Tiago Videira - Instrumental Fado (Complete With Appendixes) PDF

Instrumental Fado:
A Generative Interactive System
Tiago Miguel Gonzaga Videira
Tese de Doutoramento em Media Digitais
Junho, 2015
UT-Austin-Portugal Program
Doctoral Thesis
Instrumental Fado - A Generative

Interactive System
Author: Supervisor:
Tiago Videira Dr. Jorge Martins Rosa
A thesis submitted in fulfilment of the requirements

for the degree of Doctor of Philosophy
in Digital Media
in the
Faculdade de Ciências Sociais e Humanas

Universidade Nova de Lisboa
This work was funded by FCT – Fundação para a Ciência e

Tecnologia in the framework of UT Austin-Portugal Digital Media
program, grant [SFRH / BD / 46136 / 2008].
June 2015
“O fado?
É uma maneira de lançar a voz”
António Victorino d’Almeida

UT-AUSTIN PORTUGAL PROGRAM
Abstract
Doctor of Philosophy
Instrumental Fado - A Generative Interactive System
by Tiago Videira
Fado was listed as UNESCO Intangible Cultural Heritage in 2011. This dis-
sertation describes a theoretical model, as well as an automatic system, able to
generate instrumental music based on the musics and vocal sounds typically asso-
ciated with fado’s practice. A description of the phenomenon of fado, its musics
and vocal sounds, based on ethnographic, historical sources and empirical data is
presented. The data includes the creation of a digital corpus, of musical transcrip-
tions, identified as fado, and statistical analysis via music information retrieval
techniques. The second part consists in the formulation of a theory and the cod-
ing of a symbolic model, as a proof of concept, for the automatic generation of
instrumental music based on the one in the corpus.
KEYWORDS: fado, ethnomusicology, music information retrieval, style imita-

tion, algorithmic composition, generative music
UT-AUSTIN PORTUGAL PROGRAM
Resumo
Doctor of Philosophy
Instrumental Fado - A Generative Interactive System
by Tiago Videira
O Fado é considerado, pela UNESCO, património imaterial da Humanidade

desde 2011. Esta tese descreve um modelo teórico, assim como um programa
informático, capaz de gerar música instrumental baseada nas músicas e sons vo-
cais tipicamente associados à prática de fado. Apresenta-se uma descrição do
fenómeno do fado, das suas músicas e sons vocais, baseada em fontes etnográficas
e historiográficas e em dados empı́ricos recolhidos. Nestes dados inclui-se a criação
de uma base digital contendo transcrições musicais identificadas como fado, bem
como a sua análise estatı́stica utilizando técnicas de musicologia computacional.
De seguida procede-se à formulação de uma teoria e programação de um modelo
simbólico como prova de conceito de um sistema automático de geração de música
instrumental baseada na música constante na base de dados prévia.
PALAVRAS-CHAVE: fado, etnomusicologia, musicologia computacional, com-

posição algorı́tmica, música generativa
Acknowledgements
I still remember the day I was welcomed into this project. Professor António
Câmara enthusiastically reminded us of how a PhD was a journey and it was all
about enjoying the ride and not so much about the results. Because he believed
that if we committed ourselves to this path with our heart and soul and gave the
very best of us, in the end we would become leading experts in our areas. After
all we would spend five years of our lives researching, reading dozens of books and
articles, networking and meeting new people, and hopefully having international
experience by traveling abroad, attending conferences and developing new skills.
I have to say that at this point this is exactly how I feel. First, I do not feel
this is the end. I believe my work is a permanent work in progress, and that I
will have many more intriguing tasks to perform and a myriad of new journeys to
make. But I also believe I am a new person. I have grown in ways I would never
have imagined, and I have developed a critical sense of the world I did not know
it was possible. By the simple fact of embarking in this journey and having spent
two years, mostly on my own, living in Austin, I have developed a whole set of
new skills and I have learned countless ways of solving the plethora of challenges
I had to face.
I have also met new people who taught me invaluable things, shared cultural and
creative experiences and put me to the test in many ways. I was under incredible
pressure in countless occasions and saw myself on the verge of the unknown many
times. Often, I had to figure out on my own how to come out of hard situations,
other times I was able to count with the aid of many wonderful people who helped
me along the way.
This dissertation is just the reflection of those five years. It is a vast text I
have written to present many of my findings: to share the survey of many books
and articles, to present many calculations, assertions, but also limitations and
doubts. Many times I thought I was going in the wrong direction and I had to
improvise. Other times I have felt I did not know enough about the subject I was
reading about and had to learn, to take a crash course online or to go and find
and talk with people who knew better. No one is alone, and although most of this
dissertation is based on a very lonely journey, the truth is that most of it is built
over the shoulders of many others who have worked before me. Therefore there
ix
are numerous quotations and credits throughout the dissertation – everything that
was already done I felt I could and should use. I was advised to do so, as it would
make no sense to double the work or to reinvent the wheel. I admire those persons
and give full credit to them for all their findings. At times I felt I was just solving
a puzzle, because all those credited authors had done a terrific work unveiling data
but had done so in different times and places. My task was simply to reorganize
it, patch it up together, and make sense of it all in a new and refreshing way.
Sometimes I have also felt very insecure, because there were so many possible
approaches to the same problem. And history could not help me – I was dealing
with a new area, but with several methodologies deriving from different schools
of thought. I have met remarkable scholars who gave me valuable insight and
were leading experts in their areas. However, they did not understand much of
other subjects that I also needed to cross. Therefore, many times I have felt I was
overlapping several different worlds and wondered if my peers would ever be able
to understand or make sense of this interdisciplinary approach.
At one point I also felt I would not make it. There was simply too much work
to do, too many new things to learn, to dissect, to present, to experiment with.
At one point, I realized I was so overwhelmed that I was using my dissertation
as a metaphor to understand human behavior as a whole. At that point I simply
had to back off and see the big picture – and understand how this was just a
journey, and that I had no individual responsibility to fully model and understand
the world or even human kind as a whole. So I had to refocus on my tiny problem.
So, my individual limited perception is that I am just presenting my original work
at this point in time not because it is completed or it is perfect, but because it
has been five years and my financial aid has long ended. I believe at this point my
work already demonstrates how I have learned to do research, to critically think
and, mainly, to solve very different problems applying high-level reasoning and
scientific methodologies. If my hard work is recognized by my peers, then I intend
to proceed my endless journey searching for the ultimate truth, if there is one, and
start a new chapter of my life.
At this point, then, I wish to acknowledge and thank with all my heart to all of
those who made this journey with me and, somehow, helped to make it possible.
First, and most important I thank my family. My mother, father, grandmother,

and my recently late great-aunt, who were the true pillars always there for me
with unconditional love. Secondly, my cousins, my godmothers and godfathers
who were always there when I needed to talk.
Also, I thank my friends. They know who they are, because I talk to them
regularly. The ones who were always ready to hear my calls when I was afar, or to
listen to my rants via Internet, or in person when possible; to give me emotional
aid and comfort, to go on walks with me, or simply to share incredible cultural
experiences, both in Portugal and in the USA.
I also thank my professors – a group of remarkable scholars who aided me in

several ways. Some of them, like professor Tomás Henriques, Filomena Molder,
Carlos Guedes and Alberto Pimenta were inspirational in the sense that they
molded my character in the previous years and were pillars and references in my
thought. Others, like António Tilly, João Soeiro de Carvalho, Salwa Castelo-
Branco, Maria de São José Côrte-Real, Nuno Correia, Robert Hatten and Russell
Pinkston were present in several occasions during my present work, sometimes
even being unaware of it, giving direct insight. It is not always about quantity
but quality. There were occasions when a simple ten minute conversation had the
power to open a Pandora’s box, unleashing many thoughts or references that were
the start for a whole new chapter or path of learning.
I have to thank my peers – too many to mention, some with minor interven-
tions, others with full-credited quotations in my text. Some of them close friends,
but these ones, having direct influence on my text, gave me valuable information
and suggestions: Elad Liebman, Peter Stone, Carlos Teixeira and Manuel Dias,
scholars from the scientific world. Fábio Serranito, philosopher and competent
philologist, with his infinite patience, is my regular proof-reader and arbiter ele-
gantiae, correcting many of the grammatical flaws my text sometimes presents.
And finally, with most direct influence, I thank my Supervisor – Jorge Martins
Rosa and my tutor who received me in Austin, Bruce Pennycook who were there,
many times at a distance of a simple email, with a wise advice and a critical look
on how to proceed and what to do. Again, sometimes is not about quantity but
about the quality, timing and pertinence of the advice.
Contents
Abstract v
Resumo vii
Acknowledgements ix
Contents xiii
List of Figures xix
List of Tables xxiii
I On the musics and sounds of fado 1

1 Introduction 3
2 Objectives 7
3 Methodologies 9
3.1 Ethnomusicology and Historiography . . . . . . . . . . . . . . . . . 9
3.1.1 An operative model to study the musics and sounds . . . . . 17
3.1.2 Consilience . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2 Ethnographic and Historiographic Sources . . . . . . . . . . . . . . 20
3.2.1 Critical Sources . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.2 Fieldwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.3 Recordings and Media . . . . . . . . . . . . . . . . . . . . . 26
3.3 Systematic Musicology and Symbolic sources . . . . . . . . . . . . . 26
3.3.1 How social class shapes a canon . . . . . . . . . . . . . . . . 28
3.3.2 The Database and the urtext Corpus . . . . . . . . . . . . . 31
3.3.3 Uniformity Operations and the Edition Corpus . . . . . . . 33
3.3.4 The Melodies-only Corpus . . . . . . . . . . . . . . . . . . . 35
3.4 Music Information Retrieval . . . . . . . . . . . . . . . . . . . . . . 36
3.4.1 Testing the internal consistency of a corpus . . . . . . . . . 40
xiii
Contents
3.4.2 Test Implementation . . . . . . . . . . . . . . . . . . . . . . 43

3.4.3 A method for comparative data visualization . . . . . . . . . 47
4 On the concept of fado 51

4.1 Modinhas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.2 The influence of the Dictatorship . . . . . . . . . . . . . . . . . . . 54
4.3 Fado as Space and Identity . . . . . . . . . . . . . . . . . . . . . . . 57
5 Analysis 63
5.1 The Narrative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.2 Prosody . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.2.1 Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.2.2 The rhythm of the melody . . . . . . . . . . . . . . . . . . . 69
5.3 Structure of the text . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.4 The vocal group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.5 The instrumental group . . . . . . . . . . . . . . . . . . . . . . . . 76
5.5.1 The structure of the accompaniment . . . . . . . . . . . . . 78
5.6 Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.6.1 Internal complexity . . . . . . . . . . . . . . . . . . . . . . . 83
5.7 Tempo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.8 Rhythmic Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.8.1 Note density . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.8.2 Strongest Rhythmic Pulse . . . . . . . . . . . . . . . . . . . 93
5.8.3 Second Strongest Rhythmic Pulse . . . . . . . . . . . . . . . 94
5.8.4 Harmonicity of Two Strongest Rhythmic Pulses . . . . . . . 95
5.8.5 Strength of Strongest Rhythmic Pulse . . . . . . . . . . . . 96
5.8.6 Strength of Second Strongest Rhythmic Pulse . . . . . . . . 97
5.8.7 Strength Ratio of Two Strongest Rhythmic Pulses . . . . . . 98
5.8.8 Combined Strength
of Two Strongest Rhythmic Pulses . . . . . . . . . . . . . . 99
5.8.9 Number of Strong Pulses . . . . . . . . . . . . . . . . . . . . 100
5.8.10 Number of Moderate Pulses . . . . . . . . . . . . . . . . . . 101
5.8.11 Number of Relatively Strong Pulses . . . . . . . . . . . . . . 102
5.8.12 Rhythmic Looseness . . . . . . . . . . . . . . . . . . . . . . 103
5.8.13 Polyrhythms . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.8.14 Rhythmic Variability . . . . . . . . . . . . . . . . . . . . . . 105
5.9 Note Duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
5.9.1 Variability of Note Duration . . . . . . . . . . . . . . . . . . 107
5.9.2 Maximum Note Duration . . . . . . . . . . . . . . . . . . . . 109
5.9.3 Minimum Note Duration . . . . . . . . . . . . . . . . . . . . 110
5.9.4 Staccato Incidence . . . . . . . . . . . . . . . . . . . . . . . 110
5.9.5 Average Time Between Attacks . . . . . . . . . . . . . . . . 112
5.9.6 Variability of Time Between Attacks . . . . . . . . . . . . . 113
5.9.7 Average Time Between Attacks For Each Voice . . . . . . . 114
xiv
Contents
5.9.8 Average Variability of Time Between Attacks For Each Voice 115
5.10 Harmonic Progressions . . . . . . . . . . . . . . . . . . . . . . . . . 116
5.11 Tonality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.11.1 Strong Tonal Centres . . . . . . . . . . . . . . . . . . . . . . 126
5.11.2 Dominant Spread . . . . . . . . . . . . . . . . . . . . . . . . 127
5.12 Modulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.13 Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5.13.1 Most Common Pitch . . . . . . . . . . . . . . . . . . . . . . 132
5.13.2 Primary Register . . . . . . . . . . . . . . . . . . . . . . . . 134
5.13.3 Importance of Bass Register . . . . . . . . . . . . . . . . . . 135
5.13.4 Importance of Middle Register . . . . . . . . . . . . . . . . . 136
5.13.5 Importance of High Register . . . . . . . . . . . . . . . . . . 137
5.14 Pitch Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
5.14.1 Most Common Pitch Prevalence . . . . . . . . . . . . . . . . 139
5.14.2 Most Common Pitch Class Prevalence . . . . . . . . . . . . 141
5.14.3 Relative Strength of Top Pitches . . . . . . . . . . . . . . . 142
5.14.4 Relative Strength of Top Pitch Classes . . . . . . . . . . . . 143
5.14.5 Interval Between Strongest Pitches . . . . . . . . . . . . . . 144
5.14.6 Interval Between Strongest Pitch Classes . . . . . . . . . . . 145
5.14.7 Number of Common Pitches . . . . . . . . . . . . . . . . . . 146
5.14.8 Pitch Variety . . . . . . . . . . . . . . . . . . . . . . . . . . 147
5.14.9 Pitch Class Variety . . . . . . . . . . . . . . . . . . . . . . . 148
5.14.10 Most Common Pitch Class . . . . . . . . . . . . . . . . . . . 149
5.15 Melody . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
5.15.1 Pitch-Class distribution . . . . . . . . . . . . . . . . . . . . 154
5.16 Melodic Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
5.17 Comparative features based on Melody . . . . . . . . . . . . . . . . 156
5.17.1 Average Melodic Interval . . . . . . . . . . . . . . . . . . . . 156
5.17.2 Most Common Melodic Interval . . . . . . . . . . . . . . . . 157
5.17.3 Distance Between Most Common Melodic Intervals . . . . . 158
5.17.4 Most Common Melodic Interval Prevalence . . . . . . . . . . 158
5.17.5 Relative Strength of Most Common Intervals . . . . . . . . . 160
5.17.6 Number of Common Melodic Intervals . . . . . . . . . . . . 161
5.17.7 Amount of Arpeggiation . . . . . . . . . . . . . . . . . . . . 162
5.17.8 Repeated Notes . . . . . . . . . . . . . . . . . . . . . . . . . 163
5.17.9 Chromatic Motion . . . . . . . . . . . . . . . . . . . . . . . 164
5.17.10 Stepwise Motion . . . . . . . . . . . . . . . . . . . . . . . . 164
5.17.11 Melodic Thirds . . . . . . . . . . . . . . . . . . . . . . . . . 165
5.17.12 Melodic Fifths . . . . . . . . . . . . . . . . . . . . . . . . . . 166
5.17.13 Melodic Tritones . . . . . . . . . . . . . . . . . . . . . . . . 166
5.17.14 Melodic Octaves . . . . . . . . . . . . . . . . . . . . . . . . 168
5.18 Melodic Contour . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.18.1 Direction of Motion . . . . . . . . . . . . . . . . . . . . . . . 170
5.18.2 Duration of Melodic Arcs . . . . . . . . . . . . . . . . . . . 171
xv
Contents
5.18.3 Size of Melodic Arcs . . . . . . . . . . . . . . . . . . . . . . 171

5.19 Position of the final tone . . . . . . . . . . . . . . . . . . . . . . . . 173
5.20 Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.21 Comparative features based on Dynamics . . . . . . . . . . . . . . . 175
5.21.1 Overall Dynamic Range . . . . . . . . . . . . . . . . . . . . 176
5.21.2 Variation of Dynamics . . . . . . . . . . . . . . . . . . . . . 177
5.21.3 Variation of Dynamics In Each Voice . . . . . . . . . . . . . 177
5.21.4 Average Note To Note Dynamics Change . . . . . . . . . . . 178
5.22 Pitch Bend Messages . . . . . . . . . . . . . . . . . . . . . . . . . . 179
5.22.1 Glissando Prevalence . . . . . . . . . . . . . . . . . . . . . . 180
5.22.2 Average Range of Glissandos . . . . . . . . . . . . . . . . . . 180
5.22.3 Vibrato Prevalence . . . . . . . . . . . . . . . . . . . . . . . 182
5.23 Embellishment, ornamentation and vocal effects . . . . . . . . . . . 183
5.24 Accent and enunciation of consonants . . . . . . . . . . . . . . . . . 189
6 Embodiment 191
6.1 Fado as dance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
6.1.1 The hypothetical links with flamenco . . . . . . . . . . . . . 197
6.2 Fado feeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
7 Fado as Memory 207

7.1 Desfado . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
8 The Creative Processes 211
9 Conclusions 215
10 Future Work 221
II Automating the Composition Process 223

11 Introduction 225
12 State of the art 229

12.1 A survey on algorithmic composition . . . . . . . . . . . . . . . . . 232
12.1.1 Symbolic artificial intelligence . . . . . . . . . . . . . . . . . 234
12.1.2 Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . 241
12.1.3 Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . 243
12.1.4 Complex Systems . . . . . . . . . . . . . . . . . . . . . . . . 244
12.2 Advantages and Disadvantages . . . . . . . . . . . . . . . . . . . . . 245
12.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
13 The model 249
14 Methodology 253
xvi
Contents
14.1 Interactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

14.2 Work Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
15 Implementation of the Model 261

15.1 First Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
15.1.1 Seed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
15.1.2 Text . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
15.1.3 Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
15.1.4 Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
15.1.5 Example of a single section . . . . . . . . . . . . . . . . . . . 268
15.2 Second Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
15.2.1 Harmonic Progressions . . . . . . . . . . . . . . . . . . . . . 280
15.2.2 Ostinati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
15.3 Third Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
15.4 Fourth Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
15.5 Fifth Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
15.5.1 Rhythmic variations . . . . . . . . . . . . . . . . . . . . . . 299
15.5.2 Melodic variations . . . . . . . . . . . . . . . . . . . . . . . 302
15.5.3 Mixed Variations . . . . . . . . . . . . . . . . . . . . . . . . 303
15.6 Global Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
15.6.1 Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
15.6.2 Tempo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
15.6.3 Groove . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
16 Templates for rendering the MIDI files 311
17 Conclusions 313
17.1 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316
18 Future Work 323

18.1 Regarding the model . . . . . . . . . . . . . . . . . . . . . . . . . . 323
18.2 Regarding Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
A Supplemental Material 343
xvii
List of Figures
3.1 Complementing Fieldwork . . . . . . . . . . . . . . . . . . . . . . . 15

3.2 Operative Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.1 Possible binary rhythms for a textual line . . . . . . . . . . . . . . . 71

5.2 Semiographic schemes . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.3 Initial Tempo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.4 Ostinati . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.5 Guitarra countermelodies . . . . . . . . . . . . . . . . . . . . . . . . 91
5.6 Note Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.7 Strongest Rhythmic Pulse . . . . . . . . . . . . . . . . . . . . . . . 94
5.8 Second Strongest Rhythmic Pulse . . . . . . . . . . . . . . . . . . . 95
5.9 Harmonicity of Two Strongest Rhythmic Pulses . . . . . . . . . . . 96
5.10 Strength of Strongest Rhythmic Pulse . . . . . . . . . . . . . . . . . 97
5.11 Strength of Second Strongest Rhythmic Pulse . . . . . . . . . . . . 98
5.12 Strength Ratio of Two Strongest Rhythmic Pulses . . . . . . . . . . 99
5.13 Combined Strength of Two Strongest Rhythmic Pulses . . . . . . . 100
5.14 Number of Strong Pulses . . . . . . . . . . . . . . . . . . . . . . . . 101
5.15 Number of Moderate Pulses . . . . . . . . . . . . . . . . . . . . . . 102
5.16 Number of Relatively Strong Pulses . . . . . . . . . . . . . . . . . . 103
5.17 Rhythmic Looseness . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.18 Polyrhythms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.19 Rhythmic Variability . . . . . . . . . . . . . . . . . . . . . . . . . . 106
5.20 Average Note Duration . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.21 Variability of Note Duration . . . . . . . . . . . . . . . . . . . . . . 108
5.22 Maximum Note Duration . . . . . . . . . . . . . . . . . . . . . . . . 110
5.23 Minimum Note Duration . . . . . . . . . . . . . . . . . . . . . . . . 111
5.24 Staccato Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
5.25 Average Time Between Attacks . . . . . . . . . . . . . . . . . . . . 112
5.26 Variability of Time Between Attacks . . . . . . . . . . . . . . . . . 113
5.27 Average Time Between Attacks For Each Voice . . . . . . . . . . . 114
5.28 Average Variability of Time Between Attacks For Each Voice . . . . 115
5.29 Distribution of Harmonic Progressions . . . . . . . . . . . . . . . . 123
5.30 Distribution of tonalities . . . . . . . . . . . . . . . . . . . . . . . . 125
5.31 Strong Tonal Centres . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.32 Dominant Spread . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
xix
List of Figures
5.33 Initial tonality versus first relevant modulation . . . . . . . . . . . . 129

5.34 Initial tonality versus Final tonality . . . . . . . . . . . . . . . . . . 130
5.35 Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
5.36 Most Common Pitch . . . . . . . . . . . . . . . . . . . . . . . . . . 133
5.37 Primary Register . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
5.38 Importance of Bass Register . . . . . . . . . . . . . . . . . . . . . . 135
5.39 Importance of Middle Register . . . . . . . . . . . . . . . . . . . . . 136
5.40 Importance of High Register . . . . . . . . . . . . . . . . . . . . . . 137
5.41 Interval between strongest pitches . . . . . . . . . . . . . . . . . . . 140
5.42 Interval between strongest pitch classes . . . . . . . . . . . . . . . . 140
5.43 Most Common Pitch Prevalence . . . . . . . . . . . . . . . . . . . . 141
5.44 Most Common Pitch Class Prevalence . . . . . . . . . . . . . . . . 142
5.45 Relative Strength of Top Pitches . . . . . . . . . . . . . . . . . . . 143
5.46 Relative Strength of Top Pitch Classes . . . . . . . . . . . . . . . . 144
5.47 Interval Between Strongest Pitches . . . . . . . . . . . . . . . . . . 145
5.48 Interval Between Strongest Pitch Classes . . . . . . . . . . . . . . . 146
5.49 Number of Common Pitches . . . . . . . . . . . . . . . . . . . . . . 147
5.50 Pitch Variety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5.51 Pitch Class Variety . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
5.52 A fado archetype . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
5.53 Pitch-Class distribution in melodies . . . . . . . . . . . . . . . . . . 154
5.54 Average Melodic Interval . . . . . . . . . . . . . . . . . . . . . . . . 157
5.55 Most Common Melodic Interval . . . . . . . . . . . . . . . . . . . . 158
5.56 Distance Between Most Common Melodic Intervals . . . . . . . . . 159
5.57 Most Common Melodic Interval Prevalence . . . . . . . . . . . . . . 159
5.58 Relative Strength of Most Common Intervals . . . . . . . . . . . . . 160
5.59 Number of Common Melodic Intervals . . . . . . . . . . . . . . . . 161
5.60 Amount of Arpeggiation . . . . . . . . . . . . . . . . . . . . . . . . 162
5.61 Repeated Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
5.62 Chromatic Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.63 Stepwise Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
5.64 Melodic Thirds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
5.65 Melodic Fifths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
5.66 Melodic Tritones . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
5.67 Melodic Octaves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
5.68 Direction of Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.69 Duration of Melodic Arcs . . . . . . . . . . . . . . . . . . . . . . . . 172
5.70 Size of Melodic Arcs . . . . . . . . . . . . . . . . . . . . . . . . . . 172
5.71 Overall Dynamic Range . . . . . . . . . . . . . . . . . . . . . . . . 176
5.72 Variation of Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.73 Variation of Dynamics In Each Voice . . . . . . . . . . . . . . . . . 178
5.74 Average Note To Note Dynamics Change . . . . . . . . . . . . . . . 179
5.75 Glissando Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . 181
5.76 Average Range of Glissandos . . . . . . . . . . . . . . . . . . . . . . 181
xx
List of Figures
5.77 Vibrato Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
6.1 Fado feeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
9.1 Alternative notation to better convey “fado feeling” . . . . . . . . . 219
12.1 Fernández and Vico classification system . . . . . . . . . . . . . . . 233
13.1 A fado model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
17.1 Model Evaluation with human subjects . . . . . . . . . . . . . . . . 319

