"Ethnic" Music in the Balkans: Identity, Similarity and Classification Norms

Abstract

In extended regions, where several "ethnic" groups have been living together for many centuries, apart from systematic instruction, given at schools or universities, not always fixed boundaries develop for ideas, customs and social behavior. This is also true for music, and generally speaking for arts, whose intellectual achievements transcend human societies and cultures. In networked communities, the information superhighway brings together musical activity reflecting spheres of influence in the synchrony or diachrony of their ongoing social development. This research examines the features involved in setting up Metrics and Norms for Identity and Similarity, along with Classification methods for SouthEastern European and Eastern Mediterranean tunes as they have been traced in a survey in the Balkans.

Full text

“Ethnic” Music in the Balkans: Identity,
Similarity and Classification Norms
D. Politis1, R. Tzimas1, N. Paris2, V. Aleksić3
1Dept. of Informatics Aristotle University of Thessaloniki, Greece
2Dept. of Music Science and Art, University of Macedonia, Thessaloniki, Greece
3University of Kragujevac, Faculty of Technical Sciences, Čačak, Serbia
dpolitis@csd.auth.gr; tzimasr@gmail.com; nepa@uom.edu.gr; veljko.aleksic@ftn.kg.ac.rs
Abstract—In extended regions, where several “ethnic”
groups have been living together for many centuries, apart
from systematic instruction, given at schools or universities,
not always fixed boundaries develop for ideas, customs and
social behavior. This is also true for music, and generally
speaking for arts, whose intellectual achievements transcend
human societies and cultures. In networked communities,
the information superhighway brings together musical
activity reflecting spheres of influence in the synchrony or
diachrony of their ongoing social development. This
research examines the features involved in setting up
Metrics and Norms for Identity and Similarity, along with
Classification methods for South-Eastern European and
Eastern Mediterranean tunes as they have been traced in a
survey in the Balkans.
Keywords—Comparative Study of Musical Features, Modes,
Genres, Scales, Organizing Principles.
I. INTRODUCTION
In recent times, music distribution has reached a highest
point of influence, as cost-effective business models
overwhelmingly supply virtualization platforms, i.e.
Software-as-a-Service networks, providing endless
streams of melodious hearings. Some of these function as
on-demand payable amenities, while others base their
revenues on public media-like promotional campaigns [1].
Even further, unauthorized reproduction of artistic
works, usually developed on parallel with software piracy,
has made it possible for developing countries to be
provided with amenities, like music streams, that
otherwise would be unaffordable, and thus, limited in
circulation for them [2].
This form of communication reaches large numbers of
people, and is vigorously added to the existing triad for
music distribution: radio, television and digital discs sales.
However, while some methods for transmitting music,
like RF broadcasts, are subject to geographical restrictions
due to physical or imposed controls by regulating
authorities, others have a clearly transnational range,
operating across state boundaries and nationalities. Not
withstanding the commercial model for music distribution,
contemporary music thrives via various alternative
channels of communication, albeit sometimes piracy is
promoted as an essential factor. Therefore, somehow, an
extensive beyond border-lines continuum is shaped for
musical exteroception.
In the region of the Balkans, it is not unfamiliar to have
in concerts concurrently traditional Slavonic music
patterns accompanied by 19th century Western European
polyphonic choral renditions, Byzantine scales interfering
with maqams, microtonal scales from antiquity revived by
contemporary composers of electronic music, to mention a
few styles of popular art [3].
Musical depth in its diachrony mingles with apparent
chartbuster trademarks, like Spotify, Amazon, e-Bay,
YouTube or iTunes, which contribute to the excessive
increase in the supply chain for goods and services over
the Internet. For practical reasons, all these providers need
to classify, often in an arbitrary or controversial way, their
commodities available on demand.
Music can be categorized into different genres in a
multitude of ways. There are commercial and academic
approaches for such a classification. Genres (from French,
meaning “kind” or “sort”, originating from Latin GENUS
and Greek ΓΕΝΟΣ) were consistently used for many
centuries to classify in ordered sets of similar works
literature pieces along with other forms of art and
entertainment [4]. As new forms of music are invented,
genres may be altered, discontinued or mixed together to
produce new forms [5]. It is also possible for opuses of
music to fit into two or more categories.
As a result, the way that music is classified may be
different when encountered as a commercial activity,
strongly biased by the production tactics of the prolific
music industry, and when academic criteria are applied
In academic terms, for instance, Beethoven's Op. 61
and Mendelssohn's Op. 64 violin concertos are considered
part meronyms of the same genre, but somehow different
in form [6]. Therefore, apart from the genre classification,
we may have hyponyms for style and form classifiers. Not
all scholars agree on how classification should be
administered, or how different is style from genre, but
more or less it is accepted that they rely on common
“basic musical language” characteristics.
Even further, technological vendors and some scholars,
having experience in music which has proliferated in the
Western world [7], have developed criteria in response to
demand by audiences and producers that live in such
countries, and consequently characterize all other forms of
music outside their paradigm as “ethnic” (Figure 1).

