MARCELLO SORCE KELLER

SEGMENTAL PROCEDURES
IN THE TRASMISSION OF FOLK
SONGS IN TRENTINO [1]

SONUS - VOL. 8 - No. 2 - SPRING 1988

INTRODUCTION

This article intends to show how recent research in Italian folk music allows further insights into the process of oral composition. It does so by focusing on an important aspect of the process: melodie variation. In fact, in the domain ot folk music the study of variation is equivalent to the study of composition in art-music since the folk artist almost never sets out to consciously perform a new piece. It is by progressive variation of a given melodic line, performance after performance (by the same person, or several), that folk musicians compose. In doing so they have in mind not the desire to be creative or innovative but, rather, that of producing a convincing performance.
Expressions such as transmission and variation, in the field of folk music, simply reflect two aspects of the compositional process. This is a process that even in the realm of European art-music is progressively being recognized as chiefly a transformation process rather than one involving what in the Romantic era was referred to as invention or creation.
I will discuss one particular instance of variation-composition encountered in Trentino, a Region in North-Eastern Italy. It shows, in my opinion, how even in a relatively well known area, one may still find traditions where the musical process unfolds in manners that are more complex than formerly assumed.

THE SEGMENTAL PROCESS
The folk music of Trentino is mostly vocal, choral, and rich in narrative songs. [2] The melodies are made up of phrases some of which constitute a small repertoire of what I call segments. Segments may migrate from song to song, regardless of their length. Exceptionally, they may be as long as the melody itself or, more commonly, a four-bar phrase, or a two-har semiphrase.
The dynamics of segment circulation are cmplex because, giben their variable length, they may cause change at different structural levels in host songs. When a segment coincides with an entire melody, then we simply have a wandering tune associating with several texts (not a rare case in European folksong) . Naturally, a new text forces a melody to adapt by fragmenting or prolonging note values and, also, by rearranging some of its stress points. The other case are more intriguing because four and two-bar segments qualify for a particular position only within a song: incipit or beginning, middle part, or ending cadence. What makes the difference here is contour, rhythmic pattern, and most of all whether the tonic or the dominant is emphasized.
A few segments fare from song to song with great frequency. What propels their mobility is, at least in part, the interchange between the solo and the choral singing of the sane songs. Choral singing is by far the most common in Trentino. Its texture is often made up of a primo (usually, but not always, carrying the melody), and a secondo producing intervals of a third, sixth, and rarely a fifth. In addition one may have a third, lower part, with few, slow, repeated notes. Whenever someone sings solo, he/she often offers a synthetic version of the tune or, to put it better, a juxtaposition of segments taken from both primo and secondo as heard in some choral performance of the same song. [3] Segments from both parts, that is, are put together to form a continuous melodic line which may effectively be performed as such. [4] In addition, just as a segment way move from one part to another, it may also migrate from song to song.
The evolution of songs in Trentino, therefore, is larqely the result of the progressive replacement of segments. A segment inserted into a tune always replaces another one that is then dropped. At this point the received segment adapts to the text-line it finds. Because of this adaptation note values may be fragmented, or lengthened, and contour tself, eventually, somewhat altered. Of course, a melody may tolerate only so many substitutions before losing its identity and turning into a different one, and many tunes eventually do so.
This process taking place in Trentino has parallels elsewhere. Duncan Emrich, for instance, speaks of slippage, (or spillage) referring to the transfer of lines and, sometimes, of entire stanzas (with or without their music), so frequent among ballads in the Anglo-American tradition. [5] Segmental substitution as encountered in Trentino is, however, somewhat different because, unlike slippage, it only pertains to the melody of songs; and because it is much too frequent to be simply the result of chance (slips of memory, lapses, etc.). I suggest that the singers of Trentino unconsciously practice segmental substitution as a means of generating variety and, at the same time, maintaining a considerable deqree of uniformity throughout their whole repertoire. Naturally, fortuitous slippage, the formulaic composition discussed by Albert B. Lord, and segmental substitution, are all elements of the multifaceted process of oral composition which involves both text and music.
Let us now examine a few typical examples from Trentino. In Ex.1 there are three cases of tunes whose first phrase is almost identical, at the very beginning, and then progressively diverse. Exs. 2 and 3, likewise, show two other sets of tunes that share the same incipit whereas Ex. 4 shows the similar central phrase (or, rather, variants of it) of otherwise different tunes. Differences in melodic contour and rhythm are here largely the result of text underlay.
Ex. 5 shows several related tunes; we might call it a tune family if, beginning with the third bar these tunes were not progressively diverging. Exs. 5a and 5b are variants of the song «Ero un povero disertore». We may easily see how similar they are. In the last three of both the melody of 5a resembles the countermolody of 5b. Tunes c,d,e,f,g,h and i of Ex. 5, show how transformations of this incipit have become part of songs which are melodically different from «Ero un povero disertore» as they progress beyond the initial two bars. Let us divide, for the sake of this analysis, their first phrase into two parts. The first, up until the middle of the third bar, undergoes minimal change from one song to another. The second part, from the middle of the third bar to the end of the song (of which only two bars are here reproduced) , functions as a bridge towards the remainder of the tunes, their original part in a way (the one that as there before the incipit was borrowed from elsewhere).
All these incipits are clearly very homogenous. Tunes 5e and 5f, in particular, are identical but support a different text. The few minor differences between them, occurring from stanza to stanza (only one appears in my short example), mostly repeated notes, may again be accounted for by the necessity of accomodating words of diverse syllable count at corresponding locations within the melody.
Since the story is what really matters to the folksingers of Trentino, identical tunes with different texts are considered altogether different songs. This is twice the case among the examples presented so far. In 5e and 5f there is virtual melodic identity, whereas 5h and 5i are variants of the same tune. Exs. 1 to 5, therefore, illustrate the spreading of related melodic segments an two levels: that of a short motif, and that of entire tunes associating with unrelated texts.

