Robert W. Gordon and the Second Wreck of "Old 97"

Robert W. Gordon and the Second Wreck of "Old 97"
by Norm Cohen
The Journal of American Folklore, Vol. 87, No. 343, Archer Taylor Memorial Issue (Jan. - Mar., 1974), pp. 12-38

[Fantastic article by Norm Cohen on the "The Wreck of the Old 97." I've edited the article once, might be some typos on my part. The footnotes are at the end.

I'm including the MP3 Robert Gorden recorded in 1925 of "Old 97" by Charles Noell and the two main influential 'country" recordings. R. Matteson 2011]

Click to Listen: "Old 97" by Charles Noell Gordon Collection

Click to play: Wreck on the Southern Old 97- Henry Whitter

Click to Listen: Wreck on the Southern Old 97- Vernon Dalhart 1924


Robert W. Gordon and the Second Wreck of "Old 97"
NORM COHEN

ROBERT WINSLOW GORDON came closer than most of his peers in gathering a national folksong garland. [1] "His collecting activities during the 1920s were prodigious. From 1923 to 1927, he edited a column in Adenture and accumulated a file of nearly four thousand letters, almost everyone with one or more song texts, from some two thousand correspondents across the country, Between 1922 and 1929, his field expeditions, the most ambitious up to that time, netted over one thousand cylinder recordings of traditional singers and a large quantity of additional unrecorded texts. A single three-month field trip in 1926 yielded nearly seven hundred items. His sea chantey collection held some 1300 songs, almost four hundred of which were from oral sources.[2]

As first archivist of the Library of Congress Archive of American Folk Song, Gordon began building that repository with his own private collection as a nucleus. While archivist he acquired and indexed several major manuscript collections totalling some two thousand additional songs, and he enriched the Archive with an impressive collection of pulp publications, such as songsters, songbooks, and broadsides, that most other folksongs cholars of the day would have regarded as next to valueless.

Strange to say, Gordon never published a collection of his materials as did many of his contemporaries during the 1920s. Strangely, he never contributed a single article to any of the established folklore journals. Students of American folksong interested in his work have only his columns in Adventure, a series of articles in the New York Times, and an essay on Negro spirituals as the complete corpus of his publications.[3]

Although his scope was remarkably broad, both geographically and stylistically, he seems to have been particularly interested in native American balladry. Of the song "Frankie and Johnny" he wrote, "I possess over 90 separate and different versions together with more than twenty-five accounts, most of them inaccurate as to its origin and history."[4] His files held large quantities of material relating to "Casey Jones," and apparently he was preparing to publish a study of the "Titanic." Yet, today, folksong students must learn of his work chiefly through brief references and footnotes in the work of other scholars.[5]

Why did Gordon never publish a case study based on his researches? I have heard only one explanation offered: G ordon was too much of a perfectionist who was never satisfied with what information he had at hand." One can sympathize with the reason if indeed it was the reason-only to a limited extent: no folksong study can ever be absolutely complete and to hold publication on that account is surely unrealistic.

That Gordon was an energetic, thorough, and perspicacious worker can be demonstrated by the mass of materials related to one ballad that he studied for several years, "The Wreck of the Old 97." In this article I would like to discuss the career of that ballad and Gordon's investigation of its history. As necessary background I shall review the early history and development of the ballad out of a complex of nineteenth-century folk and popular ballads, and how the song came to be a million-selling record hit worthy of fifteen years of investigations, lawsuits, and counter-suits. The remainder of the article will focus on Gordon's efforts to establish a uthorship o f the ballad and on the way in which the courts handled the evidence of a folklorist and teams of lawyers to pass judgment on the claim of copyright infringement.

Because much of the story has been well told by others, only a skeletal account of the wreck itself will be given here.[7] Number 97 was a fast mail train on the Southern Railway that ran between Washington and Atlanta from December 1902 to January 1907. On Sunday, September 27, 1903, because of various delays, "Old 97" reached Monroe, Virginia, 165 miles south of Washington, about an hour late. At that station the crew was changed, and the engineer who took command was Joseph A. Broady, nicknamed "Steve" after the vaunted Daredevil Steve Brodie who leaped from the Brooklyn Bridge on a bet in 1886 and lived to collect.

Broady, who had come to the Southern Railway only a month earlier from the Norfolk and Western, was unfamiliar with the route of "Old 97" and highballed it southward trying to make up some of the lost time. Near Danville, Virginia, was Stillhouse Trestle, a wooden bridge spanning Cherrystone Creek seventy-five feet below. The trestle was preceded by a curve and a descending grade, a combination that made it a danger point. Broady came down White Oak Mountain toward the trestle too fast, and the locomotive, together with the five wooden cars behind it, flew off the rails and hurtled into the ravine below, completely destroying the train, killing nine persons (including engineer, both firemen, conductor, and flagman), and injuring seven others.

The history of the ballad, however, begins long before Broady plunged "Old 97" into a cow pasture beneath Stillhouse Trestle. It begins in 1865, when Henry Clay Work, one of the leading popular composers of the 1860s and 1870s, wrote "The Ship That Never Return'd." Since the development of the later parodies (for want of a better term, I use this word in the classic sense of any literary imitation, humorous or otherwise) is more easily followed if Work's song text is available for comparison, I give here the original words:

On a summer's day when the wave was rippled
By the softest gentlest breeze,
Did a ship set sail with a cargo laden
For a port beyond the seas.

There were sweet farewells-there were loving signals,
While a form was yet discern'd,
Though they knew it not, 'twas a solemn parting,
For the ship-she never return'd.

Did she never return? She never return'd-
Her fate, it is yet unlearn'd;
Tho' for years and years there were fond ones watching,
But the ship-she never return'd.

Said a feeble lad to his anxious mother,
"I must cross the wide, wide sea,
For they say, perchance in a foreign climate,
There is health and strength for me."

'Twas a gleam of hope in a maze of danger,
And her heart for her youngest yearn'd;
Yet she sent him forth with a smile and blessing,
On the ship that never return'd.

"Only one more trip," said a gallant seaman,
As he kiss'd his weeping wife,
"Only one more bag of the golden treasure,
And 'twill last us all through life.

Then I'll spend my days in my cozy cottage,
And enjoy the rest I've earned;"
But alas, poor man, for he sail'd commander
Of the ship that never return'd.[8]

Like many of Work's songs, including "Grandfather's Clock" and "Come Home, Father," "The Ship That Never Return'd" achieved wide-spread popularity, eventually entering the folk tradition to enjoy a life that far outlasted the fifty-six years extended by the copyright law. The tune changed somewhat, although that does not alter our story. A sure indicator of a song's popularity is the occurrence of parodies, and several appeared based on Work's ballad in the 1880s, if not earlier. The only one for which the date and author are documented is "The Hand-Car That Never Returned," subtitled "A Parody on 'The Ship That Never Returned,'" copyrighted in 1888 by its author, Alexander Malin, and published on a 6" x 81/2" broadside. The first verse of this parody, which describes a handcar that leaves the rails at a trestle and hits a passing ship in the waters below, is as follows:

On a Winter's day, as the Train was whistling through the "Transcontinent Galate,"
A Hand Car started, with its burden laden, over Section seventy-eight.
There was many a joke a mong the "Jerrys," for their thoughts were unconcerned,
They little knew 'twas their last sad voyage on the Hand-Car that never returned.[9]

Whether the author had created the first in a series of railroad parodies on Work's song, or whether he already had knowledge of an earlier train parody, we do not know. From an artistic point of view it might make a more interesting theory to suppose that there had already been train parodies and that Malin had composed a clever amalgam of the two themes in a kind of double parody.

Evidently, the earliest train parody of Work's song was "The Train That Never Returned."Although composer and date are unknown, if we can believe Stout's informant, it was current at least as early as 1888.[10] A rather disjointed text of this song was printed in Railroad Magazine with the note that the song was reprinted "from a song sheet issued in about 1900 by the blind singer O. F. Saunders, Burlington Jct., Mo." "'Other versions can, by virtue of various distinguishing textual features, be grouped into two categoriesd istinctf rom one another and also distinct from the Saunders/Stout song, which for convenience will be called "The Train That Never Returned #1."

"The Train That Never Returned #2" is a relatively stable group of texts all of which have three stanzas and a chorus and are localized to the C&O Railway.[12] Since the earliest of these was published in 1927, with no indication of the date of first appearance, we cannot be sure that "#2" predates" The Wreck of the Old 97." It is distinguished by the opening line, "I was going 'round the mountain one cold winter day," which is distinct from the opening lines of "#1" and "#3," all of which are something like "On a cold winter night when the smoke was curling." "The Train That Never Returned #33" consists of several versions all of which have distinct localizations and not quite perfectly parallel verses; they seem to be more closely related to each other than to the other families.[13] All these songs have a distinctive line about painting the town red, although in different cases the town is Corbin, S omerset, or Pinole.

