The use of the capabilities of computer technologies has long been used for the classication, systematization, and analysis of musical texts based on the material of folklore and vocal-instrumental genres. However, such research was not conducted on musical samples of medieval sacred monody, although this material allows us to establish the formative intonation patterns of European music. The study of the selected repertoire of the Lyubachiv irmologion of 1674 with the help of digital tools is a means of building intonation and lexical dictionaries of Ukrainian culture of the 17th century.
Ukrainian art is vividly represented in the liturgical plane, in particular in the medieval monody – unison liturgical singing. This is an iconic singing repertoire, preserved in handwritten notated collections of the 16th-18th centuries – irmologions, which, in addition to being purely liturgical, also served as a singing textbook for all Ukrainian children. Therefore, the number of irmologions grew rapidly, and the musical motifs and poetic vocabulary of the repertoire were quickly mastered. In the future, this significantly influenced the content and form of numerous literary, artistic, and musical genres of Ukrainian culture.
For research, we chose the Lyubachiv irmologion of 1674, created by Pavlo Smerechanskyi
[
Attempts to combine music and mathematics or computer analysis approaches are not new
[
The bibliometric analysis consists of a quantitative evaluation of the publication activity on
the subject, followed by the qualitative analysis of the publications [
The Scopus indexing database was chosen as the source of trustworthy, peer-reviewed works. The “monody” was used as a search word, and searched within Article title, Abstract, Keywords sections (all results are true on May, 31st, 2024). As a result, 169 papers were found. As the next step, the papers from russian federation and in russian were excluded. Of 153 papers remaining, 69 papers from the 2014-2024 period were analyzed. Of them 14 were excluded, as they were not connected to music or monody. Fig. 1 shows the dynamics of the publication activity since 2014 (RQ1), and presents the geographical distribution of papers being published on the monody study (RQ3).
Among 56 analyzed papers, 69.3% had Arts and Humanities as a subject area, 26.7% were of
Social Sciences, 2.7% were of Computer Science, and 1.3% of Psychology (RQ2). As mentioned,
the Computer Science subject area was 2.7% of the papers, which is two papers. In 2017 Italian
scientists adapted a linguistic technic of stylometry to attribute medieval liturgical monody
with a quantitative approach [
It is obvious, that information technologies are not widely used for monody analysis, which can be explained by the lack of interdisciplinary cooperation between musicians, musicologists, and IT specialists. The current work aims to fill in the gap of interdisciplinary approaches between music and information technologies and uses the Ukrainian monody as the research object (Figure 1).
Documents per year
Documents by country or territory 4 3 7 10
10 5 6 3 3 2 2 8 5 5 5 4
Let us consider the steps of the algorithm of monody structure study with the use of
information technologies, and as the object of research, we will use the Lyubachiv irmologion,
namely, the eight songs of the first Mode. We shall indicate the functionality or type of the
corresponding information technology in Table 1. The input data for initiating the process of
studying the monody structure is a digital music sheet with a score in linear notation. Such
music sheets can be in the form of a digital photo of a book page (Figure 2а), in the form of an
editable music score, typed with a score writer [
Step description
tal music score camera, scanner If the music score is edita- Score editing software Online/desktop software, ble, go to Step 5, other- paid. wise go to Step 2.
If the music score can be Music optical character Desktop software, mostly recognized, go to Step 4, recognition (OCR) paid, or with limited time otherwise go to Step 3. software of free access.
Create the music score, go Score writing software Online/desktop software, to Step 5 paid.
Recognize the music score Music OCR software Desktop software, mostly paid, or with limited time of free access.
Convert music score into Score editing/writing Mostly paid, machine-readable format software online/desktop version Perform quantitative Data analysis software. For data analysis, designanalysis ing new software might be expensive. On the other hand, it might be extremely difficult to find existing data analysis software with specific functions, to be re-purposed to the music score analysis.
Generally accessible, free Perform qualitative analy- PC, text editor, etc.
sis. End of the algorithm
According to the algorithm from Table 1, if the input information is the PDF file of a music
score in a linear notation, as in Fig. 3, the algorithm`s steps sequence will be the following:
Step 0 -> Step 1 -> Step 2 -> Step 4 -> Step 5 -> Step 6 -> Step 7, and information technologies
to be applied are the PC, score editing/writing software, data analysis software, text editor.
For the input data as in Figure 2(b) (provided by the author, Natalya Syrotynska), in Step 4 the
Capella-scan software [
In the MusicXML format, the description of each note depends on its complexity. The elements <note>…</note>, <pitch>…</pitch>, <step>…</step>, <octave>…</octave>, <duration>… were used to describe most of the notes of the studied monody in MuscXML. </duration>, <voice>…</voice>, <type>…</type>. Its explanation and short statistics on the converted into machine-readable MusicXML format odes are in Table 2.
