Bluegrass Messengers

Moses Platt and the Regeneration of Barbara Allen
 

Moses Platt and the Regeneration of "Barbara Allen"
by Charles Clay Doyle and Charles Greg Kelley
Western Folklore, Vol. 50, No. 2 (Apr., 1991), pp. 151-169

Moses Platt and the Regeneration of "Barbara Allen"
CHARLES CLAY DOYLE & CHARLES GREG KELLEY

Early in his collecting forays through penitentiaries in the middle South, John Lomax discovered a septuagenarian black man with a repertory of oral verse equal in length to The Iliad (Lomax 1947:179).' In 1933 Moses Platt, nicknamed "Clear Rock" for reasons obscurely connected with a triple-murder, sang for Lomax and his recording machine an extraordinary version of "Barbara Allen" (Child 84):

[Listen: Moses Platt]

1. O Bobby cried, poor thing she moaned-
O what whooping and hollering!
I know you're sick, you're very pale;
Young man, you're slowly dying.

2. Go bring me back my pony here;
I'll send away my servant;
Go bring my love a-here this day,
And let her gaze upon me.

3. O the more she looked when she first goodbye,
She looked right deep with sorrow:
I know he saw [?] my own true love,
True willful Barbara Allen.

4. Now the loved one died on a Christmas day;
My loved one died one Monday;
They buried them both underneath the grave,
And they placed the soil upon them [him?].

5. Now coffee grows on white oak trees,
And the river flows with brandy;
And the sweetest love that ever I knew,
And it was poor Barbara Allen.

6. Now plant this grave a bramble briar,
On Rosy's grave plant a briar;
So they grow until they reach the sky,
When the knot was tied in Ohio.

7. O let the world know that I died for love,
I died right weeping and moaning;
I said farewell when the corpse come along,
True willful Barbara Allen.

8. Now sixteen horses they may ride,
O bring me back my lover;
The loved one's gone to meet the grave,
And it's fare-you-well, Barbara Allen.[2]

More than half a century later, informal experiments with several undergraduate classes (University of Georgia, 1986 through 1990) reveal a strong consensus on two points: (1) as sung in Platt's baritone voice (without instrumental accompaniment), the ballad is strikingly beautiful; (2) lacking familiarity with other versions of the ballad, no one would be able-either by repeated listenings to the recording or by study of the transcription-to infer from Platt's rendition the traditional story-line of "Barbara Allen."

The text presented above comes from the widely available Library of Congress phonograph album titled Versions and Variants of "Barbara Allen" (c.1964), edited by Charles Seeger-an album of great usefulness for classroom purposes. The album-jacket designates the presentation of Platt's performance as "complete." In fact, the text has been edited in ways that are worth noting, though they do not much affect the present analysis.

On Lomax's unedited field-recording, AFS 201 A1 (LWO-4844, reel 15B) in the Archive of Folk Culture at the Library of Congress, when Lomax requests the ballad, Platt begins with a fragment of a stanza:

-on no river;
Don't bury me on my place I'll stay,
And it's fare-you-well Barbara Allen.

This is probably not an actual false start; Platt sings softly, as if merely to "find" the tune, the pitch, and the words. So probably Seeger was justified in editing out that introductory fragment as extraneous to Platt's intended performance of the ballad proper. Having prepared himself, Platt proceeds to sing with confidence stanzas 1-7 as transcribed above.

After stanza 7 he hesitates, and Lomax prompts him: "Now 'sixteen horses'." Then Platt sings stanza 8 and seems to be finished when Lomax prompts him again: "Where's your 'sixteen pretty maidens'?" Platt starts singing "sixteen maidens" but cannot go farther, and that recording session ends.

How the old British ballad, beloved of Samuel Pepys and Oliver Goldsmith, found its tortuous way to the Central State Farm in Sugarland, Texas, we cannot intimate. But the remarkable changes wrought upon the ballad in its progress across oceans and generations and cultures are ponderable, and pondering those changes can illustrate some important features in the evolution of ballad texts.

The nature of change in the process of oral transmission is one of the oldest, most fundamental, and most controversial issues that folklorists have addressed. In songs, where the prosody and the tune can assist memory and limit the possible range of verbal alterations, variations become especially noticeable and significant. Of course, changes can result from mere forgetfulness or confusion on the part of a performer, but they can also indicate the performer's conscious or unconscious attempt to rationalize, improve, or adapt a prototype text.

Early students of balladry like Sir Walter Scott (and even Francis James Child) generally assumed that current versions of a ballad represent degenerate legacies of some pristine ur-form, which sometimes would have been preserved in 17th or 18th century printed texts but often was lost beyond recovery.