The other important member of the complex family of predecessors to "The Wreck of the Old 97" was probably titled "The Parted Lover," although it has also been remembered by the titles "The Face That Never Returned" and "Lovers Parted." We can be certain that it predated 1903 because Charles W. Noell, who wrote one of the ballads about "Old 97," stated that he had set his poem to the tune of "The Parted Lover." The first seven stanzas of the song tell of a lovers' quarrel, on account of which the young man left his lover and rode (or sailed) away to a foreign country. His heart was still filled with pain; likewise, she turned pale and, paincrushed, bathed her pillow nightly in tears. Then the concluding moral:

Now, young men and maids, from my song take warning
Or your hearts will break with pain.
Never speak harsh w ords to a faithful over
Or he'll leave you to never return.[14]

There were probably numerous other parodies of Henry C. Work's 1865 composition in circulation by 1903, but I will mention only one more for it too is a train song. In 1927, Fiddlin' John Carson of north Georgia recorded "Did He Ever Return?" a fragmentary ballad that began:

Oh, what did the bad boy say to his mother,
When he swung to the side of a train?
"Some day I know, I will be a conductor,
On that fast mail train."

Did he ever return, no he never returned,
And the poor boy must be dead.
His body was found on a cold iron railing
And the snow had covered his head.[15]

Most singers learn the bulk of their repertories in their youngery ears. Since Carson was thirty-five-years old at the time "Old 97" flew off Stillhouse Trestle, it is likely that the ballad he imperfectly remembered predated that accident. It is evident, then, that half a dozen or more parodies on "The Ship That Never Return'd" were in oral circulation in the southern mountains in 1903; several of them involved trains. It is not unlikely, therefore, that different persons should write ballads about "Old 97," all independently using the same tune--even though such an occurrence would generally be taken to indicate that the different ballads its borrowed from one another.[16]

Precisely how many ballads were penned after the accident of September 1903 and who the various authors were is not known; what facts we do have at our disposal are largely due to Gordon's researches. During the summer of 1923, while teaching at the University of California at Berkeley, Gordon took over the column "Old Songs Men Have Sung," a regular feature of the pulp magazine Adventure. During the preceding February, his predecessor in charge of the column, Robert Frothingham, received a letter from a West Virginia correspondent requesting the words to "a railroad song about a wreck on the southern road.' Old 97 pulled in to Monroe Virginia twenty minutes behind her time[;] they gave him his orders saying Stephen your way behind.'[17]

In his column for August 20, 1923, Gordon printed this verse and asked readers familiar with the song to submit the complete text. Soon he had received three responses, and in the January 30, 1924, issue he printed a composite text of fourteen stanzas that he had collated from the three versions.[18]

In 1925, Harvard University appointed Gordon a Sheldon Fellow for the collecting of folksongs. In October he was on his way to Asheville, North Carolina, to collect material when he happened to see a Greensboro newspaper article recounting the story of the wreck of the Southern's fast mail in 1903. His appetite for information on the subject having been whetted by his earlier correspondence, Gordon stopped at the Greensboro newspaper office to search their files for further information. There he talked with newspapermen about the song and was referred to two men, Charles Noell and Fred Lewey, who, he was told, could give him more information. Gordon secured Noell's address in Greensboro and, on October 15, visited him and recorded his version of the song. Charles Weston Noell, seventeen years old at the time of the accident, told Gordon that he had written the song about the wreck.[19]

That same evening Gordon drove to Concord, North Carolina, where he found and interviewed Frederick Jackson Lewey and recorded his version of the ballad as well. Lewey asserted he had composed the ballad. Evidence that G ordon later gathered led him t o the conclusion that indeed Lewey had first begun the ballad about the wreck. To this point I shall return below.

Early in the summer of 1925, before he went on his field trip, Gordon had opened negotiations with the Victor Talking Machine Company to see if they might be willing to underwrite part of his collecting in the hope of obtaining recordable material and if he might be able to act as folksong consultant to them. Gordon argued that Victor needed an expert to advise them in matters concerning copyrights and authorship of folksong material. He felt that "the company was treading on very dangerous ground in certain instances where to copyright was, say the least, extremely questionable. I knew that in a number of cases the firm was paying royalty to unscrupulous pretenders who had no [vestige] of right in the texts they sold; and I knew that in other cases there were ample grounds for suit for infringement if only the facts happened to fall into the hands of the right parties." [20]

There was some truth to Gordon's contention, yet I wonder if he appreciated the enormity of the problems he was skirting. The new field of hillbilly phonography was turning up much traditional material, including some that, although firmly in oral tradition, was still protected by legitimate copyright. Occasionally, the authorship of such pieces was unrecognized; not infrequently, performers or writers or A&R (artist and repertory) men were claiming new copyright on such material.

On the other hand, there were many instances of copyright claims on material unquestionably in the public d omain.[21] For example, at the time Gordon wrote the above statement, he might have referred to "The Letter Edged in Black," an 1897 popular hit still protected by copyright when it was recorded by Vernon Dalhart without author or composer credits in 1925. In the category of copyright claims on public-domain material could be placed Carson Robison's "Naomi Wise."

It seems clear now that, if a song protected by copyright is collected from oral tradition and commercially published, the original copyrighth older is entitled to compensation. If, however, the collected version is sufficiently different from the copyrighted one, then someone (collector, informant, and record company executive have on occasions tepped in here) is entitled to an arrangemenct opyright. Similarly, a new arrangement on an old public-domain song can be copyrighted. The important question in both instances is how much change must be made before the extent of originality is sufficient to qualify the new version as a new composition or a new arrangement.

With regard to such practices as Robison's rewriting an old song in the public domain, a pertinent decision was handed down in the case of Wihtol v. Wells. Plaintiff had written and copyrighted words and music of a religious song, "My God and I," using as the tune a Latvian folk melody he remembered from his youth. Defendant had published and copyrighted another song with same title, different words, and substantially same tune. Plaintiff sued for plagiarism. The lower court agreed with the defendant that "the tune of the song is incapable of being protected by copyright. "The appeals court reversed the decision because t he plaintiff had sufficiently a dded to a public-domain tune." It was original work on plaintiff's part when, some thirty years later, he devised a calculated m elody score thus putting it in shape for all to read." He had made a sufficient original contribution to justify copyright.[22] Of course, the problem is always to decide what constitutes o riginality a nd how much of it is required; there is no formula that provides the answer to this issue.[23]

To a folklorist, there is probably something unsavory about the entire preceding discussion of copyright. From his perspective, a folksong passes from the original creator (I ignore the discarded theory of communal creation) through a succession of others, each possibly modifying the song in some slight way before passing it on to the next member of the community. It is generally not regarded as normal that one link in this chain of transmission change the song in such a profound manner as to confer upon the responsible individual the privilege of copyright. I suspect, however, that this is an instance of facts being ignored in the face of theory. Folksong scholars who do search for evidence of folk creativity do find it, and I am convinced that it lurks undiscovered in more cranniest han we have searched. In other words, folksongs do not necessarily change in a gradual, continuous fashion, but can change discontinuously and precipitously.

The real question is whether such discontinuous changes constitute evidence of originality in terms of copyright law, or whether, as some would argue, even drastic changes in a folksong can be understood in terms of folk conventions, cliches, and motifs-so that change is still not proof of originality, but only of the vitality of a folk tradition.

Gordon's interest in these matters extended beyond the boundaries of the continental United States. In the letter to Richards, quoted a bove, h e noted, "I still am a bit amused over the 'Didn't He Ramble' matter. I can trace that song for over 125 years, -not of course in the modern version. But the modern version comes from an extremely indecent folk version, which, in turn, goes back to a decent 'author' version, and this, in turn to still earlier folk."[24] Gordon would have been admirably qualified for the role of consultant to the Victor Company; few scholars in 1925 appreciated as he did the role of the fledgling hillbilly and race recording industries or were as familiar with such facets of native American folk music as broadsides, sheet music, songsters, newspaper columns, and phonograph records.

Gordon's overtures to Victor were carefully considered but, by a narrow vote, turned down. A few days after this outcome was made known to him, Gordon received from Victor's legal department a request for "something of the history of the tune of the song 'Ships that Never Return.' "[25] Gordon, believing that Victor was trying "to obtain for a two-cent stamp what [the company] was not willing to pay for,"[26] ignored the query. A follow-up letter seven weeks later, bearing the notation "Re-'Wreck of the Southern Old 97,' " reiterated the request,[27] but Gordon still refused to reply. Not until November of the following year did Gordon find out, indirectly, that Victor's one hand did not know what the other was doing and that the request for information had come through channels completely separate from those that had decided upon his proposal to act as consultant.

Gordon apologized to several of Victor's executives and was again on good terms with the company, but his involvement in the story of "The Wreck of the Old 97" and its court trial was not resumed until 1929. Why, one may well ask, was Victor, in July of 1925, suddenly concerned with the origin of "The Ship That Never Return'd?" The catalyst of the sequence of events that led to the inquiry was Henry Whitter. Born in Grayson County, Virginia, near Fries, Whitter made his living in the cotton mills, although he also played banjo, violin, guitar, and harmonica. Whitter's role in the genesis of the hillbilly phonograph industry has been documented by Archie Green; suffice it to note here that, in December of 1923, Whitter journeyed to New York and waxed, among other selections, "The Wreck on the Southern Old 97."[28] His recording was released the following January.