(a)
(b)
Figure 1. (a) A well-appreciated commercial scheme for classifying music;
image cropped from social media (b) an Academic approach in
discriminating Genuses and Modes by comparing Scales. The intervals per
octave are denoted in echomoria.
Figure 2. The diatonic scale of Byzantine Music compared to
chromatic and enharmonic scales. Intervals denoted in echomoria.
There seems not to be a dynamic tool to segregate what
is common between the music of Balkan countries,
Middle East countries or countries around the
Mediterranean basin [7]. When Far East traditions are
encountered, matters become more complicated. For
example, researchers are systematically investigating what
common exists between a song accompanied by a bass in
one case and tambour in another. They also try to encode
what is the bias exercised by the musical “language” used,
along of course with the phraseological attributes of the
natural language used and the idiolect features it may
mold in singing [8].
Therefore, apart from the overwhelmingly obvious
characteristics, that have to do with the CMN-based
classification schemes, there other criteria that aid the
distributors and the public to make sense out of
“unpredictable art” and not pack it in the loose, shapeless
category of “ethnic” [9].
It is true that up to now most contemporary music
produced originated from the “Western” world, not to say
the English speaking part of it. It is also obvious that
styles like “Hip-hop” or “Rap” are highly influential
among young audiences and thus have been adopted by
the community of “ethnic” music as well. However, it is
not the same to have the Hispanic version of it or the
Middle East variants of it.
Even further, as modern technology facilitates the
production and distribution of music, it seems that most
arrangements will be sooner or later produced by the
“other” world, whose tradition was styled as “ethnic”.
This paper presents examples of Balkan Music, along
with the scheme of melodies used in the greater region of
South-Eastern Europe and East Mediterranean, that
provide intuitive incentives for more accurate classifiers,
based on Identity and Similarity characteristics.
II.
P
ROBLEM
F
ORMULATION
Music, in its generalization and in a global level,
comprises a complex phenomenon, which cannot be
uniquely described, due to its inherent acoustical
variegation. This diversity in the characteristics of the
auditory effect is due to the geographical dispersion, the
musical culture of the peoples, the historical periods of
music, the languages used and other issues, which are
studied extensively in ethnomusicology.
One of the greatest differences spotted between the so-
called “Western” and “Eastern” Music is related to the
musical scales and modes that are used in both cases.
Eastern modes (e.g. Oriental Music, Byzantine Music,
etc.) have a very wide range of microtonal intervals and
therefore strong use of music chromaticism, both in
notation and during performance [7].
This is pictorially explained in Figure 1(b): the
chromatic scale of Western Music, in its most decorative
form, as far as semitones are concerned, is parallelized, in
terms of similarity, with the most diatonic of the
Byzantine Music scales, that of the 8
th
Mode [10]. The
strong lines denote the fixed notes and the intervals
formed between them, in echomoria, rather than in cents.
The full octave thus extends to 72 echomoria, having as
equivalent 1200 cents.
The double-edged arrows denote pitched toned
incompatibilities for specific notes (E with Βου and B
with Ζω) [10].
The dotted lines denote the alterations of natural pitches
that may appear as accidentals of various forms. This
comparison is not derogatory to Western Music; it rather
reflects the fact that in “West” instrumental symphonic
music has prevailed, demonstrating the zenith of its
influence with polyphony and contrapuntal masterpieces,
while in “East” the epicenter remains the interminably
evolving singing voice. For example, this microtonal
nature is evident as everyday practice in Middle East,
when the muezzin’s voice is coming atop the minarets
fives times announcing the call to prayer in a melodic
way. Furthermore, the wide range of “exotic” microtonal
fluctuations seems quite logical to contemporary
researchers, since the mathematic modeling of music
allows the extensive use of sound frequencies and
intervals, from both electronic musical instruments
(except for those tuned and restricted according to the
equal temperament system, like the piano) and the en
masse singing voice recordings, dynamically variable due
to human physiology deviations from typical norms [11].
On the other hand, this microtonal autonomy has been
significantly reduced in Western music as, apart from the
tonic notes involved, in conventional harmony, only two
“modes” have been established as dominant (major /
minor) yielding the apposite scales. In East, however,
there is an astounding variety of scales and modes, as seen
in Figure 2 for Byzantine Music [9][10][12].
This particular simplification is usually only adopted in
notation (Common Music Notation – CMN), since in