17.2 Overall Human versus Computer composed samples . . . . . . . . . 320
xxi
List of Tables
3.1 Classification with 10 taxonomies . . . . . . . . . . . . . . . . . . . 41

3.2 Classification with 39 taxonomies . . . . . . . . . . . . . . . . . . . 42
5.1 Distribution of Sections . . . . . . . . . . . . . . . . . . . . . . . . . 83

5.2 Statistics of Tempo . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
5.3 Main accompaniment figure present in each transcription. . . . . . . 89
5.4 Statistics of Note Density . . . . . . . . . . . . . . . . . . . . . . . 92
5.5 Statistics of Note Duration . . . . . . . . . . . . . . . . . . . . . . . 107
5.6 Harmonic Progressions of section A. . . . . . . . . . . . . . . . . . . 119
5.7 Harmonic Progressions of section B. . . . . . . . . . . . . . . . . . . 120
5.8 Harmonic Progressions of section C. . . . . . . . . . . . . . . . . . . 121
5.9 Harmonic Progressions of section D. . . . . . . . . . . . . . . . . . 122
5.10 Statistics of Tonal Centres . . . . . . . . . . . . . . . . . . . . . . . 126
5.11 Types of Modulations . . . . . . . . . . . . . . . . . . . . . . . . . . 128
5.12 Statistics of Range . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
5.13 Statistics of Pitch Prevalence . . . . . . . . . . . . . . . . . . . . . 139
5.14 Most Common Pitch Class . . . . . . . . . . . . . . . . . . . . . . . 150
5.15 Distribution of the most common pitch class . . . . . . . . . . . . . 150
5.16 Statistics of Pitch Variety . . . . . . . . . . . . . . . . . . . . . . . 153
5.17 Statistics of Melodic Intervals . . . . . . . . . . . . . . . . . . . . . 155
5.18 Statistics of Melodic Motion . . . . . . . . . . . . . . . . . . . . . . 156
5.19 Prevalence of Melodic Contours . . . . . . . . . . . . . . . . . . . . 169
5.20 Statistics of Melodic arcs . . . . . . . . . . . . . . . . . . . . . . . . 170
5.21 Statistics of Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . 175
15.1 Most common forms in the corpus . . . . . . . . . . . . . . . . . . . 266
xxiii
Part I
On the musics and sounds of fado
1
Chapter 1
Introduction
One utopia, and a strong force driving this project, is the understanding of the
creative process human beings go through when making songs. How and why does
it happen? Is it a craft? Is it something reproducible? Is it something that can
be learned and taught? It is something similar to a cook inventing a new dish and
then writing down a recipe? Would it be possible to systematize the process of
making a song in such a way? To reduce it to an algorithmic task, possible to be
coded to be performed by a computer?
These questions come a long way, since graduation, when studying musicology
and analyzing music in symbolic terms. They continued during Masters when the
work of other scientists, concerned with the same problems, and using computers
to emulate musical styles were being studied. And naturally they still remain a
vastly complex, open problem that is being addressed, one step at a time.
There are many folk, sung story-telling traditions like ballads, romances, blues,
mornas, country or even rap, all of them, that could make excellent case studies, as
a starting point for our bigger task. While they share many common traces among
themselves and with other past and present traditions, there is one in particular
which is a very peculiar and unique way to tell a story with very distinctive
features: Portuguese fado.
Fado is considered Intangible Cultural Heritage by UNESCO, since 2011, so we

have had a strong motivation to pursue this research within our interests (song
making) and at the same time providing a helpful contribution to the preservation
and memory of this performance practice.
3
Part I. On the musics and sounds of fado
In the first part of this dissertation, we are going to show how the concept of
fado historically emerged, what it represents, and how it has changed. Moreover,
we will offer a detailed, holistic and systematic characterization of the musics
and sounds associated with it (both instrumental and vocal), providing a series
of factual and empirical descriptors. We will discuss some values associated with
fado through a systematic analysis of the performance practice itself, as portrayed
by its performers, audiences and scholars, along the course of history, and offer a
detailed characterization of its traits. A combined interdisciplinary methodology
following the lines of ethnomusicology (ethnographic and historiographic methods)
and psychology of music has been used.
We have complemented the analysis with the use of computational musicology

on empirical data. We have conceived a musical corpus with 100 transcriptions,
identified as fado, found in the written sources. These transcriptions are piano
reductions, adapted for the domestic market, of both the instrumental accom-
paniment and vocal line sung, the vocal line being reduced to an instrumental
version. This corpus was edited and made available as a digital database1 . This
database consists of the musical scores, MIDI files, analytical, formal and philolog-
ical commentaries, as well as slots for relevant information (sources, designations,
date(s), authorship(s)) for each fado. The creation of this new digital object is
relevant for archival and patrimonial purposes. We have applied music informa-
tion retrieval techniques, followed by statistical procedures, on the corpus, in order
to identify some patterns and rules shaping its characteristics. Our results and
conclusions allow a better understanding of what fado is and open doors regarding
the construction of a theory, parametrization and modeling of the music and vocal
sounds associated with it, for pedagogical, patrimonial and composition purposes,
namely automatic generation of similar music.
The second part of this dissertation deals mainly with musical informatics.
After providing a state of the art, we conceive and describe a model based on the
previous section: a digital system, capable of generating new instrumental music
(based both on the instrumental and vocal line usually present in fado practice, the
vocal line being reduced to an instrumental version), following the processes and
rules previously found. We also discuss and present ideas for automatic evaluation
of the system and future work to be done.
1
http://fado.fcsh.unl.pt
4
This project possesses not only ethnographic, but also pedagogical relevance:
it contributes to a better understanding of what fado is, or can be according to
certain contexts, and the knowledge about how to compose and perform music
typically associated with it. Furthermore, it led to the creation of code and a
proof of concept program that may result in an application capable of generating
and playing royalty-free music, which can be used in relevant contexts. It could be
commercialized and mass distributed in the entertainment industry, as a learning
tool in an academic environment, or even be used in social and institutional con-
texts that usually lack the conventional appropriate settings to aesthetic fruition
– namely as background music.
5
Chapter 2
Objectives
In this part we will show how fado is a multidimensional concept, represent-

ing many different stances according to time, space and context and its constant
mutation and evolution through time. Moreover, we will offer a detailed, holistic
and systematic characterization of the musics and sounds associated with it (both
instrumental and vocal), using a series of factual and empirical descriptors that
will aid in reaching a definition, which will be the base for subsequent modeling.
There are some very good historiographic and sociologic descriptions, and even
several sources with musical transcriptions and phonograms. Those sources have
been surveyed and further interpreted and integrated in a new presentation, using
new angles and perspectives. We are reinforcing their analytical depth, and com-
plementing them with some empirical data. The musical dimension of fado will
be further explored and characterized.
Several sources on fado discuss and approach the concept as an absolute, fixed
category, containing in itself a series of performative or contextual phenomena.
The legitimation of a category often occurs when it is claimed by primary or
secondary sources or in the media. Sometimes it can be artificially created and
incorporated in the daily lexicon of a community by institutional enforcement, but
often the opposite occurs. The practice precedes the theory and the category is
created when, in a given cultural or spatial context, a certain consensus already
exists in the mind and memory of the people – either the practitioners, or the lis-
teners. The process of creation of categories seems to be dynamic and incremental.
It is known that nothing is born out of “cultural synthesis” or “spontaneous gen-
eration”; there always must be an individual authorship for any creation, that
7
might be followed by a certain oral or social chain of transmission, tangible fixa-

tion, or a “butterfly effect” in order for that creation to have impact, and establish
itself as canonical, or else it is simply forgotten. Often, certain phenomena have
local impact and ephemeral duration and then eventually vanish, others remain
alive in the memory of the people. Institutional legitimation is important, and
sometimes a written description by a relevant person, or the propagation in the
media is enough, especially after most part of the population having access to its
content. Fado, as a performance practice, exists in the written sources since the
eighteenth century; however, little or nothing can be assumed or claimed before
that, and even during that period, so this is a wide and fertile terrain for spec-
ulation. It does not seem prudent to assume that simply because there is not a
written source, therefore it does not exist. We find equally imprudent to assume
that, since something is claimed, then it is extensible and generalizable to a whole
context, culture, community or nation. As a rule of thumb, this chapter aims to
describe fado as a holistic concept. We will not exclude the possibility of some
aspects not being considered and one might speculate about their existence. It
seems plausible that in each moment in a given space and time someone might
have considered fado to be something very unique, and personal, given their own
life experience. In a certain context, fado might have had a different meaning
than the one more currently in use, even inside that same community, perhaps
just locally, and episodically. Perhaps one day someone decided to call fado to
a medieval monody. And someone else to a recitation accompanied by an accor-
dion. Would these hypothetical occurrences be less fado than a maxixe sung with
African percussion by Mariza1 ? Sometimes entire groups of people or even entire
communities seem to recognize as fado certain practices that others would never
do. Intergenerational disagreement is also common. Therefore it seems very diffi-
cult to establish boundaries regarding categories or limits regarding concepts. It
also seems fairly arbitrary to try to explain a dynamic phenomenon that seems to
not be self-contained and that continuously changes its own values. Still, follow-
ing the idea of Jordania (Brown & Jordania, 2011), in order to understand any
phenomenon and describe it, one can begin by general tendencies, archetypes and
stereotypes. We will try to do it, as rigorously as possible.
1
Famous fado diva, internationally recognized, and often categorized as a world-music singer.
8
Chapter 3
Methodologies
3.1 Ethnomusicology and Historiography
The field of Ethnomusicology seems to be one in a constant struggle for its own
definition. While there seems to be a central and unified notion of “fieldwork”
as the main methodology to aggregate most of its scholars and researchers, their
own views and paradigms have been changing over time. There seems to be a vast
trend that Ethnomusicology studies music in and as culture. At present, the core
departs from a tripartite model proposed by Merriam (Merriam, 1964) around the
ideas of “concept” (what music is, what music means in a given context/culture),
“behavior” (how the people actually behave, based on the concept, their inter-
actions), and “sound” (the sonic outcome resulting from their behavior). This
model is not linear, it encompasses constant back and forth dynamics, so that
sound might also shape behavior and behavior shape concept, overtime, based
both on the reaction of audiences to performances and external influences. This
model was discussed and improved by Rice on several occasions (Rice, 1987, 2003,
2013). Rice proposed a more complex model in an attempt to refine the processes
through which the sound relates to human behavior. He kept the core of Merriam’s
described as “analytic procedures” but put emphasis on what he called “forma-
tive processes”, the ways in which people historically construct, socially maintain,
and individually create music. The ways in which these issues are approached
varies. While many use a scientific worldview, a notion that data is collected “in
the field” and then analyzed “in the lab”, and universal laws extracted from it,
Geertz brought another paradigm into the discussion. He challenges the scientific,
9
objective worldview. Furthermore, he claims that it is impossible to fully under-

stand another person, that human behavior is far too complex to be reduced to
universal principles. And he believes that science does not explain anything, only
outlines (Geertz, 1973, pp. 5-26). Therefore, instead of trying to aim for prescrip-
tive universal laws, one should look instead for the understanding of meaning, for
observing what is happening in reality.
“The capacity, variable among peoples as it is among individuals,

to perceive meaning in pictures (or poems, melodies, buildings, pots,
dramas, statues) is, like all other fully human capacities, a product of
collective experience which far transcends it, as it the far rarer capacity
to put it there in the first place. It is out of participation in the
general system of symbolic forms we call culture that participation in
the particular we call art, which is in fact but a sector of it, is possible.
A theory of art is thus at the same time a theory of culture, not an
autonomous enterprise. And if it is a semiotic theory of art it must
trace the life of signs in society, not in an invented world of dualities,
transformations, parallels and equivalences.” (Geertz, 1983, p. 109)
Many ethnomusicologists adopted this hermeneutic: thus, an experiential and

subjective point of view. Therefore, one is trying to understand the values that
shape the musical practices.
Bruno Nettl defines ethnomusicology both as the “the study of music in cul-
ture” and “the study of the world’s musics from a comparative and a relativistic
perspective”. He stresses how ethnomusicologists try to understand how a society
musically defines itself, its musical taxonomies, tonal systems and theories; the
social and aesthetics functions of music in that culture: what it does and how it
should be; how a society changes its music and how it deals with external influ-
ences: the processes that encompass change and its relationships with the others.
The way in which such research should be conducted is mainly by carrying out
fieldwork. This implies being in intensive contact with a small, but relevant, group
of informants and performers, by opposition to large surveys or systematic anal-
ysis of big data (Nettl, 2005, pp. 12-13). John Blacking was one of the strongest
advocates of this kind of method and spend several years in Africa conducting
studies on the music and culture of the Venda people. His book “How musical
10
3.1. Ethnomusicology and Historiography
is man?” (Blacking, 1973) is still a seminal text in the area, and it was of great
influence to our own ideas.
At present, this method is widely recognized in academia, and it is fundamental

to understand some idiosyncrasies of particular individuals, their beliefs and the
way they perform their musical practices. However, it does not provide enough
insight into the general trend of an entire culture, neither does it grasp the big pic-
ture if one aims at a global comparison with other cultures. It is impossible to fully
understand the dynamics of thousands or even millions of people by questioning
just a few dozens.
The task of studying the world’s musics by themselves can even be more daunt-
ing if one thinks that there is not even a unified concept of what music is. The
concept of music is seen as broader than the mere “sound” or “notation” so typical
of the Western art music traditions. The attitudes of people towards the music are
seen as part of the concept as well: both of what they think music is, what power
and function it has, as well as their behavior and every action that is needed to be
performed in order for the sound to get produced and communicated (Merriam,
1964, pp. 32-33)(Nettl, 2005, p. 24). Bruno Nettl argues how the ethnomusicolo-
gists usually try to observe all phenomena that somewhat relate to that concept:
one studies what the members of a given culture claim it is music, even if it only
vaguely resembles that concept to us, as also one studies what one thinks music is
according to our own assumptions. After all, “it’s almost impossible to get away
from ethnocentrism, that it’s in the nature of culture to be ethnocentric” (Nettl,
2005).
According to Nettl, music scholars believe that
“when music is produced (in any sense of the word), something new
is being created” and ethnomusicologists have to deal with this “new in
a sense generally understood by them and new also within the specific
cognitive framework and understanding of its culture.” (Nettl, 2005,
p. 28)
In any given culture it is virtually impossible to understand exactly how new

music comes to live in the mind of a musician, however it usually arises as “the
result of the manipulation and rearrangement of the units of a given vocabulary,
of hard work and concentration” (Nettl, 2005, p. 29). In the tradition of Western
11
art songs, this manipulation usually occurs in a symbolic level, with the use of
musical notation and this process is seen as the composition, which is separated
from the translation of material in sound. In this case the musical units are
generally small and may even be “atomic” units such as a note or a grain. On the
other hand, typically in folk and traditional music, these units might be samples of
canonic corpora of a musical tradition like the Persian radif, or the jazz standards.
In many cultures people simply learn the repertoire from their parents, friends
or even teachers and masters, and introduce minor variations when executing it.
Over time, the variants may become more and more inventive, but usually they
are recognizable as forms of the original. “New songs might be composed, but
they were cast in the rhythmic, melodic, and formal mold of songs already known”
(Nettl, 2005, p. 32). This practice inevitably leads to an increasing corpus of songs
that are relatively alike, all derived and similar to one another. This is extremely
common in several musical traditions, namely the European-folk traditions and
fado is no exception, because the same creative model applies. The distinction
between composition and improvisation is clearly blurred in these cases and Nettl
prefers to consider these axes as continua. In each musical tradition the ones that
are most innovative, prolific or inspired and are able to excel in creating new songs
and positively contributing to the creation or expansion of a given repertoire are
usually seen as “geniuses” and highly regarded (Nettl, 2005, pp. 38-41).
On a related discussion, Negus presents the idea that
“there is no simple correspondence between an ethnic group and

mode of cultural expression. Musicians and their songs cannot sim-
ply express, reflect or represent a particular group of people. (. . . )
the cultural correspondences are established through particular social
practices and political activities. (. . . ) There is a process throughout
particular musical codes, signs and symbols become used and claimed
as expressions of particular social and cultural identities. It consists on
the social interactions, relations and mediations (many time political)
that occur around and across the music as it is created.” (Negus, 1997,
p. 122)
Furthermore,
“There is no straightforward or intrinsic link between the lives of

fans, the meaning of musical texts and the identity of a particular
12
artist. Songs and musical styles do not simply ‘reflect’, ‘speak to’ or
‘express’ the lives of audience’s members or musicians. A sense of
identity is created out of and across the processes whereby people are
connected together through and with music.” (Negus, 1997, p. 133)
Negus also explains how music gains new fans and new identity meanings:
“Once in circulation, music and other cultural forms cannot remain

bounded ‘in’ any one group and interpreted simply as an expression
that speaks to or reflects the lives of that exclusive group of people.
In addition, the idea that music can in some way express the unity of
a particular group of people must continually confront the apparent
disunities and differences that can be found in the world.” (Negus,
1997, p. 121)
The traditional fieldwork method generally used, by itself, we believe, is unable

to explain the way the values are themselves shaped and how some other higher-
level constraints directly shape the musical practices regardless of values (meaning
they are cross-cultural traits). What we are proposing is a reflection on how
an interdisciplinary approach making use of knowledges of other areas (namely
psychology of music and systematic musicology) can complement the model and
improve our understanding of both the shaping of musical practices and human
behavior in general. While we perfectly understand and agree, to some extent,
with many of Geertz definitions and challenges, we assume that science has made
a great deal of progress since the seventies, and the scientific and positivist methods
have indeed contributed to build a corpus of knowledge around the ideas of some
universal or quasi universal laws around human behavior. At present we do believe
that, contrary to what Geertz claims, science is able to explain some things, or
at least offer some very probable causal relations for many phenomena shaping
musical practices.
The recent studies in Embodied Cognition (Leman, 2008, 2012) depart from
the idea that all human beings have something in common: a body. Therefore,
there are inter-specific and universalistic traits deriving directly from that fact.
Namely, the presence of a heartbeat, at around 60 beats per minute, and there-
fore, how the concept of musical tempo is shaped. This is a clear example of how
the perception of a musical parameter is directly shaped by a body characteristic
13
not dependent on cultural traits. The study of universals in seen by Nettl as a set
of concentric circles, the definition of music itself being the larger one containing
the essential characteristics of the musical utterance. Then, the traces of total
musical cultures or musics are contained within the second and finally statistically
relevant traces of those musics are contained within the third. Music is claimed to
be not a universal language, but the musics of the world are not as different from
each other as the languages of the world are (Nettl, 2005, pp. 48-49). Since the
experience of learning music, performing music and communicating music is an
embodied experience, there are many more commonalities than one might think.
Derived from our body, many gestural and motor actions are also universal and
their understanding is also cross-cultural (Brown & Jordania, 2011). Therefore the
study of those common traces and their prevalence is something worth pursuing –
even if one finds that there are many, many exceptions and several ramifications,
the mere existence (or absence) of statistically relevant traces in a reasonable num-
ber of musics or even within some subgroups of musics might reveals us something
about the nature of what it means being a human in a certain context.
On the level below the body, we propose what we call “the language” layer. At
present, and based on recent neuroscientific and psychological studies, it is almost
consensual that language shapes the way human beings create and articulate sym-
bols and concepts, and therefore, it is in the base of how one thinks. Culture, as
defined by Geertz (Geertz, 1973), is the set of knowledges and interactions shared
by a certain community. The values are therefore shaped by language. There
are no values if one does not have a language to express and think about them,
neither there can be values if there is no communication between the members of
any given culture.
Furthermore, language also shapes directly the musical practices in cross-cultural

ways. The members of a given linguistic community will share common traits, even
when they belong to different cultures. As Patel demonstrated, when infants grow
their brains create discrete “boxes” out of the realm of the continuum (Patel, 2008,
2014). The natural rhythms and prosody of a native tongue will directly influ-
ence the musical rhythms human beings use to create their vocal melodies. In the
same manner, as already noticed by Leonard Bernstein (Bernstein, 1976) and Pat
Pattison1 (Pattison, 2010), there are harmonic functions in speech, and therefore,
1
http://songwork.com/video-lectures/harmonic-functions-speech, accessed June, 6,
2015
14
language also shapes the intervals used in the vocal melodies. These constraints,
we believe, will then directly shape choices governing melodic creation, overtime,
even in the absence of lyrics. Only below the level of language we envision the
values. The level of conventions, shared grammars, imagined communities. It is
widely known how this level indeed shapes musical practices and this has been
studied for decades now (Blacking, 1973, Nettl, 2005). What we propose then,
is to complement this level with the addition of two layers of information, that
are also shaping the musical practices, in ways that cannot be efficiently studied
by merely fieldwork. The ways the body directly shapes the musical practices is
something that has been explored by neuroscience and the psychology of music,
mostly through the use of empirical research (Deutsch, 2012, Zagorski-Thomas,
2005, 2012). The ways language directly shapes the musical practices has been
almost neglected, but only empirical study can do the task as well, and neuro-
science has been pursuing that road. The way the body shapes language, and the
ways language shapes values we would say is out of the scope of musicology, and
its study should be of concern to understand human behavior and human agency
in general.
Figure 3.1: Complementing Fieldwork.
Of course, there are also body actions mediated by culture, namely gestural
tropes, as studied by Hatten (Hatten, 2004, 2006, Robinson & Hatten, 2012), and
there are, of course, language traits also mediated by culture, namely the thematic
15
content of lyrics and programatic associations (for instance, Pastoral Symphony

or The Firebird ). But these are already included in the actual framework of study
of ethnomusicology.
What we are aiming at are the specific cross-cultural traits that shape the
musical practices that fall out of this realm. Due to its cross-cultural nature and
their existence in large scale in human populations, and taking into account the
advent of globalization and internet, we believe they can only be studied through
the use of systematic methodologies and the analysis of big data. Namely resorting
to statistical methods (descriptive, comparative and inferential) with the aid of
computers, and visualization of information techniques aiming to spot recurrent
patterns.
The proposed framework also assumes bi-linearity, that is, overtime, musical
practices also shape the body and the values, and in the same way, values also
shape language and the body, and language also shapes the body. It is a constant
feedback loop, back and forth, in a complex network of interactions. Our proposal,
however, is to focus mainly in the interactions directly related with the musical
practices, as we believe, they truly represent the concern of ethnomusicology and
of musicologists in general.
An introductory analogy by Seeger uses the variety of ways in which a banana

can be sliced to illustrate how several perspectives are required to understand a
concept, and even so some aspects of it will not be attained (Seeger, 2002). In a
certain way, when talking about music one is metaphorically talking about human
nature, about ourselves (Videira, 2011). One cannot be fully objective due to the
cognitive constraints of being human-in-the-world, the limitations of language and
the physical constraints of being trapped inside our human bodies in a given space
and time. Therefore, one is doomed to be limited, reductionist and imperfect.
Assuming the limitations and flaws of every research we believe that, within the
constraints of time and resources we have, the best approach to the problem is
to accept certain assumptions and build on top of them – namely the concept
of fado. While many could argue that we should go into “the field”, mix with
the communities and study the musical practices and therefore infer what fado
is, we argue that that kind of work has already been done by other scholars and
that is not our main goal. If one is to use the same metaphorical language of
Seeger, what we are trying to do is to pick up several banana slices cut by other
scholars, use other kinds of knives to make some slices ourselves, perform chemical
16
and analytical tests on the peel and speculate about its aroma and processes of
reproduction, hoping that in the end we finally get an idea of what a banana is
in order to replicate one. We cannot truly assert that the combination of several
different sources and methodologies will result better than using an established
method from beginning to end, but at this point we believe the more information
and angles one has, the better equipped one is to solve any problem regarding that
same issue.
3.1.1 An operative model to study the musics and sounds
The interdisciplinary framework envisioned aims at a scientific framework, en-