Not for very long was Whitter's recording the only one available describing the accident on the Southern Railway. In April, Ernest Thompson, a blind singer and musician from Winston-Salem, North Carolina, recorded the song for Columbia. Thompson's text included all five stanzas of Whitter's version but began with a different stanza-proof, if any was needed, that Henry Whitter was not the only person in 1924 responsible for keeping" Old 97" in the public memory. Also late in April, George Reneau, a blind singer and guitarist from Knoxville, Tennessee, journeyed to Aeolian Vocalion's New York studios to make his first recording. It was a "cover" of Whitter's first disc, and his text of "97" was the same as Whitter's.[29]

Whitter recalled that he was paid $25 for each song he recorded in 1923 and a percentage in royalties. He did not recall how much he received in royalties, and we have no knowledge of how widely his first record sold, but one interested listener was Vernon Dalhart. Dalhart, one-time popular singer turned hillbilly recording star, was given a copy of Whitter's disc during the summer of 1924. He copied the words (as best he could understand them) and recorded the ballad for Edison on May 14, 1924. His rendition was released on Edison 51361 during the following July or August. What had prompted Dalhart, who had established himself first as a singer of light opera on stage and later as a recording artist of popular tunes, to turn to hillbilly music is still open to question. In any case, the Edison disc and then cylinder releases old sufficiently well to prompt Dalhart to persuade Victor to let him record the ballad for them.

This was eventually agreed to and the song, waxed on August 13, was released on October 3, 1924. Although "The Prisoner's Song," which was coupled with "97," was the cause of the phenomenal success of that disc, the fact remains that Victor 19427 sold over one million copies in the next three years.[30] The records old so well that in March 1926 Victor called Dalhart back to the studio to re-record the songs on 19427 by the new electrical process. The new "takes" were issued on the same release number, 19427.

On October 1, 1924, just three days before the Victor catalog supplement announcing the new release, # 19427, was issued, the song "The Wreck on the Southern Old 97" was copyrighted by F. Wallace Rega (a publishing house owned by OKeh's executive, Fred Hager), with words and music credited to Henry Whitter (E600188). A second copyright was taken out on November 1 by the General Phonograph Corp., parent company of the OKeh label, still with words and music credited to Whitter (E602770). Some time later, OKeh filed a claim against the Victor company for copyright infringement. Doubtless in the hope of finding a way to circumventa ny obligationst o OKeh, Victorb egana searchf or antecedents to the song to invalidatet he copyright.[31]

Thus it came a bout that they contacted Robert W . Gordon, noted folk music authority at Harvard. Receiving no reply from him, they set about pursuing their own investigations in North Carolina, practically crossing paths with Gordon, who had just visited and recorded Noell and Lewey. Victor's field work was less decisive than Gordon's had been. Their lawyers later claimed in court that Victor's representatives had scoured the countryside seeking information and publishing notices in newspapers that Victor was searching for the author of "97" so they could pay him royalties. They found no one.

Therefore, they evidently decided to make peace with the OKeh people, and on April 8, 1926, Victor paid F. Wallace Rega $3500 for all rights to the ballad. On April 26, 1926, Henry Whitter sold to Hager for $1000 all his rights, including mechanical, to the song, except for a percentage of royalties on any discs sold by OKeh's new owner, the Columbia Phonograph Company. The sequence of these two transactions is, to say the least, curious.[32]

On March 1, 1927, a notice in the Richmond, Virginia, News Leader read:
"The Wreck of the Old 97. When did the wreck of Old 97 happen and how many were killed? If we could locate an answer to this question we could win enough money to retire in peace."[33] The following day the editor summarized details of the wreck and noted, "Offers of money by the Victor Company, as royalty on the record, are said to have met with no takers."[34] Three days later, one David Graves George wrote to the News Leader a letter stating, among other things, "I with others composed the poetry of 97." The letter was published in the March 7 edition. [35] S ometime thereafter, George appeared at the office of the News Leader and requested the loan of the letter, which he never returned. When it was demanded and handed over at the trial, it was seen that the phrase "with others" had been erased and replaced with the word "alone."

On March 16, George opened negotiations with Victor, claiming his rights to royalties; Victor refused to settle with him, and on October 10, 1928, a Bill of Complaint was filed. In April of 1929, Victor again turned to Gordon for help. In view of the evidence he had gathered, the company decided that Lewey and Noell had defensible claims to authorship and purchased from them their rights to the song for $100 each in March and May 1929, respectively. Perhaps Victor felt that settling these accounts would be of value to them in the pending trial with George, and they were most anxious to obtain from Gordon documents substantiating the claims of Lewey and Noell.

On April 15, Gordon's services were officially requested to aid Victor in defending their suit against David Graves George, and for the next two years Gordon spent much of his spare time on the case. (Gordon was careful to write all his correspondence in the case from his residential address, rather than his business address at the Library of Congress.) He estimated that during that period he spent in excess of 1000 hours in such work, more than 480 of which were spent away from Washington on specific requests of Victor.[36]

The results of Gordon's labors formed the basis of RCA Victor's defense at the trial."' Here follows the early history of the ballad, largely as Gordon reconstructed it based on his own investigations and the results of pre-trial yestimony taken during 1929. At the time of the wreck of "Old 97," Fred Jackson Lewey was eighteen years old and was living in Danville; he was one of the first on the scene after the accident. About a week later he began writing a song about it. It took him about two months to complete the ballad, by which time he was living in Lynchburg. While in that city, he sang it in the presence of several men, including Charles Noell, at the home of Noell's fiancee, who lived next door to the residence where Lewey was staying. Lewey gave the words to the song three times during the investigations: on October 14, 1925, to Gordon; again on the following day, after Gordon had played Noell's recording for him; and in April 1929, in Greensboro, North Carolina, at pre-trial hearings. The texts of the three versions differed but slightly. Below is the text as transcribed by Gordon from the October 14, 1925, recording:

Click to Listen: "Old 97" by Charles Noell

Last evening I stood on a mountain
Just watching the smoke from below;
It was springing from a long slender smokestack
Way down on the Southern Road.

It was Ninety Seven, the fastest train
That the South has ever seen;
But she run too fast that fatal Sunday evening,
And the death list numbered fourteen.

Did she ever pull in? No she never pulled in,
Though at one forty-five she was due;
For hours and hours has the switchman been watching,
For the fast mail that never came through.

Well, the engineer was a brave, fast driver
On that fatal Sunday eve,
And his fireman leaned far out in Lynchburg
Waiting for the signal to leave.

When he got the board, well, he threw back his throttle
And although his air was bad
People all said when he passed Franklin Junction
That you couldn't see the men in the cab.

Did he ever pull in? No he never pulled in, etc.

There's a mighty bad road from Lynchburg to Danville,
And although he knew this well
He said he'd pull his train on time into Spencer
Or he'd jerk it right square into hell.

When he hit the grade from Lima to Danville
His whistle began to scream;
He was found when she wrecked with his hand on the throttle
Where he'd scalded to death from the steam.

Did he ever pull in? No he never pulled in, etc.[38]

The only significant change in later renderings was in the first line: on October 15 he sang, "On a bright Sunday evening"; and in court he recited, "On a bright Sunday morning." Lewey recalled for Gordon one stanza that he said he had written but discarded because it was unsatisfactory:

Did he ever come back? No, he never came back
It's told with a soft, hushed breath-
His poor little wife fell back and went crazy
When the news came home of his death.[39]

Lewey stated that he set his poem to the tune of "The Ship That Never Return'd." Another of Gordon's informants recalled Lewey as a lad who spent much of his time sitting around singing and playing the guitar.[40] Charles Weston Noell, seventeen at the time of the accident, was a cotton mill worker until 1923, when he became a policeman. He played guitar, mandolin, banjo, and organ. He claimed to have written the song himself, but since it was reliably reported that as early as 1904 he and Lewey used to sing and play it together, it is easy to conclude that their respective versions could not have been completely independent.[41] According to his testimony, not long after the wreck he submitted a copy of his song to the Mill News, a Charlotte, North Carolina, newspaper, but it was not published, because the editor did not believe it was of sufficient interest to the mill workers.[42] The text Noell gave Gordon in 1925, the text submitted at the Danville pre-trial hearings on April 4, 1929, and the version he recited orally at the hearings on April 5, 1929, all differ from one another considerably. I give here the first of these three, presumably the most reliable version of the early form of the song he composed or sang:

Come all of you fellows and gather around me,
And a sad sad story to hear;
All about the wreck of Old Ninety-Seven,
And the death of the brave engineer.

At the Washington Station on that bright Sabbath morning,
'Twas just at the rising of sun;
When he kissed his wife, said, "My children, God bless you,
Your father must go on his run."

Steve Broady was the engineer,
And a brave, brave man was he;
For a many poor man have lost his life
For the railroad company.

Ninety-Seven was the fastest train,
That was ever on the Southern Line;
All the freight trains and passengers had to hold for Ninety-Seven
She's compelled to be at stations on time.

At Monroe, Virginia, he received his orders,
Saying, "Steve, you are way behind;
This is not Thirty-Eight, but it's Ninety-Seven,
You must -put her into Danville on time."

