How to digitalize the folk song archives?

Lozanka Peycheva, Grigor GrigorovInstitute of Folklore

Bulgarian Academy of Science

Goals (1)

• Description of the formal features of the folk songs as musical and verbal works.

• Choice of basic classifiers to be introduced into the developed digital library with the aim to allow combined search along different criteria. The choice should be made with a few important things in mind:

– the users’ necessities; – the condition of the physical records and the availability of necessary

data;– the overall expedience – the labor consumption compared to the

expected benefits from the digital library. – an expert assessment defining whether the different classification

data should be used in automated search engines or non-automated indexes.

Goals (2)

• This presentation is an attempt at expert assessment of the different indexing criteria that will be used in a planned digital library, which will be developed by a team of scientists from the Institute of Mathematics at BAS and the Sofia University St. “Climent Ohridsky” for a project called “Information technologies for representation of Bulgarian folk songs in a digital library of music, notations and lyrics”

• The sample subject for the suggested classifiers in this text is the unpublished fundamental volume by Todor Dzhidzhev containing 1100 songs recorded in the 1960s in Thrace.

Necessary Limitations (1)

• The short format of this text imposes certain limitations on the depth of the presented content.

• The expert content of the chosen classifiers will not be explained, nor will be discussed the issue of what scientific inquiries could be done using the digitalized songs.

Necessary Limitations (2)

• This text does not discuss: – the relationships between poetry, music and dance; – the musical-folklore genres and their relation to the folklore

situations; – implicitly from a musicological aspect: the formation of

modes, structuring of musical forms, verticality of the musical process.

• The first thing that would draw the interest of any user of the digital library is the available data about the recorded informational objects (meaning the songs). This will be the gathered field research data about the digitalized songs, which will provide the most accurate information about the songs.

Information on the Records (1)

• Each folk song possesses identifiers which specify the time and place of its gathering. Some of these are:1. signature (a unique number of the song);2. informant (person who conveyed the song to the

folklorist);3. birth date of the informant;4. recorder (name of the folklorist who gathered the song);5. time of record (date when the song was gathered);6. settlement (village in which the song was gathered);7. municipality (governing the village);8. administrative region (in which the village was located);9. region of folk dialect;10. the song genre according to the informant/recorder

Information on the Records (2)

• Assessment: Among the listed identification attributes of the recorded songs, those who represent important data that will be useful as a search criteria and means of analysis are 1) the information about the geographic regions where the song has been popular (i.e. the settlement, municipality, district, folk-dialect region), which will allow for territorial distribution of the folk examples; and 2) the time of recording, which will allowing for historical/chronological research of the folk tradition.

• The rest of the meta data could be displayed in non-automated indexes.

Musical Aspects – Theory (1)

• Musical Characteristics• Music combines a couple of elements of

diverse nature: 1. rhythm (the temporal aspect of music); 2. melody (the pitch in music, intonation and tone

content); 3. polyphony (cases of diaphony, polyphony,

harmony).

• Each one of these components can be described according to many different classifiers.

Musical Aspects – Theory (2)

• Rhythm is not a specifically musical concept, it can be presented as an universal concept, an universal cosmic category and is interpreted in an exclusively broad way. [Velcheva, 2007: 49, 55]. According to ancient peoples rhythm is the male principle in music (it was thought to have great magical power), while intonation (tone contents) is thought to be the female principle, which obtains the forms endowed by rhythm [Dzhudzhev, 1980: 81].

• Rhythm is: 1. an organizer of time; 2. a primary morphological feature; 3. the common thing that unifies all – the melody, the poetical text, the

dancing steps, the working and ritual gestures; 4. an organizer of the tone content; 5. a natural course of coherence („regularity”) in the ordering of time

measure;6. the basic and initial factor of music formation.

Musical Aspects – Theory (3)

• Metrum is a form of organization of the musical rhythm – a means (general measure) of measuring and coordinating of the rhythmic durations and accents.

