
140 Robin Hood Rescuing Three Squires I, III
155 Sir Hugh 1, Ila
156 Queen Eleanor's Confession I, 1Ia, III
162 The Hunting of the Cheviot l
163 The Battle of Harlaw III
164 King Henry Fifth's Conquest of France I
167 Sir Andrew Barton I, IIa
170 The Death of Queen Jane I, III
173 Mary Hamilton I, IIa, III
178 Captain Car III
182 The Laird o Logie IlI
188 Archie o Cawfield I, III
Volume IV
199 The Bonnie House o Airlie I, III
200 The Gypsy Laddie I, IIa, III
201 Bessy Bell and Mary Gray I, III
209 Geordie I, IIa, III
210 Bonnie James Campbell I
212 The Duke of Athole's Nurse II
213 Sir James the Rose I, III
214 The Braes o Yarrow I, IIa, III
215 Rare Willie Drowned in Yarrow I, IIa, III
90
AN ORAL CANON FOR THE CHILD BALLADS
Child No. Title Basis
216 The Mother's Malison III
217 The Broom of Cowdenknows I, III
218 The False Lover Won Back I, III
219 The Gardener III
221 Katharine Jaffray I, III
222 Bonny Baby Livingston III
226 Lizie Lindsay I, III
228 Glasgow Peggie III
231 The Earl of Errol III
232 Richie Story III
233 Andrew Lammie 1, II
235 The Earl of Aboyne IIl
236 The Laird o Drum I, III
238 Glenlogie III
240 The Rantin Laddie I, III
243 James Harris (The Daemon Lover) I, IIa
245 Young Allan III
248 The Grey Cock I
251 Lang Johnny More III
264 The White Fisher III
Volume V
272 The Suffolk Miracle I
274 Our Goodman I, IIa, III
275 Get Up and Bar the Door 1, III
277 The Wife Wrapt in Wether's Skin I, IIa, III
278 The Farmer's Curst Wife I, IIa
279 The Jolly Beggar I, III
280 The Beggar-Laddie III
281 Keach i the Creel I, III
282 Jock the Leg and the Merry Merchant III
283 The Crafty Farmer I, III
285 The George Aloe and the Sweepstake I
286 The Sweet Trinity (The Golden Vanity) I, IIa, III
287 Captain Ward and the Rainbow I, Ill
289 The Mermaid 1, III
293 John of Hazelgreen I, III
299 Trooper and Maid I, III
It goes without saying that for the requirements discussed above I am
strictly indebted to, and seriously dependent upon, the three general
91
J. BARRE TOELKEN
reference works cited. To the extent that they are accurate, and to the
extent that my own judgment on inclusions is sound, the above listing
comprises a basic and dependable canon. I have avoided the use of
materials which seemed to me incomplete, or on which I was not able
to exercise informed judgment. For example, I have no way to authenticate
or to question the many recent recordings of ballads and songs
made in the British Isles by such people as Jean Ritchie, Alan Lomax,
Seamus Ennis and others. Most of these records are a joy for the folklorist
to possess and use in class, and 1 would not for a moment question
the reliability of the collectors themselves. I feel less confident about the
singers, some of whom sound quite polished and sophisticated in their
delivery. What are their credentials as links of the traditional process?
Is there a distinction between Mr. Ennis as traditional singer and Mr.
Ennis as collector? Are Mr. McColl's numbers on the big Riverside
collection all from tradition or were some set to music to help fill the
record? One hopes that these and similar questions may have favorable -
or at least conclusive - answers, but until someone can clear this data
for us (someone, say, like Kenneth Goldstein, who knows the records
and knows his folklore), section III (English versions), above, must
remain dependent on the meager support of Greig/Keith.