FINAL CONSIDERATIONS
It is not unusual to find in Italy musical traditions that are very local and, at the same time, very distinctive. Trentino is one such case. From its examination we may draw that this body of folk music, like others elsewhere, does not exploit all the compositional possibilities theoretically available but, rather, specializes in a few and develops them to a high degree of sophistication. The recognition of which these procedures is adopted tells us something about the quality and degree ot innovation a particular community allows and favours. The people of Trentino like a certain amount of melodic innovation but they also like their songs to be constructed in a similar fashion. They enjoy a melodic climax, as well as caesuras and cadential points. That is why itinerant segments are either incipits, middle parts, etc. The contrary is known to be true of other Italian repertories. [6] That of Trentino is one more case to suggest that variation through gradual, progressive, alteration of a single melodic line may be much less common in orally transmitted music than it was usually thought. The question to be answered at this point, therefore, is why do we perceive as linear the end result of a process that is, on the contrary, segmental.

NOTES
[1] This article combines material discussed in a paper entitled «Kaleidoscopic and Segmental Composition: a Look at Oral Transmission in Italian Folk Music» (Midwest Chapter of the American Musicological Society, Chicago, September 22, 1984) and in another entitled «Procédé segmentaux dans !a transmission des chansons populaires dans la region Italienne du Trentino» (Seminaire Europeen d'Ethnomusicologie, Paris, October 23, 1987).

[2] For information about folk music in Trentino and the surrounding region see M. Sorce Keller «Folk music in Trentino», Ethnomusicology, Vol. XXVIII, No. 1, Jan 1984, pp. 75-B9 and, by the same author, «Life of a Traditional Ballad in Oral Tradition and Choral Practice», Ethnomusicology, Vol. XXX, No. 3, Fall 1986, pp. 449-469.

[3] The primo is the top voice while the secondo sings below. Sometimes, however, the accompanying voice sings above, rather than below the primo.

[4] It is as if the choral version of a song that can be also performed solo were the result of durchbrochene Arbeit.

[5] Duncan Emrich, Folklore on the American Land, Little, Brown and Company, Boston-Toronto, 1972, pp. 18-19.

[6] See, for instance, Giovanni Giuriati, «Un procedimento compositivo kaleidoscopico: la tarantella di Montemarano». Culture Musicali, I (1982) , No. 2, pp. 19-72, and F. Weis Bentzon, The Launeddas, a Sardinian Folk instrument, Acta Ethnomusicologica Danica, Akademisk Forlag, Copenhagen, 1969, 2 vols.