In that tradition of scholarship, one of the most important discussions of ballad change was presented in 1928 by Reed Smith in a chapter titled "The Road Downhill":

It is the penalty of oral tradition that it can exist only where literacy does not. The two vary inversely: as literacy waxes, oral tradition wanes; and as it wanes, it is automaticallyc onfined to lower and lower levels of talent and intelligence. In our day, owing to the universal spread of educational opportunity and cultural advantages, oral tradition has almost reached the vanishing-point in quantity, and has almost touched bottom in quality. [55]

Smith did allow, as an exception to the modern state of decay, the lore of "the Negroes, whose cultural status in certain sections still permits oral transmission to function vigorously and effectively" (55). The importance of Smith's discussion is not in the posited correlation between literacy and orality per se (he was certainly mistaken about that) but rather in his hypothesis that extensive transformations in the plot, setting, characterization, and tone of ballads begin with misunderstandings of individual words, followed by the application of folk-etymology to the particular instances of confusion. Thus, for example, the archaic "cried out amain" becomes "cried out Amen"; a "weapon knife," heard pronounced in an unfamiliar way, becomes a
"wee pen knife" and then a "little penny knife." The process often affects proper names: so Barbara Allen occasionally gets rechristened "Barbara Ellen" or even "barbarous Ellen."

When, over time, numerous features of a ballad have undergone the process of folk-etymology, "the separate errors merge into a kind of pervasive degeneration that lowers the whole ballad to a most lamentable level," which Smith called "organic perversion" (62-63). Despite the emphasis on degeneracy and perversion, Smith's analysis was exceptionally enlightened in its unmystical willingness to imagine in detail how alterations come about through the cumulative process of individual performers in the act of hearing, interpreting, and rendering their interpretations.

As late as 1957 an unregenerate degenerativist, Stanley Edgar Hyman, declared:

What has happened to the Child ballad in America, in sum, is that it has become inadequate narrative, aborted drama, happy-ending tragedy, corrupt and meaningless verbiage, and bad poetry in general. Some of this may be the effect of transmission in time, which seems to degenerate and deteriorate folk literature wherever we observe its effects (39).

Most modern scholars, in contrast, have been hesitant to equate alteration with deterioration, recognizing not only that ballads must change to remain socially functional under new cultural circumstances but also that the folk esthetic, "re-creatively"w orking upon a ballad over the course of time, can effect artistic improvements sloughing off the ugly and superfluous while retaining or enhancing or introducing elements that possess beauty and power.

The most influential modern explanation of the process by which oral transmission often improves the "artistic" quality of a ballad is Tristram P. Coffin's 1957 essay-often reprinted-" 'Mary Hamilton' and the Anglo-American Ballad as an Art Form."

Coffin's point of departure was the regenerative afterthought in a statement by John Robert Moore published in 1916: "After a painstaking study of the subject, I have yet to find a clear case where a ballad can be shown to have improved as a result of oral transmission, except in the way of becoming more lyrical" (Moore 1916:400; italics added). [3]

Coffin hypothesized that ballads enter oral tradition from literary or subliterary sources. Then the "frills" of the style and some details of the action are "worn away by oral transmission" (1977 [1957]: 166). Then, in the "final stage the ballad develops in one of two ways. Either unessential details drop off until lyric emerges, or essential details drop off until only a meaningless jumble, centered about a dramatic core, is left" (167). For Coffin this "emotional core" was the essential element of a ballad-plot, action, setting, characterization,a nd theme all being (more or less) dispensable. As the inessentials erode away through forgetfulness, unconscious editing, or conscious recrafting, the ballad becomes a lyric poem.

Therefore, the early ballad texts so admired by literary anthologists do not represent the pristine ur-form of composition, before the deteriorating process of oral transmission commenced. Rather, the esteemed texts were transcribed at a moment when the erosive process  of transmission was (so to speak) complete, having maximally polished the ballads to the emotional core and rendered them lyrical.

Coffin was writing in the heyday of the "New Criticism," and for him (as for the New Critics at large) lyrical was a term of great and general approbation. Two paradoxes suggest themselves in Coffin's use of the term. First, he meant by it (in part) the verbal text as considered separately from the tune-a usage which, of course, is not unique to Coffin. In that sense lyric means not a song to be sung (as if to the accompaniment of a lyre) but rather a poem to be read or recited. As Coffin was possibly aware, his account explains how folk ballads sometimes develop toward forms that meet the standards of  literary critics-not how oral transmission shapes ballads in ways that will be most appealing to folk performers and their intended audiences.
After all, much of the non-hypothetical history of ballad transmission has chronicled the artistically retrograde influence of tawdry, sentimental broadsides, songsters, and commercial recordings upon the "pure" oral tradition.

Secondly, like the New Critics at large, Coffin particularly valued dramatic qualities in "lyric" ballads (and, implicitly, in other genres of literature and folklore). According to Coffin, even when essential details of a ballad have been ground away by oral transmission-leaving "only a meaningless jumble" -the jumble will still be "centered about a dramatic core," which appears to be the same as the "emotional core" of a ballad-cum-lyric. [4] By his use of the generic designations, then, Coffin made the ultimately-polished ballad either a poem to be read or a drama to be enacted in the imagination-but not a song to be sung to a real audience.