The above 135 oral ballads constitute roughly forty-four per cent of
the total Child canon. Of the 170 remaining ballads, 153 have no American
or British ballad variants at all outside Child and his sources. Most
of the 135 oral ballads fall toward the beginning of the collection, the
ratio of oral to total ballads in each volume being: I - 37/53; II - 39/60;
III - 14/75; IV - 29/77; V - 16/40. Most of the non-oral ballads are
toward the end and in the Robin Hood category, where Child himself
admitted he was stretching matters to include texts. This clear indication
that the beginning of Child's collection, particularly the first two
volumes, is the richest in oral ballads is further reinforced by the following
figures:
a) Of the 113 ballads in Volumes I and II, seventy-six (sixty-eight
per cent) are oral; of the 192 ballads in Volumes III to V, only fifty-nine
(thirty per cent) are oral.
b) Of the 113 ballads in Volumes I and 1II, forty-five per cent exist in
American versions, while only twenty-five per cent of the 192 ballads in
Volumes III to V are found in American tradition.
c) In Volumes I and II, twenty-eight ballads (twenty-five per cent)
92
AN ORAL CANON FOR THE CHILD BALLADS
have no American or European ballad variants outside Child, while in
Volumes III to V, 125 (sixty-five per cent) have none.
The common ballad image of a girl sewing (often a silken seam)
actually appears in only eighteen of Child's ballads; of these, fourteen
are classifiable as oral according to the criteria set forth above. The
remaining four may be accounted for in a number of ways: (a) they
may be ballads which were at one time oral and have left no traces on
the basis of which we might authenticate them; (b) they may be ballads
made up by literate ballad writers drawing some of their images from
folklore; (c) they may be broadsides in which certain phrases were used
in an attempt to be appropriately folksy. Of these fourteen oral ballads,
twelve appear in Volumes I and II of Child. Other ballad "formulas"
seem to follow roughly the same pattern of distribution in Child's
volumes (though it has yet to be demonstrated that the formulas commonly
found in the oral ballads are metrical equivalents - a theorem
characteristic of the Anglo-Saxon formulaic studies).Perhaps more important to the strictly poetic elements of ballad study
is the observation that the usage, meaning, and rate of occurrence of
folklore motifs differ rather sharply between the oral and non-oral groups.
To take a simple example first, thirty-two oral ballads make reference
to the color green - especially in clothing - compared to thirteen nonoral
ballads which do so. Of the thirty-two which make this reference,
the use and/or appearance of the color is followed by death or misfortune
in thirty cases. Since there are no doubt several oral ballads classified
perforce with the non-oral ballads due to a lack of evidence on their
behalf, and since the motif here referred to appears among the thirty-two
oral ballads in sixty-seven separate versions while it is mentioned in
only one version of each non-oral ballad, the actual difference is even
more striking than it might initially appear. Clearly, the color green
had a viable figurative utility in oral tradition which was never wholly
perceived - or adopted - by the more sophisticated ballad writers.
Ten Child ballads20 utilize the image of a plant (flower, nut, fruit)
being plucked or cut in contexts of seduction; apparently the action
functions as a metaphorical parallel to, or dramatic antecedent to, the
act of deflowering which takes place in the main action of the ballad.
This relationship is reinforced in one of these ballads by the direct
statement of the young girl so deprived:
20 13, 14, 39, 41, 43, 52, 110, 112, 155, 199.
93
But he's robbed me o my maidenheid,
The flower o my bodie.21
Since these ten ballads are all oral, we may speculate that the poetic
elements they hold in common might well constitute a body of traditional
references shared by oral singers and their audiences. A close examination
does, in fact, reveal a singular coherence in the way this motif is presented
in the ballads. In seven versions of "Babylon" (14) now available to me,
the sisters go out to pick some flowers by the bank of a river. Immediately
they are accosted and attacked by a robber who, it later is found, is their
brother. In "Tam Lin" (39), Janet pulls or picks flowers in eleven of
Child's versions. In all cases where the ballad is not fragmentary, she
is approached at once by Tam Lin, a fairy changeling, who then seduces
her. In "Hind Etin" (41), the girl pulls a nut in versions A and B, whereupon
Hind Etin appears and seduces her. "The King's Daughter Lady
Jean" (52) features a heroine who in version A pulls the nut and sloe, in
version B picks a nut, and in C and D picks a flower; in all cases a young
man appears who promptly seduces her. Picking nuts is also followed
by seduction in version D of "Gil Brenton" (5), but the ballad is not
counted here among those whose versions typically employ the motif.