compassing several disciplines, to study the musics and sounds associated with the
musical practices. The way to do this, using the tools most hard-sciences use, is
the gathering, preparation, analysis and manipulation of data.
Figure 3.2: Operative Model.
Data about the musical practices should be gathered either in the field or
through historical research or both. When using fieldwork methods typically the
data assumes the form of video footage, sound recordings or transcriptions. In
the case of practices from the past, historical research in archives or personal
collections might also unveil transcriptions of music in notated forms, iconography
or textual descriptions. Either way, the first step is to obtain this data, as well as
relevant information about its context.
17
The second step is the digitalization and preparation of the data. With digital-
ization one typically aims at obtaining fac-similes, in the case of transcriptions, or
digital audio and video files in the case of recordings. The main goal is to preserve
and have redundant copies representatives of something as closest as possible of
the original source. The task following this, is the conversion of the digital raw
data to symbolic formats, suitable to be edited, parsed and analyzed. In the case
of the transcriptions, a fac-simile is really a photograph, an image format. The
main goal is to convert it into symbolic format. This is done by transcribing it
into notation software to create an urtext. This urtext can then be exported into
many different formats for subsequent use. A good practice should be to keep re-
dundant copies in several formats, preferring open ones, since proprietary formats
might expire2 . After data has been saved and archived one can proceed to prepare
it. This process, in the case of transcriptions, typically involve critical editing
methods, imported from philology. In the case of recordings the situation might
be more complicated but often it involves the restoration, editing and trimming
of the audio in order to isolate the relevant sections or parameters one desires to
study. Often, in some musicological projects, the goal is achieved at this stage,
since the preservation and publication to the community is the only objective.
The third step envisioned is the analysis of the data. This involves two stages.
The first one is descriptive and aims to make sense of the data and convey it in
terms of numerical or categorial data. Therefore one can use analogical methods
like musical analysis involving segmentation, harmonic analysis, textual descrip-
tions, tagging. More recently, when the data evolves big corpora, music informa-
tion retrieval techniques are used and the digital files are parsed automatically
generating huge files. This new data can then be analyzed through descriptive
statistics, using statistical packages, and tables and charts produced for better
visualization and comparison. The second stage involves interpretation: making
inferences and establishing correlations in order to generate models for the data.
In the case of a set of musical transcriptions, the analysis of the data typically
involves the formal and harmonic analysis of the corpus, a systematic analysis
via music retrieval techniques and the generation of tables with several different
features. These features can reflect many of the parameters by which one char-
acterizes music, like pitch or rhythmic statistics, melodic contour, note density,
2
At present, exporting the urtexts as XML, MIDI and PDF seem good options and most
notation programs are able to do this. Video and audio formats should be also saved as lossless
formats like WAV, AIFF, or lossless compressed ones as FLAC or M4A.
18
range, etc. . . Comparative statistics can be made with other corpora, or simply
inferential statistics by trying to correlate some features with other ones and cre-
ating models. Many musicological projects conclude at this stage when they aim
at the characterization of a certain genre or practice, or simply comparing a certain
corpus of songs with another, or even, to simply ascertain the internal coherence
of the initial corpus.
The fourth stage can also involve two steps or two different approaches accord-
ing to the desired intention. The first one aims at explaining the musical practices
analyzed, described, and eventually modeled, in the previous step. In that case,
one tries to produce a descriptive theory. A theory that will explain the practices
as they occurred and were observed up to that point. Using the theory one should
also be able to recreate or create new musical works that would be similar to the
ones observed. The second one is a more creative approach and aims to produce a
prescriptive or predictive theory. Speculative by its nature, in this case, following
it one should aim at create new musical works that will, hopefully, innovate the
musical practices studied, trying to anticipate the changes they will suffer. In a
global world and knowing that musical practices are dynamic and ever feeding
themselves, in a feedback loop, we would say that such approach, more than an-
ticipate might indeed suggest or force the change, in that given practice, once the
original practitioners acknowledge it.
According to this holistic vision the ethnomusicological work is only complete

when one indeed has an explanatory theory for the musical practices studied, as
well as a suggestions on how and why might they change and in which direction,
as well as the ability to reproduce, recreate or innovate them. Therefore, the
ethnomusicologist is envisioned as a humanist, a social scientist, an empirical one
(conciliating the views of Rice and Blacking), and, furthermore also an active
participant (conciliating the view of Mantle Hood).
3.1.2 Consilience
Having a unified, holistic methodology might seem impossible, especially in

the short run. However, good practices might be established in order for results
to be easily reproducible, and furthermore, to be comparable between them all.
Also, the modular approach proposed in the operative model also allows that a
researcher can work in a team, leading only the research up to a point, and then,
19
the problem to be continued by another researcher or another team, in another

time and part of the planet. The main goal is to allow for the maximization of
resources in the long term and an approach to maximum consilience between the
several areas of knowledge.
In order to achieve that, and as each ethnomusicological project, involves the

interaction and study of human beings in their context, what often lacks is data
about those contexts. Therefore what is also proposed is that in every research,
either fieldwork, with single informants or small groups, either laboratory ex-
periments, using empirical methods, standardized metadata retrieval should be
conducted.
Metadata, here, refers to a set of informations related to contextual parameters

that will allow researchers to compare, integrate and replicate studies that are
apart in time and place. Since fieldwork is usually a time consuming, lonely task
to be conducted on a small group of people, it is impossible to retrieve statisti-
cally significant data out of those samples and generalize it for a given population
or linguistic community. Therefore, some of the most important parameters to
be retrieved could be native language of the persons involved, cultural commu-
nity, degree of education, social status, age and musical values. By having more
complete sets of data it would be possible to make them comparable to a certain
extent and retrieve big data from the smaller sets. This would allow scientists to
make comparative and inferential statistics out of several fieldwork studies, and
could lead to what happens in the hard sciences: bigger, more complex projects,
involving teams from several research centers, spread throughout the world, study-
ing several musical practices, involving several human beings, and comparing the
results cross-culturally, using methods from several scientific areas.
3.2 Ethnographic and Historiographic Sources
Many written testimonies of fado have been produced throughout recent his-
tory. We have novels, memories, personal letters and periodicals referencing fado
and sometimes describing contexts, performances or situations. We also have sev-
eral kinds of essays, papers, dissertation and books. We have engaged in a critical
reading of these sources to understand the difference between subjective, personal
observations and speculation, on the one hand, and, on the other, the ones that are
20
3.2. Ethnographic and Historiographic Sources
academic or use some kind of scientific method (the later ones are rather scarce).
Also, we were aware of the several different approaches and methodologies used:
informal descriptions and opinions; sociological, philosophical and theoretical de-
scription of contexts; historical and political readings of the reality; theoretical
and analytical assumptions regarding some aspects of the phenomenon; surveys
on past sources or ethnography and field work, etc. . .
There is also a huge amount of musical transcriptions and scores. It was im-
perative to understand their context, who wrote and collect them and with what
objective. We have also applied analytical and philological treatment in order to
depurate and obtain the most faithful material.
There are also the living and dynamic sources: informants, who are themselves
performers, and living testimonies of the phenomenon nowadays. Also, we have
had access to digital sources and recordings, although not in a systematic way, but
as a consequence of living in the era of Internet: videos, TV shows, CDs, interviews,
websites, and a whole lot of disperse information that we stumble upon every day
and which is very useful to access and document whenever relevant or needed.
The interpretation and validation of all these sources is of extreme importance

since there are huge differences between them and not all of them have the same
reliability. These sources were used not only to understand what fado is, but also
to outline a coherent, complete and up-to-date definition of the phenomenon in
all its layers and facets.
3.2.1 Critical Sources
Most nineteenth-century references to fado appear mostly in novels, travel di-

aries, pamphlets, newspapers and informal essays. Although some of these sources
were consulted, namely (Almeida, 1893, Balbi, 1822, Freycinet et al., 1826, Or-
tigão, 1898, Peixoto, 1896), we have had access to their content mainly through
quotations in more recent surveys and historiographies. Most of these sources are
not scientific, nor have in mind fado as a direct object of observation. The first
academic reference relevant to our study is an entry on the musical dictionary of
Ernesto Vieira (Vieira, 1890), which provides us with a descriptive model that will
be analyzed further down the thesis.
21
The beginning of the twentieth century brought two new books that provide
an extensive account of the phenomenon: História do Fado [The History of Fado]
written by Pinto de Carvalho (Carvalho, 1903) and A Triste Canção do Sul [The
sad song from the south] by Alberto Pimentel (Pimentel, 1904). We consider
these books as foundational sources for the study of fado, and, according to the
standards of the time, we would consider them as academic studies that try to
present the history of fado, its myths and origins, as well as descriptions of the
practice, main lyrical themes, and relevant figures.
Two more books from the first half of the twentieth century are of great impor-
tance to fado history and testimony: Fado, Canção de Vencidos [Fado, the losers’
song], a series of eight lectures broadcast in the national radio presenting fado
in a negative light (Moita, 1936), actually acquired extreme relevance, because it
was so detailed in scrutinizing the traits and characteristics of fado, as well as its
history, that the result is a relevant description of the values associated with it, as
well as the surrounding society. As a reply to the former, another book, defending
fado, Os ı́dolos do Fado [Fado’s Idols] (Machado, 1937) acknowledges its virtues
in a very biased and passionate way, by quoting a series of epoch personalities,
and by conducting a biographical survey of most of the relevant performers at the
time. These two works account for a rich testimony of the period and are valuable
sources to understand how fado was seen and described at the time. After this
period we have almost a fifty years gap with few relevant publications in the area.
Some new works arise, namely (Barreto, 1959, Branco & Barreto, 1960, Osório,
1974), but they should not be taken as relevant sources since they do not add spe-
cific value, compared to more recent ones who quote, interpret and contextualize
them.
The last two decades have seen an unprecedented rise in the number of scholars
and social scientists interested in the areas of popular music, musical industry and
ethnomusicology in general, and fado in particular. This has led to the flourishing
of a considerable amount of academic studies regarding the phenomenon in several
of its facets, and to a vast number of publications of relevant interest to the area,
namely peer reviewed articles and academic dissertations. Some of these works are
grounded on historiographical resources from the past, already highlighted, and
most of them also collect dispersed information from the other sources (namely
literary works and newspapers) and try to contextualize and interpret them, and
are thus more useful for the contemporary researcher.
22
In Brazil, José Ramos Tinhorão found what he thinks is a missing link, and
published a detailed theory explaining how, in the end of the eighteenth century,
there was a performance practice called fado, which included dance as a strong
component (Tinhorão, 1994). He strongly believes that Lisbon’s fado is derived
and intrinsically connected to this older Brazilian dance. His thesis is widely
accepted by academics and has been incorporated and legitimized from then on
in most subsequent studies, having found resistance only among non-scholars.
When Lisbon was the European Capital of Culture, in 1994, a large exhibition
showcasing fado was displayed in the Ethnology Museum. This event led to the
publication of a catalog, Fado: Vozes e Sombras [Fado, Shadows and Voices],
trying to show what had been done up to that date, containing many relevant
texts describing the performative practice (Brito, 1994). In that same year Ruben
de Carvalho (Carvalho, 1994) published a book on the same theme, as well as
Barreto (Barreto, 1994). Later on, Salwa El-Shawan Castelo-Branco published a
whole book regarding Portuguese music with a chapter entirely dedicated to fado,
which is a very detailed and systematic description of the performance practice,
similar to the one already written in the catalog (Castelo-Branco, 1998). An
entire hardcover deluxe collection, commemorating one century of fado, in seven
volumes containing photos, transcriptions and essays was also published (Guinot
et al., 1999).
Both the Master’s thesis and the Ph.D. dissertation of São José Côrte-Real,
and the respective papers resulting from them, (Côrte-Real, 2000, 2002, 2010,
1991), are very important and detailed systematic studies on fado, as they contain
both ethnographic interviews with migrant communities and performers (including
Amália Rodrigues), as well as relevant documentation concerning the cultural
policies of State during the Portuguese dictatorship period (1926-1974).
Rui Vieira Nery, son of the prominent fado guitarist Raúl Nery, is a key present
day fado scholar. Not only is he a prolific scholar, but he has also collaborated in
many initiatives regarding the divulgation of fado, namely the approved candidacy
for UNESCO Intangible Cultural Heritage. His publications are detailed histori-
ographical surveys and hermeneutics on the past sources. His books Para uma
História do Fado [A History of Portuguese Fado] (Nery, 2004), Pensar Amália
[To Think (about) Amália] (Nery, 2010c), the entry on the Portuguese Encyclope-
dia of twentieth century Music (Nery, 2010a), the collection of essays portraying
23
fado, mainly through its lyrical content, in the earlier years of the Portuguese Re-
public and during the First World War (Nery & Mateus, 2012), are very relevant
sources since they portray, probably, the best overall view regarding the many
facets of the phenomenon of fado up to date.
Daniel Gouveia is not a scholar, but a fado musician and composer, and someone
who has lived all of his life among fadistas. He provides us with a very personal
testimony on the practice, some passionate considerations, and a useful inventory
of poems and repertoire (Gouveia, 2010). Alberto Sardinha, also not a scholar, but
a passionate fan, unsystematically recorded live testimonies from local people from
rural areas. He compiled his set of highly speculative theories and observations
regarding the origin of fado (Sardinha, 2010a). Both non-academic works were
reviewed by Rui Vieira Nery in passionate discussions on the press, praising the
work of Gouveia and highly criticizing Sardinha’s (Nery, 2010b, Sardinha, 2010b).
Regarding the issue of the historical origins and myths about fado, an earlier essay
was also published by Maria Luı́sa Guerra (Guerra, 2003).
A state-of-the-art generation of researchers, many of them with connections to

INET-MD, the Institute of Ethnomusicology of the New University of Lisbon, has
been publishing relevant information concerning fado in papers, studies, essays,
catalogs and dissertations. Among them we highlight Pedro Félix (Félix, 2013),
who writes on fado in the musical theater, Pedro Russo Moreira and João Leitão
da Silva, who wrote very relevant chapters concerning issues related with fado,
musical industry and cultural policies in their Ph.D. dissertations (Leitão da Silva,
2011, Moreira, 2012). We can also include in this set of works the dissertations of
Ricardo Fonseca (Fonseca, 2011) and Liliana Rodrigues da Costa (Rodrigues da
Costa, 2012) who approach fado from a socio-economical point of view, but also
provide interesting data that can be further analyzed from other perspectives. Also
relevant is the work of Sara Pereira focusing on the iconology of fado (Pereira, 2015)
and Pedro Pavão dos Santos on the issue of identity (Santos, 2015).
We also have the work of scholars who approached fado from the perspectives of
space and identity. Michael Colvin studied the neighborhood of Mouraria (Colvin,
2008), while Richard Elliot dissects how “fado acts as a cultural product reaffirming
local identity via recourse to social memory and an imagined community, while
also providing a distinctive cultural export for the dissemination of a remembered
Portugal on the global stage” (Elliott, 2010). Kimberly Holton, a Portuguese
American, provides us with a historiographical survey (Holton, 2006).
24
An original empirical study regarding the acoustic and phonatory characteri-

zation of fado voice was also recently published (Mendes et al., 2013).
3.2.2 Fieldwork
Systematic fieldwork using mainly an emic approach was conducted in Lisbon,

since 2009, taking notes from informants, attending performances and reading
and paying attention to the media. Especially relevant was the contribution of
a group of fado performers in the context of tertúlia [informal gathering], typical
restaurants, and touristic venues3 .
A pioneering anthropological-sociological fieldwork on fado was conducted in

the late seventies in the neighborhood of Alfama. Although it was not intentionally
designed this way, in the end it led to a most interesting anthropologic-sociological
study with many consequences for the understanding of the musical phenomenon
(Costa & Guerreiro, 1984). We would also like to highlight recent academic disser-
tations, mainly conducted in Lisbon, representing serious and rigorous fieldwork.
They constitute a body of knowledge that is not only relevant because of the
dissertations in themselves, but also mainly because they contain data, descrip-
tions, and ethnographic interviews available to be read, compared, studied and
submitted to new and further interpretations. Among them we have the studies
by Oriana Alves and Liliana Sebastião (Alves, 2003, Sebastião, 2009), from the
New University of Lisbon. Also Lilly Ellen Gray, which offers us a valuable testi-
mony and ethnographic view on fado, containing much valuable information about
its music and vocal sounds to be considered (Gray, 2013). Lilly Ellen Gray also
published other valuable papers regarding the same theme (Gray, 2007, 2011).
Soraia Simões, researcher and author, has also been conducting a series of
ethnographic interviews with musicians, composers, instrument builders and per-
formers with the aim of having a relevant contemporary sound archive. Part of this
archive, called Mural Sonoro, is available online4 . She also recurrently moderates
lectures, debates and informal gatherings around relevant musical themes, namely
around fado. Several of these reunions were attended. The available material is a
valuable primary source for the study of emic perspectives on fado during our age.
3
The youtube channel “Tertúlia” contains recordings of some live performances of the group:
https://www.youtube.com/channel/UCJlUZ6RIRgbL7iHfbNdfJhA, accessed June, 6, 2015
4
http://www.muralsonoro.com, accessed June, 6, 2015
25
3.2.3 Recordings and Media
A systematic study of the early phonographic recordings is yet to be completed.

Portugal does not have a National Sound Archive and the materials are not easily
accessible, that being a work in progress. It is expected that in the near future
the situation will be solved, since members of INET-MD like António Tilly and
Pedro Félix are currently working on the issue. One of our main goals was to work
on the symbolic written sources, therefore we have only used as references the old
recordings presented in the CDs that come along with books like the ones from
Paul Vernon (Vernon, 1998), Daniel Gouveia (Gouveia, 2010), Alberto Sardinha
(Sardinha, 2010a), and Leonor Losa (Losa, 2013) along with random old digitized
recordings that could sometimes be found on the Internet. These early recordings
were extremely important as comparative references not only for one to be able to
imagine the actual sound of nineteenth century scores, but also to perceive change
during the twentieth century.
Two important series were broadcast by RTP (Portuguese national television)

recently: Trovas Antigas, Saudade Louca, a history of fado, (Ávila, 2010) and
Estranha Forma de Vida, a history of Portuguese Popular Music from the twentieth
century, both with archive footage, interviews and relevant information. We also
had access to several similar references and testimonies in an informal way, being
constantly and randomly exposed to TV shows and often old archive images and
recordings with relevant information broadcasted. Some of them are available
online and can be consulted.
3.3 Systematic Musicology and Symbolic sources
In order to fully characterize the musics and sounds of fado we consider the
study of past musical examples to be a fundamental source. Since fado perfor-
mance pre-dates the era of sound recordings the study of symbolic notation is
required to access the music of those periods. There are several transcriptions
of fado on musical notation, the most recent being the 180 fados transcribed by
António Parreira, on the so-called “traditional” canonical repertoire fixed in the
beginning of the twentieth century (Parreira, 2014). However, we have decided
to go to older sources. We have compiled a representative corpus of one hundred
samples to perform our initial research. The Cancioneiro de Músicas Populares,
26
3.3. Systematic Musicology and Symbolic sources
compiled by César das Neves and Gualdim Pais (Neves, 1893), contains several
traditional and popular songs transcribed during the second half of the nineteenth
century, 48 of which categorized as fados, by the collectors, either in the title or in
the footnotes. This set was complemented with a transcription of Fado do Marin-
heiro, provided and claimed by Pimentel to be the oldest fado available (Pimentel,
1904, p. 35). We have also had access to the digital version of the collection of
musical scores from the Theatre Museum in Lisbon, comprising more than one
thousand individual popular works that were previously scanned by researchers
from INET-MD. The first 51 categorized as fados were selected in order to com-
plete the corpus. Therefore, we have obtained 100 transcriptions representing
works composed roughly between 1840 and 1970.
All this data is not only relevant for the understanding of the rules and patterns
that generate the typical music associated with fado practice, but also as cultural
heritage archive and source material for future projects. Having this in mind,
and thinking how a new digital object could arise from this data, a collaborative
team was created in Lisbon, in articulation with Professor Vı́tor Badalinho and
some of his master students of New Media and Web Practices: Cândida Gouveia,
Ricardo Correia and Rita Pereira. The team has developed a website hosted in
the New University of Lisbon server5 , allowing the database soon to be available
to the general public, as well as allowing sorting and searching relevant content
using several criteria. The database content is provided in the appendixes of this
dissertation as well.
We estimate that there are at least 200 more works categorized as fados to be
found in the collection. Future work will be transcribing and editing the remain-
ing scores categorized as fados and enlarge the database. We have chosen these
particular sources because they were readily available and still not yet analysed
or digitalized. Many more could follow.
These transcriptions were meant to be played by upper class girls in their homes
and salons. The distinctive accepted social model at the time was the one imported
from Paris and thus it was of the utmost importance for them to learn French and
to play piano at a basic level (Nery, 2010c). According to Leitão da Silva, the
notion that the musical scores were meant for the domestic entertainment market
led the collectors to present them according to the conventions of commercial sheet
5
http://fado.fcsh.unl.pt
27
music editions at the time, namely by adapting their harmonies and instrumenta-
tion to those conventions, softening their lyrics, and dedicating each piece to an
aristocratic lady. Furthermore, the choice of repertoire to be transcribed also priv-
ileged the texture of solo voice, despite many traditional Portuguese songs (from
Alentejo and Minho, for instance) using polyphonic ones (Leitão da Silva, 2011,
pp. 257-258). In a certain way, and following a phenomenon common to other folk
genres across Europe, social class shaped a canon (Negus, 1997, pp. 73-74), and
the mere study of the music of rural populations through the means of notions of
scientific objectivity reflected the idiosyncrasies and values of the collectors and
transcribers more than those of the popular repertoire itself.
3.3.1 How social class shapes a canon
Negus reviews an interesting theory developed by Dave Harker, following the

tradition of Marxist analysis, on how social class seems to be shaping a canon.
Harker sees mediation as a phenomenon by which
“particular people passed on songs they had taken from other sources,
in the form of manuscript or print, but that in the very process of so
doing their own assumptions, attitudes, likes and dislikes may well have
significantly determined what they looked for, accepted and rejected.
Not only that, but these people’s access to sources of songs, the fact
that they had the time, opportunity, motive and facilities for collect-
ing, and a whole range of other material factors will have come into
play (. . . ) while we cannot ‘read off’ what a person did with songs
from, say, their class position, it is still the case that their social ori-
gins, education, occupation (or lack of it) and so on were obviously
connected with how they felt, thought and acted in relation to songs”
(Negus, 1997, p. 69).
It follows the concern on how the mediators
“bourgeois song collectors systematically neglected the music of

working-class people and constructed and then managed to institu-
tionalize a gentle, pastoral, ideologically biased ‘tradition’ of what he
calls ‘fakesongs”’ (Negus, 1997, p. 69).
28
If we extrapolate this to the Portuguese case and collectors like César das Neves,
then we have an interesting issue to take into account: how representative is this
collection? How “ideologically polluted” might it be? How can one guarantee that
it is not the case that an entire canon was shaped by their hands? It is exactly
because of these issues that one has to deal with written sources (musical scores)
with a great deal of care. Some of these issues are also discussed in a recent study
about the folklorist Armando Leça and his work in the Portugal of the first half
of the century (Pestana, 2012).
According to Leitão da Silva,
“The heterogeneous techniques used in musical transcription point

to the ambiguous space that traditional music occupied. On the one
hand, the study of the music of rural populations relied on notions of
scientific objectivity in which the collector’s agency is effaced. Con-
versely, the larger segment of musical transcriptions of these repertoires
shared the conventions of sheet music editions (in their harmonization
and instrumentation), a fact that points to its integration as part of the
market for domestic entertainment” (Leitão da Silva, 2011, p. 257).
Following his reasoning, Leitão da Silva, states that
“According to Pestana, the collection and publication of the reper-

toire contained in the Cancioneiro that was drawn from rural con-
texts was underpinned by a nationalist ideology and supplied a version
of these songs that was accommodated to urban audiences. Conse-
quently, it is possible to discuss the Cancioneiro as a symptom of a
process in which the vernacular was aestheticized in order to integrate
the conventions of the entertainment market. In this process, one of
the criteria in selecting repertoires collected in rural areas was the mu-
sical texture. Despite including songs from many Portuguese regions
the Cancioneiro consisted exclusively of songs for solo voice, glossing
over the vocal polyphonic practices of several of the country’s provinces
(such as Alentejo or Minho), and emulating the conventions associated
with the commercial edition of sheet music. Moreover, each song was
dedicated to an aristocratic or bourgeois woman/girl, which indicates
the publication’s intended audience as well as reinforcing the parallel
29
between singing and playing the piano with the feminine segments of
several strata of Portuguese society” (Leitão da Silva, 2011, p. 258).
According to Negus’ review on the history of ballads in Western Europe,
“the oral tradition had not been quite as anonymous, collective and
immediate as has often been portrayed. There was often a distinct
division of labour between song composers and performers. During
the Middle Ages, professional singers had frequently sold their ser-
vices particularly to the nobility and clergy. Hence, the oral tradition
had not simply entailed a fusing of creation and reception in a pas-
toral communitarian world. Instead, there were many notable indi-
vidual performers, whether the aristocratic and trained bards or the
jongleurs and traveling minstrels. The broadside ballad, facilitated by
the printing press but drawing on oral traditions of performance and
music circulation, was distributed by sellers who drew crowds of peo-
ple onto streets corners and into town squares and performed their
songs. These ballads covered a rich variety of subjects. There were
songs about war news, labor songs, love songs and joke songs about
human relationships. There were songs about death and murders, gal-
lows songs and tales of criminals and royalty. There were songs about
sport and prominent figures and campaigning songs that addressed po-
litical or social issues. Combining together what would now probably
be identified as news, commentary and entertainment, these ballads
often achieved higher sales than the newspapers of the time. This has
prompted some writers to suggest that the ballads were a precursor
to what become the ‘tabloid’ newspaper: in these times the populace
‘sang their journalism’.
Gradually over time the relationship between the producers and
consumers of these ballads became more organized and singers and
shouters were replaced by more specialized reciters. Towards the end
of the eighteenth century the broadsides were increasingly adopted by
professional performers and began to be produced with a more the-
atrical quality as the street performances interacted with and became
part of the music-hall tradition of entertainment. These changes were
accompanied by municipal attempts to regulate public behavior in the
30
streets of the cities. Laws were enacted that restricted public shouting
and singing and which led to a sharp decline in the trade of the street
performers. Eventually the street ballads were superseded by the for-
mal production, distribution and plugging methods that were adopted
to distribute songs by the sheet music industry. Yet, the practices
of the song peddlers and ballad hawkers were retained in many of the
practices of the boomers and pluggers who utilized public singing when
promoting songs in music halls and vaudeville theaters. The popular-
ity of sheet music peaked at the end of the nineteenth century, but the
song sheet and many of the techniques of ballad production (such as
producing songs quickly in response to topical events) were adapted
to the Tin Pan Alley tradition of ‘song factory’ production” (Negus,
1997, pp. 73-74)
3.3.2 The Database and the urtext Corpus
According to Bardin, a strong reference in content analysis methodology, the

