Parry Collection of Yugoslav Folk Music

Eminent Composer, Who Is Working on It, Discusses Its Significance

By Béla Bartók

From The New York Times, Sunday, June 28, 1942

Since I arrived in the United States in October, 1940, many people have asked me, again and again, what I am doing here. When I explain that I am studying and transcribing the Milman Parry Records, as a commission from Columbia University, it appears that scarcely any one knows about the very existence of this collection, still less about its excellence.

To begin with, it is a most important collection of folk-music, unique of its kind. The story of its origin is this:

Professor Milman Parry, a classical philologist at Harvard University, decided some years ago to go to Yugoslavia to explore existing folk poems. He went there twice with a pupil, Albert Lord, in 1933 and 1934. He was equipped with a double-disked recording apparatus and quantities of disks, and with directions and suggestions by Dr. George Herzog at Columbia University, an expert in folk-music research.

Remarkable Results

The result of these two trips is quite incredible. Dr. Parry found two poems of about the length of "The Odyssey," one of 13,000-odd lines, the other of 12,000. He also collected many shorter ones, of several thousand lines. He recorded most of these songs; the longest took more than twelve hours, not including rests for the singer (and probably for the collector).

The total of records made with ninety different singers is more than 2,200 double-sided disks. Dr. Parry also collected and recorded about 300 various other types (so-called "women's songs"), on about 350 double-sided records (some of them in Turkish or Albanian), and instrumental folk-music on eight records.

After his return in 1935 Professor Parry died in a tragic accident. Mr. Lord is preparing the texts for publication.

I heard of this marvelous collection when I visited the United States in April, 1940, for a five-week tour (my second visit to the United States). I was told that the musical part (the melodies) with a few exceptions, were not yet transcribed. One of the reasons for my return to the United States was the possibility of a careful study of this material, which I had badly missed in Europe.

No Registering Instruments

The Serbo-Croatians (and the Bulgarians) never used registering instruments when they made collections of their own folk music. (However, the use of a recording instrument is nowadays considered as a condition sine qua non in that kind of research work.) The scientific value and reliability of the existing published Yugoslav material is therefore considerably lessened. It is true there are a few commercial records of Dalmatian folk-music available in Yugoslavia; there are about fifty records of Yugoslav heroic poems in a phonograph archive of Berlin and some at Prague. But what is all that in comparison with 2,200 double-sided records of the Parry Collection! The importance of it consists first in its vastness: it is in fact the richest recorded collection of Yugoslav folk-music enumerated; it is the collection par excellence of this material.

Second, it is not a fragmentary collection, as until now the best folk-music collections in Europe have been; every song is completely recorded. We poor scholars of those Eastern countries had to economize on time, on blanks, on expenses, on everything. So we generally had to confine ourselves to the recording of the first three or four stanzas, even of ballads as long as forty to fifty stanzas, although we knew quite well that every piece ought to be recorded from beginning to end.

Third, the records are mechanically fairly good, in any case, much better than the average of our European scientific records made on the spot (in the villages). Aluminum disks were used; this material is very durable so that one may play back the records heaven knows how often, without the slightest deterioration. Sometimes the tracks are too shallow, but copies can be made in almost limitless numbers.

Records Continuous

Fourth, the records of even the longest pieces are continuous, thanks to the two disk-plates on the recording machine. Theoretically, every piece, no matter how long it be, could have been recorded without any interruption (the singer, of course, had to get a rest after a few hours' singing). I have some melancholy recollections of our worries and troubles, when, after each two and a half minutes of singing, the business had to be stopped, the ready record taken off, the new blank put on, and in the meantime, the singer generally forgot where he left off.

Fifth, there are many "conversations" in addition to the songs incorporated in the recording, talks between collector and singer concerning data connected with the song, with the singer, with the circumstances referring to the performance of the song, etc. When you listen to these "conversations" you really have the feeling of being on the spot, talking yourself with those peasant singers. It gives you a thrilling impression of liveliness, of life itself.