Bronson lists four versions of "The Baffled Knight" (112), in which
a maid escapes seduction by pretending to go along with the young man
in his design to "go and pick the plums," or grapes; she tells him to
climb the plum tree until her father passes, after which she will join him.
When he is safely up the tree, she says (in version 37):
You may eat your grapes and suck your stems
For I am going to the house, sir.
And he later reminisces:
And every time she looks at me and smiles,
It makes me think of climbing.22
21 "The Knight and the Shepherd's Daughter," 110 (1). Cf., in addition to a number
of versions giving essentially this wording, "But he's robbd me of the flowery branch,/
The flower of my bodie" (H), "He has robbed me of my maidenhead,/ The fairest
flower of my bodie" (E).
22 It should be noted here that these verses are from "Katy Mory," a ballad included
by Bronson under his materials for Child 112 because of the identical context of
thwarted seduction, though with the admission that "Katy Mory" is so distant in
wordinga nd tune "as to be scarcelye ntitled to claim relationship"( II, 547). Kenneth
Goldstein has asserted to me in private his belief that the two are entirely distinct. In
either case, the usage of the motif is our central consideration: either it is used in
distant versions of 112, or it is used in versions of the ballad "Katy Mory." Other
94 J. BARRE TOELKEN
AN ORAL CANON FOR THE CHILD BALLADS
In "The Bonnie House o Airlie" (199), Argyle, the attacker, insists that
the lady of the castle show him her dowry. She takes him to the water
side where her dowry is found hanging in "the fair plumb tree." He then
rapes her while her castle suffers an analogous fate at the hands of his men.
Fruits are evidently used as the tools of seduction, or at least of enticement,
in "Sir Hugh" (155); although it transpires the villainess is
not bent on sexual seduction, she uses traditional means (i.e., apples,
cherries, a fig, sugar) to lure the young Sir Hugh into a position where
she can kill him.
The symbolic reference to picking or pulling a flower is strengthened
in many cases by mention of the name of the flower. For example, in
several versions of "Tam Lin" the girl pulls a rose (sometimes a double
rose), a flower which has for several reasons a special connection with
women in general and with virginity in particular. The terms "rose"
and "rosebud" have long been associated with the woman, the latter
still occasionally in use in the newspapers to denote a young girl in her
prime, a debutante. Partridge, in his Dictionary of Slang and Unconventional
English, gives for "rose": "feminine pudendum; virgin," and in
his Shakespeare's Bawdy, "pudend, maidenhead.... Cf. 'He that sweetest
rose will find,/Must find love's prick and Rosalind,' As you Like It, III: ii."
Such usage by Shakespeare does not, of course, mean that the item
in question must have been folklore, but it does attest to the playwright's
assumption that the reference would have been familiar enough
to his aural audience not to have constituted an impediment either
to the train of thought or to the imagery through which that thought
was expressed. Macer, the medieval herbalist, notes that the rose is to be
used as a medicine for several ailments peculiar to women; the Brown
Collection of North Carolina Folklore (VI, 624) lists the rose as useful
in love divination; Maria Leach mentions, among other appearances
of the rose in traditional belief, its use as a device on a virgin's tombstone,
and adds that in Wales a white rose is often planted on the grave of a
virgin. The term is often found in folksong as a direct but figurative
reference to virginity, as it is in "Poor Nell," one of Sharp's collected
pieces: "My virgin rose you stole away,/O wed me sir, said she."23 The
versions, closer to the regular type of 112, use other flowers, notably marigold and
rue, as a means of reinforcing the knight's frustration.