analysis organization occurs in three chronological poles: the pre-analysis, the
exploration of the material and the consequent analysis of the results, inference
and interpretation. The pre-analysis is an organizational stage corresponding to a
period of intuitions, the goal of which is to systematize initial ideas and make them
operational leading to a scheme or a plan. Usually at this stage three objectives are
to be accomplished: the choice of the material to be analyzed, hypothesis and goals
formulation, and the elaboration of cues that will sustain the final interpretation
(Bardin, 2009, pp. 95-96).
“There are two main ways in which people approached pieces of music. One was
their overall form and the other was their melodic, harmonic or rhythmic content.
(. . . ) Analyzing the form of a new piece basically consisted of assimilating it into
one existing formal prototype or another” (Cook, 1994, p. 9). Since the content
of the database refers to the musical scores from the late nineteenth and early
twentieth century we are using as a model the first fado definition by Ernesto
Vieira, as shown in his musical dictionary:
“a section of eight bars in binary tempo, divided into two equal

and symmetric parts, with two melodic contours each; preferably in
31
the minor mode, although many times it modulates into the major,
carrying the same melody or another; harmony built on an arpeggio
in sixteenth-notes using only alternating tonic and dominant chords,
each lasting two bars”(Vieira, 1890, pp. 238-239).
This definition is our hypothesis and we have evaluated it using both philological
and comparative methods referenced in empirical musicology publications inspired
on the methods and techniques of formal content analysis and philology (Grier,
1996). We believe that our universe conforms with the general rules of exhaustivity,
representativity, homogeneity and pertinence (Bardin, 2009, pp. 97-98), hence
configuring a coherent corpus.
The technique applied consists in giving a unique identifier number to each

fado occurrence found in the sources mentioned (001, 002, 003, etc.) and then to
re-transcribe each one into a Music Notation program. In this case we have used
Finale!(MakeMusic Software). We have chosen this program because not only
it is a flexible and professional engraving software, but also because it allows to
export the files into several formats, namely XML (open binary file format that
can be read in any other music notation software, proprietary or not). Also, this
software allows us to edit the scores and export them as PDF (printable image
format), so they can be seen and analyzed in any machine, and also to export
them as MIDI files, so they can be listened to, or further computed as data.
Furthermore, the software possesses a humanizer that can be set to several preset
styles so that the rendition sounds as if it was played by humans, adding some
subtle stylistic and gestural conventions, and not be strictly mechanical.
The Database consists of several metadata fields that can be updated, trans-
formed and changed easily. Beyond the identifier number we have assigned slots
to insert common names and designations for each fado, sources where they were
found, author(s), composer(s), date(s), philological and personal commentaries,
analytical notes and formal schemes, as well as the XML, the PDF and the MIDI
files themselves, so anyone can have an instant “global view” of the fado when
navigating through the database. This database comprises what we call a corpus
of urtexts, since all fados were transcribed as close as possible to the original fac-
similes sources. No humanizer preset was used in this corpus to preserve the data
as raw as possible.
32
3.3.3 Uniformity Operations and the Edition Corpus
After the initial database was concluded, we began the next step of our method:
the preparation and exploration of the material. This preparation consisted in
standardization processes that allowed us to classify the materials as equivalents.
Usually, if the pre-analysis operations is successfully concluded, the analytical
stage is nothing more than a systematic and fastidious process using previously
defined procedures (Bardin, 2009, pp. 99-100).
The first method applied, in this case, was a philological method, borrowing
from textual criticism techniques for clearing up the document from textual errors
and incongruences. There were several ways to do that, but the best one was
perhaps an adaptation of eclecticism (Hartin & Petzer, 1991, pp. 47-53) and
internal evidence strategy: we have looked for evidences coming from the text
itself independently of other similar documents or from the physical characteristics
of the document. As a matter of example, we can say that recurrence within the
document was a very good clue: if there are two or more similar or mirrored
melodic themes, we expected them to behave in the same way; but if in one or two
occurrences of the same theme we found certain notes, and on the other we found
a different note, and especially if this note was outside the scale, we have assumed
that this note was out of context and it was a textual error. We have also applied
external evidence regarding not exactly identical documents, but documents that
were expected to be in the same cultural tradition and respect the same tonal
conventions. In our specific case, one can assume that the fados from this period
all behave tonally and with very simple harmonic functions, so when in a given
document a note or a chord is found that does not conform to this tradition, and
if this note and chord is not recurrent within the document, there is no reason to
believe that this was an avant-garde fado, but instead that it is a textual error. We
have applied the same principles regarding rhythms and even structural forms: we
expected phrases of four or eight measures, for instance, so when we saw a phrase
with nine measures in between others of eight, we assumed that it was a textual
error.
Several fados were notated in quaternary meter or in binary meter but with
durations doubled, originating sections of sixteen bars long, therefore we have
standardized all transcriptions into the same time signature and beat unit as
Ernesto Vieira’s model. In this case, binary tempo with a quarter note as a unit.
33
At this stage we have defined our context and our recording units (Bardin, 2009,
pp. 104-108). Since we were not sure of the ontological dimension of a fado,
our context units were not the fado itself, but each one of the occurrences in the
literature per se, even if there were two or more with the same designation or with
very similar content. Those were the best references available and, at the same
time, the most reliable units that contained a micro-universe that could be studied
and analyzed. The recording units, on the other hand, varied accordingly to our
hypothesis.
The first one we tested, Ernesto Vieira’s definition, called for the comparison of
phrasal units (melodic contours) that became symmetrical and comparable within
the universe of each occurrence, so the best method was to consider each one of
these “melodic contours” or musical phrases, that usually were correlated directly
with lyrical content6 , our recording units.
We have then analyzed each occurrence in the database and performed formal
processes conducting to the stylization of each one, turning them into a structural
matrix. Much of this process can be described formally using specific techniques
coming from musical analysis, namely
“Semiotic analysis: analyzing a piece of music semiotically means,

first, chopping it up into units possessing some degree of significance
within the piece; and, second, analyzing the way in which these are dis-
tributed throughout the piece, with a view to discovering the principles
that govern this distribution” (Cook, 1994, p. 151).
“Semiotic analysis proceeds through a number of stages and (. . . )
the initial stage is segmentation. (. . . ) Really it is patterns of recur-
rence that determine where one motivic unit ends and the next begins;
recurrence is in other words the principal criterion on which the process
of segmentation is based” (Cook, 1994, p. 152).
6
When we are dealing with a musical score, which has its basis on a narrative (we must take
into account that the music is the background of a story, a tale, a fate, a fado), the lyrical content
is a crucial element to define the musical structure, at least of the vocal part. It seems clear by
now that each musical phrase, or melodic contour is based on a lyrical line following prosodic
rules. So, the recording units are, in fact, connected to the poetic structure of the fado, but
as we are discarding this information when we are dealing with raw musical scores, sometimes
that connection seems invisible. It must be said, however, that in case of any doubt regarding
when or why a musical phrase should be segmented we have always looked up for the original
lyric associated with the source and have made our best decision taking that lyrical content into
consideration: in the fado, the melody serves the word (if yet we need to recall an old dispute
between Monteverdi and Artusi).
34
And so, we have applied this methodology to the database identifying where
the recurrences occurred (the recording units) and we were able to segment the
material identifying the formal sections of each instance. We have added double
bars in each transition point so they are visually identifiable for any reader. We
have called the corpus at this stage the [critical] “Edition corpus”, since it was
critically edited and it is the most complete and standardized possible corpus still
retaining all the original information from the sources. Since many original sources
lack articulatory or dynamic instructions they seem somewhat distant from the
actual practice and their sonic output is rather mechanical and dull. Therefore,
in this case, after some trial and error testing, the fados were rendered using the
humanizer preset “marching band”, in the expectation that these MIDI files in
containing subtle articulatory and gestural variation better reflected a human per-
formance. We have also used this version of the corpus in many of the calculations
and as a source for values and comparative analysis.
3.3.4 The Melodies-only Corpus
While the Edition Corpus is perfectly adequate for comparing form, texture,
harmonic material, range, and a series of relevant parameters, it is not a good cor-
pus to actually compare the melodic material of each fado, because of disparate
criteria in formal units, recurrences, size, keys, metrics, etc. As such, we have
“cleared” the extra-material that did not conform to the basic structure – what
one could call the idiosyncrasies of each fado – namely, introductions, stylish inter-
mezzos, codas and finales, and even ornaments that were caught and written down
by the transcriber at the time. We have also eliminated all the accompaniments
(ostinati and harmonic material, eventual bass lines). Finally, we have transposed
all fados into the same key – either C Major in the case of major fados, or A minor
in the case of minor fados. When we did this we were, no doubt, cutting parts of
the fados that are essential to each one. However, we do believe that we were not
cutting what it was the most essential (for comparative purposes) and what gives
a fado its identity as a group, and plus, what should be common and comparable
between all of them. The statistical comparison between the melodic elements of
the corpus would not have been feasible without taking this crucial step. However,
we ought to pay attention to the fact that
35
“the most inductive way of beginning an analysis is to look for

patterns of recurrence. If you are going to be genuinely inductive,
then you need to begin with a relatively full version of the music (. . . )
since any more drastic initial reduction of the music will necessarily
incorporate a priori judgements as to what in the music is essential
and what is not.” (Cook, 1994, p. 208)
So, it is vital to be able to assume why and how the divisions and cuts were made in
order to preserve the integrity of the corpus and the reliability of the investigation.
“This first stage of a semiotic analysis is known as a paradigmatic

analysis because it consists of extrapolating the units of significant
structure in music. (. . . ) The second stage of the analysis is known as
syntagmatic analysis and here attention returns to the temporal aspect
of the music” (Cook, 1994, p. 165).
All these procedures were, of course, manual since no computer program up-to-
the-date was able to tell, within a MIDI file, when exactly a musical phrase begins
or ends and why a certain measure is to be seen as a textual error, for instance.
There are algorithms and programs of segmentation being done and in develop-
ment, namely see (Melucci & Orio, 2002, Paulus & Klapuri, 2006, Stammen &
Pennycook, 1993, 1994, Wiering et al., 2009), but not with the required precision
and reliability necessary to work in an effective and timely manner for our pur-
poses. However this is no problem to the trained human ear and eye, which can
easily correlate the lyrical content to the melody and also the different melodic
contours between themselves and assume and justify why the segmentations oc-
cur in those exact spots. At this stage we had a third corpus that was called
“Melodies-only”.
3.4 Music Information Retrieval
Having “normalized” and “typified” the fados (occurrences), and flattened them
into their matricial structures, we started looking for melodic or harmonic patterns
to see if there were relevant hypothesis to be tested. We then built a spreadsheet
using the relevant data from our hypothesis, using the MIDI parameters. Thus we
36
3.4. Music Information Retrieval
entered a new stage of our method: the treatment of the material already prepared
in the previous stage.
The first step of the treatment was the codification. It corresponded to a trans-
formation – made according to specific rules – from the raw data, which allowed
the material to be represented in units that would describe precisely the charac-
teristics lying in the material (Bardin, 2009, pp. 103-104). In our specific case, the
codification was made using MIDI language. The international MIDI code con-
verts all the parameters that lie in the fado transcriptions, in the literature, into
numbers. With those numbers it was possible to perform all the statistical opera-
tions needed to infer patterns, rules or laws that govern the style with which they
were created. We have no illusions that the MIDI code obtained omits vital pa-
rameters that usually appear in a human performance and that were surely present
in the fado performances of the nineteenth century (namely the gestures). “Any
transcription, especially a transcription as drastically simplified as conventional
notation, constitutes an interpretation of what is heard” (Cook, 1994, p. 225).
So, we were working with a defective and subjective source of the performance
being studied, however, MIDI codification strictly respects all the parameters that
were in the literature occurrences. And since we barely had anything more than
that, we were not losing any more information by introducing the codification at
this stage. We were then apt to use a comparative method (Cook, 2004). In the
comparative method
“the analysis was based on the relative frequency with which differ-
ent intervals occur in a single musical line. (. . . ) The other important
thing to consider in making an analysis of this sort is the selection of
the basic data. (. . . ) the sample must be reasonably large” (Cook,
1994, p. 189).
We have already mentioned the importance of the corpus and the preservation of
some analytical principles when of its constitution (exhaustivity, representativity,
homogeneity and pertinence), but one cannot stress enough that if the corpus was
not well formed or if it lacks sufficient data, the validity of the results would be at
risk. So it was absolutely crucial to know when and why to stop the building of
the corpus (since the job must not be endless also) in order to achieve optimum
results of the statistical operations performed.
37
This method has a past history. For instance, John Blacking applied it to the
Venda Songs, a tribe in Africa (Blacking, 1973).
“He compares the various songs of this repertoire with each other in
the hope of discovering both what it is that they all have in common,
and the rules of transformation in accordance with which the same
underlying structure can result in a large number of apparently quite
different songs” (Cook, 1994, p. 204).
We also have to take into account, for instance, the studies of Kolinski regarding
melodic contour and its models to statistically characterize popular musical genres
(Kolinski, 1965, 1973, 1982).
“It discovers an underlying structure in relation to which the mu-

sical surface can be viewed as a kind of elaboration. What is more, it
is objective in a way (. . . ) that there are clear and explicit rules for
doing the analysis so that any two people should come up with the
same result” (Cook, 1994, p. 208).
Also, regarding melodic contour we should take Adams (Adams, 1976), Morris
(Morris, 1993) as references besides, of course, the extensive tonal theory of per-
ception developed by Lerdahl and Jackendoff (Lerdahl et al., 1996) whose theories
led to in-progress software development of applications capable to make an ana-
lytical approach with MIDI files and real sound samples by the Kinetic Controller
Driven Adaptive and Dynamic Music Composition Systems project (Guedes et al.,
2009). All these approaches of the methodology seem valid with their merits and
disadvantages, and we were influenced by several combinations of them while test-
ing and experimenting various parameters within the fado occurrences. We have
also read and been influenced by the methods and specific techniques of statis-
tics applicable to music described by Beran (Beran, 2004), a general reference, as
well as the Empirical Musicology compilation (Clarke & Cook, 2004) and several
articles from the Journal of New Music Research.
As soon as the database was concluded, there was a long period of experimenta-
tions and tests implementing several of these analytical techniques and statistical
operations in order to make enough inferences to provide us with a relevant set
of rules that govern fado. The simplest and most immediate choice was to use
38
a spreadsheet and manually retrieve information that we believed could not be

easily retrieved automatically – namely the formal and harmonic analysis and the
segmentation process much prone to contextual issues leading to ambiguity and
errors. Therefore, these two sets of operations were done manually using all our
previous skills in music theory and analysis.
Following a suggestion of Elad Liebman, we then explored the several Auto-

matic Music Information Retrieval tools available in order to find other suitable
alternatives for other kinds of parameters. These were the MIDI Toolbox (Eerola
& Toiviainen, 2004), available to install in Mathematica, or the Music21 7 , a toolkit
for computer aided musicology assembled by MIT, that is fairly accessible to use
and program in Python. We started programing with Music21. However, after
an initial period of experimentation, we were not satisfied with the steepness of
the learning curve versus the efficiency of the results, and instead discovered the
set of tools developed by Cory McKay, the Bodhidharma and the JMIR (McKay,
2004, 2010). The Bodhidharma was of particular interest and ended up being our
choice based on usability, ease of the learning curve, flexibility and customization
of the tests and corpus and ability to import data and export results – not only the
program has a very intuitive and nice graphic user interface, but it is also able to
import and perform calculations natively in MIDI files without further conversions.
Furthermore, it allows us to decide precisely which series of tests and parameters
we can calculate, and export them instantly into our spreadsheets. It can also
be trained to perform automatic music classification based on personalized corpus
using specific characteristics that we can determine, which is a very useful feature
both to verify the consistency of the initial corpus and to perform some evaluation
tasks. Having all this in mind, we decided to invest our time in this particular
program and most of the tests and features we have calculated were retrieved us-
ing this software. The precise algorithms are described on McKay thesis, his text
being our source for the definition of each parameter (McKay, 2004).
Already in the final stages of our work, we have had access to a new software,
MeloSpyGui 8 , which is a library, implemented on top of music21, with a friendly
graphic-user interface. This library currently being developed in the core of Jaz-
zomat Research Project, directed by Dr Martin Pfleiderer, at the Hochschule für
Musik Franz Liszt Weimar (The Liszt University of Music Weimar), is able to
7
http://web.mit.edu/music21/, accessed June, 6, 2015
8
http://jazzomat.hfm-weimar.de/, accessed June, 6, 2015
39
retrieve several musical features from monophonic melodies in both EsAC and
MIDI formats. We have successfully tested and used it with the “Melodies-only”
corpus (since it is unable to handle polyphony) mainly to complement our work
with additional features regarding melodic contour.
3.4.1 Testing the internal consistency of a corpus
In order to evaluate the consistency of the repertoire contained in the corpus

we have inspired ourselves both on a study conducted on Cretan Songs, using
music information retrieval systems (Conklin & Anagnostopoulou, 2011), and on
very recent attempts at automatic music classification of fado based on sound
recordings (Antunes et al., 2014). We have devised an objective and reproducible
test to perform evaluation tasks using the same software we have used to retrieve
information: Bodhidharma.
Cory McKay’s software Bodhidharma is able to extract up to 111 different fea-

tures from MIDI files. These features are divided into seven different categories:
instrumentation, musical texture, rhythm, dynamics, pitch statistics, melody and
chords. Instrumentation retrieves information relative to the possible 128 pitch-
instrument patches plus the 47 percussion instruments in the percussion key map,
available in any general MIDI type 1 file. With this class of features it is possi-
ble, for instance, to know the prevalence of certain timbres over others. Musical
texture features complements this set. One is able to retrieve data on channels
and tracks, and thus infer statistics on voices and how they behave between each
other. Rhythm features represent a set of operations along the time dimension.
The software is able to calculate statistics on pulse, note durations and densities.
Dynamics refer to features that rely on the parameters associated with note ve-
locities and volume channel messages. Pitch statistics are a set of features based
on the possible 127 pitches each note-on can have. All notes derived from channel
10 (usually associated with percussion) are ignored. With this information one
can extract information usually related with tonalities and scales. While these
statistics infer information on pitches as a whole, the category of melodic features,
on the other hand, provides us with information on the horizontal dimension –
the relation between those pitches, namely intervals, type and direction of motion,
and melodic arcs. This later set is complemented by another dealing with infor-
mation on the vertical dimension of pitches, namely the type of chords used. Not
40
all features have been implemented and some of them are multidimensional which
causes difficulties in running tests with them. For that reason we have opted to
choose only the one dimensional features that have been implemented, in a total
of 71. We refer back to chapter 5 of Cory McKay’s dissertation for a complete list
and discussion of all these features (McKay, 2004, pp. 55-68).
Cory McKay’s software was tested using two different classification taxonomies.
First, a reduced one for faster evaluation purposes and for comparative purposes,
with other similar studies that were also based on rather small classification tax-
onomies. Afterward, an extended one, more suitable for real-life purposes. The
reduced taxonomy comprised originally three root categories: jazz, popular and
western classical. Each of this subdivided into three leaf categories. In total the
reduced database consisted of 9 taxonomies with 25 recordings each, for a total
of 225 MIDI files. For our purposes we added the fado category (and respective
MIDI files) where we thought it would make sense, inside the popular category,
next to country and rap, resulting in a total of 10 taxonomies for classification.
Table 3.1: Classification with 10 taxonomies
Jazz Popular Western Classical

Bebop Rap Baroque
Jazz Soul Punk Modern Classical
Swing Country Romantic
Fado
The extended classification taxonomy comprises a total of nine root categories

and several sub-categories to a total of 38 taxonomies, to which we have added the
fado one. This database has the same 25 recordings for each taxonomy leading to
a total of 950 MIDI files, plus the fado MIDI files we have added.
According to Cory McKay:
“the expanded taxonomy developed here (. . . ) made use of an

amalgamation of information found in scholarly writings on popular
music, popular music magazines, music critic reviews, taxonomies used
by music vendors, schedules of radio and video specialty shows, fan
web sites and the personal knowledge of the author. Particular use
was made of the All Music Guide, an excellent on-line resource, and of
the Amazon.com on-line store. These sites are widely used by people
41
Table 3.2: Classification with 39 taxonomies
Rock Jazz Rhythm and Blues