actual performances, the deviations from notation are
obvious and in many cases overshadow the writings on the
sheet. This can become obvious from the comparison of
two audio performances of the same song from different
artists. Although both versions are based on exactly the
same western music notation (they can even be in the
same tonality as well), the auditory and aesthetic result
can be very different for many reasons: different
chromaticism [7], different musical instruments, etc.
However, although in our era music circulates
predominantly recorded, in the majority of cases, for
centuries, music was received in written form. Therefore,
the acoustic content related with the conceptual semantics
of music (i.e. its semasiology) was inferred out of its
semiology [13].
III. INFORMATION THEORY: THE THEORETICAL
BACKGROUND
When different sets of symbols are used for music
notation, then a mathematical quantity expressing the “a
priori” probability of occurrence for a particular symbol or
a sequence of symbols may be used, as contrasted with
that of alternative sequences of basic elements.
If M different symbols are encountered, representing
the multitude of notes for a certain musical system in fixed
order 1, 2, ... , Μ, and having correspondingly
possibilities of appearance p1, p2, ... , pM within a melody,
in Information Theory terms no other constraint may be
set for the use of these symbols, at least denoting
knowledge that proceeds from theoretical deduction [14].
Shannon's equation denotes that if M different symbols are
encountered in a sequence of G musical events, then the
“entropy”, i.e. the informational content of this series of
semeiotic events is
H = 𝐺𝐼 = 𝐺∑𝑝𝑙𝑛𝑝


(1)
where I is the mean piece of information delivered by
each symbol encountered in a melodic sequence, and Gk a
constant dependent on the analysis of k-length tuples of
notes; it is also true for these symbols that
∑𝑝= 1