He climbed in his engine at Monroe, Virginia,
Saying," Fireman, it's do or die!
I'll reverse the lever, throw the throttle wide open,
We'll watch Old Ninety-Seven fly."

Steve Broady he was that engineer,
On that fatal Sunday eve;
And his fireman was leaning far out at Lynchburg,
Just waiting for the signal to leave.

When they gave him the board he threw back his throttle,
Although his airbrakesw as bad;
And the people all said when he passed Franklin Junction
That it seemed like the engineer was mad.

Steve Broady he said to his black and greasy fireman,
"Just throw in a little more coal;
And when I turn over White Oak Mountain,
You just watch my drivers roll!"

Now it's a awful bad road from Lynchburg to Danville,
And from Lima it's a three-mile grade;
It was on this grade that his airbrakes failed him,
And look what a jump she made.

Falling down this grade at eighty miles a hour,
His whistle began to scream;
He was found in the wreck with his hand on the throttle,
Where he'd scalded to death from the steam.

When the news came slipping o'er the telegraph wires,
And this is the way it read:
"That brave engineer that pulled Ninety-Seven,
Is lying in North Danville, dead."

Did she ever pull in? No, she never pulled in,
At three forty-five she was due;
It was hours and hours dispatchers were w aiting,
But that fast mail never came through.

Did she ever pull in? No, she never pulled in,
You could hear it in silent breath;
His poor little wife fell back and fainted,
When the news came home of his death.[43]

The principal differences in the version given on April 4, 1929, were the inclusion of two additional stanzas: one, after the fourth given above; the second, after the eighth:

Ninety-Seven was the fastest train,
That was ever on the Southern Line;
But when she pulled in at Monro Virginia,
She was Forty-Sevenm inutes behind.

The conductor said to the engineer and fireman,
"Don't neglect that whistle or bell;
For we must put this train on time in to Danville,
Or we will drop her right in to Hell."[44]

Noell told Gordon he had set his song to "the tune of 'The Parted Lover' and not into the tune of 'The Ship That Never Returned.' At the time of my writing this poem I did not know the tune of 'The Ship That Never Returned.' ... you can well catch my voice on this record as being a little bit lower and a little bit slower and milder than the tune of 'The Ship That Never Returned' which is directly in accordance with the tune of 'The Parted Lover.' "[45]

As mentioned above, the tune of "Ship" that has been traditional for fifty, possibly one hundred, years, is different from Henry C. Work's original melody. The familiar tune to "Parted Lover," Noell's tune in 1925, is indistinguishable from that commonly used for "Ship," "97," and all the other parodies in this family. Noell's various texts suggest that he obviously was familiar with Lewey's song but changed and expanded it considerably. Although only a few of Noell's verses correspond to lines in Lewey's rendition, we do not know how much his original version utilized verses that Lewey had composed and later rejected: note, for example, Noell's final two stanzas. He told the court that the penultimate stanza he gave Lewey credit for writing, but the final one was his own. Lewey, in 1925, before he heard Noell's recording, spontaneously recalled that final stanza as one that he had written and rejected.[46]

Lewey's song (in its final form) had a slight debt to earlier versions of "The Train That Never Returned" (compare his first stanza and chorus). Noell's ballad borrowed considerably from Lewey's but probably added much that was new. Gordon concluded that "Lewey started off the whole affair and that at a later date, perhaps about 1905-6, someone else made additions."[47] To Victor, Noell's claim was evidently as valid as Lewey's; whether a court of law would substantiate Noell's claim to authorship has never been critically tested.

As indicated above, Noell and Lewey frequently performed their song(s), either apart or together, in various towns they lived in, including Fries, Danville, and Lynchburg. In about 1904, Frank Burnett, another Virginia cottonmill hand, heard Noell singing the song to Lewey's guitar accompaniment. In November 1930 he recited for the court the song as, to the best of his recollection, he learned it from Noell and possibly from Lewey:

It was one bright Sunday morning, as I stood on the mountain
I was watching the smoke from below.
It was springing from a tall and slender smokestack
Waydown on the Southern railroad.

It was 97 the fastest mail
That the south had even seen
But she run too fast on the fatal Sunday morning
And the death list was numbered fourteen.

Did she ever pull in, no she never pulled in
At one-forty-fives he was due
It was hours and hours have the switchmen been waiting
For the fastm ail never pulled through.

97 she runs from Washington to Atlanta
She runs it rain or shine
All freight trains and passengers take the side for 97
She is bound to be at stations on time.

They give him his orders at Monroe, Virginia,
Saying "Steve, you are way behind time,
This is not 38 but its old 97
You must put her into Spencer on time."

He turned and said to his black, greasy fireman,
"Just throw me in a little more coal,
And when we cross the White Oak mountain,
You can watch old 97 roll."

It was a mighty bad road from Lynchburg to Danville
And Lima was a three mile grade,
It was on the grade that he lost his air brakes
Just see what a jump he made.

He went speeding down the grade making 90 miles an hour,
And his whistle began to scream,
He was found in the wreck with his hands on the throttle,
And scalded to death with the steam.

Now, young ladies, you must take warning
From this time now and on,
Never speak harsh words to your true loving husband
He may leave you and never return.[48]

Ten years later, Henry Whitter, a friend of Burnett's youth, heard Burnett singing it when they were both living near Fries and thereupon learned it. Unlike Lewey, Noell, and Burnett, Henry Whitter was determined to make a career out of his music, and he kept the ballad of "Old 97" active in his repertory until the day in 1923 when he recorded it for the General Phonograph Corporation. However, he changed it from the way he learned it; in about 1922, he testified, he made some slight changes and set the song to the tune of "Ship," which meant making it "more peppy."[49]

What the tune was like beforet hese changes, we can only speculate. Evidently, Whitter also shortened the song considerably, stripping off several of the opening stanzas. According to Kelly Harrell, Whitter was the first person he heard who opened the song with the verse, "They give him his orders."[50] Whitter's text, as he recorded it , follows:

Click to play: Wreck on the Southern Old 97- Henry Whitter

They give him up his orders at Monroe, Virginia,
Saying "Steve you're way behind time,
This is not' Thirty Eight' but it's 'Old Ninety-Seven,'
You must put her in Spencer on time."

Steve Brooklyn said to his black greasy fireman,
"Just shove on a little more coal,
And when we cross that White Oak Mountain,
You can watch old Ninety-Seven roll."

It'sa mighty rough road from Lynchburg to Danville,
And a line on a three mile grade;
It was on that grade that he lost his air-brakes
And see what a jump he made.

He was going down grade making ninety mile an hour
When his whistle begin to scream,
He was found in the w reck with his hand on the throttle
And was scalded to death with the steam.

(So come) you ladies you must take warning
From this time, now and on,
Never speak hard words to your true loving husband,
He may leave you and never return.[51]

It is readily apparent that Whitter's text consists of the last five stanzas of Burnett's, with slight modifications. Burnett's differs from both Lewey's and Noell's, principally by the inclusion of the final stanza borrowed from "The Lover That Never Returned." Thus, if Burnett's recollection of the song he sang some two decades earlier is correct, Judge Avis was wrong when, in his decision, he attributed the addition of this stanza to Whitter. It might also be mentioned that Whitter's shortening of the text had nothing to do with the time restrictions of a 10-inch 78 r.p.m. disc: his rendition consists of ninety seconds of singing and one hundred seconds of instrumental breaks.

One other commercial recording artist learned his version directly from Lewey or Noell. Kelly Crockett Harrell, who lived in Fries until 1925, worked in the textile mills from 1903, first as a weaver, then as a loom fixer, then as loom mechanic. Born in 1889, two miles from Mack's Meadows, Virginia, his recording career began in mid-1924 with recordings made for the Victor Talking Machine Company. In August of the following year, for OKeh's recording team in Asheville, North Carolina, Harrell recorded, among other songs, "The Wreck on the Southern Old 97" to Henry Whitter's guitar and harmonica accompaniment. This recording was made at the request of the OKeh people, who, for some reason, wanted a 12-inch disc version of the song.[52] Harrell had first heard the song in 1904 in Fries from Lewey, and then from Noell. He testified that the two men sang different versions and that they kept changing their versions.[53]

Since Harrell's version was intended for a 12-inch 78 r.p.m. disc, he was allowed time for a song longer than the other early recordings, all of which were issue on 10-inch discs. Except for the rearrangement of a few verses and one additional chorus, Harrell's text was very similar to Burnett's. The additional chorus he insisted was Lewey's:

Did she ever pull in? No she never pulled in
And that po' boy he must be dead;
Oh yonder he lays, on that railroadtr ack
With the car wheels over his head.[54]

This chorus is very similar to that of Fiddlin' John Carson's "Did He Ever Return?" quoted earlier, which suggests that Lewey had borrowed it from oral tradition. In the spring of 1931, after numerous delays, the trial of George v. Victor Talking Machine Co. finallyc ameb eforet he court. In retrospect, David Graves George could have sued Victor on two different grounds. He could have acknowledged that there was a song about Old 97 in public domain (or written by someone identifiable) but that he wrote another song or re-arranged the one in public domain, adding sufficient original m aterial to qualify him for some copyright protection. In that case, the defendant (Victor) would owe him royalties if he could show that the Victorr ecordingw as basedo n his version (rather than on the underlying public- domain song). On the other hand, George could have claimed that he wrote the original song about Old 97, that the song had not previously been published so that he owned common-law copyright, and therefore Victor's use of the song was an infringement of his common-law copyright. This claim would have some justice even if the recorded version was sufficiently different from George's s ong to qualify as a new arrangement; George would have been entitled to compensation as original author.