• Metro-rhythm is associated with the periodical measurement (regularity) of time and with the accents in the rhythmical movement.

Musical Aspects. Rhythm (1)

• The main metro-rhythmic song features could be described by means of the following basic classifiers: 1. Measure (expresses the times of the metrum with number of beats in

a single rhythmical entity. It is written with digits in the beginning of the note score)

2. Time (the single rhythmical entity of which a melody is consisted – melodies with consecutive times of the same length are called measured and melodies without time are called non-measured)

3. Rhythmical scheme (forms the skeleton and outline the rhythmical graph of the song melodies)

4. Rhythmical cores (combinations of short and long times in elementary rhythmical groups called metric steps)

5. Rhythmic stereotypes (established patterns, which have been in the consciousness of an ethnical community for centuries)

6. Rhythmic genealogy (classification of the songs according to their similarities on the scale of homorhythmics – heterorhythmics, which can be used for research on their evolutional development)

Musical Aspects. Rhythm (2)

• Assessment: Among the listed classifiers, time and measure ought to be included as classifiers in the digital library, as the only objective indicators, because their ordering is independent from the subjective interpretation of the future users (the subjective interpretation is unavoidable with any kind of systematization – creation of systematic catalogues, indexes, etc.) and so their value can be given unambiguously.

Musical Aspects. Melody (1)

• Melody is related to the pitch in music. According to Svetlana Zaharieva “The relationships between different pitches, which have crystallized in melody, are bearers of musical specifics and are considered the “purest”, most specific aspects of music from the musicological perspective, because they are just musical, without non-musical sides, which can be found in rhythm for instance, from the perspective of the poetic or dancing formation” [Zaharieva, 1979: 54].

• We can say the following about the musical folklore melodic line: 1. It is a bearer of intonation of musical type; 2. It is clearly outlined with its relief, extreme pitches (the initial, the top

and final tone); 3. It manifests as intonation models – individualized and structurally

outlined melodic-linear entities; 4. In each village (micro-dialect region) dozens of such intonation

models used to be performed.

Musical Aspects. Melody (2)

• To achieve searching and classification (grouping) of melodies according to some of their similar (identical) intonation features, musicologists have outlined several sound categories, describing the following elements of melodics: 1. Tone volume (ambitus). (the interval between the highest and lowest

tones in the melody)2. Sound order. (all tones of a melody, sorted in ascending or

descending order according to their pitch)3. Melody schemes (the melodic (intonation) schemes of the song)4. Components of the melody line (pre-thematic elements): sound

linearity, skip –gradualness. 5. Melodic cores (little, but individually outlined tone combinations, which

have their own expressive appearance) 6. Music formulas and themes (available patterns, inherited from

tradition, complete structural units with specific musical expression)7. Melodic stereotypes (ornaments, cadence stereotypes, introductions

and endings).

Musical Aspects. Melody (3)

• Assessment: Classifiers in the digital library among the indicated components can be the following: tone volume (ambitus) and sound order. The grouping/classification of folk songs according to tone volume (ambitus) and sound order is independent from subjective interpretations.

Implicit Verbal Data. Context of Song Performance (1)

• The meaning of a folk song arises at the crossing point between her text and the context in which it is used. Because the song is not just a text, it is a statement, verbal and musical act.

• In the opinion of Roman Jakobson, the statement contains 6 components: sender, receiver, context (time and space), message (text), channel (i.e. type of the statement) and code (the language on which the statement has been made).

Implicit Verbal Data. Context of Song Performance (2)

• In order to describe all these components as non-crossing sets many classifiers are necessary. The analysis shows that the description of the specific features of the sender and the receiver need at least 5 classifiers, of time and space – 5 classifiers more, the channel needs 3 more and the code specifics – 3 more classifiers.

• The complete description of context is impossible: because of the frequent lack of data, and for the sake of expedience – the labor would not be compensated by the benefits in effect.