(2)
How this concept works out may be demonstrated with
the use of a very simple “alphabet”, comprising of two
“letters”, namely 0 and 1. If G cells are encountered in a
sequence, Ν0 may be assigned to 0 and Ν1 assigned to 1,
so that Ν0 + Ν1 = G. The probability that a certain cell may
contain 0 is Ρ0=Ν0/G and the probability that it may
contain 1 is then Ρ1=Ν1/G.
The number of possible ways to arrange these G
consecutive cells is P = G ! / (N0 ! N1!) (3)
Out of these numerous “messages” that may be formed
by arranging the letters of the “alphabet” along these G
cells, entropy is related to the uncertainty the observer
senses out of a series of notes, describing a melodic line.
In music, it is obvious that notational symbols are not
arbitrarily put together, but rather they are ingredients and
constituents with artistic arrangement, i.e. they are
successive applications of “functions” that contribute to
the formation of harmonized sequences.
Conditional entropy Hcond is a measure of how
“bounded” is the melodic message, that is what is the
probability of a musical symbol aj to appear, if a sequence
of k symbols has preceded, arranged within the G melodic
cells. This set {pi} of k symbols {ai1, ai2, ... , aik}
imposes some kind of restriction on what the k+1 symbol
of the sequence may be.
If p(aj/ai1, ai2, ... , aik) is the probability that symbol aj
may appear if the tuple of k symbols {ai1, ai2, ... , aik} has
prevailed, then
𝐻 =
−∑𝑝(
 𝑎, 𝑎 , . . . , 𝑎 , 𝑎) . 𝑙𝑜𝑔𝑝(𝑎/
𝑎, 𝑎 , . . . , 𝑎 )(4)
in bits/symbol [14].
During the transmission of these symbols, the melodic
source may be found in different states. Compound
entropy is a measure of the ability of a melody to shift into
various r sequences of symbols, i.e. to form various
characteristic compositions, different one from each other.
𝐻 = − ∑𝑝(𝑘)𝑙𝑜𝑔𝑝(𝑘)

 =
− ∑𝑝(𝑎

 , . . . , 𝑎 )𝑙𝑜𝑔𝑝( 𝑎 , . . . , 𝑎 )
(5)
in bits/k symbols [14]
where 𝑝(𝑘)
is the appearance-probability of the ki
state.
This approach yielded notable analyses till the turn of
the millennium, where music was more circulating in
written form than as an audiovisual event. Indeed, in our
times music is disseminated without any delay or
difficulty in multiple forms (lyrics, scores, videos, audio
files, karaoke, MIDI, audio tracks, etc.). Therefore,
analyses where then more “alphabet” oriented.
To explain this situation, one may use the parallelism of
the Slavic languages. It is evident that this linguistic
family is distinct from the others surrounding it
geographically, and that there are clearly many common
elements between its members when the verbal
communication is encountered. However, when written
communication is taken into account, the family of West-
Slavic languages (Polish, Czech, Croat, Slovak, Slovene
and Sorbian), which is using variants of the Latin
alphabet, is quite different from the East-South subgroup
(Russian, Ukrainian, Byelorussian, Bulgarian,
Macedonian) in its semeiotics. Many times no one-to-one
correspondence may be found, not only to morpheme
level, but sometimes even between characters representing
sounds used in speech. Serbian, even though not any
more directly linked to Serbo-Croat, is a special case of
one language, more or less with its variants and dialects in
neighboring countries, written in both the Cyrillic and
Latin alphabets.
Consequently, there is an obvious shortage of symbols
representing sounds between members of this linguistic
family group, and some kind of symbolic augmentation is
employed, leading to disambiguation for the exact
pronunciation. This happens to many other languages as
well; however, it is characteristically obvious in the
enclave of Slavic languages acoustics.

