Part 3

There are literally thousands of songs in the books searched. In the _Original Sacred Harp_ alone there are 609, and the _Hesperian Harp_ holds 677. And while other books are slimmer and duplications from book to book are numerous, it must still be quite evident that it was no easy task to identify just the songs I was after. At times I had to apply a number of criteria. Often the folky nature of the text pointed to an equally folky tune. There was another hint sometimes in the name given as that of the composer of the song. When I met with the names Moore, Walker, Chapin, Breedlove, White, Carrell, Davisson, Hauser, McCurry and a number of others, in the upper right corner of the song page, then I was practically certain that the song on that page was usable. For the men in question were, in reality, not composers. They were recorders and arrangers of unwritten music.[10]

When an example of the old unwritten music made its way into the authorized church hymnals--as happened to a restricted degree from fifty to seventy-five years ago--it was called a "Western Melody" or a "Southern Melody." Such designations became another reliable token of folk source.

More important than any external indications in determining whether I was dealing in a particular instance with a folk-tune, was the character of the tune itself. The ability to recognize a folk-tune comes to the student of such music gradually, somewhat as does the recognition of a strange language or dialect. It came to me that way; but after assembling my tunes I felt that their general folk character might to some degree be reduced to a set of definite traits. I therefore reexamined not only my own melodies but also those far more numerous tunes in the secular collections of Sharp and others, for such characteristics as tonal trend, rhythmic trend, tonality (modal character), and musical form. Since there is no available definition of a folk-tune and since probably no succinct one can be made, I am hoping that my deductions in the following paragraphs as to some earmarks of American folk-tunes may be helpful to others who are interested in our traditional melodism, as they have been helpful to me.

Tonal Trend, Tune Families

The very beginning of a folk-tune has characteristic marks. The first accented note is usually the tonic of its scale. In almost all cases this first-accent note is preceded by an up-beat note which also is usually a tonic. The upbeat note coming second in frequency is the lower 5 of the scale, with the higher 3 even less often thus employed. The interval, if any, between the up-beat and the first accented note is thus either an ascending fourth, an ascending third (in those cases where these first two notes are 1 and 3) or descending third. And these intervals, though small, are often broken or bridged by an unaccented intervening note. Tunes beginning with an interval of a fifth (ascending 1 to 5 or descending 5 to 1) are quite rare. Common folk-tune beginnings are thus:

As to melodic trend within the body of the tune, I shall speak only briefly. It is a broad subject, too broad to be discussed adequately in this connection. A survey of my tune-thematic card catalog reveals, however, a few characteristics of this melodic trend. The first is that the tunes assume usually an initial upward trend. Another is that the steps or intervals employed are small, predominantly seconds, thirds, fourths, and fifths. Greater intervals are found however at the juncture of two phrases. From these observations we may assume that the American folk-singer does not like big intervals.[11] This assumption, based on recorded tunes, is strengthened when one listens to folk singing and notices their anticipatory slides or scoops in approaching a tone that is only a little higher or lower than the one just sung, a practice which may be interpreted as an anticipation of, and an attempt to master, that which is vocally difficult. But while the individual jumps from note to note are not as a rule great, the pitch compass of the entire tune is often surprisingly wide. The melodies usually end in a descending cadence to the tonic.

Along with the great variety in form which we meet among American folk-tunes, there are certain melodic formulas which seem to be favorites and reappear with unimportant variations as the tonal vestment of many different songs, so many indeed that they might well be looked on as wandering tunes (reminding one of the familiar wandering stanzas in folk texts) or, since they are not identical from song to song, tune families.

In the present collection I have come upon six tune families of different sizes and have named them in each instance after the song which seems to be the most representative member of the family. They are the 'Lord Lovel' family, cast in the ionian mode; 'I Will Arise', aeolian and ionian; 'Hallelujah', mixolydian; 'Kedron', aeolian; 'Babe of Bethlehem', dorian; and 'Roll Jordan', ionian. The tunes in this collection and elsewhere belonging to the 'Lord Lovel' family are listed under the song 'Dulcimer'. Those belonging to the other families are listed under the songs for which the family is named.

Metrical Patterns

In the matter of metrical patterns we find also a variety, and favorites. We have noted the almost universal use of the up-beat. The up-beat initiates two different rhythmic trends, one of which is the iambic, the prevalent one in American folk-tunes: ^2/_4 * | *'. This two-part type of accent unit (of notes or syllables, whichever way we approach the matter), while occurring in series of four and three, as we have seen, may be found occasionally also in twos, fives, and sixes. Indeed the folk-tunes not infrequently show a refreshing independence of the demands of perfect quadraticality. The other rhythmic trend initiated by the up-beat is the less often used one made up of three-part units, which appear either in three-four time, ¾ * | *' * or slow six-eight time, ^6/_8 * | *' *. With more notes (syllables) in this single amphibrachic unit, the series of such units grows naturally in syllabic length. It often outgrows thus its function as a mere melodic phrase and tends to assume that of the melodic sentence. A fine example of this is in Sharp's recording of the 'Cherry-Tree Carol'.