Classic Rock Bop Blues
Blues Rock Bebop Chicago Blues
Hard Rock Cool Blues Rock
Psychedelic Fusion Soul Blues
Modern Rock Bossa Nova Country Blues
Alternative Rock Jazz Soul Funk
Hard Rock Smooth Jazz Jazz Soul
Punk Ragtime Rock and Roll
Metal Swing Soul
Western Folk Modern Pop World Beat
Bluegrass Adult Contemporary Latin
Celtic Dance Tango
Country Blues Dance Pop Salsa
Flamenco Pop Rap Bossa Nova
Fado Techno Reggae
Smooth Jazz
Western Classical Country Rap
Baroque Bluegrass Hardcore Rap
Classical Contemporary Pop Rap
Early Music Traditional Country
Renaissance
Medieval
Modern Classical
Romantic
with many different musical backgrounds, so their systems are perhaps

the best representations available of the types of genres that people
actually use. These two sites are also complimentary, in a sense. The
All Music Guide is quite informative and well researched, but does not
establish clear relationships between genres. Amazon.com, in contrast,
has clearly structured genre categories, but no informative explanations
of what they mean.” (McKay, 2004, pp. 81-82)
The discussion of musical categories vastly transcends the scope of this work, so
we are using these taxonomies as they have been presented, and using them as
a starting point to demonstrate the potentials of our methodology. For a more
in-depth discussion of musical categories we refer to the work of Fabian Holt (Holt,
2007).
42
3.4.2 Test Implementation
Intuitively the corpus seems rather dispersed since 48 fados provide from the
same homogeneous source (César das Neves’ Cancioneiro (Neves, 1893)) and all
others are individual folios transcribed by several different people, spanning a
rather large period. So in order to obtain a consistent training set from these two
different universes our training sample consists of the first quartile of the corpus
(roughly half of the fados from the Cancioneiro) and the third quartile from the
same corpus (half of the fados from the individual folios). By having this criteria,
we consider to have obtained a coherent sample representing the entire corpus.
In the first trial we have trained 50 MIDI files from the Edition Corpus using
the implemented, one dimensional, 71 features. The trial was executed using
ten taxonomies (the original nine taxonomies from the database of 225 MIDI files
assembled by Cory McKay plus the new fado taxonomy, with 50 recordings, created
under the sub-category folk in pair with country and traditional). The resulting
classification of the remaining fados was a success of 100%. All fados from the
second and fourth quartile of the database were successfully classified as fados
achieving very high scores (min 53,8%, max 98,4%, avg 83,4%). These outstanding
results led us to think that they were actually too good to be true and we suspect
that some bias coming from encoding might have led to this outcome: namely all
trained fados in the Edition Corpus were encoded using Finale!“marching band”
humanizer style, most of them using the same 2/4 time signature, a base tempo of
64 beats per minute with little fluctuation, and very few dynamics when compared
to a probably much wider range of options coming from the rest of the MIDI files
of the database. If we interpret critically the results of the previous study on
automatic classification of fados performed on sound files (Antunes et al., 2014),
we have a precedent of fado being easily recognizable by a very distinctive set of
specific features, namely instrumentation and range. While these characteristics
might seem to help the task to be easily performed, in fact, in the end they do
not help that much because we do not extract valuable information regarding the
possibility of fado having unique melodic, rhythmic or harmonic traces beyond
those features.
In order to overcome this problem, our solution then was to try to eliminate the
strong bias those specific features bring. So, we decided to use the urtext corpus
instead with all the small errors, incongruences, and different time signatures it
43
contains. We also encoded it without any humanizer preset. And finally we

decided to train the database using only 49 features: the one dimensional ones
directly linked to rhythm, melody and harmony deliberately discarding all features
related to timbre, instrumentation, pulse, dynamics and even time signature. Our
main concern was to indeed try to understand if the main parameters involved in
the original transcriptions, by themselves, could represent a corpus of a coherent
genre and if fado could be, in the extreme, recognized and classified only based on
rhythm, melodic and harmonic features.
For the second trial we repeated the process of using 10 taxonomies and the
same 225 auxiliary samples used in the first trial, but this time with 50 recordings
from the urtext corpus and the lesser amount of features. The result was still sur-
prising: 48 fados were successfully classified as fados, which indeed indicates the
strong coherence of the genre despite our intuition of a very fragmented corpus,
with still very high scores (Min 54,2%, Max 97,0%, Avg 80,7%). The two exem-
plars not classified as fados were clearly outliers for very good reasons. Number
076 is one of the fados originally in 3/4 and, as it resembles a light waltz configura-
tion, was classified as baroque with a score of 31% or romantic with 23,3%. These
low values, none of them achieving even near the threshold of 50% just show how
the system was confused by this particular piece, mixing traces from several gen-
res. When one looks at this particular score, in its urtext version, one can easily
understand why the system was lead into classifying it under the umbrella of the
classical genre. The second outlier was work number 091, which was classified as
bebop with 42%. The low score indicates also a great uncertainty about this par-
ticular work. When one looks at the score, one easily understands the ambiguity:
it is a fado with very peculiar and complex contrasting rhythm figurations, chro-
matic passages, ornamental figurations, and it uses at least four different voices.
The sophistication of this particular arrangement (which is also one of the most
recent ones) when compared to the rest of the corpus clearly singles it out. Still,
we have to add that both these works must have something of fado about them,
since they were not clearly classified as any other genre, remaining in a kind of
sweet ambiguity.
In our third trial we opted for refining the process and used the same urtext
corpus with the same 49 features based only on rhythmic, melodic and harmonic
traces. However, we used 39 taxonomies and the total database of 950 MIDI files
plus 50 fados as a training set. We wanted to know if increasing the complexity
44
and array of possibilities the system would get confused by them and the peculiar-
ities of the database would start to be singled out more often. The results show
that 45 fados were still successfully classified as fados with very relevant scores
(Min 46,3%, Max 96,8%, Avg 75,7%). Although a success rate lower than in the
previous tests (as expected), this seems to show how consistent the corpus is after
all, contradicting our initial intuitions. Two of the fados, works 044 and 050, were
wrongly classified as renaissance with high scores (65% and 75%). Our best guess
for this fact is probably the relevant percentage of ornamental figurations and em-
bellishments appearing in these scores. Works number 091 and 097 were classified
as unknown. This means that no taxonomy was able to gather any relevant score
and just shows how incredibly ambiguous these works are – number 091 being
recurrent in the previous test too. Number 088 seems also with a very strange
and ambiguous classification – the software chooses to classify it as ragtime with a
score of 51% but also as fado with 86,3%. This apparent contradiction comes from
the fact that fado is a sub-category from the larger category folk, while ragtime
is a sub-category of jazz. When the system assumes a first classification under
the umbrella of jazz, then is has difficulties in finding a suitable sub-category. On
the other hand, if the system as a second choice opts to classify it as folk, then it
clearly nests it on fado. The same ambiguities seem to arise with 079, since the
systems classifies it as fado with 87% but points out how ragtime has a strong
score of 63,3%. A closer look at the scores of these works reveal how, in fact,
many idiosyncrasies typical from ragtime are present, hence the ambiguity. We
also have to point out how work number 041 also created doubts: although it was
classified successfully as fado it only achieved a score of 51%, displaying a second
choice of country blues at 37,7%.
In all of these trials one can notice how the fado corpus is actually larger
than the other taxonomies. Since the original database was provided with 25
samples for each taxonomy we also wondered what would happen if we trained
the fado taxonomy with 25 samples as well in order to not create a bias in this
respect. Therefore in our fourth trial we kept the same conditions of the third
trial. However, we trained the classifier with 25 random fados from the set and
then asked the software to classify the remaining 75 ones. We were expecting that
this could add some confusion since now there was much less initial information
available to create a fado profile. If indeed the corpus was not that much consistent,
among itself, the results could be more varied. However, we felt that this step had
to be taken in order to eliminate any bias created by over-representation of fado in
45
the training stage. The results of this trial almost confirmed our expectations: the
system was indeed much more confused and was much less effective in recognizing
as fado the 75 remaining samples. Nevertheless, the classifications did not spread
randomly among other taxonomies; instead, they were highly concentrated around
very precise taxonomies. The system successfully classified as fado 37 samples
without any doubts (49,3% of the total), and additionally was very divided on 14
ones classifying them either as fado or as country blues (18,7%).
This double classification happened because none of the taxonomies got a score
above 50%, with two or more of them being very similar. Sample 022, for instance
got a score of 43,9% as country blues and 44,9% as fado, while sample 033 got
scores of 41,3% and 41,2% in those same taxonomies. If we add up the 14 samples
that were classified as possible fados to the 37 ones, we get a total of 51 samples
successfully classified (68% of the total). Then we have 20 other samples classified
as country blues representing 26,7% of the total. If we add up the 14 cases where
the system could not decide between fado and country blues, we get a total of 45,3%
of the samples being possible country blues songs. If we consider both clusters, we
realize that both categories encompass 57 possible cases (76% of the total), which
indeed shows that there is some consistency among the corpus, even if it is not the
ideal one. In 11 samples the system returned unknown (14,7%), meaning that no
taxonomy could create a relevant score. This shows that although these samples
might not have been recognized as fados, they were not recognized as any of the
other possible taxonomies. This either leads us to believe that some unique traces
were not grasped by the training set and that they are only available in these
particular samples, or perhaps that they are indeed peculiarities of fado and some
of these samples were needed as trainers. As we have seen in the previous trials
with more samples this was not a relevant problem. Five samples were classified
as ragtime (6,7%) which is also a trait present in previous classifications and not
a surprise. Finally, sample 088 was classified as romantic (at 40,3% score) and
sample 054 as either bluegrass (32,8%) or country blues (32,8%). We might add
that this last trial represents what our intuition expected from this corpus – most
of the works being slight ambiguous and presenting many traces from several
different styles, and thus confusing the software. However, what the tests have
shown is that our intuitions were indeed exceptional and that the rule was the
software having no doubt whatsoever about the true classification of the works
present in the corpus, even when using a non-formatted and non-homogeneous
version and only relying on harmonic, melodic and rhythmic features. We consider
46
the deviations on the last trial were rather homogenous instead of dispersed, and
the confusion between fado and country blues is rather expected given that we
are only analyzing rhythmic, melodic and harmonic features. These two musical
practices share common traits since they pertain to the umbrella of folksongs,
what we would call a “daily life singing stories” tradition. We speculate that if we
had taxonomies for ballads, romances, modinhas or mornas the confusion might
add up. Also the confusion between ragtime might be expected since all the fado
samples are written as piano solo folios following much of the same conventions of
the ragtime piano solo folios of the time.
In conclusion, we believe the method proposed, although experimental, has

shown to be useful as a way to know more about the internal consistency of
any given corpus and also its “families” and look-alike taxonomies. Moreover,
it reinforces our idea that it is also useful as an evaluation tool: it may allow
us to know if any sample modeled out of a given corpus will be very different
from the samples of that same corpus. If the database is large enough and has
a relevant number of taxonomies, and if any given sample is very different from
the taxonomies trained, then it will either be classified in the wrong taxonomy
or be classified as unknown. What we would propose as future work, then, is
a reinforcement of the database. In an ideal scenario, many more samples and
taxonomies would be present in order to have a more adequate representation of
the world’s musics.
3.4.3 A method for comparative data visualization
By themselves, the statistics retrieved on the fado corpus are useful to under-
stand the constraints regarding genre and are a resourceful material that provides
concrete values for the generative model. However, they lack contextual informa-
tion in relative terms. It is not easy to make interpretations or draw meaningful
conclusions until we have comparative references to guide us. Therefore we took
advantage of the database provided by Cory McKay and decided to use it as an
experimental sample to represent the world’s musics. We are aware of the limi-
tations and the narrow scope of a sample comprised by 39 taxonomies and 1050
samples, but on the other hand, given our limited time and resources, it seemed
to be a good compromise for a starting point.
47
While we already had some preconceived ideas about fado and its place within
the context of the several world’s musics, we completely lacked empirical data to
support our thoughts. Our idea was to design a methodology based on what nat-
uralists were doing when collecting different animal species and observing them
and comparing each other. These ideas of using recordings of music and compar-
ing them has a long tradition in Ethnomusicology, but the problem is that most
of it was done manually, without the resource of computers. The emergence of
computers and the possibility of having large databases of music in digital format
allowed the resurgence of systematic ways of analyzing music.
With this experience, very much based on what is done in the exact sciences,
we hope we have achieved a systematic, reproducible way to show the relevance
of the data gathered, and also empirically demonstrate what the place of fado is
in relative terms.
The first step was to use the software Bodhidharma to retrieve the feature data
from all the MIDI files. All the MIDI files were already tagged and organized ac-
cording to their name and taxonomies. In order to better detect and visualize the
possible bias created, we decided to include the raw data categorized as “fado” and
then include the edited data (the homogenized corpus following Ernesto Vieira’s
model) as well, categorized as “Fado Edition”. This way we could also see, in rel-
ative terms, to what extent the encoding and normalization process would change
the perception of the corpus. Once the data was retrieved, it was copied into a
spreadsheet where it was further trimmed – namely all the multidimensional fea-
tures (for complexity and visual reasons) were ignored. Now the main question
was what could be done with such a large amount of data? A spreadsheet con-
taining thousands and thousands of numbers is nearly useless if one is unable to
make sense of the data, therefore we had to devise a plan to be able to visualize
the data in an almost immediate and meaningful way.
The clean and organized data was then prepared to be imported into R!, a
statistical open source programming language. The intention was to find a nifty
way in which to display the data that would be visually informative in comparative
terms: we were going along the same way naturalists look into a board with the
pinned butterflies and can immediately see their differences and similarities. We
wanted to look at a graphical display and immediately grasp what would be the
relative place of fado among all other taxonomies, but at the same time understand
how the other taxonomies relate among themselves. We also wanted to have all
48
relevant information present in order to understand each unique taxonomy in case

it was needed. Therefore, side-by-side box-plots of all the taxonomies was the ideal
solution. We decided to sort the box-plots by the median value in order to convey
the relative place of each taxonomy visually in an immediate way. The decision
between the median and the mean was not an easy task, because sometimes one
does not know for sure what the most meaningful value is. While in most exact
sciences the arithmetic mean is actually the norm, among human artifacts outliers
are frequent and can greatly distort some parameters. Hence, we have decided to
sort the taxonomies based on the median but we have included the mean value
as well. By having both values included in the graphic, and paying attention
to its differences, one is also able to deduce information about possible outliers.
In order to better understand the significance of the position of fado regarding
the remaining recordings we have also decided to provide lines representing global
median and mean values. Notice that since fado is over-represented in this sample,
not only because it has four times more recordings, but it is also doubled because
it has two corpus, we have excluded the values of fado for the calculation of the
global median and mean in all cases.
Since all fado samples in this corpus are encoded as solo piano pieces the statis-
tical comparison of features relative to instrumentation and musical texture made
no sense in this context, since they do not represent the usual and most common
practice. Therefore, they were ignored in this particular descriptive experience.
Following this method we were able to generate a series of figures, representing the
empirical data, that will be extensively used along the dissertation to illustrate,
reinforce or disapprove the characterization of the musics and sounds of fado.
49
Chapter 4
On the concept of fado
Fado is a complex “total-art form”. One cannot just have the lyrics, or just
the music, or just the performer. One needs to have the whole set of elements
for fado to happen. Both lyrics and music are regarded as a base for the fadista
to create on top of them. Manuela de Freitas points out that the ancient Greeks
provide the foundation for this distinction in the concepts of Tragedy and Music,
fado resembling Tragedy, since it is something more than music (Freitas, 2013).
Fado can be understood only in the context of an empathic performance, of a
narrative that has to be embodied, suffered and then conveyed to an audience
through the use of several resources that we will try to analyze further down the
text, involving the styling of a melody, several body gestures and postures and a
fluent interchange with the musicians. The audience is then expected to empathize
and experience the emotions, affects and moods being communicated to them. It is
because of this whole set of characteristics that need to be present that it is argued
that recordings are not the same as live fado, because recordings cannot capture
the image and all the language conveyed by the body. Through the recordings we
lose valuable information.
Seeing fado as a “tragedy” or a theatrical display of the narrative, the unique-

ness of each performance, extensive to the exclusivity of the repertoire, is another
value largely pointed out, especially among older fadistas. “Fadistas talk about
singing their lives and their stories as individuals, in their own style, with sin-
gularly marked voices” (Gray, 2013, p. 28). They claim, with some nostalgia,
that in the old days each fadista was unique, had his or her own identity and
nowadays all are copy cats. Currently, it is fairly common for several fadistas to
51
sing the same fados, using the same lyrics, and many times imitating the gestures
and styling of other famous fadistas. This is a typical convention among popular
music in general, with the practice of covers, and imitation, of popular artists and
their repertoire. This consequence seems to emerge from a shift in technology
and the massification of popular songs through the means of digital reproduction
(Tschmuck, 2010). Although this has not happened just to fado, fado as a prac-
tice, however, has been extremely affected by it. While until the beginning of
the twentieth century fado was orally transmitted, and there was a convention
and an ethical code that fadistas sang their own narratives, after the advent of
sound recording things began to change. Many new fans and practitioners started
learning the repertoire directly from the recordings and used them as a fixed ref-
erence, as an urtext, the “correct version”. And while in the “old days” it was
an important value to be unique and original, nowadays some fadistas and fans
already protest when they actually hear deviations from the recorded versions,
which seems to be a paradox. The situation today is that we can imagine a con-
tinuum of possible performances, having the maximum number of fixed elements
in one side and a maximum improvisation in the other (using a concept developed
by Salwa Castelo-Branco (Castelo-Branco, 1994, p. 133)), and while nineteenth-
century values seemed to privilege maximum improvisation, nowadays there are
clearly some fans, especially among the younger generations that seem to lean to
the other side. The evolution of the repertoire by itself, with the experimentation
with new structures, and new kinds of pre-composed music, since the fado-songs
presented in the theaters, were already contributing to this shift. It was a slow
and dynamic evolution, in a continuum where all possible combinations still exist.
Just the distribution and how they are perceived, and valued, is changing.
4.1 Modinhas
According to Rui Vieira Nery, in the transition of the eighteenth to the nine-
teenth century, the dominant genre cultivated in Portugal, using Portuguese lyrics,
was a generic type of sentimental song called “modinha”. The term modinha re-
ferred to a series of diverse practices performed in distinct social contexts – it
was applied both to some erudite written songs, with complex melodies and harp-
sichord accompaniment, or to the peasant songs from the Lisbon outskirts, or
even to the popular ballads carried by the blind itinerant beggars. In these latter
52
4.1. Modinhas
cases, the songs were usually described as being very simple melodically and har-
monically, as having a very melancholic character, and as being performed with
a large degree of improvisation, usually accompanied by a viola [the emic term
for the acoustic guitar]. There was no clear boundary between all these perfor-
mance practices, since they coexisted in the same time and space, and they were
constantly inspiring each other (Nery, 2004, p. 31).
When Nery analyses some examples of modinhas published in that period he

recognizes the typical alternation between tonic and dominant, which already oc-
curred in the lundum, and a sad, fatalist and melancholic themed lyricism. The
most curious fact, however, is that in two different printed versions of the same
modinha “Cruel Saudade” authored by Manuel José Vidigal, what varies is the
ornamentation of the vocal line, which shows how it was clearly not fixed and,
in some way, demonstrates how the improvisatory character was already present
(Nery, 2004, p. 33). We have very few documentation regarding transcriptions of
these practices, since they were mainly transmitted orally. However, we have lit-
erary testimonials and descriptions. Nery quotes William Beckford, who in 1787,
describes the modinha as being “an original sort of music different from any I
ever heard, the most seducing, the most voluptuous imaginable, the best calcu-
lated to throw saints off their guard and to inspire profane deliriums” (Beckford
& Alexander, 2005, p. 54). Nery also points out how many modinhas being sung
in the streets of Lisbon were coming from Brazil, due to the intense commerce
and social relationships between Portugal and the colony. Sometimes the music
was very simple, other times poets like Tomás António Gonzaga or Caldas de Bar-
bosa would lend their texts to erudite local composers, like Marcos Portugal, who,
would then try to compose more complex songs, nevertheless, searching inspira-
tion in the popular melodies. These refurbished and re-elaborated versions would
then reach the salons and higher classes, since they suited the new audience and
were an acceptable repertoire for the young ladies to play and sing them at the
piano.
This curious mix between the popular and the erudite draws the attention of
the foreigners who wrote about it: Nery quotes William Kinsey stating that the
Portuguese should not try to imitate the Italian style arias because the modinhas
and their own popular songs were good enough to elicit tears and strong emotions
among the audience. Nery also draws our attention to the fact that already in
these descriptions one could see a certain concern for the national themes and the
53
importance of a “national identity” portrayed by the Volksgeist – what emerges

from the people (Nery, 2004, pp. 32-35). These facts just show us that, although
the word “fado”, according to our known sources, was not being used at the time
to describe these phenomena, everything else was already in existence. This just
reinforces the idea of how the category was not simply shaped out of thin air,
but that it was dynamically attributed to a series of contexts that were already
taking place. We have very little doubts that some of these modinhas, especially
the popular ones, improvised with a guitarra, and sung over a melancholic text,
had plenty characteristics of the early fados, even if they lacked the name.
4.2 The influence of the Dictatorship
The dictatorship that governed Portugal between 1926 and 1974 was responsible
for huge transformations in fado. Fado was shaped and typified as a social ritual
representing not only proper entertainment, dignified for touristic purposes, but
also a form of nationalistic propaganda. The regime clearly understood how
“music is a form of ‘affective communication’ and “is able to con-

vey meaning and values (. . . ) which can shape patterns of behavior
imperceptibly over time until they become the visible background of
real political activity. (. . . ) music works as a public form of knowl-
edge and mode of understanding which is shared by vast numbers of
people across the world, those of us who know that, while music is
not a ‘universal language’, the meanings of musical sound are not that
‘indecipherable’ ” (Negus, 1997, p. 221).
During the 1930s the cultural policy was to try to sanitize and minimize the impor-
tance of fado in society. Luiz Moita prepared a series of eight talks, broadcasted
in the Emissora Nacional, and later published in a book entitled Fado: Canção
de Vencidos (Côrte-Real, 2000, p. 100). However, by
“Revising and interpreting exhaustively the literature on Fado, Luiz

Moita helped not only to concentrate on the one hand and to broad-
cast the information about it on the other hand, but also and more
important yet, to establish a number of ‘truths’ mainly related to the
54
4.2. The influence of the Dictatorship
history of Fado, that ultimately served – against his will – to enhance

its significance as an emblem of national identity” (Côrte-Real, 2000,
p. 158).
Following the lines of other nationalist policies in Europe, António Ferro, the
culture and propaganda minister, endeavored to manipulate the several musical
expressions, from art music to popular music. “In all three areas the policy was
to revive the past, obeying the dictator’s orders for national reconstruction based
on ‘the glorious past of the nation”’ (Côrte-Real, 2000, p. 84). One of the main
initiatives was the national contest A Aldeia Mais Portuguesa de Portugal (The
Most Portuguese Village of Portugal), involving villages from all over the country.
The main idea was to create a model for an idyllic rural past through which all
Portuguese citizens could identify themselves and bond as a unified group. The
various criteria to observe dealt with
“natural landscape and topographic physiognomy, furniture and

domestic implements, clothing, popular arts and industries, trading
forms, means of conveyance, poetry, tales, superstitions, games, songs,
music, dance, theater, festivities and other usages. (. . . ) This event
represented the engagement of the regime in the creation of a living
model of preservation of the past in which people were treated as mu-
seum objects, (. . . ) as actors of characteristic or ‘typical’ plays, or
even as toys or ‘puppets’, easy to manipulate” (Côrte-Real, 2000, p.
88).
At around the same time, a vast ethnographic project, headed by Armando Leça,
was being supported with the goal of collecting the music from the rural areas, the
true Portuguese music, headed by Armando Leça. While Armando Leça completed
the project and collected around 500 songs (sound recordings and transcriptions),
the project was not concluded nor edited, probably because the regime could not
find a convincing way to make the results stand out as a popular product, pleasant
for the urban masses, according to their own intents of manipulation (Pestana,
2012). On the other hand, the regime was successful in dignifying fado by applying
censorship to the lyrics and requiring the professionalization of the practitioners
and the legalization of the casas de fado [fado houses]. Gradually, the practice
became more and more standardized according to certain conventions, and even
55
the repertoire began to stabilize. Most of the fados we can find in the transcriptions
from the nineteenth century disappeared and a new derivative corpus of fados
emerged. Nery describes all this process and claims that this formalization took
place around 1926-1945 (Nery, 2004, pp. 188-220). The systematic transcription of
this repertoire was made recently by António Parreira (Parreira, 2014). Although
the regime during this period never expressed any positive discrimination or even
a clear, deliberate intention of promoting fado as the National song, the truth is
that with its laws, rules and interventions, in fact contributed to its emergence.
After the Second World War, fado became gradually more and more important
due to broadcasting. The Emissora National at that period had several orchestras,
two of them dedicated to popular and folk music and a school of singers. Côrte-
Real explains with great detail how the radio “inundated the audience with Fado
transmissions” (Côrte-Real, 2000, p. 159). It was a fertile period for the spirit
policy 1 instituted by António Ferro, in an indirect way, when both the radio and
television were not only broadcasters but also content producers since they could
manipulate and control exactly what kind of content the population would have
access to. On the other hand, in the streets, since many casas de fado were
being controlled and the repertoire had stabilized in a sanitized version, and the
performers were professional, the practice became accessible to virtually everyone,
from the lowest classes to the elites. Some remarkable performers could be seen
and heard by millions with the aid of technology, and so they achieved the status
of stars or divas, and, with the massification of recordings the new generations
would learn the repertoire following their traits and gestures. Amália Rodrigues
at this point became a name of surmounting importance.
Amália Rodrigues, the most famous Portuguese diva, saw herself as a music-hall
singer; she sang basically everything and every musical genre without discrimina-
tion. Fados, however, she found to be pretty dull and repetitive at the time, and
so she felt the need to reinvent the genre, to make changes into the repertoire and
make it “nobler” and “more erudite”. Therefore, she chose to sing literary poems
instead of the typical strophic narratives and surrounded herself with schooled
composers to make new music to sing those poems. Alain Oulman was one of
her favorite composers, and one that composed new songs and melodies shaped
to take the most advantage of her versatile voice. “Both metric and the harmonic
1
The literal translation of the concept “Polı́tica do Espirito” as employed by Côrte-Real.
56
4.3. Fado as Space and Identity
language used by Oulman are deeply innovative regarding the traditional conven-
tions of the genre” (Nery, 2010c, p. 74). Thus, the new songs were not very well
received among the musicians and traditional fado listeners at first. José Nunes,
one of Amália’s guitarra players, used to say mockingly before the rehearsals of the
new repertoire “Well, let’s hit the Operas then!” (Nery, 2010c, p. 75). However,
Amália did not think the same way. She stated:
“The guitarra players, in fact, had to learn all those new harmonies
composed by Alain, which had nothing to do with fado, because fado is
harmonically poor. And I sang them because for me they were fados.
The nobility within them is what matters. Because fado is a condition
of the soul, a song that means fate and carries that meaning inside. If
Alain’s music was deprived of fado I would not sing it” (Nery, 2010c,
p. 75).
This seem to point out to a vision of fado being “something” that can be recognized
in a music or a poem: another layer of information that Amália was empathetically
recognizing as fado and then embodying it in her performance.
By the time of the Revolution what had not been initially planned as such, had
happened – “Fado represents a paradigm of the use and manipulation of expressive
behavior by a political regime” (Côrte-Real, 2000, p. 97).
4.3 Fado as Space and Identity
The transformations and evolution of fado during the twentieth century led to
the concept becoming more abstract. Several sources identify, equate and asso-
ciate fado with a series of philosophical problems, concepts and issues that vastly
transcend music, dance or literature. Context seems to be everything – and also
who, where, when and how. The fact that the nationalistic propaganda of the
dictatorship worked is documented in several sources and it is still visible nowa-
days, a little bit everywhere. It is interesting how the image of Portugal and fado
projected into foreign cultures is so embedded by these myths and conceptions
invented and shaped by the spirit policy of the Estado Novo. One should not be
surprised, then, when the first foreign academic papers and studies referencing
fado reveal a concern about the city of Lisbon (space) and the character of the
57
citizens and their inner feelings towards themselves in the world (identity). Fado is
seen and perceived as a representation of the city of Lisbon2 , and of the Portuguese
people themselves, the association with melancholy and longing, the reference to
the untranslatable word “saudade” and to the daily life of the people in the typical
neighborhoods of Lisbon. Some informants refer to it as “what happens” after a
certain social context – people gather to eat and drink, lights dim, someone sings
a melancholic song; fado “happened”. In a certain way, the original meaning of
fado, “fatum”, was expanded and reinforced, gaining a weight and a preponder-
ance in certain contexts that clearly transcends the performance practice. Fado is
thus seen as an abstraction, a symbolic and philosophical concept.
“Dantas’ representation of Severa’s death, in his novel and in his