Figure 4. Musicians performing the ode “Isaiah Dance” in various
modes. Left, Group I performers, right, a Group II one.
Figure 3. The 9
th
ode of the Resurrection Canon for Mode 5 in Byzantine
Music Notation and CMN transcription.
IV.
E
XPERIMENTAL
R
ESULTS
As hitherto explained, there is a considerable multitude
of scales used in modal music in the Balkans and Eastern
Mediterranean, apart from the usual major and minor
ones. Not all musicians are trained to reproduce this
versatile music literature adequately; as a mater of fact,
few can shift from one musical paradigm to another.
For testing the acoustic properties and qualities of the
singing voice, as information received out of semantics, a
liturgical tune performed widely in the Balkans and
Eastern Mediterranean was used: the 9
th
ode of Mode 5,
“Isaiah Dance”, in “Heirmologic” style. In its
contemporary written form it has been excerpted from
memorable editions of the “Heirmologion” in 1825 and
1901, and it is attributed to Petros the Peloponnesian
(1730-1778). Scholars reasonably induce that it has been
performed more or less the same melodic-semiotic way in
sermons like matins, weddings and ordinations since the
early 17
th
century [15].
The melody was critically edited by Prof. N. Paris, and
was also transcribed to CMN semeiotics [16]. Experts in
music were used to perform it.
In terms of Information theory, the performers decipher
a series of G notes to describe the combining vocal or
instrumental sounds that produce the melodic form. Not
all (musical) “alphabets” have one-to-one correspondence
in reflecting particular quantities, like pitches, or qualities,
like expressions of stress. As it happens with spoken
languages, there is a multilevel set of forms, especially in
vowels, where in some languages certain combinatory
sounds, phenomenally alike, may be perceived as
monophtongs, diphthongs or even triphthongs. Also, as
for writing purposes languages may employ syllabic or
logographic signs, in the same sense music semeiotics
may be pitch oriented, as is the case with CMN, or to form
a Delta system, as is Byzantine Music [11].
Many times the score itself, leaves room to the singer
for more added value in his performance by patterns of
intonation, prosody and “qualitative” patterns of stress,
that have no exact equivalents in CMN and Western
European “diatonic” mentality.
This research involves three levels of expertise
involved.
A. The “scholars”
They are the ones that may compose melodies. They
attempt to express as better as they can, in writing, the
melismatic nature of an artistic arrangement, like a song.
Their accomplishments are not particularly probed in this
survey - rather they are taken axiomatically. (Figure 3.)
B. The “performers”
They are skillful in interpreting, vocally or
instrumentally, a piece of music. They are experts in
“translating” the semeiotics of melodies to accomplished
performances. Their skills are easily recognizable. They
are capable of quickly comprehending the score; in some
cases they were able to give a rendition of it as soon as
they had seen it in the computer screen (Figure 4).
They acted modally in three clusters.
The first cluster, aka Group I, performed the melody
following the Byzantine Music semeiotics, in their
original form, for the vocal part. For the instrumental
rendition, the CMN score was adapted to the maqam style
closest to the vocal performance and was adequately
performed by an accordingly tuned tambour (Figure 4).
The second cluster, Group II, performed the melody
instrumentally, using a bass and CMN (Figure 4).
However, before playing the tune, they had listened many
times the “original” performance, as Group I recorded it.
In the places where they sensed a discord, they rather
played the music by ear than by the exact CMN dictation
so to pertain the pervading mood. When using a
keyboard, they exerted pitch bendings to “correct” the
discorded tones.
The third cluster, aka Group III, to avoid this obvious
dissonance of the written forms, as far as this specific ode
is concerned, employed computer technology to reproduce
instrumentally the tune. Instruments like harp or piano
where used. Then, the singers hearing with headphones
the CMN tune, they performed accordingly the vocal part,
following the notation rather than the original hearing, as
recorded by Group I.
Obviously, the renditions of Groups I and III are the
ones semeiotically correct. For this purpose, F0 was
calculated, with an autocorrelation method, using
MATLAB. The exact comparison of the fundamental
frequency curves between them reveals the difference in
styles of execution (Figure 5).
Indeed, the melodic curves reveal many characteristics
of voicing that are inherent to the way of each musical
system used. Of course, some phenomena are not related
with F0 per se; for instance, the “chirp” like deviation
seen between 450 and 550 ms in Group III's performance,
is due to the emphatic way that the singer performed the
otherwise unvoiced alveolar fricative /s/.
It also becomes evident, that Group I's singing is
producing continuous sounds, and the alteration from one
vowel to the other is performed in a combinatory way that
keeps track of the “qualitative”, vocalization expressive