If, by mischance, the process of transmission results in neither the lyric nor the dramatic, according to Coffin, the ballad becomes "a nonsense song." That category Coffin exemplified by certain nurseryrhyme versions of old ballads and by "the amazing Texas version of Child 84, 'Boberick Allen'" (1977 [1957]:167), namely a two-stanza text collected by Mary Virginia Bales (1928:111-12) from a black student at Terrill High School in Fort Worth, presumably in the 1920s:

When I wus but a girl sixteen,
I wus in love with Boberick;
De othah girls did not see
Why he did always follow me.

He walk to town an' den right back,
To see if I wus on his track,
But he could neber fin' me dah,
Becuz I wus away somewhah.
His name wus Boberick Allen.

Coffin, in a more recent commentary, remarked that this text "reflects a complete degeneration" of the ballad, that here "loss of detail combines with other forces of degeneration" to produce an "extremely corrupt and nonsensical" song (1977:85 and 11).

If terms like degeneration could be used non-judgmentally, there might be no objection to calling the Fort Worth text of "Boberick Allen" degenerate. Certainly it has lost most of the accustomed stanzas and episodes and situations, even transferring the traditional name of the female protagonist (and regularly the title given to the ballad itself) to the male victim. Although, as the original collector noted, "the girl is still the disdainful one" (Bales 1928:111), the use of the first-person voice radically changes the point of view and alters the "drama" that prevails in most texts of the ballad.

The fact is, of course, that epithets like degenerate, decayed, perverse, and nonsensical do imply value judgments, and those judgments are unfair. They disparage the performer or informant-in this case the unnamed black teenager from Fort Worth-as well as that individual's family and associates. If singing and hearing the two stanzas about the unrequited love of a boy named Boberick brought pleasure or meaning into what must have been a hard life, then who are folklorists to pass judgment?

From the purely analytical standpoint, such judgments are inappropriate for another reason: they risk confusing the transcribed text with an authentic performance, or the scholarly reader (or collector) with a participant in the folk culture.

The point is that apparently fragmentary, eroded, or otherwise truncated versions of a ballad may, as actually performed within a group, function as narrative. That is to say, if a story is well-known to both performer and audience, the detailed recounting of events and circumstances may cease to be obligatory. Thus the five-stanza Maine text of "Mary Hamilton" (Child 173), Coffin's exemplary centerpiece of reduction to a "lyric core" (in contrast with the 18 stanzas of Child's A-version or the 23 stanzas of the B-version), does not necessarily replace narrative with lyricism; rather, it may imply or allude to the entirety of plot, setting, and characterization that would have been familiar to Mainelanders from other sources. [5]

An extreme way of putting the point might be that within the communal consciousness of a folk group resides the macrotext of a ballad, from which an individual performance may optionally select and re-sequence stanzas (or, if this macrotext is not actually formulated into stanzas, then scenes, episodes, or speeches), depending on the personality or disposition of the singer, the occasion, the time available, the audience's mood or attention-span, and other "contextual" considerations.

If such an explanation does not fully account for the Fort Worth text of "Boberick Allen," it does seem suitable for another African- American version of "Barbara Allen," collected by Roger Abrahams (1968:93-94) on the West Indian island of Nevis in the early 1960s:

As I was going along the road
I heard the church bell tolling.
As I was going along the road,
I see the coffin passing.

Oh, lay it down, oh, lay it down,
That I might gaze upon him.
Oh, lay it down, oh, lay it down,
That I might gaze upon him.

Barbry Alone, Barbry Alone,
A kiss from me will save you.
But a kiss from me you will never get,
Not if your heart were broken.

A young man died for me today,
And I may die tomorrow.
A young man died for me today,
And I might die tomorrow.

Barbry Alone, Barbry Alone,
A kiss from me will save you.
But a kiss from me you will never get,
Not if your heart were broken.

Comparing this presentation with what might be called the normative sequence of events and scenes in "Barbara Allen," we find the performer leaping into the middle of the story, focusing on Barbara in the presence of the young man's corpse, with the third stanza repeated as the fifth by way of a sort of refrain-harking back (perhaps in Barbara's anguished recollection) to the bedroom scene, in which she had rebuffed the man's plea for life-saving solace. Abrahams mentioned "the confusion of pronouns" in this text and alluded to other West Indian renditions in which the sex of the villain" Bobby" becomes indeterminate. In one variant... the singing persona laments the inability of "Bobby" to rise from the dead to save her, while in another ... the singer is a man who says, while gazing on the coffin: "A young girl died for me today, / I would die for him tomorrow." [1968:94] So, the male identity of the Fort Worth Boberick Allen is not wholly aberrant after all. [6]

In fact, however, the pronouns in the quoted text from Nevis are not confused at all; quotation-marks could easily be inserted that would show the consistency. Rather, what we have here, as in the Fort Worth version, is the female protagonist herself in the role of narrator, intermittently quoting herself-and possibly quoting the young man as well, depending on how we interpret the reiterated appellation "Barbry Alone."