23 Reeves, Idiom, p. 172.
95
J. BARRE TOELKEN
rose is also of course connected strongly with the Virgin Mary in Christian
imagery and symbolism.
Observation of this consistent relationship between plucking fruits,
picking flowers, branches, etc., and seduction as a commonplace in
the oral ballads allows us, among other things, to support Tristam Coffin's
hypothesis that the motive for the murder in "Edward" is not an early
example of attempted conservation of natural resources (most American
versions giving as the reason for the fratricide "the breaking of a little
bush"), but that it has something to do with seduction - one brother
killing the other over a girl.24 The figure of speech may indeed be a
kenning, as Coffin suggests, for it calls its referent something it is not:
a girl is not a bush. But beyond that, in the light of comments above,
it is a kenning with symbolic overtones that would be familiar to a
traditional, that is, aural, audience. Neither must the reference be seen
as euphemistic, since other ballads exist which deal rather baldly with
incest. Rather, it may be treated as a traditional metaphor, in spite of
the possible objections of those who consider any subtle figurative expression
in a ballad suspicious.
Although plucking flowers and plants occurs in several ballads in
which childbearing, not seduction, provides the focal point, none of
the ballads I have classified as non-oral uses this motif in any way in
versions available to me. On the basis of this sharp distinction, I feel
we may theorize with a great measure of confidence that plucking flowers
is an action used figuratively in traditional ballads as a symbolic unit
which is often reinforced by direct reference to specific flowers and plants
which have traditional associations among a folk audience. These elements
may be used by a traditional singer to manipulate the texture of his ballad
in such a way as to enhance its meaning and impact while maintaining
a strict economy in wording.
Some years ago a view was put forth by Scott Elliott that pulling or
plucking plants in the Child ballads compares with the breaking of the
bough in the sacred grove of Diana at Aricia; thus, he held, such an
action must be taken as a sign of trespass, and the young man who
starts up immediately in several of these ballads ("Tam Lin" furnishes
a good example) is to be construed as the guardian of a sacred grove.
24 Coffin brings in further evidence to suggest an incestuous relationship between
the girl and one or both of the brothers. See "The Murder Motive in 'Edward,'"'
Western Folklore, VIII (October 1949), 314-319.
96
AN ORAL CANON FOR THE CHILD BALLADS
In the first place, it is extremely doubtful that the rural ballad-makers
and singers would have known of such a sophisticated and far-off
concept (to say nothing of its language); neither is there evidence that
they had such a tradition of their own. Further, Tam Lin reveals during
the first dramatic interchange with Janet that he is a mortal who has been
captured and put under a spell by the Queen of Fairies. He needs to be
unspelled, and it is not at all rare in folklore to find that the unspelling
procedure involves not only dangerous tasks but sexual intercourse
(as it obviously does in "King Henry" [32] in which the key verse is
represented in Child's collection by a row of noncommittal asterisks).
Janet's action of plucking a rose at the beginning of "Tam Lin,"
then, may be seen as a symbolic device aimed at the audience to establish
a frame of reference for seduction and unspelling, not at Tam Lin
himself for the purpose of conjuring him up as a sacred guardian, a
role which, after all, has little if anything to do with the narrative. Once
such an evocative motif as picking flowers is seen as a consistently
operative element in traditional narrative, based on commonly shared
traditional references among the audience, there is little need for us to
turn to Strabo (ca. 58 B.C., our source for the sacred grove of Diana at
Nemi) or, for that matter, to Robert Graves, present guardian of the
sacred grove.25 With the determination of an oral canon for study, and
with the close investigation of elements common to that canon, our
answers are nearer at hand and more reliable.