play, followed by the refrain from Sousa do Casacão’s ‘Fado da Sev-
era’ has made an impact on the fado novo of the 1930s to 1970s by
presenting the fadista’s death as a national concern. Dantas identi-
fies the fadista as a national figure, because of her associations with
the national song, and thus he has made the Mouraria relevant to
Portuguese outside of Lisbon. (. . . ) Thus the Fado ‘Maria Severa’
perpetuates Dantas’ association between Severa’s death, the death of
Mouraria, and the death of the Fado, to blame the regime for the loss
of tradition, a consequence of the annihilating [of Mouraria neighbor-
hood] progress” (Colvin, 2008, p. 67).
Colvin describes how Mouraria in the nineteenth century was already a web of
social contrasts, and thus the municipality of Lisbon decided to demolish it, along
with the historic riverside in an attempt to hygienize it. The idea was mainly to
create a new image of Lisbon proper for tourism (Colvin, 2008, pp. 69-70).
“The nostalgia of fados novos set in the pre-Republican Mouraria

blurs epochs in the Fado’s diachronic evolution. Anachronisms and
ellipses confuse decades, as musicians of the early twentieth century
become contemporaries of fadistas of the mid-nineteenth. A vague
longing for a lost Mouraria dominates as literary and legendary allu-
sions subordinate historical accuracy. (. . . ) And the fadista resigns
2
One could also claim the particular case of Coimbra and the progressive autonomization of
a second genre there, however this was a much later phenomenon with particular idiosyncrasies
beyond the scope of this thesis. Furthermore, it was observed nowadays Lisbon fadistas singing
repertoire with Coimbra style interchangeably.
58
him- or herself to visiting the ghosts of the alluded Mouraria by con-

sciously reconstructing its nineteenth-century avatars in verse. (. . . )
We toss about the generic expressions severas, marialvas and vimiosos,
thus cementing the artificial folklorization of a fadista Mouraria derived
from a literary referent grounded in historical/biographical testimony.
Novos fadistas of the twentieth century appropriate the history of the
Fado to serve their propagandistic interests; lyricists and fadistas re-
construct an idealized Mouraria characterized by cohabitation between
aristocrats and prostitutes and, seldom, noblewomen and ruffians: a
quarter in which the social hierarchy of a pre-Republican Lisbon does
not exist” (Colvin, 2008, pp. 73-74).
Related to this aspect and to how Mouraria was revamped, Negus also notes how
“Local city councils in various parts of Europe have also considered

using or have used music making as part of a strategy for regenerat-
ing neglected urban areas and to try and bring people back to city
centers, particularly during the night. (. . . ) through the allocation
of resources, local councils have been able to support a diversity of
different interest groups which have in turn been able to create their
own momentum and generate alternative types of cultural practice on
a small scale. In contrast, national governments, faced with the pres-
sures of the nation-state system, have often been far more exclusivist
(nationalist) in orientation” (Negus, 1997, p. 216).
In the end, “songs and music accumulate and connect with new meanings and
beliefs as they pass through time and travel to different places” (Negus, 1997,
p. 195). And the same performance practice earns new symbolic meanings and
functions depending on who is performing, who is listening and where and why
and how.
Colvin thus explains how fado continuously references an idyllic past, a “com-
munal longing”, an expression of an eighteenth-century Lisbon (Colvin, 2008, p.
77). He also describes how the aristocracy, who hitherto had disdained the fadista
class, out of curiosity and caprice invited them into their salons (Colvin, 2008, p.
78). The aristocratic appropriation of fado was another factor to help promote
its reputation and consolidation as national song in the mid-twentieth century.
59
Nevertheless the consequence of its cross-class popularity was a modification of

the practice, since the singers were no longer fadistas. Quoting Osório, the experi-
ence became more staged: “The fado is mutilated and ceases to express a human
condition, a concrete way of life. Its setting becomes more and more artificial,
because the human sub-class that once expressed it starts to disappear” (Osório,
1974, p. 119). In Osório’s opinion, tourism deals the Fado its most wounding, but
not fatal blow, and thus reduces the song to “a wailing, mawkish postcard (. . . )
a whining chant” (Colvin, 2008, p. 84).
The fact that the practice of Fado seems so connected with the notion of a
particular space and of a conveyed typical “Lisbon sound” or “Portuguese sound”
raises important issues. Negus addresses and reviews several studies regarding
the problem of geographies, namely the “Liverpool” and the “Miami” sound, and
proposes three distinct features to approach the issues related to “place-specific
music”: first, how the material circumstances in particular places (such as the
characteristics of the population, entrepreneurial activity and communication net-
works) provide possibilities for and hence contribute to the production of specific
sounds. Second, some of the ways that particular music instruments, rhythms
and voices (using a particular language, accent and words) can be employed to
communicate a symbolic sense of the identity of a place. Third, how listening
involves the recognition and interpretation of how a place is signified musically
and how this is often related to the way that music can be used to construct a
sense of “spatial rivalry” (Negus, 1997, p. 189). As Negus points out, based on the
“Liverpool sound” study, several sounds around the world are connected to a cer-
tain geography through the building of certain narratives and social experiences.
Often musicians build a sense of spatial rivalry through music by contrasting the
characteristics of their own sounds with those from other places. They also ex-
perience the music as “their own” and impose a meaning on the music that is
grounded more in inter-places rivalries than actual sounds. The sounds from a
certain place are explained, by the musicians, through a series of vague and often
mystical quotes and very romanticized and intuitive discourses (Negus, 1997, p.
185).
In Portugal, similar studies have been conducted by anthropologists, sociol-

ogists and ethnomusicologists providing relevant and insightful material. Costa
and Guerreiro, conducted a vast study dedicated to the Alfama neighborhood,
in Lisbon, and realized that one could say that fado, within its limited array of
60
musical compositions, known by everyone, in those neighborhoods, functions as a

support to the broadcasting of all kinds of messages: either the daily news or the
apology of certain social positions, being them conservative or progressive. It is
central to the understanding of fado to see it as a means of making possible the
generalized expression of the lower classes and, on another level, as a means of
ideological propaganda. In Alfama, the ideological expressions conveyed by fado
can only be perceived by the mediation of a core of significations linked with the
social life of the neighborhood and its intricate conflictual-symbiotic relationship
between the rich and powerful, on one side, and the urban plebs (deeply associ-
ated to pre-capitalist activities), on the other. All of this is present in the contrast
that structures the majority of the amateur fado heard in Alfama: the feelings of
rebellion against the rich, but also the envy of a similar situation; the secret (but
sometimes assumed) hope of individual social ascension to a privileged life situa-
tion. In the contrast lies a mourning over a life full of sadness, poverty, insecurity
and dependency, but also the inability to conceive a global change in the social
structure. One desires a better place in the social hierarchy, and this feeling is re-
flected in several distinct ways, either in real life or in the imaginary life expressed
by fado. One of the current practices currently observed is to highlight values like
dignity and courage as means of compensating for the lack of money or political
power. All of this is associated with the fatalism of those unable to conceive an
autonomous project of social change (Costa & Guerreiro, 1984, pp. 184-185).
Therefore, “the enduring appeal of songs through time should not be under-
stood as ‘constant and fixed, not open to an infinite range of interpretations, but
as being founded upon acts of appreciation and judgement that occur within def-
inite social relations’ ” (Negus, 1997, p. 194). In this context, “hegemony” may
be defined as
“how dominant groups not only seek to maintain power through

coercion and persuasion, but also attempt to gain consent for their
leadership by adopting a range of cultural symbols and then connecting
these with their own political and ideological leadership. Hegemony is
actively won and maintained by accommodating and incorporating a
range of dissenting or potential oppositional beliefs and by redefining
these in relation to a particular agenda” (Negus, 1997, p. 195).
61
And so we can talk of a myriad of hierarchies among fado, in both listeners and
performers. There is the political regime expecting to bond and unify the masses
to better control them. There is the level of the street fadistas, the casual dreamers
that only want to express their own feelings and tell their own stories; but also
the professionals that intend to earn money to live and do not even have to feel
what they are singing, being, in a sense, actors. There are also the aristocrats
desperately trying to escape the boredom of their lives, or simply wanting a pretext
to mingle with other social classes, or even just having an attraction for the rustic
and the exotic; but also the tourists coming from afar and expecting authenticity,
an expression of the genuinely local.
James Félix advocates two central ideas regarding authenticity in fado: the
first is that there is no absolute standard or measure of authenticity, but it is
ascribed on a case by case basis and is a personal judgment which may differ from
one individual to the next. A second idea is that, while authenticity is important,
it does not mean that the inauthentic is of lesser value, but rather they serve
two different functions, and as such cannot be judged in the same way. James
Félix argues that the terms “amateur” and “professional” should not be taken as
opposites, but actually they describe two different orders of things: amateur being
more a state of mind and belief. Amateurs performs fado because they believe in
what they are doing and every fado for them is a personal thing, whereas being
a professional simply means they get paid for it, or they do it for money. In that
sense, it is possible to be both professional and amateur at the same time: if paid
artists really feel strong emotions during one public performance of a certain fado,
then they would be both, however when they sing it the next night, if they do not
feel it, but just sing it faking the emotions, then they are just being professional
(Félix, forthcoming).
62
Chapter 5
Analysis
The characteristics of fado derive and are constrained by the abilities and lim-
itations of the performers, the instruments and the contexts in which the practice
occurs. If one takes into account that in nineteenth-century Portugal most of the
population was illiterate, poor, and without access to music education, then fado
is a reflection of these initial conditions. As soon as conditions change, namely
performers belonging to higher social classes, being schooled, living in different
spaces and using other musical instruments, the musical practices also change,
mirroring this fact. We will describe how these initial conditions shaped the music
and detail its traits, and then how change over time modified the characteristics
of the repertoire.
A fado performance can be systematically analyzed by breaking it down into its

core elements and observing the sequence of its events. Typically one has a core
textual and lyrical element, a narrative, in the form of a poem, which is the fado
itself. That fado is then presented to an audience by a performer, usually singers
that identify themselves as being fadistas, as they embody fado. Those fadistas are
usually supported by other elements, which will provide an instrumental support to
their theatrical display. The observation and characterization of all these elements
in the chain of events will highlight all the relevant codes, conventions and values
associated with the practice. The analytical categories developed by Alan Lomax
in the “Cantometrics” were used as inspiration (Lomax, 1978).
63
5.1 The Narrative
At present there is vast agreement among researchers that “fado was always
poetically constrained. The text was often improvised either by one performer
or by two challenging each other” (Nery, 2010c, p. 65). Alberto Sardinha and
Gouveia (Gouveia, 2010, p. 75) argue that fado was a poetic genre “consisting
in a pitiful narration about the life of a certain character” (Sardinha, 2010a, p.
59). Fado in Portuguese means literally “fate”. A fado is a destiny, an outcome,
and it could be defined as a story unfolding through several stanzas and, as such,
it would have affinities with the traditional Portuguese ballads in their literary
form. Sardinha states that “people from rural areas think that fado and tradi-
tional ballad are one and the same” (Sardinha, 2010a, p. 51). And that would be
the reason for the name, for the origin of the designation “fado”. Usually it would
be a sensationalistic or fatalistic story, episodic, eventually historical, concerning a
certain character. Based on both this assumption and the arguments of Carolina
Michaelis de Vasconcelos, Sardinha argues that fado as a literary genre has its
roots on the fifteenth or sixteenth century (Sardinha, 2010a, p. 187). Moreover,
he claims that due to its nature (narrative and mainly oral), being created by and
for the people, within the lower classes and with themes without erudite relevance,
it would be a genre largely ignored by academia and the higher classes. There-
fore, one would not find references to this phenomenon in the literary sources.
of the period. Also pointed out by António Tilly1 , the designation “fado” could
be a broad concept meaning “ballad”, “song” or any narrative being performed,
regardless of its musical structure. These traditional ballads, at first spread by
jesters, would later be carried throughout the country by the blind and the men-
dicant wanderers, who would sing and narrate them, selling flyers with the tales
and performing in hostels in exchange for food and sleep.
“After listening, wondered or horrified, to the latest news on crimes,

disgraces, or extraordinary successes and phenomena sung by the blind
musicians; after buying from them the flyers with the narratives, the
people, relying on their musical memory, would repeat these ballads
in their daily life: during labor (. . . ) and at the evening gatherings”
(Sardinha, 2010a, p. 93).
1
Private conversation, 10/04/2014
64
5.1. The Narrative
These events are also confirmed by other sources (Côrte-Real, 2000, p. 168), and
they are congruent with the other European folk song traditions exposed by Ne-
gus, and already discussed in a previous section. Sardinha also analyzes a relevant
statement from César das Neves in his Cancioneiro: “This fado, from the mid-
nineteenth century, is the primordial type of the popular weeping fados, more to be
listened to as a ballad than to be danced”. According to Sardinha, this statement
is relevant because it is the recognition by a nineteenth-century musicographer
that there is a primordial kind of weeping fados, having a popular, “anonymous”,
musical origin based on oral transmission. Furthermore, it reinforces the idea of
an association between this kind of weeping popular fados and the traditional
ballads, and also the idea that there were fados to be danced and others to be
listened to (Sardinha, 2010a, pp. 213-214). Although there seems to be no dis-
pute regarding the existence of these practices, Sardinha seems to be the only
one claiming the label “fado” to define them derives directly from the narratives.
Following this assumption, fado would then primarily be a literary genre. The
issue resides, then, in understanding the true importance of that poetic-literary
dimension and up to which point it defines the concept. Tinhorão, however, ar-
gues that dance and musical accompaniment were primordial and that the lyrics
come from the previous tradition of fandango improvised singing, where the per-
former creates a narrative on the spot. The most used formation would be stanzas
with heptasyllabic lines, sung alone or in a kind of provocative contest between
two performers (Tinhorão, 1994, pp. 69-72). Quoting Armando Leça, he finds a
parallel between the development of the lyrics in fado and in traditional ballads
since the sixteenth century, forcing the melody to be symmetrical and with a very
slow and dragged pace allowing the performer time to invent a coherent rhyme
(Tinhorão, 1994, p. 76). He also associates the themes sung with the evolution of
the social profile of its practitioners: from very low class (prostitutes, slaves, etc.)
to the labor classes, further stating that until the mid-nineteenth century the nar-
ratives were mostly positive, pragmatic and humoristic, while after the transition
their character became much more realistic and fatalistic, mirroring a certain kind
of self-consciousness of the people in relation to their own daily life. It was this
thematic transition that led to the independence of a sung fado from its previ-
ously danced version (Tinhorão, 1994, p. 79). Rocha Peixoto describes how most
themes used in the traditional ballads come from either the Romans and Celts
or universal feelings, and thus, claims that fado is the only thing that genuinely
represents Portugal and everything that is Portuguese. All those typical mournful
65
themes (love, good, evil, death, passion, revenge, etc.), are sung under the same
rhythm, using a poor music of slight variations2 (Peixoto, 1896, pp. 10-11).
Among the players, fados are known by their melody names, regardless of the
lyrics. The general audience and, sometimes, even the singers do not know the
name of the melody that they are singing, instead identifying the fado by the name
of the poem. Lyrics used to be exclusive to each fadista, sometimes original, other
times created on the spot. Some were famous for being able to create coherent
narratives, of several stanzas, covering whichever topic, lasting two or three times
what is today considered the usual duration of a typical fado (Nery, 2010c, pp.
111-112).
Contemporary authors claim no new literature for fado is available nowadays,

in the sense that the themes expressed (love, passion, betrayal, affects in general,
politics and social situation of the country) are recurrent. This lack of thematic
change can be confirmed comparing the themes and lyrics of nineteenth-century
(Carvalho, 1903, Pimentel, 1904), early twentieth-century (Guimarães, 1938), and
contemporary fados (Gouveia, 2010, Nery, 2004). And yet this is valued, since
it fulfils the expectations of the audience, and it is the core of what allows each
one of us to create empathy with the performance. However, even if the themes
remain the same, the lexical field and erudition of the words has broaden and since
the middle of the twentieth century it is possible to find erudite poems circulating
side by side with the popular fados.
5.2 Prosody
One aspect that seems to gather enormous consensus among fado performers is
the value attributed to the way one articulates the text. When we read many of
the ethnographic resources, namely the personal interviews gathered through the
2
“No nosso romanceiro nada ha cujo thema nao seja celtico, romano ou universal. (. . . )
Portugal tem pois e apenas, de genuinamente seu, o fado; o fado para a folia, para o amor,
para a amargura e até para a morte, em choradinho z’i á beira do sepulchro! N’um mesmo
schema metrico, de norte a sul, d’antes, hoje e sempre, o povo enquadra todas as suas ideias e
sentimentos, todos os factos, n’essa melopeia derrancada que só pode gestar-se n’um paiz que
nunca foi mais que uma ruina, raro com lampejos de uma opulencia fruste. Ignez de Castro e a
Severa, o bem e o mal, o rosto da lua e as vozes do echo, além-tumulo e a redemção, a paixão,
a desdita, o ciume, a vingança, até o Pobre Portugal, tudo se canta n’um mesmo rythmo, uma
musica de pequenas variantes, alcanceada, gemebunda, irreparavel. Os que nao cantam, sentem,
ouvem com um prazer morbido, interpretam os sentimentos no quadro ineluctavel d’esta logica.”
66
5.2. Prosody
decades, and see the opinions among the fadistas, we cannot help noticing how
the idea of a good, clear diction is praised: this gives an extreme emphasis to how
important the story being told is, and how the music is just a mere vehicle for the
word. Moreover, not only is the way one says the text important, but also how
one divides the text. Alfredo Marceneiro is constantly praised and referred to as a
master, not because of his timbre or vocal range, which was extremely limited, but
because of his inventiveness and the way he divided and stressed the syllables. In
other words, the value of prosody or, more specifically, the rhythm of the melodic
line plays a crucial part in defining the genre among its own performers. If there
is a doggerel line, with the wrong metric accent, it will ruin the cadence, the flow.
The vocabulary, the lexical field, does matter. According to Daniel Gouveia,
fado needs to be popular. Not all poems are suitable to be fado lyrics. If the
popular masses cannot understand it, then it is not fado. Fado can be high-art,
there are some beautiful poems that can be sung as fado, but not all of them are
suitable. Also, some words do not belong to the fado universe, some words just
sound inadequate, as if out-of-context or anachronistic in a fado lyric, for instance,
a word like “telephone” (Gouveia, 2010). Among the values stated by lyricists and
composers, the Lisbon accent was regarded as an important one, as well as the
intonation – the way in which one conveys the narrative, the weight the words
carry and how one perceives meaning through them. If one perceives a rural or
foreign accent in fado, it is not fado anymore, it loses something. Amélia Muge,
for instance, stated that her song sung by Camané sounded like fado because he
puts a weight in the words that she, herself, does not put. In her version it sounds
“light, carefree”, in his version it sounds “heavy, worn down by suffering3 .
5.2.1 Language
One interesting issue that arises with language can be exemplified with a case
study: the project Fado Novato4 . Fado Novato [Novice Fado] are a group of three
American musicians who seek to learn and perform fado. They are document-
ing every step of their evolution along the way. There are two men, one playing
guitarra and the other viola, and one woman singing the vocals. When we listen
3
Debate fado(s): Escritas e Autorias, moderated by Soraia Simões, promoted by Mural Sonoro
in 15/12/2013.
4
http://fadonovato.com/, accessed June, 6, 2015
67
to this project, we see almost no problem with the musicality of the instrumen-
tal parts – they imitate at a very good level the instrumental arrangements from
the recordings they are learning from and even the timbre of the instruments is
quite accurate. In some recordings the guitarra is actually a Bouzouki and the
instrument is close enough to provide an acceptable rendition of the mood it is
supposed to convey. On the other hand when we listen to the vocals we find them
hilarious, distractive, and they totally ruin the experience for us. We cannot say
that we are listening to fado at all, because all the emotions, feelings and the
context associated with the traditional experience we are used to know changed.
And they changed because of the language. Even though the singer sings in Por-
tuguese, the accent and prosody sound very strange. This is a big issue regarding
the Portuguese language spoken in Portugal. Contrary to what happens in Brazil,
where the language is very open and musical, the Portuguese from Portugal is
spoken with many closed vowels and with the end of the words almost vanishing
into thin air. The sound becomes muddy and blurry and almost indistinguishable
for a non-native. These are characteristics intrinsic to our prosody and they can-
not be changed (Frota, 2000). Pattel already analyzed these problems in a wide
volume on music, language and the brain (Patel, 2008), and explained how human
beings are heavily constrained in their sound and musical perceptions according
to the native language they learn since childhood. It is as if one gets locked into
a system of sounds and therefore other extraneous sounds and subtleties nonex-
istent in one’s mother tongue remain out of their reach and perception after a
certain age. An accent is basically the result of these constrains: after one learns
and incorporates their mother tongue into their brains and phonetic system, ev-
ery other language one learns afterwards will be spoken according to these basic
guides and cues one already possess. Therefore Spanish people speaking English
will sound very characteristic because they are speaking English with the stresses,
rhythm and prosody of the Spanish language. This is certainly true in the case of
Fado Novato’s vocalist, and most foreign vocalists in general – they are unable to
pronounce the words correctly, to provide the adequate stresses in consonants and
vowels and to make a correct division of words. Most of the time they just sing
with a very strange accent. Since the correct diction and prosody of the narratives
of fado are of paramount importance for most of the listeners, as already stressed,
the lack of this feature harms the performance irrevocably. Now, what seems
of great interest in this experience of bi-musicality is that the fado experience is
only ruined for Portuguese natives. For anyone else in the world, the experience
68
5.2. Prosody
will still sound as genuine as any other Portuguese singing fado. We have asked
some foreign friends informally what they thought about Fado Novato (one of our
friends even understanding a bit of Portuguese) and they could not find anything
wrong with the performance. Although we cannot say we have conducted any
methodical study or have a reasonable sample to test this hypothesis, the truth is
that we can speculate with a fairly reasonable expectation that the experience of a
performative genre that depends so heavily on the perception of the language, and
the correct accent, is only jeopardized when the listeners are native speakers and
the performers are not. This assumption is important because it leads us to think
that the cues that may identify the phenomenon of fado may vary accordingly
to the language of the speaker. In this specific case, we dare say that some fado
performances will not be recognized as fado by Portuguese native speakers, while
being cheered and applauded as genuine or acceptable performances by foreigners.
Of course, we see none of this as being a problem specific to fado. The same will
happen with any performative practice that depends heavily on some national
characteristics best recognized by natives. There are certainly similar problems
with American country music sang with British accent, for instance, deemed as
“ridiculous and laughable” by an American friend, or even all the couleur locale
references used in many classical Western works that, for sure, must seem also
inappropriate in the cultures they are supposed to reference or portray. In the
end, the issue of bi-musicality (Hood, 1960) and language problems just reinforce
the idea that fado is mostly a Portuguese phenomenon because of some difficulty
for a Portuguese native to give credit and recognize a traditional performance by
anyone who is not Portuguese5 . And therefore the issue of the “imagined commu-
nity” (Anderson, 2006) and of a shared experience among its constituents makes
total sense in this case.
5.2.2 The rhythm of the melody
Many practitioners recognize that, in any given traditional fado, the sequence
of harmonic progressions is fixed, as well as most of the basic underlying metrical
structure, but the melodic line is not, simply following the harmonic contour. “In
addition, certain aspects of poetic form in relation to which lyrics can be set to
5
Although Gray claims to have had some success after intense learning within the Portuguese
community and grasping what she believed to be most crucial cues regarding the accent (Gray,
2013, pp. 48-69).
69
specific traditional fados are also fixed, governed by rules regarding the number of
stanzas, number of lines, and, sometimes, the syllabic count and rhyme scheme”
(Gray, 2013, p. 145).
The rhythm of the melody can be thought as having two layers: an abstrac-
t/notation layer and a performative one. The first one is actually what one sees
in most of the musical scores, in the transcriptions and it has been described by
Frederico de Freitas (Freitas, 1973), for instance. Freitas presents a semiographic
scheme for fado and confronts it with the lundum and habanera, stating its differ-
ences. He describes this rhythm as being usually a seven-note pattern (a feature
directly imported from the prosody, since the fado lyrics usually are heptasyllabic
lines), written out in a 2/4 time signature, consisting of an eight pick-up note,
and then two eight notes on the first beat and a group of three syncopated notes
on the second beat, and a final note finishing the phrase on the downbeat of the
following measure. Although a very interesting scheme and, in a way, a useful
description, it is not entirely congruent with the data. In our musical phrases,
gathered in the corpus, it is possible to observe all possible combinations of mu-
sical durations arising, as long as they are somewhat consistent with the prosody
and emphasize the stressed open vowels. Inspired by the ideas of Nketia regard-
ing some African traditions, that “on the phonological level, the syllables of the
words of text lines, which correlate with pulses or beats in music, are the basic
constituents of structure” (Nketia, 2002, p. 148) and that “the rhythm of songs
reflects, to a large extent, the natural grouping of the words in a text line” (Nketia,
2002, p. 149), we propose to analyze the rhythm of the melody in fado relating it
directly to prosody and as a consequence of the lyrics, rather than building up a
mere mathematical archetype.
Our own observations of fado suggest that the rhythm of the melody is a direct
consequence of prosody and lyrical structure and that the syncopation identified
by Freitas is, on the other hand, a consequence of a gestural shaping discussed
further ahead. Fado usually employs heptasyllabic lines, although up to twelve
syllables per line have been used, to be mapped onto the strong beats of two 2/4
bars. “The aggregate duration of the syllables of the text line must correspond to
that of the musical phrase. A number of adjustments are made if there is disparity
in the length of the musical and verbal units of structure” (Nketia, 2002, p. 152).
The majority of European Portuguese words are stressed in the penultimate sylla-
ble; also, most monosyllables, namely articles, conjunctions and prepositions are
70
5.2. Prosody
performed as non-stressed words. Therefore, statistically, one seldom finds a stress