According to such statistics, an assumption can be made. Being abstract, chant`s music
score can be considered as a text, and each note can be treated as a word, as in linear notation
it is surrounded by spaces. We do agree with [
Following such an assumption, the AntConc software [
The AntConc is a free text analysis software, which was used in various research, such as
[
From Table 2 analysis, all the notes have the voice equal to 1, so we can omit this information. The note duration and type are equivalent (960=Half), and the numerical value can be omitted. Octave is only 4 or 5, so we shall code it with Four or Five, respectively. Following these simplifications, we will code every note MusicXML format into single-line form, for example, the note from Fig. 4 will be coded into BFourHalf.
After notes of all odes of the Irmose`s first mode were coded, the files were saved in .docx format and then uploaded to AntConc software. Figure 5 shows the File View tab of the AntConc, with eight odes uploaded, and Ode Nine`s content is mapped.
It can be seen, that the Ode 9 has 95 tokens which are notes, and 20 different notes are in this Ode.
Opposite to the language, the single note analysis will not be informative, and AntConc
has a tool to analyze the sets of words (notes here). The N-grams tab allows analysis of
ngrams, the chain of n notes, that occur more frequently than once. For justification regarding
the length of the alalysed chain, we will turn to O. Tsalay-Yakymenko [
The calculated results indicate that the most frequent trigram is
AFourQuarter BFourQuarter CFiveHalf, in linear notation, see Figure 7.
Among the AntConc functions there are no relevant tools to analyze the structure of the monody, namely, are there any organization or arrangements of musical elements in an ode. To perform such analysis, the NoteFinder software was designed. Functional demands for such software were the following: • • • • • •
Files in MusicXML format as input data.
Analysis of multiple files.
Allows uploading music sequences to search for.
Allows searching for different music sequences simultaneously.
Allows visual representation of the search results.
Saving both visual and quantitative information.
Such software was developed in python Jupyter Notebook (.ipynb) format but was packaged as .exe program, which means for software to work correctly, the archive should be unpacked, and .exe file should be run.
The NoteFinder software has no dependencies on the operating system. Minimum hardware requirements:
Processor: AMD Athlon X2 2.8 GHZ, Intel Core 2 Duo 2.4 GHZ.
Ram: 2 GB.
Video card: DirectX 10.1 (AMD Radeon HD 6450, Nvidia GeForce GT 460).
DirectX: version 11.
Network: No internet connection is required, software runs locally.
As sequences to analyze, we used top-10 frequent trigrams, as it was found using AntConc software (see Figure 8). Found sequences were created in MusicXML format.
Users can upload numerous music sequences, and then choose which of them to analyze, if not all. In Figure 9 the main window of the designed software is presented, with a list of files uploaded (to the left).
The results of the analysis present the number of musical structures (called constellation) occurrences in each Ode. Each constellation is marked with a different color, as mentioned in a legend on top of the window. The charts represent the organization of the constellation in a music score of the Ode. Fig. 10 presents a closer view of Ode 3, Ode 5, and Ode 6 structure by NoteFinder. From the visual analysis, it is seen, that three of Odes have a rather similar beginning.
From Figure 9 it is seen that Ode 7 has the least number of the constellation in it, and this is the only constellation. These results can be also retrieved from the .csv file, which NoteFinder exports (see Figure 11).
Combining visual and numeral results, it is seen, that the most various Odes are 1st, 5th, and 9th, and they have the most numerous appearances of constellations. It can be commented as the first, middle and the last Ode are as if they are framing the Mode.
This paper presents an algorithm for IT support of a process of monody structure analysis. Such analysis involves knowledge of information technologies and musicology, which makes such research interdisciplinary. While studying the monody with adapted software (AntConc) and software specifically designed for such analysis (NoteFinder), it is too early to draw any conclusions on the monody's structure, as only one Mode was analyzed, which is a limitation to such research. Nevertheless, this paper presents a viable approach, the continuation of such research depends on the volume of the analyzed material. To overcome the mentioned limitation, authors are working on widening the music scores in machine-readable format to be analyzed with the NoteFinder software. The current task is the creation of an electronic mapping of one of the Ukrainian monody's artifacts, the Lyubachiv irmologion, to be accessed freely, in various digital formats:
Book`s pages as the pictures, in image file format.
Odes as a sequence of pictures, as an image archive.
Odes in score-edited format.
Odes in interchangeable, machine-readable format (MIDI, MusicXML).
Odes texts in text editable format.
Being able to analyze and draw reasonable, quantitatively proved conclusions, the monody researchers will add to deepening the sphere of values and ideas about the essence of Ukrainian identity, to the preservation of medieval national artistic heritage, and to expand educational programs of various specialties, including practical and performing artistic directions. Also, the formation of a lexical dictionary of the 17th century is the prospect of further linguistic research.