One more set of examples will suffice, I hope, to suggest the value
of close study within the oral canon. Common in the Child ballads are
the many plants (typically willows, lilies, roses, briars) xhich grow on
graves, apparently in the role of metamorphosed souls. Of them
Wimberly commented, "One could not ask for better evidence of the
primitive character of ballad tradition," and ascribed their use to implicit
belief in transmigration, not to a sense of imagery.26 Child called
the motif of plants growing on graves a "beautiful fancy" which illustrates
the belief that "an earthly passion has not been extinguished by
death" (1, 96). As we shall see, the motif functions poetically, and does
so with considerably more coherence than one would normally expect
25 Or to Freudians, who could easily make wondrous - if unintelligible - castration
fantasies of this whole group.
26 Lowry Charles Wimberly, Folklore in tile English & Scottish Ballads (Chicago,
1928), p. 39, n.
97
J. BARRE TOELKEN
of a mere "fancy"; and while it certainly may be a bit of pagan residue,
its continued use suggests that singers- if not ballad critics - have considered
it an important element of ballad narrative.
Among the many ballads in this category, eight in the Child collection
concern two lovers who die and on whose graves plants grow,
eventually meeting and twining together. In these ballads several plants
are mentioned or described growing on graves, but by far the majority
of versions favors the rose and briar, usually as follows:
Lord William was buried in St. Mary's kirk,
Lady Margaret in Mary's quire;
Out o' the Lady's grave grew a bonny red rose,
And out o' the knight's a briar. ("Earl Brand," 7)
In nearly every version using this motif a verse follows which describes
the plants growing together into a true-lovers' knot. Probably for reasons
suggested above, the rose almost invariably appears on the lady's
grave, the briar on the man's. More specifically, in six ballads ("Earl
Brand," "Lady Maisry," "Lord Thomas and Fair Annet," "Fair Margaret
and Sweet William," "Lord Lovel," "Lady Alice") a total of sixty-eight
versions have the rose on the woman's grave, the briar on the man's,
and end with the true-lovers' knot, while six versions of the same ballads
reverse the sexes (usually retaining the knot). In the case of "Prince
Robert" (87), versions A and B use birk and briar and love knot, but
no connection with sex is discernible from context. (One may guess from
the relations noted in the above six ballads that the connection was
thought to be obvious.) In "Barbara Allen," twelve versions connect the
rose with the woman, eleven of which end with the true-lovers' knot
while fifty-one relate the rose with the man, briar with the woman, most
of them retaining the knot. Of these fifty-one, forty-eight are American.
With the single exception of "Barbara Allen," then, whenever a
distinction in sex is made, the preference is clearly for the identification
of rose with woman, briar with man. A closer examination of the ballad
stories here concerned will indicate further that the departure from this
preference in the case of "Barbara Allen" is probably not a matter of
degeneration. All these eight ballads share one contextual trait: the
frustration of what is known in the ballads as "true love." In each instance
the couple is kept from physical union by death; one partner
is killed, or dies of sorrow, and the other soon follows, usually on the
morrow, succumbing to a broken heart. It seems then as an adjunct
98
AN ORAL CANON FOR THE CHILD BALLADS
or complement to the theme of thwarted love that plants appear in the
conclusions of these ballads. They not only suggest stasis and repose
at the end of an often furious series of actions - an important narrative
function referred to as das Gesetz des Abschlusses by Axel Olrik - but
they become stand-in characters, performing physically what the couple
presumably wanted to but could not, and thus forming a suitable dramatic
resolution to the tensions central to the story (especially that part
of the story Coffin would call the "emotional core").
Here, in other words, the plants do not simply represent souls, for in
frustrated true love neither the audience nor the participants find much
comfort in death only - even in vegetable immortality. The resolution
lies rather in the couple's reunion by the joining of the plants in a love
knot, a device which is traditionally considered a sign of amorousness,
and because of its appearance, when loosely tied, of sexual intercourse.