at the beginning of a sentence. When trying to pair the first stressed syllable with
the first downbeat of the measure, the previous non-stressed ones will inevitably
become pick up notes, this being a recurrent trait. Most stressed syllables contain
open vowels which are, conventionally, emphasized when sung thus reinforcing a
typical feeling of syncopation. For more information on European Portuguese lan-
guage and its prosody, consult Frota (Frota, 2000) as well as Castro and Lima
(Castro & Lima, 2010).
Using an example of a heptasyllabic textual line “O fadista que é fadista” [the

fadista who is fadista] we exemplify how one could map this line into a rhythm
that makes sense.
ã 42 Œ ‰ œ œ œ œ ‰ œ œ œ œ Œ
o fa - dis - ta que'é fa - dis - ta
4
ãŒ ‰ œ œ œ œ œ. œ œ œ Œ
ãŒ œ. œ œ œ œ. œ œ œ Œ
Figure 5.1: Possible binary rhythms for a textual line.
All these three versions would be acceptable as normal speech. However, when
mapped into this musical structure, the first one generates only two stresses and
the tonic syllable “é” is elided and does not sound very natural within the practice.
The second version generates three stresses, all tonic syllables correctly mapped
onto the strong beats. Also, for reasons detailed ahead, this version would reinforce
a fado feeling and be preferred. The third version would generate four stresses,
and it would generate even more fado feeling. However, one is mapping the non-
stressed syllable “o”, onto a strong beat, which also does not sound natural. So, in
most cases the second version would be considered the “correct” one. Hence, most
lines are sung with three stresses, mapped onto the three strong beats available.
71
“Since whatever text a song maker selects to convey his thoughts

or sentiments will emerge with its concomitant prosodic features, the
choice of words and word groups whose rhythms and tones can give
a unit of structure particular character or ‘swing’ when placed in cer-
tain positions is critical for the song maker. So is the matching of a
particular pattern of rhythm or tones desired in specific positions with
appropriate texts or vocables. The tendency to repeat significant text
lines or musical phrases in the body of a song can make this distinctive
character a special source of aesthetic interest.
The characteristic ‘swing’ or distinctive feature could also be estab-
lished for whole categories of songs, particularly the songs of dances
and games with specific sets of motor beats and body movements. This
is achieved through the choice of a basic rhythmic structure that recurs
in all of the items or serves as a guide to the construction of the units
of structure. Sometimes the particular basic rhythm or its outline may
be articulated or reinforced in an accompanying percussion or other in-
strumental form. In some traditions, this relationship is so close that
particular instrumental rhythms accompany sets of songs that share
the same phraseology” (Nketia, 2002, p. 159).
The prosody of the Portuguese language also shows us that most of the stressed
syllables will have open vowels associated with them. Open vowels in fado are, by
their nature, and the performative tradition and conventions involved, emphasized
when sung, compared to closed and muted ones. So, when open vowels and stressed
syllables coincide there is a combination that will stand out by itself, and the singer
will tend to make them last longer and with louder dynamics in comparison to
the others. This will reinforce the feeling of syncopation, even if one is trying to
make all durations sound equal, because one has to shorten the neighboring notes
in order to prolong the ones that are naturally being emphasized due to prosody.
What happens in performance adds a new layer of complexity to the semio-

graphic scheme. Usually, the performers do a styling of the melody and that
includes manipulating the rhythm in a strange combination between improvisa-
tion and groove. What happens is mostly the result of playing and having fun
within the limits of the prosody. That skill is highly valued among the older gen-
erations of fadistas. The art of well “dividing” the words, highlighting the story
and emphasizing its meaning is one of the main goals of fado. The ones who are
72
5.2. Prosody
able to perform this task very well are the ones most praised and highly regarded
among their peers. What happens in real life is what one could try to describe in
notation in the following examples.
j
A
& 42 œ œ œ œ œ œ œ ‰
œ
j
B
œ œ œ œ œ œ ‰
&œ œ
3
j
C
œ œ œ œ œ œ œ œ œ ‰
3
&œ œ
3 3
j œ. œ œ œ. œ œ ‰
&œ œ
Figure 5.2: Semiographic schemes of different performances of the same

melody.
The musical phrase A should be read as the “normal” notated musical phrase,
a common vocal line, with a typical rhythm. We are using the archetype of Freitas
in this example. The musical phrases B, C and D are examples of possible rendi-
tions of the phrase A. One would rarely expect to actually find them notated in
that way, however, it is expectable and possible to hear all possible combinations
between B, C and D and even several others with durations in-between, in an
actual human performance. One should also note that the division of the beat
in two is a characteristic not as typical of fado as one might think: fado perfor-
mances are very imaginative and fluid and often one can perceive subdivisions
of three mixed with subdivisions of two inside the same measure, as shown in
the examples. We stress again that we believe most of these subdivisions are the
natural consequence of prolonging and emphasizing certain stressed syllables for
emotional effects, thus stealing duration from the neighboring ones. The values
attached to these emotional effects derive mainly from what it means to embody
a tragic narrative and can be found in many other musical traditions around the
globe, like Arabic music, flamenco or even American blues. These values also rely
73
on other embodied cues, beyond the language. For instance, “the melodic song
forms known as Arabesques and Melismas are fine examples of Arabic influence
upon flamenco. Reminiscent of liturgical chants, Melismas, or extended vocaliza-
tions of several notes to a single syllable, are a strong component of flamenco song.
Important characteristics include the extension of vowels, expression of profound
feelings, and priority of emotions over lyrics” (Ruiz, 2007, p. 75).
The slight and constant variation of the durations of the rhythm of the melody
is, then, one of the characteristics that define fado nowadays. We believe that
the semiographic information, carried in the data, represents a quantized version
of an ideal, an urtext, an archetype of a parameter that in being consistent holds
the piece together coherently. If the rhythm of the melody was purely random
and too varied, it would make no sense stylistically. Analyzing phonograms or live
performances and measuring objectively the deviations associated with certain
artists, and thereby inferring profiles, could constitute an extensive future work
project.
5.3 Structure of the text
Most fados present phrases of similar length throughout the whole song, de-
riving from lines with similar metric. Ernesto Vieira referred to fado lyrics as a
four-line stanza developed into four ten-line stanzas as a recent form, not popular,
bearing no relation to Arabic poetry (Vieira, 1890, p. 239). Melodically, it would
translate directly into three quatrains – the first eight lines being the first two
quatrains, and then a third one emerging from the last two lines repeated. Rui
Vieira Nery adds that the oldest and more frequently used model of fado lyrical
structure was a quatrain with seven metric syllables (redondilha maior ) sung over
a melody repeated as often as the quatrains of the lyrics.
During the nineteenth century the dominant practice seems to have been the
use of loose quatrains, without any apparent motivic connection between them, or
a vaguely coherent narrative (Nery, 2010c, p. 109). In the beginning of the twen-
tieth century, five and six line stanzas started to be used, and in the middle of the
century other complex forms based on free metrical lines also emerged. Gouveia
presents a detailed analysis of fado’s lyrics over the time and it is possible to see
that lines ranging from one up to twelve metric syllables have been used, the most
74
5.3. Structure of the text
common by far being seven, followed by ten, twelve or four. He also demonstrates
how stanzas virtually of all sizes have been used, but the most common are four,
five and six lines stanzas (Gouveia, 2010). A common characteristic is the fact
that into the same music one can sing several lyrics as long as they share the
same metric structure. Contrary to thematic content, there is a clear evolution
and change in the format and structure of the fado lyrics along time (with con-
sequences on the music), correlated with the erudition and skill of the poets, but
also with censorship practices during the dictatorship period, and the contexts of
fado production (Côrte-Real, 2000, Gouveia, 2010, Nery, 2004).
An evolution regarding this strophic model was the introduction of a chorus.

We do not know exactly when fados for theater with coplas and refrain, what we
called fado-canção opposing strophic fados, started being composed. However,
we know that around 1910 we find many printed examples of this new sub-genre,
mentioning the play where they were featured and the performers who became
more famous by singing them. Unlike strophic fados, these fados-canção were
always mono-lyrical, each melody was associated with one and only one poem,
unless the fado-canção itself was parodied in another play (Nery, 2010c, p. 161).
According to Pedro Félix, musical theater was the greatest factor in shaping the
repertoire during the first half of the twentieth-century (others being censorship,
the intervention of previously legitimized erudite authors, and the world-music
market). Theater acted as an important local engine for music industry in general
and for the phonographic industry in particular, because it promoted singers,
authors and repertoire. Theater changed fado structurally because, due to the
show format and conventions, the typical song formula with alternating verses
and chorus was adopted, instead of the old strophic format. The new form did
not easily allow the use of several lyrics in the same music, so it took the name
fado-song. Often within the community of fado practitioners, these fados received
designations like fado-musicado [fado put to music] or fado-próprio [fado with its
own music], meaning that one particular original music was fit into one poem. This
also means that the lyrics were associated with that particular performer, with all
the consequences that emerge from that fact, namely the usual portraying of the
performer’s photo in the musical scores and the inscription “creation by” – the
notion that although a musical work is composed by someone, it is “created” by
75
the performer6 . By doing this, and establishing a dichotomy between two formats
that reflected two ideologies of authenticity, a traditional repertoire emerged by
opposition – the fado-fado or fado traditional. By portraying and caricaturing
stereotypes, theater and operetta were also fundamental in the development of an
iconic fado. Also, the several recordings made of the fados that were sang in the
shows represent a significant corpus of the recorded repertoire of popular music
that was available locally in that period (Félix, 2013).
This evolution is important because it constrains and bonds the literary practice
to the musical practice. The narrative is no longer hierarchically superior since
the music is tailored and specific to that same narrative. The structure, changed
to accommodate a chorus parts away from a literary genre and gets closer to a
performative one. No one in their daily life communicates using choruses, repeating
the same words over and over in their speech. As Frith explains: “a song does
not exist to convey the meaning of the words; rather, the words exist to convey
the meaning of the song” (Frith, 1998, p. 166), therefore we see a reversion of
the roles derived from the own lyrical structure. Therefore, in this segment of the
repertoire it is no longer the literary genre the defining element, instead the music
is.
5.4 The vocal group
Typically, the narrative is expressed as a monophonic vocal line by a single

performer. Occasionally, the audience might interact by shouting, applauding, or
reinforcing a chorus (when it exists), in unison, with minimal blending. Seldom,
and usually in familiar contexts and amateur venues, two or more fadistas might
sing, one after the other, in a kind of a challenge (cantar ao desafio).
5.5 The instrumental group
The most common ensemble since the twentieth century consists of one classical
six-string nylon guitar (named viola, in European Portuguese) and one Portuguese
Guitar (pear shaped twelve-string lute, called guitarra). During the nineteenth
6
As kindly suggested by Fábio Serranito, in a private conversation, this could also be a
gallicism since créer means the first time a work is performed in public.
76
5.5. The instrumental group
century several written and iconographic sources show fadistas accompanying

themselves, often with a guitarra, playing arpeggiated chords. Sources also men-
tion other string instruments, the accordion or the piano (Cabral, 1982, Castelo-
Branco, 1994, Cristo, 2014, Nery, 2004, Sá de Oliveira, 2014). We speculate that
one would play with whatever was available in the given context. Due to the lim-
itation of recording technologies in the early twentieth century, often fados were
recorded using pianos and wind or brass instruments (Losa, 2013).
The instrumental group’s function is to provide a ground for the fadista to cre-
ate on top of, and its spatial organization also reflects a hierarchical organization.
Usually the instrumentalists are on the background, they are seldom mentioned
and their interaction with the audience is minimal (Côrte-Real, 1991, pp. 30-35).
Most instrumentalists are male. The singer is often the only one announced and
well known by the audience (Côrte-Real, 2000, Gray, 2013).
The viola provides the beat and the harmony guiding the singer. It is played
with a static right hand position to facilitate the rhythmic execution. The thumb
is responsible for the bass line while the three middle fingers either produce chords
by simultaneously pulling the higher-pitched strings or arpeggiate them. Usually
the fingers do not rest, meaning that the strings are not dampened after their
execution, except when a deliberate effect of staccato or dry sound is desired.
This last situation is appreciated regarding a more traditional repertoire (Sá de
Oliveira, 2014, p. 42).
The guitarra complements the voice. Whenever the voice wanders (whether for
deliberate aesthetic reasons or lack of skill), the players have to react and adjust,
back and forth, respecting each other’s roles. In nineteenth-century and in early
twentieth- century sources, mostly very simple ostinati -like accompaniments are
observed, usually by a solo instrument. However, during most of the twentieth cen-
tury and nowadays one can often distinguish three layers: a bass line, a harmonic
texture, and a countermelody. The viola plays the bass line and chords using pre-
defined stock ostinati. When available, the bass mostly plays the same bass line
as the viola. The guitarra complements the voice using stock melodic formulas.
“The guitarrista José Fontes Rocha emphasized how the most important task of
the guitarra is to fill in (preencher ) the silences opened by the fadista between
vocalizations. Each guitarrista, he said, ‘fills the space in his own way’ ” (Gray,
2013, p. 139). Armando Augusto Freire, “Armandinho” (1891-1946), significantly
77
renewed the accompaniment practice by introducing guitarra countermelodies be-

tween the sung melodic lines. They were so characteristic that in many cases they
were appropriated and integrated by the community on the identity of the fados
where they were first performed (Nery, 2010c, pp. 131-132) (Brito, 1994, Fre-
itas, 1973). His composition method was based on his ease in improvising short
melodic formulas, which he linked and manipulated until they formed a whole
piece. His most complex instrumental works are rhapsodically structured with a
set of sections apparently autonomous, but where formerly presented sections often
reappear as musical refrains. The timbral uniqueness of these guitarra gestures,
absent from other practices, makes them a key feature in recognizing fado. These
countermelodies and ornamental motifs are largely absent in the musical scores
from the nineteenth century and in the earlier recordings, however they become
rather common later.
When available, other instruments might join in, namely a bass guitar, or extra
guitarras and violas resulting in string ensembles. In larger venues, like theatres,
an entire orchestra can be used. Especially from the mid-twentieth century on,
the addition of bass guitar is rather common in larger shows, and more recently
acoustic basses and percussions are used as well, in the context of promoting fado
as a world music product.
The resulting sound is polyphonic and polyrhythmic, but with a good degree
of blending, since there is usually a perceived tune, and a steady beat that all
members coordinate with.
5.5.1 The structure of the accompaniment
The instruments follow the same simple time signature as the voice: usually
a 2/4. As mentioned before, the rhythmic figurations are independent from the
voice, as the instruments usually play ostinati or danceable patterns. The four
layers (bass line, ostinati, melody and countermelody) usually have different and
independent rhythmic patterns – the melody being the most free and syncopated,
the countermelody also rather groovy, and the ostinati and bass line more regular
to keep the beat in place. It is important to note, though, that even the steady
ostinati provided by the viola usually present a waddled feeling.
78
5.6. Form
Fado narratives are usually styled and improvised on top of stock musical struc-
tures. The first definition of these structures appeared on the musical dictionary
of Ernesto Vieira, where they were defined as
“a section of eight bars in binary tempo, divided into two equal

and symmetric parts, with two melodic contours each; preferably in
the minor mode, although many times it modulates into the major,
carrying the same melody or another; harmony built on an arpeggio
in sixteenth-notes using only alternating tonic and dominant chords,
each lasting two bars” (Vieira, 1890, p. 238).
Usually music theory is a consequence of a given practice and here, once again,
that seems to be the case. When Ernesto Vieira writes this definition it seems
clear that he was observing and describing his own contemporary reality. This de-
scription seems very pertinent and somewhat in agreement with the data observed
at the time, if we use the 48 songs labeled as fado that appear in the Cancioneiro
of César das Neves and other similar period sources as a reference. That being
said, we do not know exactly the criteria used by César das Neves to transcribe
the works in Cancioneiro or how faithful his transcription was. Our analysis has
shown that, although there is an attempt at standardization, some works reveal
shortcomings, deficiencies and small textual incoherences. We also do not know
exactly how far away the transcriptions are from the reality they mean to portray.
Given the highly erratic and improvisatory character of the melodies, we do not
know exactly how the fixation corresponds to the contextual reality. Was it the
result of only one of the many possible interpretations of a given singer? An “aver-
age” or combination of several of them? Is it possible that different interpretations
of the same melody originated different transcriptions? One can only speculate.
However, the fact is that, among all of them, there is a considerable complexity
fluctuation, which might just show the evolution of the practice during the period
it documents.
5.6 Form
Ernesto Vieira’s model applies fairly well, but usually doubled. Given one
section to be the mapping of two lines of the lyrics, in question-answer format,
79
over four bars using a given harmonic progression, then, Ernesto Vieira’s eight bar
model would be an AA or AB case.
An introduction (I) is present in 72% of the fados from the corpus. Some have
a small introduction (two to four measures) with chordal material, while others
present longer ones, most of the time with motives drawn from the main melody;
some fados have an eight bar introduction that is, most of the times, a variation
of the first section of the fado, but can also be independent material standing by
itself.
The first section (A) is representative of the two first lines of the lyrics, in
question-answer format. It consists of a four bar section, following the tonic-
dominant-dominant-tonic harmonic structure, with two usually arch-shaped or
undulating melodic phrases – each one being a sung line of the text – on top of it
(a1a2). Most of the times this section is immediately repeated with very small or
no variation, resulting in an 8 bar section (AA′ ) = [(a1a2)(a1′ a2′ )].
Following this, we have a second section (B), which represents the remain-
ing lines of the stanza. It usually keeps the two bar arch-shaped or undulating
melodies, in the question-answer model (b1b2), however, most of the times, with
harmonic variation on the first phrase (b1). The harmony can go virtually any-
where: the data shows that although many fados indeed keep the tonic-dominant
shape, or move to the second or fourth degree, others modulate to secondary dom-
inants before returning to the dominant-tonic of the original key. This four bar
module is usually also repeated, and so, the fado archetype presents itself as a
sixteen measure structure (AA′ BB ′ ) = [(a1a2)(a1′ a2′ )(b1b2)(b1′ b2′ )].
This archetype is the most frequent pattern, found in 23% of the cases. The
AABB formula is then repeated as many times as there is text available to be
sung.
A coda (O) is present 33% of the time. When there is one it can be as simple as a
two bar chordal cadenza or as complicated as a whole eight measures instrumental
section based either on the fado melodic material or on the introduction, when the
introduction stands by itself.
Most of the fados transcribed in the Cancioneiro follow this simple structure,
which is basically a small variant on the model as defined by Ernesto Vieira:
[(I)AA′ BB ′ (O)].
80
5.6. Form
When we analyze fados from other sources, however, we realize that this variant
can present itself doubled. A second fado immediately follows the first one. Usually
the two fados are related by tonality – the second fado is in the relative key
of the first one (A minor to C major, for instance), or on the parallel key (C
minor to C major). There is usually no other relationship between the material
of the two fados, that could stand by themselves coherently, besides the lyrical
structure, which usually implies that the notated rhythm of the melody (hence a
very important unifying parameter) may remain constant.
Sometimes an intermezzo (Z) linking the two fados might be present (it happens
11% of the time in the database). This intermezzo can be a chordal cadenza on its
own, or can be related to either the introduction or the coda. Therefore we have a
more complex archetype that defines the form of many of the fados found in the
sources:
[(I)AA′ BB ′ (Z)CC ′ DD′ (O)]
Since this modular archetype is very flexible as it is entirely derivative from

the lyrical content, and the lyrical content is itself flexible, performers often opt
to repeat verses or to add or omit verses or to change the structure of the poem.
Therefore we can find many variations and exceptions to the structural model.
It is possible to find fados with twelve bar structures based on ABB or twenty
measures based AABBB. It is not unusual either to find combinations between
the melodic motives themselves: ABA′ B ′ .
Much rarer, but still possible, is the combination or alteration inside the basic
modular four measure structures. Sometimes instead of the two bar phrase model
antecedent-consequent, depending on the lyrical structure, it is possible to find a
section (A) with six bars, because it is formed out of three arch-shaped phrases,
usually one of them, being the repetition of a previous one (a1a1′ a2) or (a1a2a2′ ).
This may happen when the singer decides to literally repeat a line of the stanza.
Furthermore, we must remember that these archetypes and the sources most of
the time reflect only the performance of some of the lyrical content, assuming that
all the remaining stanzas will be performed using the same structures. Therefore,
in the final realization of one fado, one can have something as complex as this:
[(I)AA′ BB ′ (Z)CC ′ DD′ AA′ BB ′ (Z)CC ′ DD′ AAB ′ (O)].
81
The interplay between AB, where the first half of B represents a harmonic shift
regarding A (in the simpler and older forms), is in more modern fados replicated
on a higher level. As in nature, it is fascinating how self-similarity and self-
referential processes repeat themselves in nested structures. Therefore the same
idea of harmonic contrast already present in AB, appears when the “modern”
transcriptions have as its core a structure of sixteen measures ABAC. We can
establish a parallel between this structure and the previous one presented if we
see the “old A” as AB and the “old B” as the AC.
Continuing this reasoning leads us to think that if we want new sections,

an “old C” for instance, one can be made up out of DE. Therefore the “sim-
pler form” in old fados ABABCC, often appears in modern fados replicated as
(ABAC)(ABAC)(DEDE).
So, the scheme presented above can be also lead to more complex structures
as:
[(I)(ABAC)(A′ B ′ A′ C ′ )(DED′ E ′ )(Z)(ABAC)(A′ B ′ A′ C ′ )(DED′ E ′ )(O)].
We can understand how this particular scheme is in fact a variant of
[(I)ABABCC(Z)ABABCC(O)], just “doubled”.
Most of the times the lyrics and their internal intricacies will decide the shape
of the final structure; but also the will of the performers, if they decide to repeat
or omit some verses, and even their desire to show off (by vocalizing or humming
without singing lyrics, for instance). The same fado, the same musical transcrip-
tion, can, therefore, have multiple structures when performed live.
Having all these possibilities in mind we realize how complex and different all
forms can appear, at first, to one’s eye or ear, when, in fact, they are all simple
variations of the same model.
Overall, in terms of number of sections, the distribution is as follows:
There are no fados with one single section, and 34% of them have two sections
(A and B), this being the most frequent case. Paired structures (and thus sym-
metry) are preferred since 25% of the fados have four sections (A, B, C and D),
compared to only 14% having three (A, B and C) and 15% having five (A, B,
C, D and E). Longer fados seem to be rare, since only 8% have six (A, B, C,
D, E and F ) sections, and 4% have more than six sections. Overall, the cases
82
5.6. Form
Table 5.1: Distribution of Sections
Number of Sections %
1 0
2 34
3 14
4 25
5 15
6 8
7 or more 4
Total 100
Introduction 72
Intermezzo(s) 11
coda 33
with two and four sections alone account for 59% of the cases, clearly defining the
archetypes, and confirming both Ernesto Vieira’s model in its simple and doubled
versions.
5.6.1 Internal complexity
There is a gradient mirroring several degrees of internal complexity among