As Abrahams remarked, in the quoted text the ballad "has become almost a lyric lament" (1968:93). It shows several of the predictable effects of oral transmission over an extensive time-and an extensive cultural space as well-including the wonderful folk-etymologizing of the traditional name given to the woman, slighted, alienated, and bereft. [7] But the transformation of a ballad into a lyric does not of itself cause the song to cease functioning as narrative.

Also in reference to African-American lore, Martha W. Beckwith made a related point in 1924. She noted the tendency among West Indian performers to render a ballad in the form of a cante fable, with verses strung "upon a connecting thread of prose to carry along the action," or else to condense a ballad into "the elliptical form characteristic of the Jamaican lyric song"-even though under special conditions (namely, in the presence of the collector's microphone) she found the same informants quite capable of singing full narrative versions (Beckwith 1924:455 and 457). So mere forgetfulness is not always a sufficient explanation for a seemingly fragmentary or disjointed or eroded or lyrical performance of a ballad. The performer and the audience would actually discern a coherent narrative, if not within the stanzas as overtly performed then behind them, so to
speak.

Abrahams has applied the term "creole performance event" to the way in which, in certain Caribbean variants of Old World texts, "the ballad story is altered in formulation and fitted into a different community, into its traditional repertoire and its structure of sentiments" (1987:130-31).

The process has analogs in the lore of the United States. D. K. Wilgus and Eleanor Long, discussing the so-called "blues ballad,"
have noted "the near-total suppression of narrative sequence in favor of a series of comments upon a story which must in large part be inferred from these comments" (1985:439). The intent is "to sing a story not by directly relating that story but by celebrating it; not by following a chronological sequence, but by creating a sequence of concepts and feelings about it. The story is there, but not explicitly delineated" (439).

In other words, the total story is a given, and the ballad performer, secure in his knowledge that his audience is familiar with the referent, ... selects a scene or a number of scenes for presentation and, freed from the necessity of presenting circumstantial detail, concentrates on  conventionalized dramatic scenes, oblique delineation of character through action and speech, and lyrical comment by characters or the narrator. The technique causes problems for the outsider, the investigator
for whom the ballad was not intended (443).

While giving several examples of nineteenth-century broadsides that have been "recomposed into" blues ballads, Wilgus and Long do not seem to acknowledge that the same could happen to Child ballads entering the oral traditions of African-American groups. [8] Perhaps because blues ballads so often share with "vulgar" broadside ballads "the unabashed expression of attitude and emotion in subjective commentary upon the event" (439), blues ballads have seemed inhospitable ground for the transplanting of Child ballads, with their characteristic restraint and understatement.

Although the resemblances are suggestive, it would be imprudent to insist that Moses Platt's version of "Barbara Allen" be called either a blues ballad or a creole performance event. For one thing, the term creole implies African connections that would be hard to establish. For another, Platt's tune is not especially "bluesy"; in fact, it is remarkably "regular" as compared to the extraordinariness of his verbal text-a fact that may explain why the most comprehensive surveys of variants have neglected really to notice this text.

Bertrand Bronson (1962:321-91), tallying 198 variants of "Barbara Allen," did not mention Platt's, despite its accessibility in the
Library of Congress Archive of American Folk Song. Charles Seeger's extensive essay "Versions and Variants of the Tunes of 'Barbara Allen'," written to accompany the Library of Congress phonograph album Versions and Variants of "Barbara Allen" (c. 1964), gave a simplified transcription of Platt's tune (Seeger 1966:127) but otherwise made only the oddly understated remark that Platt's performance (along with one other on the album, each referred to merely by number) is "not, strictly speaking, typical of traditional country singing" (128). [9] Ed Cray's taxonomy of verbal texts (likewise prepared in connection with the Library of Congress album) also failed to notice the distinctiveness of Platt's rendition (1966:164-67). [10]

Whether typical of African-American singers or not, Platt was decidedly willing to adapt his performance of ballads to the audience and the circumstances. Lomax (1967:179-80) reported hearing Platt sing "a new version of The Old Chisholm Trail, an endless ballad describing the experiences of a band of cowboys" on a cattle drive. In one episode a cowboy is "thrown high and left hanging on a limb.... Clear Rock sang four stanzas describing this incident and then ended his song."

"That's rather hard on that cowboy," I suggested, "to ... leave him sprawling on a limb in what is doubtless an uncomfortable position." "Lemme git him down, boss; I'll git him down!" And at once he sang in perfect tune:

'Cowboy lyin' in a tree a-sprawlin',
Come-a-ti-yi-yippyy, ippy yea, yippy yea,
Come a little wind an' down he come a-fallin',
Come-a-ti-yi yippy, yippy yea. [11]

Like the West Indian performers described by Beckwith and Abrahams, Moses Platt would certainly have assumed the familiarity of his audience-which included not only Lomax but also Platt's fellow inmates, who in some cases had been listening to his songs for years with the plots of both "The Old Chisholm Trail" and "Barbara Allen." For Lomax, Platt performed at least three quite different versions of "Barbara Allen":

... he afterward s ang "Bobby" Allen, as he called the old English ballad, true to tune, but hopelessly mixed with a famous cowboy song entitled  "The Streets of Laredo." His song buried "Miss Allen" in a desert of New Mexico with six pretty maidens all dressed in white for her pallbearers, though there seemed water a plenty for a rose and a briar to grow over her head till they met and "got twined in a knot and couldn't grow no higher." In other rendition he buried "Bobby" Allen out of Dallas with the mourners "hollerin' and a-squallin' " (Lomax 1947:179).