(It is still fairly common among children as an obscene string-game.)
Since the plants here actually take part in the dramatic denouement,
their function may be seen as thematic as well as symbolic, and their role
is further extended in some variants to allow for still another stroke of
frustration when someone (the girl's father, a parish priest, a clerk, a meddlesome
old woman, or a cold wind) cuts them down again. In one version
of "Earl Brand":
But bye and rade the Black Douglas,
And wow but he was rough!
For he pulld up the bonny brier,
And flang't in St. Mary's Loch.
It is both interesting and significant to note that it is the male's plant
which is considered at fault here, in good part a sign in and of itself of
the sexual overtones of the image.
The notable departure in "Barbara Allen" from the sexual connections
seen in the above ballads is significant to our examination. We note
that there seems to be no confusion about where the rose belongs;
in the other ballads the rose is predominantly on the girl's grave, while
in "Barbara Allen" it is predominantly on the man's. Now "Barbara
Allen" is the only one of these eight ballads in which one of the lovers
is consciously at fault in the other's demise. Barbara has wilfully rejected
Sweet William's plea to return his love, and has told him in effect to go
ahead and die. This he promptly does, of course, and it seems likely
that ballad singers along the line of transmission have felt it desirable to
99
J. BARRE TOELKEN
capitalize on this aberration by providing for a sense of justice at the
end of the story, placing the inhospitable briar on the grave of the guilty
Barbara, the rose on the grave of her innocent victim who died in the
flower of his youth.
Since most of these versions which connect the rose with Sweet
William are American, we may postulate further that among other
things at work here is a general de-emphasizing in American tradition
of the specific connections between plants and sex. Yet the American
versions of the other ballads discussed here retain the connection, suggesting
that the tradition is far from dead, and indicating strongly that
there remains a consistency in the employment of plants to denote sex
in one instance and to imply judgment in another. On this basis, we may
presume that ballad singers have worked from a body of traditional
commonplaces which may be utilized variously in various poetic contexts,
and that, at least in the instances discussed here, there is a fair amount
of conscious artistry at work, not simply knee-jerk reference to a stock
of familiar stanzas. If it is indeed just a matter of borrowing, it is a
selective and coherent process, and is demonstrably satisfying on poetic
grounds to a traditional audience, which would share and understand -
perhaps almost subliminally - the connotations therein. It seems to
me that these important aspects of balladry are not adequately explained
in terms of chance, degeneration, or technical proficiency with metrical
formulas.
All the ballads under discussion here are, or have been, current in oral
tradition; conversely, among those Child ballads which I have for convenience
termed "non-oral," only a few use plants in any way organically
connected with plot or theme. I conclude, then, that plants represent
one of perhaps many fields of traditional reference which have allowed
oral ballad singers and makers to use directly understandable figurative
language which drew on the audience's own cultural background of
superstition, popular belief, and folk metaphor for its impact and meaning.
Since these figurative elements were used in common by many balladmakers
and singers, they do represent a traditional (i.e. culturally shared)
battery of poetic devices which may be seen as analogous to the metrical
formulas and themes discussed by Parry, Lord, and Magoun. But
where these scholars stress metrical equivalents and the individual
singer's virtuosity in handling them, I here stress poetic units from tradition
which might be called "textural formulas," and call attention to
100
AN ORAL CANON FOR THE CHILD BALLADS
the fact that they are evocative figurative units with associations communally
shared. And instead of starting with a formularized concept of
what a formula is and where it appears and applying that concept to
all texts, I have here established first the oral nature of the texts and
have then tried to see what elements, if any, are operative therein.
These observations are intended to augment and qualify, rather than
to refute, the formulaic study of poetry. Nonetheless, it does seem to me
that the concentration on technical matters in oral poetry has in general
tended to de-emphasize, even to denigrate, the poetic concerns, so that
while we admit to certain beauties in the ballad, or in Beowulf, say,
we are encouraged to talk about them in terms either of "primitive
fancy" or of unintentional effects accidental to metrical demands, singing
rules, lapses of memory, limitations of subject matter, and other parapoetic
considerations.