But while the vast majority of folk-tunes follow one or the other of the above described patterns, we must remember that metrical precision or mechanical adherence to any formula is the least of the folk's concerns. Indeed, we should be justified in assuming such exactness, as seen in text lines of carefully measured lengths and in perfection of rhyme, to be sure signs of individual creative participation; whereas greater freedom and variability in tune and text aspects are obviously characteristic of the folk's vocalism.

Scales, Modes

The folk-tunes of America are not, in the main, built up on scales of the diatonic major and minor systems which, as is well known, have assumed their present form under the demands of _harmony_; but on a modal system which grew out of _melodic_ exigences long before harmony made its conquest of the music of western civilisation. Nor do the folk-tunes of this country make use of all the tones of even these modal scales. They often employ but five or six of the seven available tones, leaving characteristic gaps in such scales.

American folk-tune collectors have had their troubles in the interpretation of modal melodies. I have had mine. Even such a life-long student of these things as Cecil Sharp met many a knotty problem in classifying his Appalachian tunes. In view of this difficulty I called on Hilton Rufty, a thorough musician and a reliable authority in the folk-music field, to lend a hand in the modal classification of these tunes. He generously acceded to my request; and by the time my requests had ceased and before his generosity had been exhausted he had checked or corrected all my modal classifications of the tunes in this collection. In making clear Mr. Rufty's effective and practical method of identifying the character of tunes I shall reproduce his Classification Chart and quote here his explanation of it.

Rufty's Classification; Chart of Tunes

In identifying the modal character of the "gapped" tunes I have deemed it advisable to proceed by an entirely arbitrary method, free from any sort of theoretical connotation. Should a missing tone be presupposed to make either a major or minor, perfect or imperfect, interval with the tonic, there arise at once ambiguities of modality. For purposes of harmonic treatment it is quite necessary to decide upon which particular mode a gapped tune suggests, but in studying the purely melodic aspects it is reasonable to accept the tune as an entity, considering it in its actual tonal structure and not with regard to its possible modal permutations. To accomplish this purpose I have evolved a chart, based on methods used by Cecil J. Sharp in his _English Folksongs from the Southern Appalachians_, which for the great majority of the tunes in this collection is an adequate system of classification. The arrangement of the chart is very simple: there are five columns, each beginning with one of the five pentatonic scales. Immediately below each pentatonic scale are four hexatonic scales which are formed by the addition of the missing tones, singly and in their variable positions. The system permits these variables to be read in terms of natural and flatted tones. Lastly in each column are three regular heptatonic modes which are the outgrowth of supplying both missing tones simultaneously and in variable combination. The gaps in the pentatonic and hexatonic scales are indicated by slurs and the numerical positions from the tonic of the missing tones. The supplied missing tones are indicated by black notes, and in fitting any given tune to any scale on the chart I have endeavored where possible to let these black notes indicate the weak tones. Since it was possible, so far as the _actual tonal structure_ of the tunes was concerned, to have a choice in the placing of them, the device of indicating weak tones was a happy solution to a more careful classification. Above each tune in this book I have indicated the modal and, following this and in parentheses, the tonal pattern of the tune with the heptatonic scale as a norm, that is, treating gapped tunes arbitrarily as broken-down heptatonic tunes. A Roman numeral indicates a major or perfect interval with the tonic; an Arabic numeral a minor interval. In event of augmented or diminished fourths or fifths I have used conventional signs. A gap is indicated by a dash.

As a practical example of classification let us take at random, say, 'Weeping Savior', a song of the present collection. Counting the tones of the melody we find six with the sixth degree missing. We observe that the tune has a major second, minor third, perfect fourth and fifth, no sixth, and a minor seventh. By transposition we see that from a standpoint of the tonal pattern alone the tune can be listed either as Hexatonic, Mode 2, A or Mode 4, b. But the examination of structural detail shows clearly that 3 being a strong tone and 2 being decidedly weak gives preference to the first classification under Mode 2.

While pentachordal and hexachordal tunes (which do not conform to this system of classification) may be perfect entities, I have, nevertheless, for purposes of uniformity classified them on a heptatonic basis, that is, as heptatonic tunes with the sixth and seventh, or seventh alone, missing respectively. Similarly, while it is somewhat tautological to say, for instance, a tune is heptatonic ionian, I have prefixed the term heptatonic to facilitate identification and to balance the constantly recurrent pentatonic and hexatonic.