Quite possibly, within Platt's repertory there existed no actual standard or definitive text of the ballad, no formulated basis from which variations were generated in individual performances. The case is normally very different with white singers in America and the British Isles, who often insist that they never vary their performance, notwithstanding evidence that they do (Gerould 1957 [1932]:163-88).

Both of the versions Lomax referred to differ considerably from the one we have transcribed off the Library of Congress album. The variations are far too extensive to result from mere lapses of memory. Our text contains only a shadow of the reported contamination from "The Streets of Laredo" (Laws B1), possibly detectable in the clause "they placed the soil upon them" (stanza 4) and in the procession of "sixteen horses" (stanza 8)-the latter incident sung, we should recall, in response to Lomax's prompting, which was probably based on his having heard the contaminated version mentioned in the quotation.[12]

The most remarkable importation into our version is the passage in stanza 5 from the widely collected play-party song "Pretty Little Pink," most versions of which contain a stanza like this:

Coffee grows on white oak trees,
The river flows with brandy-O.
Go choose some one to roam with you,
As sweet as 'lasses candy-O (Bodkin 1937:164).[13]

Platt incorporated the first two lines into a stanza that serves as a sort of interlude-marking the passage of time in the narrative and affording a moment of psychological respite-between the burial of the lovers and the germination of the symbolic plants on their graves:

Now coffee grows on white oak trees,
And the river flows with brandy;
And the sweetest love that ever I knew,
And it was poor Barbara Allen.

Rhetorically, the imported imagery of the first two lines furnishes a periphrasis of the idea of sweetness that is made explicit in line 3, anticipating the reposeful and forgiving image of the twining plants. Platt handles the twining-plant scene most unusually:

Now plant this grave a bramble briar,
On Rosy's grave plant a briar;
So they grow until they reach the sky,
When the knot was tied in Ohio.

"In Ohio" is almost certainly a folk-etymologizing of the phrase no higher, frequently found in texts (such as the one by Platt that Lomax mentioned) which say the plants grew till they "couldn't grow no higher." Less easy to explain is how one of the graves came to be occupied by an interloper named Rosy, but the fact that Platt planted briars on both graves, instead of the traditional one briar and one rose, left a rose in search of a referent, which evidently has become the woman herself.

Of itself, the importation of stanzas or episodes from other songs is common enough in the history of ballad transmission; in fact, the customary rose-and-briar ending of "Barbara Allen" is a late adaptation from older texts of "Lord Thomas and Fair Annet" (Child 73) or "Fair Margaret and Sweet William" (Child 74) or other heartless-love ballads. Versions of "Barbara Allen" can also be found containing lines appropriated from "Sir Patrick Spens" (Child 58) and "James Harris" (Child 243) and "Pretty Peggy O." [14] Especially for a singer with Platt's capacious repertory, the ad hoc merging of different songs might have been well-nigh inevitable.

Nor, as we have seen, is there anything extraordinary about the alteration of individual words through folk-etymology. Americans singing "Barbara Allen" have always found opaque the terminology of Old World texts that bury one of the lovers in the choir of a church (the other at some distance away in the churchyard); so a whole array of rhymes for briar have been pressed into service to help rationalize the detail: besides various manipulations of nigh her, by her, or beside her (which undermine the emphasis on distance between the grave sites), sometimes a lover is buried under a spire or a tower, on the morrow, in a graveyard that is higher, in a bower, in a square, illogically in a squire, unprettily in the mire, superfluously in attire (or, as grotesquely transcribed in one instance, in a-tire); even the resort to Ohio is not unique. [15]

The properties of Platt's version that are distinctive, that most sharply differentiate it from traditional performances by Americans of English or Scottish descent, are its disregard for linear narrative sequence, the shifting point of view (including recurrent first-person intrusions), and the emotionalism.

Although the plunging in medias res is commonly found in Child ballads generally, such a beginning is decidedly atypical for "Barbara Allen," which most frequently opens with one or more "frame" stanzas that distance the story from the place and time of performance: in another town "where I was born," in " the merry month of May" of an unspecified past year, and the like. A typical version next zooms in on the bedroom where the young man lies sick from love and rejection, and he sends for Barbara.

Platt has dispensed with the formulaic frame opening altogether, commencing with Barbara's arrival at the bedside. In more conventional versions, great dramatic impact can result from the stark unemotionalism of Barbara's observation, "Young man, I think you're dying": a great difference from the crying and moaning and whooping and hollering in Platt's rendition.