Figure 6. The melodic ending of the first word to phoneme /a/. Upper and lower
graphs: the time series of Group I and III utterances. Middle graph, F0 estimated
with an autocorrelation method for Group I (yellow curve) and Group III (green
curve).
Figure 5. The first word of the ode, as performed by Group I
(upper part) and by Group III (lower part). The time series and the
calculated F0 curve of each recording are presented.
nature of Byzantine Music semeiotics [11]. On the
contrary, when CMN style performance is given
eminence, vowels rather stick to the pitch levels of the
note executed than following the oscillating characteristics
of the singing voice; as a result there are no obvious
“passing” paths from one pitch to the other.
Apart from the partial discord in pitch levels, for B
(Figure 1b), it seems that notes are perceived in CMN
performance style as clearly sustainable voicing quantities
at a specific pitch level [12], while in Byzantine Music
they are expressive voicing curves anchored around some
basic tonal levels. Alterations are more detailed, and they
are more oriented to forming specific accentuated
intervals (Figure 1b, Figure 2), some times smaller than a
semitone, than hitting a specific note by any means.
In Figure 6, this specific difference in performing
mentality is demonstrated. The transient suffix /a/ of the
word /
I
sa
I
a/ is demonstrated; it is not the ending of a
melodic phrase, but merely a word ending.
Apart from the differentiating characteristics spotted
thus far, it is obvious the disparity in the sustainability of
the phonation, in ms, albeit the fact that both groups
pronounce /a/ quite energetically, characteristic to Greek
style articulation [17].
Furthermore, it is obvious that note B is somewhat
lowered, when performed the Byzantine Music style,
while for Group III performers it seems that it cannot
deviate from the CMN dedicated key signature, and form
a smaller interval A-B.
It is an obvious point of discord.
C. The “wide public”
The “performers”, being experts in the field, where able
to specifically detect the difference between the various
renditions. More or less they spotted the points
emphatically presented in Figures 5 and 6.
In order to detect if the “wide public” responds in a
similar manner to the rendition of the song, a survey was
conducted aiming to decipher the sensitivity and the
specificity of the final recipients of music, the so called
“wide public” [18].
It is apparent that audiophiles do not have the same
subtlety in musicological matters as performers do.
However, the degree of sensitivity demonstrated by
listeners is critical for the way that music is accomplished
as an artistic product.
For this purpose, a group of some 40 audiophile
Computer Science students, aged 21 to 25, was used to
diagnose the reaction of the general public to the
differentiating versions of a song. From this audience, 14
members were female and 26 were male.
They were quite proficient, as they were actively
engaged in computer music production and distribution
practices. Nevertheless, not all of them had profound
musical skills, and therefore they were quite a
representative pool for the reactions of the section of the
contemporary community that is actively engaged in
listening to music.