Despite the fact that the ballad-makers may have been innocent of a
critical jargon which might have given a name to the device I here call
"textural formula," the modern critic and ballad scholar need not apologize
about discussing metaphor, image, symbol, theme, and figurative
language in the ballads as long as he is willing to find out what associations
such things as plants, colors, birds, and animals had in the traditional
backgrounds of those through whom the ballads have passed (both
singers and audiences), and as long as he limits his study to those ballads
which did in fact pass through or among a group of people which recognized
these associations. The establishment of an oral canon for the
Child ballads will hopefully provide a reliable corpus of texts on the
basis of which further generalizations can be made relative to the oral
poetry of the traditional ballad.27
University of Oregon
Eugene, Oregon
27 There is the further, more exacting, task of determining which of Child's texts
under any given oral ballad are more reliable as documents of tradition. The present
study has dependedo n statisticallys ignificantg roups for its generalizationsb, ut any
more particular study of a single Child ballad text should begin by the establishment
of that text's oral provenience. As this article goes to press, D. K. Wilgus advises me
on the basis of his own researches that since the Barry version of "Prince Robert" (87)
is only a trace, and since the Combs variant is highly suspect, the ballad should be
excluded from the present canon in the lack of other supporting evidence. Wilgus
does not argue that the ballad is not, or has not been, oral, but rather that the evidence
in its behalf is at present too slim. Certainly the matter should be settled on this and
other ballads before generalizations about "the ballads" can be safely put forth.
--------------------------------
Footnotes:
1. For a cross section of the nature and scope of this topic and examples of particularly
incisive applications of the theory to early English literature, see: Milman Parry,
"Studies in the Epic Technique of Oral Verse-Making: I. Homer and the Homeric
Style," Harvard Studies in Classical Philology, XLI (1930), 73-147; Albert B. Lord,
"Homer and Huso: II. Narrative Inconsistencies in Homer and Oral Poetry," Transactions
and Proceedings of the American Philological Association, LXIX (1938), 439-445;
Francis P. Magoun, "Oral Formulaic Character of Anglo-Saxon Narrative Poetry,"
Speculum, XXVIII (1953), 446-467; Stanley B. Greenfield, "The Formulaic Expression
of the Theme of 'Exile' in Anglo-Saxon Poetry," Speculum, XXX (1955), 200-206; J. J.
Campbell, "Oral Poetry in The Seafarer," Speculum, XXXV (1960), 87-96; W. A.
O'Neil, "Another Look at Oral Poetry in The Seafarer," Speculum, XXXV (1960),
596-600. These and other studies in the oral-formulaict heory are admirablys ummed
up and evaluated with reference to similar approaches to the traditional ballad by
Bruce A. Beatie, "Oral-TraditionaCl omposition in the Spanish Romanceroo f the
Sixteenth Century," Journal of the Folklore Institute, I (1964), 92-113. See also Jones
and Friedman, cited below.
2 It would be grossly unfair to suggest that the investigations of the oral formula in
English poetry have been made wvithouct lose study of the English texts and cultural
contexts; rather, it is the way in which poetic phenomena from outside the province
of the texts are used to describe the poetic characteristicso bserved therein, and the
way in which the data are interpreted according to a preconceived, inflexible concept,
that bear further scrutiny.
3 James H. Jones, "Commonplace and Memorization in the Oral Traditions of the
English and Scottish Popular Ballads," Journal of American Folklore, LXXIV (1961),
97-112. See also the rebuttal of part of Jones' thesis by Albert B. Friedman, "The
FormulaicI mprovisationT heoryo f Ballad Tradition- A Counterstatement,"J ournal
of American Folklore, LXXIV (1961), 113-115.