In a normalized sequence of events, Platt's second stanza (at least the second half of it) would precede the first. In fact, resequencing the stanzas in the following way will permit a reader to glimpse in chronological order several scenes, actions, and speeches that Platt has selected for highlighting:

Stanza 2: The young man sends for Barbara.
Stanza 1: Barbara arrives, observes the man's condition, reacts to it and comments on it.
Stanza 3: After she departs, [having been overtaken by the cortege bearing the young man's corpse] Barbara is smitten with remorse.
Stanza 7: Barbara professes her love and resolves to join him in death.
Stanza 4: Barbara dies, and the two lovers are buried.
Stanza 6: The symbolic plants spring from the graves and intertwine.
Stanza 5 and stanza 8 have no evident narrative import.

Aside from the unsequencing of the narrative, what is likely to confuse unwary listeners to Platt's rendition is the abruptness with which the singer shifts from the third-person to the first-person, with the first-person speaker sometimes reporting the words of a character and sometimes presenting the narrator's or singer's own response to the material. For example, the first two lines of stanza 1, though hardly objective, are immediately recognizable as third-person narration.

Then, with no overt signal of the change, the next two lines are giving Barbara's words at the bedside, and suddenly the lines following are words that the young man must have spoken prior to the preceding quotation from Barbara. In stanza 4 we must wonder whether the phrases "my loved one" and "the loved one" are synonymous or antithetical; likewise with "my lover" and "the loved one" in stanza 8.

The first-person voice in stanza 5, we at first assume (since both characters have already been located "underneath the grave"), belongs to the singer himself, proclaiming that "the sweetest love that ever I knew" was Barbara. But did he know Barbara? Are the phrases "my own true love" (stanza 3), "my loved one" (stanza 4), and "my lover" (stanza 8) perhaps uttered by the singer in propria persona? I n any case, even if (in this version) we do not observe Barbara being disdainful or spiteful, what are we to make of the reference to her sweetness? Such questions, of course, are beside the point. The point, which the singer conveys with great lyric power, is that after all the destructive pride, animosity, and sorrow are spent, love has once
again resumed its normal, desirable state of sweetness. It does not matter whether the final stanza conveys Barbara's lamentation over lost love or the young man's anachronistic farewell to Barbara or the singer's parting tribute to young love's impetuosity and folly. The net effect (to borrow the felicitous phrasing of Wilgus and Long) is to celebratet he love that the story epitomizes. By the lyricism of the performance, the sorrows of that love have been transmuted into something rich and strange. However thorny the experience of love may have been, somewhere in the sky above Ohio-or perhaps above Sugarland, Texas-a transcendent knot is tied. Even from the drabness and brutality of a hard-time penitentiary, one is able to glimpse a world where coffee grows on white oak trees and the river flows with brandy.

Not only, then, did Moses Platt in his singing of "Barbara Allen" create a beautiful and functional piece of poetry. He also exemplified  some complex ways in which ballad transmission can bring about textual change. Besides the process by which folk-etymology alters individual words or phrases and the process by which forgetfulness or editing erodes away the ugly, the superfluous, and the lyrically nonfunctional, more extensive and more fully-conscious processes of re-creation operate. Quite conceivably, those processes are more prevalent in African-American performances of Child ballads than they are in performances by English speakers of British descentthough the point cannot be safely assumed at this time. The example
from Platt does show that those processes-similar in many ways to what Abrahams called "creole performance" -are not limited to Caribbean folk groups.

The processes include the optional omission and resequencing of episodes, as well as the rapid and unsignaled shifting of point of view, the ad hoc importation of snatches from other songs, and highly emotional reactions expressed by the performer and projected upon the characters. Assuming that the audience for which the ballad is being sung already knows the full story, the result will be in no way incoherent or confusing. Rather, if the performance is skillfully managed, the result is an exhilarating celebration of the realities and sorrows and possibilities of life.

University of Georgia Indiana University
Athens, Georgia Bloomington, Indiana

Footnotes:

1. Lomax chronicled the tours in chapter 5, "Penitentiary Negroes"( 1947:134-164) and chapter 6, "Iron H ead and Clear Rock"( 165-188). Also, some of Lomax's radio broadcasts given in the 1940s, collectively titled The Ballad Hunter: Lectures on American Folk Music (part 10, side B of the fifth disk), give reminiscences of and performances by Moses Platt and his colleagues. On the importance of Lomax's prison-collecting, see Wilgus 1959:185-186. The present authors thank Bess Lomax
Hawes and Joseph Hickerson for their advice.

2. Transcribed from Versions and Variants of Barbara Allen c. 1964:B19.

3. Ballad scholars never mention a slightly earlier commentary in which Moore specifically delineated the situation that Coffin would examine (of course, Moore, unlike Coffin, saw the process as a deterioration in quality):

In many of the fragments, the story has been lost so completely that only a name or two serves to associate these fragments with the complete ballads. In such cases, there is a marked tendency for these chips to lose the chief characteristicos f the old block, and to become lyrical in character. It is the story that seems to drop out first; it is the situation, with the lyrical comment upon it, which remains. [Moore 1914: 394]

4. A term that Abrahams and Foss (1968:20) have used in passing, attention-center, might be an apt replacement for both of Coffin's terms. To some degree Abrahams and Foss pondered the stylistic and generic gradation of ballad texts from narration t o purely dramatic dialogue to "lyric" (37-47).