An examination of these spiritual folk-tunes reveals a great predominance of gapped scales. Only 23 per cent of them use the full seven-tone series; 44 per cent are hexatonic; 23 per cent are pentatonic; and seven tunes use only from 1 to 5 of their scale.

The incidence of the different modes has been impossible to ascertain. We are sure of a mode, as Mr. Rufty has noted, only when the scale tones are all represented in the melody. Proceeding however in questionable instances according to the more or less clear modal _implication_, I have found that about 52 per cent of these tunes may be interpreted as ionian (major), about 30 per cent as aeolian, 7.5 per cent each as dorian[12] and mixolydian, and three tunes as phrygian.

I leave the interpretation of the significance of these figures to others. I venture to suggest however that they will be found to indicate a survival of gapped and modal tunes that is unique in the folk-music of today among peoples of European stock.[13]

A modally constructed tune is, as I have indicated, almost sure to be a folk-tune. And if a melody shows the characteristic gaps, its folk nature is quite assured. Indeed, the complete filling-in of the gaps, creating two half steps, is a sign, though not always a sure one, of art influence.[14]

The above paragraphs show in a general way a few of the more important and evident _features by which American folk-tunes may be recognised_.[15] Their presence or absence in specific cases has helped me to decide as to the fitness of a tune for acceptance into this collection.

_Tunes of Religious and Worldly Folk-Songs Compared_

I have indicated above (page 6) that many of the present tunes were borrowed outright from secular folk-songs. The tune-to-tune relationships were discovered to some degree, as I have indicated, by accident. A spiritual tune would remind me of a secular one. I would look it up in Sharp or elsewhere, verify the relationship, and note it under the proper song in this collection. Such accidents, however, account for but comparatively few of my related-tune discoveries. In most instances they came to light as the result of a methodical comparison made possible by my having catalogued my spiritual folk-tunes and a large number of secular folk-melodies. I shall not go into a detailed explanation of this cataloguing method here, chiefly because it is one which, though it answered my own purposes well, would probably be found inadequate as a tool for students of comparative melody in general. I shall say merely that the catalog was a card index of tune beginnings, all transposed to a key which had two flats as its signature. The arrangement was based on the scale position or relative pitch of the first few tones. At the beginning of the catalog were those tunes which began on _b_-flat, then came those beginning on _c_ and so on. The arrangement among those tunes beginning on any one tone, followed the same pitch sequence, from lowest to highest, taking into consideration the second, third, and more notes of the tune beginnings where necessary. That is, my lexicographical arrangement was like that of the dictionary, but with notes on a regular staff taking the place of letters, and with the scale steps taking the place of alphabetical sequence.

The actual working out of this scheme may be observed in the arrangement of tunes in this collection. In each of the three parts the tunes appear in their catalog sequence.[16]

Through a consistent comparison of the tunes in this catalog with those in secular tune files made on the same plan, I have been able to discover the organic relationship of upwards of 150 melodies in this collection to an even greater number of traditional folk-tunes associated with _secular_ texts. This greater number is explained by the fact that one and the same tune in this collection was often found related to a number of worldly songs. To one tune 'Pilgrim', for example, I discovered 17 secular related melodies. The relationship runs in degree all the way from one which is barely recognizable to one which consists in an almost note-for-note identity.

The catalogs were also of distinct value in bringing to light scores of interrelated tunes _within_ the collection, and thus in bringing to light the tune families mentioned on page 14 above.

The search for kindred secular tunes was most fruitful in the case of the ballads and somewhat less so for the hymns. Among the spiritual songs the search yielded surprizingly meager results. The reason lay probably in the nature of the spiritual-song tunes themselves. These tunes--whatever their source--were often altered through the arbitrary intrusion of refrains and choruses. Among these tunes, therefore, my finding of secular analogies was limited usually to melodic _parts_ instead of whole tunes.

To be sure, the tune relationships, religious to secular, which I have pointed out, touch little more than half the songs under scrutiny. But when it is taken into consideration that the related secular tunes were all found in a body of British Isles-American melodies not much greater than that of the spiritual tunes themselves, then it would not seem unreasonable to assume that a complete catalog of American worldly folk-tunes would reveal cognates to many more, possibly to all of the tunes presented here. The kinships already discovered, however, warrant the assumption that _these spiritual tunes are part and parcel of the ancestral folk-melodism of the English-speaking peoples_.

The worldly-religious tune comparison has also shed more light on the motives which led the revival folk to borrow from the store of secular melody and on the manner of that borrowing. We have indicated above our belief that one motive was the crying need for rousing and familiar tunes. Another reason seems to have been the mere fact that the borrowed tunes _were_ worldly. Worldliness was of itself an asset. Fighting the devil with his own weapons had its distinct advantages in revival technics. But just how and why a _particular_ secular tune came into the religious atmosphere is not always evident. In some instances, however, the examination of the secular original song makes this clear.