Figure 7. The evaluators' perception of the differences between the melody
interpreted with Byzantine Music scales and CMN scales.
The evaluation of the performances delivered by
Groups I, II and III is analytically presented in Table I.
The evaluators were not given a set of predefined
phrases to choose from; rather they were guided to express
themselves freely in what they sensed as “different” or
“alienating” in the renditions involved.
However, if they spotted a deviation, they were hinted
to describe it in an exact manner.
A considerable percentage could not even trace a
significant difference. The majority, nevertheless, could
sense, at least, that there was some kind of a different
tuning involved. Few, furthermore, could even exactly
spot it. These qualitative evaluation results may be
graphically observed in Figure 7.
V. CONCLUSION
Musical “alphabets” attempt to decipher a continuum,
the musical one, whose adjacent elements are in many
occasions perceptibly different one from each other, in
their attempt to describe phenomena of the singing voice
like stress - intonation, prosody, and sequences of melodic
pitches. Since these features are also language dependent,
it becomes obvious that semeiotics are reprimanded
amidst their mission to accurately describe, in scientific
terms, with sets of limited representative characters, the
complex perceptual characteristics of musical sounds.
The musical continuum of the Internet connected
societies and economies, within which hearings from a
wide gamut, in terms of synchrony and diachrony
abundantly circulate to its most remote nodes, may
provide the incentive for augmentation and merging of a
variety of signs and symbols, along with their use or
interpretation.
Genuses, Modes and Genres may not be sufficiently
decoded, if not previously recorded and, the most
important, encoded so to include most, if not all, of their
relevant characteristics, qualities and events associated
with them.
ACKNOWLEDGMENT
The authors express their gratitude to their students Y.B
(University of Macedonia), E.M., S.P., O.E., and S.G.
(Aristotle University of Thessaloniki), all skillful
musicians and music producers, for arranging the
production, along with the audiovisual staging of their
singing and instrumental rendition.
REFERENCES
[1] T. Patokos, “A new era for the music industry: how new
technologies and the Internet affect the way music is valued and
have an impact on output quality,” Panoeconomicus, vol. 55, no.
2, 2008, pp. 233-48.
[2] R. Burnett, “Music and the Internet, ” in The Handbook of Internet
Studies, Willey-Blackwell, 2011, pp. 440-451.
[3] Pennanen Risto Pekka, “Lost in scales: Balkan Folk Music
Research and the Ottoman Legacy”, Mузикологија, no. 8, 2008,
pp. 127–147.
[4] M. Mavroedis, The Musical Modes in Eastern Mediteranean Sea –
The Byzantine Echos – The Arabic Maqam – The Turkish Maqam
(in Greek), Athens: Fagotto Editions, 1999.
[5] S. Hagel, “Reversing the Abstraction of Ancient Music Theory.
The Case of the Genera,” Studien zur Musik-archäologie Orient –
Archäologie, no. 22, pp. 461–475, 2008
[6] R. Hatten, R., “Musical Meaning in Beethoven: Markedness,
Correlation, and Interpretation,” Advances in Semiotics, Indiana
University Press, 2004.
[7] V. Barsky, Chromaticism, Netherlands: Harwood Academic
Publishers, 1996.
[8] V. Stojkova-Serafimovska, “The enlarged second in the vocal
musical tradition: Byzantine or Ottoman influence (The
Macedonian Case),” in Proceedings of the International
Symposium Urban Music in the Balkans, September 28 – October
01, 2006, ASMUS, Tirana 2006, pp. 120–124.
[9] C. Cox, and D. Warner, D., Audio Cultures, Readings in Modern
Music, New York: Continuum, 2007.
[10] D. Giannelos, La Musique Byzantine, Paris: L’Harmattan, 1996,
pp. 63-75.
[11] N. Paris, and D. Politis, “Beyond MIDI: Theoretical Foundations
for the Voice-Instrument Digital Interface for Byzantine Music,”
in Proceedings of the 4th Conference on Interdisciplinary
Musicology CIM08 - MUSICAL STRUCTURE, Department of
Music Studies, Aristotle University of Thessaloniki, 3-6 July
2008.
[12] E. Field-Selfridge, Beyond MIDI. Cambridge, MA: MIT Press,
1997.
[13] E. Wellesz, A History of Byzantine Music and Hymnography, 2nd
ed. revised and enlarged, Oxford: Clarendon Press, 1962.
[14] H. C. Spyridis, and D. V. Politis, Information Theory Applied to
the Structural Study of Byzantine Ecclesiastical Hymns,
ACUSTICA, vol. 71, no. 1, pp. 40-49, May 1990.
[15] M. Alexandru, Paleography of Byzantine Music, Athens: Kalippos
Academic Electronic Books, 2017.
http://hdl.handle.net/11419/6487
[16] A. J. Swan, “New Developments in the Transcription of Byzantine
Melodies,” The Musical Quarterly, vol. 29, no. 2, pp. 254-256,
April 1943.
[17] D. Politis, G. Kyriafinis, J. Constantinidis, and N. Paris,
“Neurophysiology-based acoustic measurements of singing
contours - an empathetic interactive loudness approach,” in the
Proceedings of the International Conference on Interactive Mobile
Communication Technologies and Learning IMCL 2017,
Thessaloniki 30 November - 1 December 2017, pp. 803-812.
[18] J. Sundberg, “The Perception of Singing,” in D. Deutch (Ed.) The
Psychology of Music, 3rd Edition, London: Academic Press, 2013.
TABLE I.
EVALUATION OF SONG DIFFERENCES FROM THE “WIDE PUBLIC”
No Perception of Differentiation Quantity
1 Cannot comprehend the issue / No answer 10
2 No notable difference detected 1
3 Overall difference detected 18
4 Considerable difference detected 3
5 Can identify which quantities exactly differ 5
6 Mainly perceived differences in timbre 2
7 Mainly perceived rhythmic variability 1