5. Coffin (1957) and Abrahams and Foss (1968:48-60) have reviewed several other examples of lyrically-reduced ballads. Barre Toelken (1968:40-44) has commented on surmises by members of a folk group regarding the narratives implied by lyrical b allad f ragments. In some of those cases unlike the ones principally being examined here-the narrative may indeed have been "lost forever" from the recollection of the folk group (Coffin 1977 [1957]:172).

6. The sex-change is what has occasioned the most comment about the Forth Worth version. Such a change, however, had been on record for more than a decade when Bales published her text. A full nine-stanza version collected from rural white people in Mississippi in 1909 tells how "A young girl on her death bed lay,/ For the love of Barbry Allen"-and throughout the narrative "Barbry" is a man (Perrow 1915:146).

7. Transcribing the same text two decades later, Abrahams (1987:115) has re-edited the recurring line "Barbry Alone, Barbry Alone" as "Barbry alone, Barbry alone," tacitly changing the supposed surname or nickname into a mere adjective.

8. In their discussion of what they proposed as Celtic analogs of blues ballads, Wilgus and Long did include translations of several Irish versions of Child ballads, among them a text that is said to be a variant of "Barbara Allen" (1985:463-464).

9. The basis of Seeger's study was 76 whole or partial recordings available in the Library of Congress's Archive of American Folk Song (now the Archive of Folk Culture), supplemented by 200-odd recordings or transcriptions from other sources.

10. There exist, it should be noted, quite "regular" versions collected from African Americans for example, Smith's F-text from South Carolina, "a negro variant, communicated in 1913" (1928:138-140); or the C-text, "as sung by the Virginia darkies" about 1912, in Davis (1929:309- 311).

11. One of Platt's versions of "The Old Chisholm Trail" is transcribed in Lomax and Lomax (1938:37-39).

12. Quite regularly the dying cowboy implores his friends to "lay [or roll] the sod o'er me" in the stanza (often occurring as a refrain) that begins, "Beat the drum slowly...." Sometimes the protagonist asks for sixteen (though more often six) cowboys to accompany his corpse to the graveyard; "sixteen" t exts appear in Lingenfelter, D wyer, and Cohen (1968:426-427) and Lomax and Lomax (1938:417-420). The Lomaxes i ncluded a text from Platt's S ugarland c olleague J ames "Iron Head" Baker that is an interesting amalgam of "The Streets of Laredo" and one of its prototypes, "St. James Hospital"; Baker's dying cowboy desires "Sixteen young whore gals for to sing me a song" (420-421).

13. Such a stanza i s also featured i n other play-party o r dance songs, including "Josey," "Susan Brown,"" Four in the Middle," a nd often one separately t itled "Coffee G rows o n White O ak T rees." Most general and regional folksong collections contain texts. See especially Botkin 1937. Thomas Talley (1922:107) collected an African-American variant that does some interesting folketymologizing:" Coffee grows o n w'ite folks' t rees." I ncidentally, Talley included t wo very different versions, neither of which he regarded as either a "dance song" or a "play song"; rather, one is classed among "love rhymes" (127), the other in a miscellaneous class of "pastime rhymes" (107). Interestingly, "Barbara A llen" itself has been reported as a play-party song (Newell 1903:78-79).

14. Versions derived from the popular Forget Me Not S ongster (by 1850, editions had issued from a dozen or more publishers in New York, Philadelphia, Boston, Saint Louis, Montreal, and probably elsewhere) often contain a stanza probably borrowed from "James Harris," in which the man seeks to lure Barbara by the offer of ships, and a stanza apparently adapted from "Sir Patrick Spens" (or one of its cousins), in which Barbara reads a letter and laughs; the prototype can be seen in Forget Me Not Songster(1 974 [1835]:142-144). Examples of derivativet exts of "Barbara Allen" (the first from an Illinois manuscript of 1866) are given by Musick (1947:210-212) and MacKenzie (1928:35-37). The Z-text in Davis (1929:340-341) contains the line "I fell in love with a lady like a dove," from "Pretty Peggy O," versions of which appear in Sharp 1932 (2:59-61). Davis's L-text of "Fair Margaret and Sweet William"(2 36-237) is a composite of that ballad and "Barbara Allen." Sharp's B-text of Child 7, "Earl Brand" (1:16-17), contains a stanza imported from "Barbara  Allen."

15. Ohio occurs as a rhyme for briar in a Virginia version of "Barbara Allen" from 1915, the E-text in Davis (1929:313-314); and also in a 1922 version of "Lady Margaret and Sweet William," Davis's O-text (1929:239). The transcription a-tire appears in the GG*-text in Davis 1960:196-198.