Had George pursued the first alternative, Victor's defense likely would have been that it had had no possible access to George's unpublished song and therefore could not have used it as a basis for the recording. If George had chosen the second option, Victor could have responded either by proving that he could not have been the original author, or by showing that, even if he were original author, someone else had written an independent ballad on which the recording was based.

As it happened, George claimed that he was the original author, and Victor chose to prove he could not have been, even though it probably would have exonerated the companys imply to have shown that there was another independent version that it had used. The company's eagerness to obtain proof of Noell's and Lewey's claims o f authorship, and its willingness to pay them for rights to the song, would have been consistent with either line of defense.

Victor's case against George rested on two types of evidence. The first, familiar enough in courtrooms, comprised inconsistent or perjured testimony. There was the matter of the altered letter to the News Leader, referred to above. There was the fact that George presented as evidence a sheet of paper on which he had traced, with carbon paper, the music from the sheet music of "Ship" and then had written in the words to that composition. He claimed he had done this in 1903 and had witnesses to attest to that claim. But an analytic chemist demonstrated that the type of carbon paper employed was not in use until long afer 1903. Furthermore, the same chemist, also a handwriting expert, testified that, based on samples of George's handwriting in 1901 and 1927, the words to the song were written in 1927.

George's daughter testified that within a week of the wreck she had sung her father's song in the home of a friend who had just obtained a new organ and wanted to try it out. But it was established later that the friend did not have an organ in her home until May 1907. A similar claim was made by another witness for the plaintiff, and again it was later shown by counsel for defense that no organ was present in the home until November 1907. Another acquaintance of George's testified that George had asked him to testify at the trial that the plaintiff had written the song for him on a certain day in a certain barber shop, but he had refused. "If I win, you'll get a piece [of the pie]," George assured him.[55] Such testimonies were very damaging to George's case.

Some of the above points were first suggested to RCA Victor's lawyers by Gordon (for example, the chemical analysis of the carbon paper). However, Gordon's primary role was to study the various versions of the ballad and to demonstrate, if possible, that on the basis of internal evidence George could not have composed, in 1903, the ballad he claimed he had.

Gordon had suggested to Victor's legal department some time before the trial that he would "gladly chart and study the differences [among the various texts] in accord with a method which often brings surprising results. This method is similar to one now recognized as standard in all studies of folk-tales. (See an article by Archer Taylor in Modern Philology, 25, May 1928, entitled" Precursers of the Finnish Method of Folk-Lore study.'") The method is most delicate and not usable except by one trained in the subject. Also there are certain differences necessary in the study of songs as opposed to tales since the laws of growth and change tend to reverse somewhat."''[56]

In court, when Gordon took the witness stand, he was asked, "Is there any accepted method used in the study of Folk Song text which enables the expert to determine the relative date of any given text?" He replied in the affirmative": The text is examined entirely on internal evidence, not taking into consideration the testimony or alleged statements of the versions, but treating them all as possible. ... They are examined internally for the presence or absence of verse, for changes or variation in phrasing, the presence of misunderstanding. . . or errors, and treated, in other words, as the philologist goes about the study of language."'[57]

What Gordon did do was to present to the court an immense chart that represented comparatively all the important texts, designed to show differences, similarities, and borrowing.[58] He established two major points. First, he noted that in four instances Dalhart had misunderstood the words on Whitter's recording, and consequently he rendered " Spencer", "Steve," "air-brakes," and "begin to scream" as "Center," "Pete," "average," and "broke into a scream," respectively. The first three of these locutions have since served as benchmarks to determine that any collected or recorded version of the ballad derives from Dalhart's immensely influential recordings. In three of the four instances, George's manuscript followed the Dalhart version. There were, however, differences between the two texts apart from George's two additional stanzas. The two versions are compared below; those phrases or stanzas in George's text that are not in Dalhart's are enclosed in parentheses:

(On a cold frosty morning in the month of September,
When the clouds were hanging low;
Ninety-Seven pulled out from the Washington station,
Like an arrow shot from a bow.)

They gave him his orders at Monroe, Virginia,
Saying, "Pete, you're (you are) way behind time;
This is (It's) not Thirty-Eightb, ut it's old Ninety-Seven,
You must put her in Center (Spencer) on time.

He looked round then to (looked at) his black, greasy fireman,
"Just shove on (And said "shovel") in a little more coal;
And (For) when we cross that White Oak Mountain,
You can watch (see) old Ninety-Seven roll."

It's a mightyr oughr oadf rom Lynchburg to Danville,
And a line on (Lima it's) a three-mile grade;
It was on that (this) grade that he lost his average,
And you see what a jump he made.

He (They) was going down grade making ninety miles an hour
When his whistle (Who when the whistle whistle whistle) broke into a scream;
He was found in the (a) wreck with his hand on the throttle,
And a-scalded (scalded) to death with the steam.

Now ladies, you must take warning,
From this time now and on (time on);
Never speak harsh words to your true love and (loving) husband,
He (For they) may leave you and never return.

(Did she ever pull in, no she never pulled in,
For hours and hours... as watching,
For the train that never pulled ...) [59]

Gordon further demonstrated that, in several other respects, the texts of Whitter, Dalhart, George, and the other commercial recordings all differed from the "'old" versions-a group of four texts that seemed almost certainly to date from before 1910.[60] For example, in the "old" versions, the road from Lynchburg to Danville is described as "bad," while Whitter, Dalhart, George, and the post- Whitter phonograph records use the adjective "rough."

In one respect Gordon's argument was weak. Although it is reasonable to assume that the phonograph versions that have the same five verses as the Dalhart recording (the Skillet Lickers on Columbia, Jones and Hare on Brunswick) were based largely, if not entirely, on the Dalhart/Whitter text, it is not reasonable to make the same assumption for the texts of Harrell and Thompson, both of which included stanzas not found at all in the Dalhart/Whitter version. In fact, testimony established that Harrell had learned the song considerably earlier than Whitter had, although such evidence is not "internal" and therefore should not have been used by Gordon in his evaluation.

Ernest Thompson was never located for interview; however, as a native of North Carolina he had probably also learned his version from oral tradition long before he recorded it. So, since Harrell said "rough road," it is very likely that the locution was in use well before the Whitter recording of 1923. An artist who has been singing a ballad for two decades is not likely to change the wording on hearing a different version." [61] In the last analysis, "internal evidence" alone does not offer a very strong defense one way or the other.

For example, internal evidence could not refute the hypothetical claim that p erhaps George had mailed Dalhart a copy of "his" text some time before either Dalhart or Whitter recorded it. Such hypothesis is, of course, improbable to say the least: two years later, perhaps yes. By that time Dalhart was deluged by requests from his fans to sing this or that song, but "97" was Dalhart's first hillbilly recording. He would not have received s ucha request in early 1924. So the imaginary hypothesis is refuted, but by external, not internal, evidence. Without the knowledge that Dalhart had copied Whitter's recording (external evidence) and the knowledge that Whitter had learned his text from Burnett, who in turn had learned it from Lewey or Noell (external evidence again), we could not rule out George's role as a link in the chain somewhere. I would not be surprised to learn that every folksong investigator who has pursued one song through all its meanderings has likewise encountered problems that could not be resolved on the basis of textual evidence alone, but required some external data: dates of recording or manuscript writing, a date of sojourn in a particular county, or a verification of who learned a particular variant from whom.

Gordon's zeal in executing his inquiry led him into many areas in addition to the collection of texts and interviews of informants. He searched thoroughly for all newspaper accounts, for train schedules, for weather reports from the day of the wreck, for interviews with survivors of the wreck. When these searches came to the attention of Victor's chief attorney for the case, Louis B. Le Duc, he wrote a letter to Gordon expressing his worry that Gordon's investigations were taking him into matters that were irrelevant to his responsibilities in the preparation of the defense."[62]


Gordon's reply stressed the necessity of knowing what the actual facts were in order to date various versions-the assumption being that "a man who had at least heard a great deal of talk about the wreck would be closer to fact than a garbled version sung at a much later date by a many who had not been in the vicinity of Danville at the time.'"[63] T herefore, Gordon argued, he had to know the actual facts of the accident as well as what was commonly said and believed at the time."[64] Le Duc replied:

... if we propose to prove actual facts connected with the wreck, you would not be allowed to do this by repeated statements made to you by witnesses; the eye-witnesses themselves would have to be put on the stand. . . . the hypothesis on which your expert conclusions rest must be founded entirely in the evidence admitted at the trial. If such conclusions depend at all on facts which have not been regularly proved, your entire testimony would probably be ruled out and lost to the defendant. Your role on the stand, therefore, will be solely that of the expert dealing with hypotheses and not at all that of the fact investigator. The several versions of the song, their dates and places of publication and circulation- together with the newspaper stories of the wreck, if you deem these importantare the facts on which I have always expected your testimony would be based in its entirety.,[65]

Thus, the very narrow type of case study, confined to purely textual matters, was all that would have been allowed in court; when Gordon assayed to make a more comprehensive study, taking external as well as internal f actors i nto account, he was warned by the lawyers to desist.