When the revivalist heard the Scottish-American sing

Will you go, Lassie, go To the braes o' Balquhidder?

he evidently saw at once the possibilities of turning the text to his own evangelistic purposes, and wasted little time in making it over into 'Sinner's Invitation:'

Sinners go, will you go To the highlands of heaven?

which he sang to the same tune.

The ballad tune to

O'Reilly on the rolling sea Bound for Amerikee

went over easily into the song which told of the Christian voyager who was 'Bound for Canaan.'

The old ballad 'Geordie' begins

As I walked over London bridge.

The revival singers took this hint, with its tune phrase, and produced, in 'Victoria':

I have but one more river to cross.

In the traditional ballad 'In Seaport Town' there is a recurring phrase:

Till at last they came to that lonesome valley.

This "valley" suggested to the religious mind the emotional depression of the almost converted mourner as well as the valley of death; and thus came into existence the beautiful spiritual 'Lonesome Valley':

You got to go that lonesome valley, You got to go there by yourself

whose tune is closely related to that of the secular song.

The 'Poor Stranger' of the English secular ballad who appeared also as "poor strange girl," a "roving soldier," and a "rebel soldier," all of whom are "far from my home," exerted both melodic and textual influence on the 'Heaven-Born Soldier' who urges his comrades to

Come along and shout along And pray by the way.

The melody which Johann Sebastian Bach, the great adapter of folk-tunes, made a peasant sing in his _Cantata_ 'Mer hahn en neue Oberkeet' spread to England and became there the setting of a number of popular texts in the first half of the eighteenth century. One of these songs, dating from 1772, was 'Farewell, Ye Green Fields and Sweet Groves' which gave birth, probably also in England, to the religious song 'Green Fields', found in every old southern fasola book. Its opening lines are

How tedious and tasteless the hours When Jesus no longer I see. Sweet prospects, sweet birds and sweet flowers Have all lost their sweetness to me.

With 'Saw Ye my True Love' as a model, the task of making the religious text 'Saw Ye My Savior,' sung to the same tune, was a grateful one.

The happy inebriate who is his own hero in 'Way Up On Clinch Mountain' is reformed and regretful in 'John Adkins' Farewell' where he gives warning to other alcoholics in the same melodic strain.

From the above examples it would seem that the secular text contained often some hint which led the religious adapter in making his new poetic lines; and that the secular tune usually followed as a matter of course.

The comparison of tunes shed no actually new light on the age of the tunes. But it made clear the fact that the folk's stock of melodies is assembled from divers times. The tunes of two songs in this collection, 'New Orleans', and 'Hark my Soul', have tonal trends strikingly similar to that of melodies found in the eleventh and early thirteenth centuries respectively. From the early seventeenth century we find 'Mourner's Lamentation' which was in those earlier times 'Wae is Me for Prince Charley', a Jacobite song about Charles II of England. 'Beggar' is a remake of 'A-Begging We Will Go' which has been traced back to 1611. 'Captain Kidd' or 'Kidd', as it is disguised in the fasola books, dates from the first part of the eighteenth century. It is significant that most of the tunes mentioned in the above paragraphs are comparatively modern in their musical aspects. This fact leads to the suspicion that the really old-sounding tunes, those in the antique modes--dorian, phrygian, and the like, especially in their gapped forms--originated in still earlier times. Here is an inviting field for the student of comparative folk-melodism.

_Conclusion_

I have been impressed, as I have come to know these tunes better, with their variety and beauty. They are believed, by the country folk who still sing them, to be "the most beautiful music on earth." When I first heard this sweeping judgment I put it down as emanating from an understandable though extravagant zeal, one which was all the greater perhaps since the singers, mostly oldsters, felt they were fighting for the very life of a dying cause. But I now see I was mistaken. The songs are living vigorously without being fought for. The country folk clearly realized--however they may have expressed the realisation--that the "good old songs" were ingrained in their racial souls and that for this reason it was the most completely soul-satisfying of all music from whatever source.

If this was and still is the firm belief of those uncounted thousands who know and sing the country songs, those who are still carrying on the tradition for the sheer love of it and the joy they get out of it; then is there not an inspiration for us? Is that picture not an incentive to look into, to learn to know this _tonal tradition_, the chief one in our ethnic background? This quest might well lead to an examination of our other _acquired_, not _inherited_, musical concepts and judgments, in search for reasons why, in acquiring them, we have ignored the simpler art of the past. And from this approach we might open the question as to whether these reasons are valid,--wise or unwise.

American folk-music, basing squarely on that of the British Isles, is purer, I assert, and more completely representative of the _peoples_ among whom it has developed, and less representative of _individual_ creative activity than is the folk-music of other Western peoples. As evidence of this I present this collection, commending it to the serious consideration of those interested in fundamental phases of American culture.