References Cited
Abrahams, Roger. 1968. Charles Walters-West Indian Autolycus. Western Folklore 27:77-95.
Abrahams, Roger. 1987. Child Ballads in the West Indies: Familiar Fabulations, Creole Performances. Journal of Folklore Research 24:107-134.
Abrahams, Roger, and George Foss. 1968. Anglo-American Fo lksong S tyle. E nglewood Cliffs, NJ: Prentice-Hall.
Bales, Mary Virginia. 1928. Some Negro Folk-Songso f Texas. Publication o f the Texas Folklore Society 7:85-112.
The Ballad Hunter: Lectures on American Folk Music. No date. By John Lomax. Library of Congress AAFS L49-53 (78-rpm phonograph disks).
Beckwith, Martha W. 1924. The Child Ballad in Jamaica: A Note upon the Origin of the Ballad Form. PMLA 39:455-483.
Botkin, Benjamin A. i937. The American Play-Party Song, with a Collection of Oklahoma Texts and Tunes. Lincoln: privately printed.
Bronson, Bertrand. 1962. The Traditional Tunes of the Child Ballads. Vol. 2. Princeton: Princeton University Press.
Coffin, Tristram P. 1957. "Mary Hamilton" and the Anglo-American Ballad as an Art Form. Journal of American Folklore 70:208-214. Rpt. in Coffin 1977:164-172.
Coffin, Tristram P. 1977. The British Traditional Ballad in North America. Rev. ed., with a supplement by Roger deV. Renwick. Austin: University of Texas Press.
Cray, Ed. 1966. Commentary on the Words [of "Barbara Allen"]. Appended to Seeger 1966:164-167.
Davis, Arthur Kyle. 1929. Traditional Ballads of Virginia. Cambridge, MA: Harvard University Press.
Davis, Arthur Kyle. 1960. More Traditional Ballads of Virginia. Chapel Hill: University of North Carolina Press.
Forget M e Not Songster, 1974 [1835]. Rpt. Norwood, PA: Norwood Editions.
Gerould, Gordon Hall. 1957 [1932]. The Ballad of Tradition R. pt. New York: Oxford University Press.
Hyman, Stanley Edgar. 1957. The Child Ballad in America: Some Aesthetic Criteria. Journal of American Folklore 70:235-239.
Lingenfelter, Richard E., Richard A. Dwyer, and David Cohen. 1968. Songs of  theA merican W est. B erkeley: University of California Press.
Lomax, John. 1947. Adventures of a Ballad Hunter. New York: Macmillan.
Lomax, John, and Alan Lomax. 1938. Cowboy Songs and Other Frontier Ballads. 2d rev. ed. New York: Macmillan.
Moore, John Robert. 1914. Omission of the Central Action in English Ballads.  Modem Philology 11:391-406.
Moore, John Robert. 1916. The Influence of Transmission on the English Ballads. Modem Language Review 11:385-408.
MacKenzie, W. Roy. 1928. Ballads and Sea Songs from Nova Scotia. Cambridge, MA: Harvard University Press.
Musick, Ruth Ann. 1947. The Old Album of William A. Larkin. Journal of American Folklore 60:201-251.
Newell, W. W. 1903. Games and Songs of American Children. 2d ed. New York: Harper.
Perrow, E. C. 1915. Songs and Rhymes from the South: Part 3. Journal of American Folklore 28:129-190.
Seeger, Charles. 1966. Versions and Variants of the Tunes of "Barbara Allen." Select R eports [o f the Institute of Ethnomusicologya t UCLA] 1 (no. 1): 120-163. Reproduced (with the same pagination) as a brochure accompanying Versions and Variants of "Barbara Allen" (c.1964).
Sharp, Cecil J. 1932. English Folk Songs from the Southern Appalachians. Edited  by Maud Karpeles. 2 vols. London: Oxford University Press.
Smith, Reed. 1928. South Carolina Ballads, with a Study of the Traditional Ballad To-day. Cambridge, M A: Harvard UniversityP ress.
Talley, Thomas. 1922. NegroF olkR hymesN. ew York: Macmillan.
Toelken, Barre. 1986. Context and Meaning in the Anglo-American Ballad. In The Ballad and the Scholars: Approaches to Ballad Study, by D. K. Wilgus and Barre Toelken, 31-52. Los Angeles: William Andrews Clark Memorial Library.
Versionasn d Variants of "Barbara Allen."Circa 1964. Edited by Charles Seeger.
Library of Congress AAFS L54 (LP phonograph disk). Accompanying brochure Seeger 1966 (including Cray 1966).
Wilgus, D. K. 1959. Anglo-American Fo lksong S cholarship sin ce 1898. 2d ed. New Brunswick: Rutgers University Press.
Wilgus, D. K., and Eleanor Long. 1985. The Blues Ballad and the Genesis of Style in Traditional Narrative S ong. In Narrative F olksong: New Directions. Essays in Appreciation of W.E dson R ichmond, ed ited by Carol L. Edwards a nd Kathleen Manley, 437-482. Boulder: Westview.