On March 11, 1933, almost two years after the conclusion of the trial, Judge John Boyd Avis of the circuit court handed down his decision that George had authored the ballad, as he had claimed, shortly after the wreck, and that the song on the Victor disc was essentially the same song, so that the company owed George appropriate royalties. In his decision, Avis first rejected the argument of the defense that George had copied his text from the Dalhart recording. The differences between the two texts were too great for that, he contended, regardless of the important similarities that were unique to those two texts. And he did not fail to notice that, for some unknown reason, when Noell recited his text on April 5, 1929, he used the word "average" instead of "air-brakes."

He rejected the argument that the repeated phrase "whistle whistle whistle" indicated copying from the recording because it occurred at a point where two blasts of a train whistle sounded in the background behind Dalhart's singing. While he respected the able testimony of Gordon and Melcher, the chemist and handwriting expert, he concluded that they "erred in their conclusions." Of the other arguments against George, he said nothing in his opinion. Avis acknowledged that Noell and Lewey could have composed songs about "97," but argued that the Victor song could not have been based e ither on one of their songs or even on some combination of both of them.

Therefore, he felt there must have been a third text preceding the Victor one, namely, George's. While I believe Avis was right in his conclusion that there must have been a third text, I believe the reasoning leading to his conclusion was faulty. His opinion was based largely on an insufficient understanding of the nature of oral transmission; he took permutations of single words as evidence for alternate sources.

For example, he felt Dalhart's "They give him his orders at Monroe, Virginia" could not have been derived from Noell's "He received his orders at Monroe, Virginia." Quite to the contrary, Avis decided that what similarities the texts of Noell and Lewey had with that of George were due to borrowings from George's prior text. Thus, ironically, Avis based his decision solely on his (incorrect) analysis of internal evidence and ignored the external data.

RCA Victor appealed the decision of the district court, and on January 3, 1934, the Third Circuit Court of Appeals at Philadelphia, in an opinion written by Judge J. Warren Davis, unanimously reversed the lower court's ruling. George's attorneys brought the case to the United States Supreme Court, which, on December 17, 1934, reversed the appellate court decision.

The successive court actions devolved upon the question of whether the decree of the district court was final or interlocutory; if it was an interlocutory ruling, the defendant (RCA Victor) had to appeal within thirty days; but if the decision was final, that time limit did not apply. Victor maintained that the ruling was final, whereas George's lawyers argued it was interlocutory. The circuit court upheld RCA; the Supreme Court, however, overturned that ruling and decreed that since the district court judgment was interlocutory the circuit court had no jurisdiction after thirty days. This reinstated Judge Avis' ruling in favor of George, which required RCA to present extensive evidence to enable determination of the extent of royalties due George. The district court, on September 15, 1938, awarded George $65,295.56. Again RCA appealed and took the case to the Third Circuit Court of Appeals, and on July 14, 1939 that court, this time with Judge Maris writing the decision, again reversed the lower court ruling. George appealed to the Supreme Court later that year for a review, but the petition was denied on November 13, 1939. A rehearing on the petition for rehearing was denied by the Supreme Court on December 11, 1939. Finally, a motion for leave to file a second petition for a rehearing was denied by the high court on January 29, 1940, thus ending ten years of litigation.

Today, over three decades later, are we in a better position to pass final judgment on David Graves G eorge's claim to authorship? The evidence presented in court does seem to have been predominantly against him, indicating clearly that he had assembled false testimony and perjured witnesses to back his claim. When the plaintiff has gone that far, one hesitates to give credence to any of his claims.

Yet, the opening stanza of George's version proves that he did know the song apart from the Dalhart and Whitter recordings. That stanza, claimed by neither Lewey nor Noell, strongly suggests that there was at one time a third ballad circulating about Old 97. We have nothing that can definitely be assigned to this hypothetical third ballad save for George's opening stanza. Everything else, except for the "Parted Lovers" moral, can be derived from Lewey's and Noell's texts.

But this third ballad must have been well known at one time, as some variant of this stanza occurs in a half-dozen different renditions. The earliest appearance in print was in 1911, when a reader of Railroad Man's Magazine from Asheboro, North Carolina, requested a ballad beginning,

On the twenty-first day of last September
The clouds were hanging low ;
Ninety-seven pulled out from the station
Like a narrow shot from a bow.[66]

A manuscript version given by Mrs. Lera McNeeley Stephens of Lynchburg, Virginia, which she testified she had written prior to 1911, began,

On a cool frosty morning in the month of September
As the clouds were hanging low,
97 pulled o utf romt he Washington Station
Like an arrow shot from a bow.[67]

A similar stanza opened Ernest Thompson's 1924 recording. Two other versions, not available to Gordon in 1930, but probably both very old, begin with a similar verse. One was collected from millhand and hillbilly musician Dorsey Dixon; the other from black street- and medicine show singer Pink Anderson. Dixon reported he had learned his version in 1906 or 1907 from string-band musicians of Darlington, South Carolina. Anderson, born in Laurens, South Carolina, in 1900, may have had his version not long after. The date in Dixon's version was "one cold November morning"; that in Anderson's was "November 29 ."[68]

These facts are pertinent because of an interesting hunch Gordon had about the origin of this particular stanza. He had been bothered by the references to cloudy or frosty weather, because September 27 , 1903, was an unusually warm and bright day. In checking through newspaper files, Gordon found that the most famous wreck that befell the Southern Railroad occurred on November 30, 1906, when "Train 37, southbound, overtook and ran into Train 33, also southbound, ten miles beyond Lynchburg..... This was a cold frosty morning at the end of November."[69]

Gordon also noted that in 1906 the government granted the Southern Railroad a subsidy for fastmail trains, which was divided b etween 97 and 37, the latter being the fast mail train. Le Duc was at first impressed by this line of reasoning, but later he thought it would complicate the testimony excessively if it were brought up, and Gordon mentioned it only briefly while he was on the witness stand. [70]

Gordon may not have noticed it, but it is intriguing that George's two witnesses who claimed to have played his new song on their new organs both obtained their musical instruments in 1907. Perhaps, then, George did write a ballad, but about this later wreck. Whether or not this last speculation is correct, I am tempted to give credence to George's claim of authorship of a ballad about a train wreck. I suggest the hypothesis that George did write a ballad that was in circulation for awhile-either in 1903 or 1906- but that later even he forgot it; so that when he heard the Dalhart record, he knew he had written something about Old 97 and concluded that Dalhart's song was his ballad. The rest was simply a matter of buttressing what he knew as a weak, though just, case. Many rural mountain folk have a different notion of what constitutes authorship of a new song. Collectors have encountered informants who asserted authorship of some song, when all they had done was alter other songs or other versions of the same song to produce a new version of the song for which they claimed authorship. The alternative to this explanation, of course, is that George fabricated the entire story.

My final thoughts on this lengthy discussion concern not Old 97 or David Graves George, but rather Robert W. Gordon himself. In the mid-1920s Gordon was ahead of most of his contemporary folksong colleagues in his awareness and utilization of mass media: witness his commandeering of the pages of Adventure magazine to collect some four thousand texts from a network of informants across the country. We do not know precisely when Gordon became aware of the folkloristic significance of hillbilly and race records; it must have been very early, however, for in 1925 he spoke with Dalhart about the latter's career-surely making him the first folklorist ever to interview a hillbilly artist. He must also have been one of the first folklorists to appreciate the complex role of the phonograph as both documenter and preserver-as well as modifier-of oral tradition.

I have already mentioned his own use of reccording equipment on field trips; apparently Gordon played a significant role in encouraging companies to develop compact and portable recording equipment that could be used for field work.[71] One can only speculate what pioneering work in the area of hillbilly music Gordon might have carried out had he not abandoned his folk music activities early in the 1930s.[72]

APPENDIX: DISCOGRAPHY OF RECORDINGS OF "'97" PRIOR TO 1931

1. Henry Whitter. "The Wreck on the Southern Old 97," OKeh master 72167-A, recorded December 12, 1923.
Release: OKeh 40015, January 1924.

2. Ernest Thompson. "The Wreck of the Southern Old '97," Columbia master 81742-1, recorded April 26, 1924.
Releases: Columbia 130-D, ca. August 1924; Harmony 5120-H (pseud. Ernest Johnson), ca. 1928.

3. George Reneau with Gene Austin. "The Wreck on the Southern 97," Aeolian Vocalion master 15054, recorded between April 15 and May I, 1924.
Releases: (Aeolian) Vocalion 14809, ca. July 1924; (Brunswick) Vocalion 5029, ca. January 1927.

4. Vernon Dalhart. "The Wreck on the Southern Old 97," Edison master 9514-C, recorded May 4, 1924.
Releases: Edison 51361, ca. August 1924; Edison (cylinder) 4898, ca. September 1924.