Implicit Verbal Data. Context of Song Performance (3)

• Considering the situation, what should we do?• Assessment: For the aims of the digital library we need a

description on two components:1. Time-and-space of performance – here are the typical genre

definitions of the songs such as: sedenkarski (sung at work-and-woo gatherings) songs, harvest songs, Christmas folk songs etc. – this kind of meta data is also most thoroughly collected.

2. Additional indicators for the performance context: “sung to a lassie”, “sung on the way”, “sung by men only” etc.

• The Problem: The two different criteria define different crossing sets of data. The decision to use them is awful from a theoretical perspective but was chosen for the lack of a better one. In spite of its obvious disadvantage it will work well, since it follows the established genre classification, which has also been based on different criteria.

Implicit Verbal Data. Content of the Song. 1. Dialogue (1)

• The content of the song consists of three components: narration (development of the action), description (description of persons and objects) and actual dialogue.

• The center of the song is the dialogue – the other components are optional. Therefore the dialogue should naturally be the focus of the description. The description of the dialogue can again be made with use of the Jakobson model.

• The description of each participant in the song’s plot needs at least 17 classifiers – they are more since the participants might be not only common people (but also saints, animals, demons etc.). And more classifiers will be needed to represent family relations, religious, ethnic and other types of relations. The specifics of the context will require at least 6 classifiers, and of the communicational contact – 3 more.

Implicit Verbal Data. Content of the Song. 1. Dialogue (2)

• Assessment: It is not appropriate to introduce these classifiers in the digital library – the colossal labor of their description will not be compensated by future benefits. Still, the information could be processed in non-automated indexes.

Inmplicit Verbal Data. 2. Formal Features (1)

• The important and useful formal features are associated with two groups of data:

• Number of verses, availability of refrain, availability of otpyavane (one singer sings a melody and another singer repeats it in answer), availability of chorus (in the more modern layers), introductory verse, structure of the verse. All of these are theoretical problems.

• Key semiotic systems – personal names (significant for the reconstructive theories), numbers, colors (significant for the volumes in symbolism of the number/color).

Inmplicit Verbal Data. 2. Formal Features (2)

• Assessment: Among the above only the verse structure is meaningful to be introduced as a classifier in the digital library. It has to be introduced as a formula (of the type 4 + 6, 6 + 6, where the first number shows the number of syllables in the first hеmistich, the second one – the number of syllables in the second hemistich, the sign “+” reflects the place of the caesura (intonation pause).

• The rest of the data can be represented by non-automated indexes.

Implicit Verbal Data. 3. Thematic focuses (1)

• When we speak about the verbal component of the digitalized folk songs, we hit the inevitable question about their content, and about the topics on which it is focused. Is it possible to order the digitalized songs according to their topics?

• It is impossible to describe the themes of a song with formal criteria. Still, stable thematic focuses can be fixed, which appear to be the backbone of the plot and the engine driving the action. Here are the most important ones: 1. praise; 2. sickness; 3. death; 4. murder; 5. flirting/wooing (in the broadest sense of the word); 6. betrothal; 7. marriage; 8. birth / pregnancy; 9. gift; 10. song; 11. crying; 12. oath; 13. blessing; 14. laughter; 15. dream / dreaming; 16. forbiddance; 17. sin according to traditional ethics; 18. plead.

Implicit Verbal Data. 3. Thematic focuses (2)

• Assessment: It is appropriate that these classifiers (and probably others, which can be added later) are introduced in the designed digital library. Availability /non-availability of separate thematic focuses or sets of thematic focuses would give comparatively good idea about the content of the song.

Conclusion

• Designing of proper classifiers that properly represent the creative principles (musical and verbal) in a specific model of the digital library will assist the searching for and finding of knowledge in specific scientific fields. This can become possible through a series of systematic classifications (which will be accomplished without file cards). These proceedings will present diverse possibilities for practical orientation around the materials available in the digital library and for systematization of the digitalized musical folklore into informational objects, described by various criteria.

• The digital library created by our team aims to present a model and a key to a series of further researches, which could be done on a more huge quantity of musical folklore material from Bulgaria.