Sixth, some of the heroic poems, or at least some parts of them, have been recorded from the same singer twice, with an interval of some days or some weeks between the recordings. This proceeding is important. The differences on the one hand and the identical parts on the other hand will show what parts of the words (or melodies) are more constant, what parts are more subject to changes, and to what degree. (The reader must have in mind that folk-songs are a living material; and, as every really living thing or being, subject to perpetual changes, preserving constancy only of certain general formulae.) Very few instances in our European collections are known where this could have been done. As a variation of this experiment, the same poem has been recorded from different singers, in order to show what are the personal traits depending on the individual singers, and what are the permanent ones, beyond the personality of the singer.

These are the outstanding qualities which raise this unique collection to a level never attained by any European collection known to me.

Certainly, it has some shortcomings, but where could one find a collection without any fault? Folk music collections inherently cannot be done without imperfections. Besides, these shortcomings are negligible in comparison with the accomplishments. One shortcoming is on a spiritual plane. According to a rule well known to research workers on folk music, the transcription of the words of the recorded song should be made immediately after finishing the record; in any case, while the singer is present and available for eventual explanations. This rule has not been observed, probably in order to save time.

The collector worked with a Croatian, coming from the ranks, but with a fairly good education and with a keen sense for folk poems. After completion of the research work in Yugoslavia, the Croatian was brought to the United States and he worked here for many months, transcribing the texts of the records. Although he had been present at the work of the recording in Yugoslavia and knew the language and its various dialects, it nevertheless happened that he could not make out, here and there, some words or some lines from the records.

This shortcoming — considering its rare occurrence — is quite negligible, and I mention it only to indicate my absolute impartiality of approach to professor Parry's great work. Had he insisted on transcribing the text on the spot, he would have lost months and months in doing so, and would probably have made only half of the recording. Even if one starts the business of collecting folk songs with the firm resolution to transcribe every line, every word of the texts on the spot, it happens — and I know from personal experience how often it happens — that for some reason or other, there is absolutely no possibility of acting according to this principle.

Recording the People

Having done one preliminary notation of the melody and the transcription of the whole text, there follows the great moment of recording. (This used to be the method of my work.) It frequently occurs that the singer involuntarily changes some of the words or lines or even she or he puts altogether different words to the melody when recording it. Sometimes the collector, who is always in a hurry, and working in a state of excitement, does not perceive the slighter differences and believes everything to be all right. Sometimes it may happen that one of the singer's relatives angrily interrupts the work and orders him or her to leave the friendly meeting immediately and go hastily home. Now, even if the collector noticed some deviations in the words of the record, he scarcely can do anything to obtain the variations on the spot.

Letting Well Enough Alone

To get the singer again would mean perhaps a loss of one or several days; so the collector leaves matters as they are and takes his booty home, only to find out that there are several words or lines in his records which he cannot understand and transcribe at all. And to tell the truth, this occurs even more frequently in — let us say my collection — than in the Parry Collection.

Another shortcoming is the voluntary one-sidedness of the material: 2,200 records with heroic poems and a little more than 200 with other types of melodies give no adequate picture of the actual frequency of the various types of melodies (some important types are missing entirely). Professor Parry's original idea was to investigate only the heroic poems and the ways and means of performance. We can be thankful to him for not having adhered to his first plan and for having recorded some examples of other types too. On the other hand, had he recorded more of the "women's" songs he probably could not have recorded so many heroic poems. This latter and more important part is as complete as humanly possible. This was his principle aim and this he achieved.

Done in the Twelfth Hour

When I stress the international importance of having at our disposal a collection of Yugoslav heroic poems as complete as possible, I must add the following observations:

The work was done in the twelfth hour. Already in 1934 and 1935 there were signs of slow but constant decline and deterioration in the life of heroic poems. Heaven knows if and how they can survive the present disastrous events at all.

These poems seem to be the last remnants of a folk usage at least several thousand years old, expressed in words and in music, leading back perhaps to antiquity, to the times of the Homeric poems. Nowhere else can be found a similar usage, not even in the other countries in the Balkans. Professor Parry, by his intuition, seized the last chance to preserve its sound image.

I am very grateful to Columbia and Harvard Universities for the privilege of transcribing and studying these songs. And all of us who are interested in Eastern European folk music will always think with deep gratitude of the late Professor Parry to whom we are indebted for this material.