5. Vernon Dalhart. "Wreck of the Old 97," Victor master 30632-1, 3, recorded August 13, 1924.
Release: Victor 19427, October 3, 1924 (superseded by No. 12; see below).

6. Vernon Dalhart. "Wreck of the Southern No. 97" (All releases issued under the pseudonym of Sid Turner), Pathe master 105510-1, recorded ca. August 1924.
Releases: Pathe 032068, October 1924; Perfect 12147, October 1924; Harmograph 970, ca. 1925; Romeo 5050, ca. May 1931; Oriole 8050, ca. May 1931; Jewel 20050, ca. May 1931.

7. Vernon Dalhart. "The Wreck of the '97," Plaza Music Co. master 5552-2, recorded ca. July 1924.
Releases: Bell 340, 1924; Banner 1531, June 1925; Regal 9829, June 1925; Domino 3501 (pseud. Bob White), June 1925; Oriole 325 (pseud. Dick Morse), June 1925; Silvertone 2701, ca. 1926; Conqueror 7067, n.d.; Clover 1694, ca. 1928. Canadian releases: Apex 8295, ca. December 1924; Apex 8428, ca. February 1926; Starr 10040, n.d.; Leonora 10040, n.d.; Microphone 22004 (pseud. Fred King), n.d.; Domino 21121, n.d.

8. Vernon Dalhart. "The Wreck of the Southern Old '97," Starr Piano Co. master 9149, recorded ca. October 28, 1924.
Releases: Gennett 5588, ca. December 1924; Gennett 3019, June 1925 (superseded by No. 13; see below).

9. Carl Fenton's Orchestra, with vocalists Billy Jones and Ernest Hare. "Wreck on the Southern Old 97," Brunswick master (number not known), recorded March 20, 1925.
Release: Brunswick 2857, April 1925.

10. Kelly Harrell. "The Wreck on the Southern Old 97," OKeh master 9279-a, recorded ca. August 26, 1925.
Release: OKeh 7010, ca. December 1925.

11. Vaughn de Leath. (Exact title not known), Cameo master 1809-B-I, recorded ca. January
1926.
Releases: Lincoln 2473 (pseud. Gloria Geer), March 1926; Cameo 875 (pseud. Anabelle Lee), March 1926.

12. Vernon Dalhart. "Wreck of the Old 97," Victor master 30632-4, recorded March 18, 1926 (supersedes No. 5 above).
Releases: Victor 19427, ca. April 1926; Bluebird B-5335, February 1934; Sunrise 3416, ca. February 1934; Montgomery Ward M-4477, Autumn 1934.
Canadian release: Victor 119427, ca. April 1926.

13. Vernon Dalhart. "The Wreck of the Old Southern 97," Starr Piano Co. master 9617, recorded ca. June 19, 1926.
Releases: Gennett 3019, ca. September 1926; Gennett 5588, ca. September 1926; Champion 15121, September 1926; Herwin 75503, ca. November 1926; Challenge 161, ca. March 1927; Challenge 320, ca. March 1928 (superseded by No. 17; see below).

14. Gid Tanner and His Skillet Lickers with Riley Puckett and Clayton McMichen. "The Wreck of the Southern Old '97," Columbia master 143786-1, recorded March 28, 1927.
Release: Columbia I5142-D, May 20, 1927.

15. Vernon Dalhart. "Wreck of the Old '97" (Grey Gull released under various pseudonyms, including Jeff Calhoun), Grey Gull master 2343B, recorded ca. early 1927.
Releases: Grey Gull 4131, ca. early 1927; Radiex 4131, ca. early 1927; Paramount 3018, ca. May 1927; Broadway 8053, early 1930 (Note: Evidently the Grey Gull and Radiex releases were superseded by No. 16; see below).

16a. Arthur Fields. "Wreck of the Old '97" (released under various pseudonyms, as shown; release dates for these are uncertain), Grey Gull master 25II-A, recorded ca. May 1927.
Releases: Grey Gull 4131 (as Arthur Fields); Radiex 4131 (as Arthur Fields, Jeff Calhoun, or Mr. X).
16b. Grey Gull master
Releases: 2511-B. Grey Gull 4131 (as Arthur Fields, or Jeff Calhoun).
I6c. Grey Gull master 25II-C.
Releases: Grey Gull 4131 (as Vel Veteran); Radiex 4131 (as Arthur Fields). Unknown take number: Van Dyke 84131 (as Mr. X); Madison 4085 (as Vel Veteran).

17. Vernon Dalhart. "The Wreck of the Southern Old 97" (Supertone release titled "Wreck of the Old Southern '97"), Starr Piano Co. master GEX-1I254A, recorded ca. May 22, 1928.
Releases: Champion 15121, July 1928; Silvertone 8141, ca. August 1928; Gennett 6654,
December 1928; Supertone 9241, ca. March 1929.

18. "Happy" Dixon's Clodhoppers. "Wreck of the Southern Old 97," Warner Brunswick master E34982-A, recorded ca. October 22, 1930.
Releases: Melotone 12052, October 1931; Polk 9033, October 1931.
Canadian release: Aurora 22030 (as by Yoder's Yokels), October 1931.

John Edwards Memorial Foundation
University of California
Los Angeles, California

------------------------
Footnotes:

1 Archie Green, Only a Miner: Studies in Recorded Coal-Mining Songs (Urbana, 1972), 90.

2 Information on Gordon's collection comes from data compiled by Joseph Hickerson of the Library of Congress Archive of Folk Song, and from a letter from Gordon to James E. Richards, Victor Talking Machine Co., November 2, 1926. All letters to or from Gordon quoted in this paper are in the Robert W. Gordon Collection of American Folksong at the University of Oregon at Eugene. I am grateful to Professor Barre Toelken for making this material available to me.

3. The New York Times series was published in the Sunday Magazine section between January 2, 1927, and January 22, 1928, slightly abridged by the editors. The unabridged series was published as Folk-Songs of America (New York, 1938) under the auspices of the WPA in mimeographed format. "The Negro Spiritual" appeared in Augustine T. Smythe and others, The Carolina Low Country (New York, 1931).

4 Letter to James E. Richards, November 2, 1926.

5. Gordon's work is discussed briefly in D. K. Wilgus, Anglo-American Folksong Scholarship since 1898 (New Brunswick, N. J., 1959), and in Green.

6. This opinion was offered by Alfred Frankenstein in an interview in his home in San Francisco on April 19, 1970. Frankenstein was a friend of Gordon's during the 1920s.

7. See, for example, Freeman H. Hubbard, Railroad Avenue (New York, 1945), 251-261; Stewart H. Holbrook, The Story of American Railroads (New York, I947), 430-432; B. A. Botkin and Alvin F. Harlow, A Treasury of Railroad Folklore (New York, x953), 449-450; Judge John Boyd Avis, "Victor Talking Machine Co. v. George," in the Federal Reporter, Second Series 69 (April-May 1934), 871-879; and Henry M. Belden and Arthur Palmer Hudson, The Frank C. Brown Collection of North Carolina Folklore, vol. 2, Folk Ballads (Durham, 1952), 512-521.

8. Henry Clay Work, "The Ship That Never Return'd," published by Root and Cady, 1865.

9 Alexander Malin, "The Hand-Car That Never Returned," 1888.

10. Earl J. Stout, Folklore from Iowa (New York, 1936), 72. The informant's name was Oren Beck.

11 Barbara Kreimer, "Information Booth," in Railroad Magazine (March 1967), p. 32.

12 In this family are the texts given in The Frank C. Brown Collection, vol. 2, Folk Ballads, 509; Ethel Park Richardson, American Mountain Songs (New York, 1927), 42; Sigmund Spaeth, Weep Some More My Lady (New York, 1927), I39; idem, Mountain Songs (Chicago, 1937), 5; various song folios published by M. M. Cole of Chicago, such as Tiny Texan, Cowboy and Mountain Ballads (I930), 52; and Opal Burroughs, Archive of Folk Song (AFS) recording # 11, 873 A-II.

13 In this family are the texts given in Geneva Anderson, "A Collection of Ballads and Songs from East Tennessee" (Master's thesis, Duke University, 1932), 213; John A. Lomax and Alan Lomax, Our Singing Country (New York, 1941), 254, "The Wreck on the Somerset Road," a transcription of Justice Begley's recording on AFS #1532; Wilma Pace, Western Kentucky Folklore Archive (WKFA) (November 1959); Mrs. Darrell Pace, WKFA (May 1962); E. N. Caldwell, WKFA (E. C. Perrow Collection, 1913); a recording by Clarence Greene entitled "Narrow Gauge Song" (Ed Kahn Collection, August 1961); and a manuscript version sent me by John V. Walker of Corbin, Kentucky, entitled "Cumberland Valley Wreck" (August 1972).

14 Brown Collection, vol. 2, Folk Ballads, 509.

15 OKeh 45176 (Master #W81728), recorded in October 1927.

16 These songs were not only locally produced. Wm. Jerome's "He Never Came Back" (1891) seems to borrow at least chorus structure from Work's original song. This on vaudeville piece was very popular stages in the I890s and, later, among hillbilly musicians.

17 Letter from Jack O 'Connor to Frothingham, February 12,

18 Adventure, 1923. January 30, 1924, p. 191.

19 Gordon North Carolina ms. # i; also cylinder records A -I, A-2, A-3. This and other field- recorded material is deposited at the Archive of Folk Song, Library of Congress.

20 Letter to Richards, November 2, 1926; see fn. 2, above.

21 The factors that can qualify a piece for being considered a public-domain work, as well as many other aspects of the problems discussed here, are treated at length in O. Wayne Coon, "Some Problems with Musical Public-domain Materials under United States Copyright Law as Illustrated Mainly by the Recent Folk-Song Revival," in Copyright Law Symposium, 29 (New York, 197I), 189-218 (available as Reprint #27, John Edwards Memorial Foundation, Folklore and Mythology Center, University of California, Los Angeles).

22 Harriet F. Pilpel and Theodora S. Zavin, Rights and Writers (New York, I96O), 170-172.

23 Alan T. Dworkin, "Originality in the Law of Copyright," in Copyright Law Symposium, ri (New York, 1962), 60-81.

24 The "modern" version Gordon refers to is probably "Oh Didn't He Ramble," by James R. Johnson and Bob Cole, written under the pseudonym of Will Handy (I902); it ultimately goes back to "The Derby Ram."

25 Letter from John Paine, July 2, 1925.

26 Letter to Clifford Cairns, Victor executive, November 2, 1926. 27 Letter from Paine, August 21, 1925.

28 Archie Green, "Hillbilly Music: Source and Symbol," JOURNAL OF AMERICAN FOLKLORE, 78 (1965), 204--228. Green notes the problem in dating Whitter's first recording session. Whitter asserted, both in pre-trial hearings in 1929 and in his song folio published ca. 1935, that he went to New York on March 1, 1923, to record. At the trial, however, Ralph Peer, who had been employed by OKeh at the time, testified that the recording had been made on December 12, 1923. Perhaps a test recording in March proved unsatisfactory, requiring a second try. The issued master number is consistent with the later date. Discographic data for this and other early recordings of "97" are listed in the Appendix.

29 Actually, the vocal on this and Reneau's other early numbers was provided by the young popular singer, Gene Austin, probably because Vocalion executives found Reneau's singing too rough to suit them.

30 This sales figure is based on information given at one of the later trials. Elsewhere I have reproduced the sales statement and discussed the various claims that have been made, running as high as twenty-five million. See JEMF Quarterly, 6 (I970), 171-173. In the same journal, 152-159, Walter D. Haden has given a new view of Dalhart's recording of "The Prisoner's Song."

31 "Until 1924 the stock defense in almost every copyright suit was that the infringed portion belonged to the public domain. A defendant invariably won his case when he showed both works were traceable to the same common source, and delving deeply enough into old records usually rewarded him." (Alfred M. Shafter, Musical Copyright [Chicago, 19391, 230.) In the 1924 "Dardanella" case, Judge Learned Hand ruled that, although testimony showed the melody was based on a very old ostinato, there was still sufficient originality to warrant protection (Shafter, 233). 32 The Victor/Hager bill of sale was reproduced in the court transcript; the Hager/Whitter one was in the possession of Whitter's son, Paul Whitter. It was copied and made available to me by Archie Green.

33 The Richmond News Leader, March I, 1927, p. 19. Somehow, this paragraph was misquoted in the court transcript, the last word being rendered as "business." The author of this curious query was never identified; George, in court, denied he had written it.

34 Ibid., March 2, 1927, p. 19.

35 Ibid., March 7, 1927, p. 17.

36 Undated document in Gordon files, University of Oregon.

37 On January 4, 1929, the Radio Corporation of America had concluded negotiations for the purchase of the Victor Talking Machine Co. (See RolandG elatt, The Fabulous P honograph [New York, 1965], 247.)

38 Gordon North Carolina ms. #4, transcribed from cylinder A-5.

39 Quoted from document in Gordon files titled "Statements of Fred J. Lewey, Concord, N.C., Oct. 15, 1925," but possibly written some time later.

40 Letter to Gordon from W. L. Plott, November 1926, quoted from a document in Gordon files titled "Statements Tending to Back Lewey's Claims."

41 Pre-trial testimony of Kelly Harrell, April 1929, at Danville; and trial testimony of Frank Burnett.

42 According to Noell's pre-trial testimony, April 1929, at Greensboro, N.C.

43 From Gordon files (University of Oregon). Gordon re-transcribedth e field recordings A -i, A-2, A-3 for the trial; his original transcription ( Gordon N.C. ms. # i), probably m ade shortly after his 1925 field trip, differs considerably from this text. Unfortunately, the cylinder recordings have deteriorated greatly with time, so that I am unable to resolve fine textual uncertainties. As far as I can tell, this later transcription is the more accurate.

44 From pre-trial hearings in Danville; the text is given in Federal Reporter (quoted in Brown Collection, vol. 2, Folk Ballads, 512-516).

45 Gordon N .C. ms. # i, transcribed from cylinder recording A -3.

46 See fn. 34.

47 Letter to E. H. Murphy, Victor Talking Machine Co., March 6, 1929.

48 Trial transcript, p. 189; from pre-trial testimony of Burnett.

41 From pre-trial testimony in Danville, April 1929.

50 From pre-trial testimony in Danville, April 1929.

51 Transcribed directly from OKeh 40015. This transcription differs in several locutions, mostly minor, from Gordon's, which can most easily be found in Brown Collection, vol. 2, Folk Ballads, 513. This points out the difficulty in transcribing recorded songs, particularly from singers who occasionally mumble and swallow their words. (It is amusing to note that the Brown Collection transcription differs in two instances from that in the Federal Reporter, from which it was presumably copied literatim et punctatim.)

52 Testimony of Ralph S. Peer.

53 From pre-trial testimony in Danville, April 1929.

54 From pre-trial testimony in Danville, April 1929.

55 Pre-trial testimony of Walter W. Rowles at Lynchburg, May 31, 1930. Also quoted in Federal Reporter, 876.

56 Letter to Murphy. March 6, 1929.

57 Trial testimony.

58 This chart, designated D-45 in the list of defendant's depositions, is missing from the trial transcripts that are on deposit at the New York Federal Record Center. I have been unable to locate any other copies of the transcript, in either the Gordon material at the University of Oregon or the Library of Congress, from RCA Victor, from Shapiro & Bernstein, Inc., present holders of the copyrights of Whitter, Noell, and Lewey, or from Victor's lawyers.

59 From trial depositions.

60 This group includes the texts of Lewey and Noell; a manuscript sent Gordon by Mrs. Ethel A. Bridges of Danville on November 14, 1926, which she claimed was written "possibly 20 years ago or longer, by Mr. John Moore"; and a manuscript given, at pre-trial testimony, by Mrs. Lera McNeely Stephens of Lynchburg, which she testified she had written prior to 1911.

61. I might note here an amusing detail, although it does not refute my statement: Noell's third rendition of the ballad, given in court on April 5, 1929, uses the locution "lost his average," not "airbrakes." This slip, probably prompted by the controversy over the terminology, was not unnoticed by Judge Avis.

62 Letter to Gordon, September 19, 1930.

63 Letter to Le Duc, September 23, 1930; probably never mailed.

64 Letter to Le Duc, October 4, 1930.

65 Letter from Le Duc, October 6, 1930.

66 Railroad Man's Magazine, 1911, p. 382.

67 See fn. 49.

68 Dixon's version can be heard on Babies in the Mill (Testament LP T-33oI, recorded August 1962); Anderson's is on American Street Songs (Riverside RLP 12-611, recorded May 1950) and also on The Blues of Pink Anderson: Ballad & Folksinger, vol. 3 (Prestige/Bluesville BV 1071, recorded August 1961). One long variant beginning with this stanza, the first line being "It was on the 22nd of last December," seems to have been stabilized in print by the 1930s, if not earlier. The identical text appears in Lonnie E. Anderson and Raymond Anderson, Red and Raymond's Song Book Featuring the Red Headed Briarhopper and His Yodelling Son Raymond (n.d., ca. 1934); in Lillian G. Crabtree, "Songs and Ballads Sung in Overton County, Tennessee" (Master's thesis, George Peabody College for Teachers, 1936), 114; in Old Time Songs and Poems (September- October 1967), 32; and in a letter to me from Marie A. Benson, July 1968.

69 Letter to Le Duc, October 22, 1930. 70 See fn. 62.

71 I am indebted to Joe Hickerson and Barre Toelken for bringing this fact to my attention.

72 I take the title of this paper from a letter of Gordon's to Miss Eva Davis, court stenographer in Danville, dated October 24, 193o0. He concluded this letter with "the hope that the fourth wreck of Old 97 may take place in Camden in November" (the place and expected date of the trial). Since I have been unable to identify with certainty the implied second and third wrecks, I have taken the liberty of altering the numerology. An abbreviated version of this paper w as read at the Twenty-First Annual Meeting of the California Folklore Society at California State College, Dominguez Hills, California, May 1973.