the analyzed fados. Among most of the Fados in the Cancioneiro, the structure
is pretty simple, plain, and with a huge number of parameters holding internal
repetition, thus gluing the form together. They denote more of an archetypical
notated piece than a literal transcription.
Among the more modern transcriptions one can find cases of the opposite,
for example occurrence 0627 , which holds a high degree of internal complexity by
means of extreme variation. This fado seems to us more of a literal transcription of
what a live performance would tend to be than a written structural skeleton. It is
possible to clearly outline the sections and perceive the phrases and melodic arcs;
however, one finds elisions forcing the overlap between the end of a section and the
beginning of the next (hence two consecutive musical sections lasting fifteen bars
instead of sixteen). That is a trait that we would imagine occurring in the embod-
iment of the performance, not among the symbolic data. Also, there is constant
new melodic and harmonic material on top of very similar rhythms and shapes.
7
Identified as“Despedida”, composer Luiz de La Cruz Quesada, lyrics by José Coelho da
Cunha, edited by Sassetti c.1920.
83
Again, this denotes, in our opinion, a styling of a repeated idea that would natu-
rally occur when performing a notated score holding internal repetition. The same
can be said of slight changes in harmony with passage tones and chromaticism,
which are mainly there either for embellishment purposes of a similar passage, or
for outlining a bass. Sophisticated elements of this kind are usually not notated,
although they often happen in live performance, especially when the musicians are
professional and tend to express external influences from other genres.
The occurrence of these various degrees of internal complexity among the data
reflect the various approaches of the composer/transcriber/arranger when notat-
ing a dynamic and live performative tradition: they may choose to make a void
archetypical skeleton and provide just a melodic idea and some chords or instead
outline a very detailed set of instructions to be strictly followed and which will
mirror conventions and past performances of this tradition. There is no “good”
or final answer and many factors influence this kind of decision: not only the
background and education of the person in question, but also the target and the
desire of making a more “difficult” piece to render in a piano, as well as the will
to simply be faithful to a certain occurrence of a live performance. So, these kinds
of details inside the data tell us more about the tradition and the person that
was responsible for the notation than they tell us about fado itself. Their mere
existence is very useful regarding the inference of shapes and gestures when lack-
ing a physical recording, for instance. The main problem, in the end, seems the
absence of a uniform way or a set of defined criteria on “how to make a proper
transcription” or of a uniform body of transcribers/composers following the same
principles. This issue was discussed further in the sub-section “How social class
shapes a canon”.
5.7 Tempo
Usually fados are categorized according to their mood, given the character of
their lyrics. So, there are sad or mournful fados and cheerful ones. Sometimes
a cheerful fado is also called marcha [march], because they are accompanied us-
ing the ostinato and tempo of a popular march. Sometimes these genres seem
interchangeable, and most fado singers also have marches in their repertoire. We
have realized (and confirmed this assumption with both fadistas and the critical
sources), that a march is basically the name given to a fado in double tempo (or
84
5.7. Tempo
cut tempo, depending on the way one sees it). Most transcriptions do not have
a precise indication of tempo, instead, most of the times, they have a subjective,
textual one. However, a precise initial tempo is a crucial numeric parameter, not
only for performance, but also as a reference for the calculation of note durations
and rhythms in seconds. The initial tempo is simply the “tempo in beats per
minute at the start of a recording” (McKay, 2004, p. 71). Therefore, we had to
assign a specific tempo to each transcription from the corpus, even when one was
absent or was subjective. Our criterion was the original guidance from sources,
complemented with our subjective internal feeling regarding the character of each
song, compared to our own experience as long time fado listeners, improvisers and
composers. Therefore the tempo parameter in this study is to be read in rela-
tive and not precise factual terms. According to the sources, most original fados
seemed to be played at around 72 to 84 beats per minute, this range averaging
at 78 and equating the indication of Allegretto (Nery, 2004, p. 82), (Vieira, 1890,
p. 55). We have tried to approximate the average of the corpus to this value.
The truth is that most fados sounded too fast and aesthetically unpleasant when
played at that same speed. Our own intuition placed most of them at 64, while a
few dramatic ones still sounded too fast and were placed at 56, and others, clearly
marches (and therefore more suited to cut-tempo) were placed ranging from 92 to
128. The rallentando, tenuto and fermata indications were added when notated,
and are responsible for the small deviations, minding that the overall corpus con-
tains the introductions, intermezzos and codas, while the corpus of melodies misses
them. We are aware that these choices somewhat condition the calculations and,
in order to reproduce them with precision, the table of tempos is necessary and
therefore made available. In relative terms, however, we believe the results reflect
the practice.
Table 5.2: Statistics of Tempo
Initial Tempo Avg Med St Dev Var Min Max Range Skew Kurt
Corpus 79.93 64.00 24.25 588.21 56.00 128.00 72.00 1.00 -0.83
Melodies-only 78.41 64.00 23.26 541.05 56.00 128.00 72.00 1.17 -0.41
We have also done a comparative data visualization of this feature regarding

other musical taxonomies (figure 5.3).
This feature is highly problematic because it is very sensitive to encoding and

transcription decisions. Basically, the initial tempo of every recording can be
85
Initial Tempo
250 Global Mean
Global Median
200
Beats per minute
150
100
50
Contemporary Country
Country Blues
Modern Classical
Adult Contemporary
Ragtime
Hardcore Rap
Classical
Soul Blues
Reggae
Funk
Psychedelic
Baroque
Romantic
Alternative Rock
Soul
Renaissance
Chicago Blues
Salsa
Blues Rock
Hard Rock
Metal
Celtic
Cool
Swing
Flamenco
Punk
Rock and Roll
Bebop
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado (Edition)
Pop Rap
Fado
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.3: Initial Tempo.
halved or doubled depending on the time signature chosen. In some cases dealing
with complex measures the multiplicative factor can vary as well. As one can see,
editorial decisions to conform most of the corpus to Ernesto Vieira’s model (thus
assigning a time signature of 2/4 to most fados) had a tremendous effect in the
relative position, because a high number of initial tempos were halved. If one looks
at how bebop is on the high part of the spectrum with values ranging above 200
one is led to think most of its values are certainly doubled. So, overall, it seems
difficult to make a comparative study and draw meaningful conclusions based on
this feature as it is. Still, looking at the relative distribution with a grain of salt,
one might also assume that the number of doubled or halved tempos in each style
are also randomly distributed and so, they might compensate each other. Taking
that assumption as an initial guide, and then discarding the edition corpus bias,
one can notice that fado is just slightly on the lower part of the spectrum with a
mild tendency to present slower songs than most styles.
86
5.8. Rhythmic Patterns
5.8 Rhythmic Patterns
The transcriptions show accompaniment figures based on alberti figurations,

corridinhos, marches, fox, habaneras, or a combination of these (figure 5.4).
The analysis of the data has shown that the accompaniment figure repeats it-
self over and over during a certain period. Sometimes it goes along throughout
the entire work (it happens in 80% of the transcriptions), but other times it just
spans through some sections, having other contrasting ostinati in the other sec-
tions (20 transcriptions show two or more different ostinati). The same ostinato
might show some variation at crucial points in each section – namely at caden-
tial points, where, often, it just pauses so that it signals the audience the end
of the section. Furthermore, it might pause randomly or present small tweaks to
increase variety. Some fados present very elaborate and elegant hybrid figurations,
including slurs and articulation marks, which might just reflect the commitment
of the transcriber to present a more idiomatic notation. Clearly there are osti-
nati that are written to be played on a piano (namely alberti and arpeggio), while
others resemble more what a viola would do (namely fox, marches, corridinhos
and fingered figurations). Usually, and again, as with many other parameters and
situations, the musical scores are very strict regarding the ostinati patterns and
just present them as an archetype. When in performance, the players sometimes
improvise, somewhat freely, and the same fado may be played with different osti-
nati in different occasions. Therefore, there is a clash between the semiographic
schemes and what one can actually hear in a live rendition.
Some of these figures might be dated or linked to very specific contexts. The
alberti and corridinho figuration are massively present in the transcriptions and
they can be heard in the early recordings, however they are rarely used nowadays.
Chula (or malhão) figuration appears as the main ostinato in only one score, but
shows up in some other scores, mainly as a secondary ostinato and not as the
prevalent one. Habanera figuration is almost non-existent as well. The ballad
figuration is often mentioned in the nineteenth-century literature, and can also be
heard in the early recordings, but today seems mostly associated with serenades
and with the Coimbra repertoire. On the other hand, we highlight the fact that
figurations like fox and bolero, almost non-existent in the transcriptions, at present
are often used in live performances and recordings, which might indicate a change
and the gradual popularization of these styles of accompaniment. Also, the name
87
A lbe rti A rpe ggio
?2 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
4 œ
A rpe ggio va ria tio n Bo le ro

œ œ œ œ œ œ œœ œœ j œœ œœ
?œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
J
Bo le ro va ria tio n Ba lla d

œœ œœ œœ œ œœ œœ œ œ gg œœœ ggg œœœ gg œœœ gg œœœ ggg œœœ
?œ œ œ œ œ œœ œ œ œ œœ œ œœ ggg œœœ
œ œ œ gg g gg
œ g
g g g
C hu la /Ma lhã o C o rridin ho

œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
?œ œ œ ‰ œ œ ‰ œ œ œ œ œ œ
J œ J œ œ
Fin ge re d figu ra tio n s Hy brid fin ge re d figu ra tio n

j
œ. œ.
. . œ. gg œœ ‰ œ œ œ
? >œ œ œ >
œ œ œ œ œ. œ. œ. œ gg œ œ ˙
Fo x
Hy brid figu ra tio n s
j j j j
œœ ‰ œœœ ‰ œœœ ‰ œœœ ‰ œœœ
? œ œ œœ œ œ œ œœ œ œ œ
œ œ œ œ
œ œ
œ œ
Fo x/Ma rc h (Po pu la r) Ma rc h
œœ œœ œœ œœ œœ œœ œœ
?œ œ œ œ œ œ œ œ
œ œ œ œ œ œ œ œ œ
Ha ba n e ra Wa ltz
œ œ œ œœ œœ œœ œœ œœ œœ œ œ
?œ ≈ œ œ œ œ œ œ œ œ œ œœ œœ
œ ≈ œ œ œ œ
3 3
3 3
Figure 5.4: Ostinati.
88
Table 5.3: Main accompaniment figure present in each transcription.
Ostinato #
Alberti 21
Arpeggio 5
Bolero 1
Ballad 5
Chula/Malhão 1
Corridinho 30
Fingered 3
Fox 1
Habanera 2
Hybrid 5
Marcha 25
Waltz 1
Total 100
of these figurations might be emic and not consensual: the bolero, for instance,
corresponds to the figure of the Cuban bolero, which is similar to a beguine or a
baião, and very different from the Spanish bolero, based in triplets.
There are two main contrasting grooves appearing in the fado accompaniments.
Probably derived from the sad military marches and sorrowful songs, it is common
to convey a processional feeling in slow tempi, with emphasis on the downbeats
and resonating chords (not dampened). On the other hand, in more vivid fados,
they are played favouring an uptempo, waddled feeling, and the chords plucked,
stacatto, and sometimes even dampened. When played the first way, the alterna-
tion between a bass note and a chord is called fox, when played in the second way
it is often called march. The figuration is the same, it only changes the tempo
and the groove. The distinction between a fado and a march seems blurred. The
repertoire is interchangeable and among the transcriptions there are many popular
marches which are considered fados. Vital Assunção (viola player, composer) told
us that a march is basically a fado in double or cut tempo. In the popular parades
and traditional festivities, with band accompaniment, the same patterns and con-
fusions occur: often the snare drum plays a processional figuration, instead of the
typical military march one (Pinto, 2004, p. 120). Many times fados are sung as
part of the marching band repertoire in those festivities (namely the traditional
neighbourhoods). Conversely, most fadistas sing popular marches in their typical
contexts. So, it seems that, basically, while many fados are not marches, most
89
popular marches are considered fados, or at least suitable to be sung by fadistas

and appreciated by the same audience, in the same contexts.
These accompaniment figures are usually learned by imitating older, experi-

enced players, sometimes in real performance situations8 . Often, also by listening
and imitating the records. They seem to depend mostly on the taste, skill and
inventiveness of the players, hence, their frequency distribution, as seen on the
table, to us, does not reflect the actual practice. Empirical experience and the
listening of performances, recordings and videos, highly suggests that fox, bolero
and march are the archetypes nowadays, among traditional players, and complex
fingered figurations, among the virtuosos.
Regarding the guitarra, examples of typical stock formulas found in performance

were described by Salwa Castelo-Branco (Castelo-Branco, 1994, p. 139), and
we present a reproduction of her transcriptions (figure 5.5), with her permission.
Sometimes they are present in the musical transcriptions, other times they are
not.
8
https://www.youtube.com/watch?v=6xg1MeCeEJY, accessed June, 6, 2015, in the video we
can see how the male viola player whispers the chord names to the female viola player who is
imitating and following him, learning in real time.
90
Gu ita r & 42 ≈ bœ œ œ œ œ œ œ œ Œ
œ œ
&≈ œ œ Œ
3
#œ œ bœ œ
œ œ œ œ #œ œ
œ
&≈ bœ Œ
5
&‰ ≈ œ œ Œ
7
R œ bœ œ
#œ
œ œ
œ bœ
œ
&‰ bœ Œ
9
Figure 5.5: Guitarra countermelodies.
91
5.8.1 Note density
The note density was defined as simply the “average number of notes per sec-
ond” (McKay, 2004, p. 70). We have calculated the average number of notes
per second in the corpus, which range from a minimum of 4,03 to a maximum
of 13,46, averaging on 7,48. If one considers only the melodies, then the average
is considerably smaller (which is logical), ranging from a minimum of 1,73 to a
maximum of 5,03 and averaging at 2,47 as shown in the table below.
Table 5.4: Statistics of Note Density
Note Density Avg Med St Dev Var Min Max Range Skew Kurt
Corpus 7.48 7.26 1.74 3.02 4.03 13.46 9.43 0.83 1.03
Melodies-only 2.47 2.25 0.66 0.44 1.73 5.03 3.30 2.15 5.37
In comparative terms with other taxonomies (figure 5.6):
Note Density
60 Global Mean
Global Median
50
40
Notes per second
30
20
10
0
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Baroque
Flamenco
Classical
Ragtime
Hardcore Rap
Romantic
Cool
Celtic
Hard Rock
Alternative Rock
Psychedelic
Reggae
Soul Blues
Soul
Blues Rock
Metal
Funk
Swing
Bebop
Chicago Blues
Punk
Rock and Roll
Salsa
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.6: Note Density.
Fado is one of the less dense genres in the database, in both versions of the
corpus, alongside with the early music genres and country blues. This particular
feature seems a useful one to characterize the practice. One contributing factor to
92
this situation might be the arrangements privileging sparse instrumentation. On

the opposite side of the spectrum we find salsa, rock and roll, punk and techno,
which are intuitively rhythmic driven genres in the sense that percussion instru-
ments are highly involved and are important in defining such genres.
5.8.2 Strongest Rhythmic Pulse
The strongest rhythmic pulse corresponds to the “bin label of the beat bin with
the highest magnitude” (McKay, 2004, p. 69).
Any song has a defined tempo, that may largely vary, and therefore that sta-
tistical value might not by itself be very relevant. However, accounting for that
variation, one still knows that if information relative to the strongest rhythmic
pulses is retrieved then one is able to draw meaningful conclusions: “the two high-
est peaks of the beat histograms tend to have particular importance, as they are
likely to represent the main beat of the music or one of its multiples or factors”
(McKay, 2004, p. 69). The relationship between those pulse values and the tempo
values will indicate the presence and relevance of groovy genres and regular or
irregular rhythms. For instance, if in a given song the strongest rhythmic pulse
(in beats per minute) is consistently half of the tempo value (also in beats per
minute) then, one deduces that the figure representing half of a beat is the most
relevant one. If this happens consistently in a given genre, then one might have
a defining feature. On the other hand, even if these kinds of features might be
elusive, or have great variability, their frequency might reveal more hints. The
numerical value of the strongest rhythmic pulse might not be relevant, however,
the frequency in which it occurs is. It will say a lot about the variability of the
rhythmic figurations present in a given song. If one, then, combines that frequency
with the frequency of the second most important rhythmic pulse, then one might
not only deepen that discussion but also infer conclusions about the complexity
of the rhythms as well (figure 5.7).
As expected, there seems to be great variability regarding this feature, espe-

cially among the taxonomies soul, alternative rock, classical, country blues, techno,
flamenco, fado and bluegrass. Fado has the second highest median value of the
list, with an overall span of values above the global mean and median. In this par-
ticular case one can see how the effect of edition strongly affects the corpus since
the edited corpus has actually the lowest median value. In its raw state it seems
93
Strongest Rhythmic Pulse

200
Global Mean
Global Median
150
Beats per minute
100
50
Country Blues
Modern Classical
Adult Contemporary
Salsa
Bebop
Rock and Roll
Cool
Chicago Blues
Swing
Ragtime
Soul Blues
Punk
Reggae
Psychedelic
Soul
Hard Rock
Funk
Renaissance
Blues Rock
Celtic
Romantic
Metal
Hardcore Rap
Alternative Rock
Classical
Baroque
Flamenco
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado (Edition)
Pop Rap
Dance Pop
Fado
Medieval
Traditional Country
Tango
Techno
Figure 5.7: Strongest Rhythmic Pulse.
impossible to discern what the strongest pulse in fado is or even if there is one.
The edited version, which is conformed to the Ernesto Vieira’s model, following
a binary time signature and a humanized performance of a median tempo of 64
beats per minute, in most recordings, reveals a scattered pulse around what would
be each eight-note.
5.8.3 Second Strongest Rhythmic Pulse
The second strongest rhythmic pulse is defined as the “bin label of the beat bin
of the peak with the second highest magnitude” (McKay, 2004, p. 70).
As observed in figure 5.8, the variability regarding this feature seems to be

even larger than the previous one. One might notice how salsa and bebop show
up again at the bottom of the list, indicating some expected consistency, since
these are usually groovy genres, but fado is now at the middle. What seems
perplexing is that both corpora actually almost coincide, hinting that the edition
does not seem to particularly affect this feature. The mean value at around 64,
however, shows that in the edition corpus the second most important pulse is
94
Second Strongest Rhythmic Pulse
200
Global Mean
Global Median
150
Beats per minute
100
50
Country Blues
Modern Classical
Adult Contemporary
Salsa
Bebop
Soul Blues
Swing
Funk
Psychedelic
Rock and Roll
Hard Rock
Soul
Chicago Blues
Celtic
Reggae
Blues Rock
Metal
Cool
Hardcore Rap
Ragtime
Flamenco
Romantic
Punk
Classical
Alternative Rock
Renaissance
Baroque
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Fado (Edition)
Fado
Pop Rap
Medieval
Traditional Country
Tango
Techno
Figure 5.8: Second Strongest Rhythmic Pulse.
indeed the beat, therefore the quarter note. On the other hand, among the raw
data it shows precisely the same value. Taking into consideration that the time
signatures are not harmonized, and represent mainly both binary and quaternary
ones, one expects both quarter notes and half notes to be considered, which is
consistent with the variability and span graphically displayed. Still, as predicted,
these seem to be rather confusing features to infer anything particularity relevant,
by themselves, and the additional information presented ahead might make more
sense.
5.8.4 Harmonicity of Two Strongest Rhythmic Pulses
The harmonicity of two strongest rhythmic pulses is defined as “the bin label of
the higher (in terms of bin label) of the two beat bins of the peaks with the highest
magnitude divided by the bin label of the lower” (McKay, 2004, p. 70). Basically
this information will confirm the importance of groove in a given style. When
the harmonicity approaches 1, it means that the two strongest rhythmic pulses
are concentrated around the same bin label. This will emphasize the amount of
scatter around the pulses, and thus might indicate rather fluctuating performances.
95
However, when that is not the case, it generally means that the second strongest
rhythmic pulse will be a factor or a multiple of the strongest one. This indicates
not only more tidy rhythmic performances but also the preference for regular or
irregular rhythms, depending if the factor is an even or an odd one.
Harmonicity of Two Strongest Rhythmic Pulses
4.0
Global Mean
Global Median
3.5
3.0
Factor
2.5
2.0
1.5
1.0
Country Blues
Modern Classical
Adult Contemporary
Salsa
Bebop
Soul Blues
Swing
Funk
Psychedelic
Rock and Roll
Hard Rock
Soul
Chicago Blues
Celtic
Reggae
Blues Rock
Metal
Cool
Hardcore Rap
Romantic
Punk
Alternative Rock
Renaissance
Baroque
Ragtime
Flamenco
Classical
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Fado (Edition)
Pop Rap
Fado
Medieval
Traditional Country
Tango
Techno
Figure 5.9: Harmonicity of Two Strongest Rhythmic Pulses.
As seen in figure 5.8, not surprisingly most genres have harmonicities of either
1 or 2. The most improvisatory and jazz-related genres tend largely to 1, while
the more conservative and regular ones tend to 2. Ragtime seems to be the most
varied including the whole range up to 4. This graphic helps clarify what happened
to fado – while the raw data conforms to a harmonicity of 2, clearly indicating a
notated tendency for a dichotomy between the half beat and the beat, the edited
version conforms to 1, which we believe better reflects the actual improvisatory
and fluctuating practice.
5.8.5 Strength of Strongest Rhythmic Pulse
The strength of strongest rhythmic pulse is defined as the “magnitude of the

beat bin with the highest magnitude” (McKay, 2004, p. 70).
96
Strength of Strongest Rhythmic Pulse

0.5
Global Mean
Global Median
0.4
0.3
Frequency
0.2
0.1
Country Blues
Modern Classical
Adult Contemporary
Bebop
Swing
Salsa
Cool
Chicago Blues
Celtic
Rock and Roll

Punk
Flamenco
Funk
Soul Blues
Blues Rock
Psychedelic
Romantic
Metal
Hard Rock
Ragtime
Soul
Classical
Reggae
Alternative Rock
Hardcore Rap
Baroque
Renaissance
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Dance Pop
Pop Rap
Fado (Edition)
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.10: Strength of Strongest Rhythmic Pulse.
This graphic (figure 5.10) reveals that the strength of strongest rhythmic pulse
is a rather important feature in the case of fado. Both the raw data and the edition
corpus show up among the highest median values, above the global median and
mean. This trend is shared by other folk traditions like country blues and hardcore
rap, but also by the early written music, which might reflect the conventions of
western written notation and how they are encoded in the MIDI files. As expected,
on the opposite side, all the improvisatory, groovy, swing-styled and jazz-related
genres show up.
5.8.6 Strength of Second Strongest Rhythmic Pulse
The strength of second strongest rhythmic pulse is defined as the “magnitude

of the beat bin of the peak with the second highest magnitude” (McKay, 2004, p.
70).
This feature (figure 5.11) mirrors the trend found in the previous one, with
the early music, country blues and hardcore rap being above median and mean
values, and the jazz-related, swing and improvisatory genres being at the bottom
97
Strength of Second Strongest Rhythmic Pulse
0.4 Global Mean
Global Median
0.3
Frequency
0.2
0.1
Country Blues
Modern Classical
Adult Contemporary
Swing
Bebop
Chicago Blues
Cool
Salsa
Soul Blues
Celtic
Rock and Roll
Blues Rock
Punk
Flamenco
Psychedelic
Romantic
Funk
Metal
Hard Rock
Soul
Reggae
Classical
Ragtime
Alternative Rock
Hardcore Rap
Baroque
Renaissance
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado (Edition)
Dance Pop
Pop Rap
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.11: Strength of Second Strongest Rhythmic Pulse.
part. The mere existence of the same trend in both features indicates that two
pulses alone seem to have too much weight in some styles while equally being less
important in others. What remains to be seen is whether in the case where these
weights are important they actually represent similar pulses, multiples, or factors
of one pulse. One can also notice how the edition process affected the fado corpus
in particular – while the raw data still displays one of the highest magnitudes, the
edited corpus approaches the global mean and median.
5.8.7 Strength Ratio of Two Strongest Rhythmic Pulses
The strength ratio of two strongest rhythmic pulses is defined as “the magnitude
of the higher (in terms of magnitude) of the two beat bins corresponding to the
peaks with the highest magnitude divided by the magnitude of the lower” (McKay,
2004, p. 70).
The relationship between the strongest and the second strongest rhythmic pulse
(figure 5.12) should show how important is a given pulse regarding all others.
Given the trends shown by the previous features one should expect an inversion
98
Strength Ratio of Two Strongest Rhythmic Pulses
Global Mean
2000
Global Median
1500
Ratio
1000
500
0
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Hardcore Rap
Alternative Rock
Ragtime
Baroque
Hard Rock
Reggae
Soul
Metal
Funk
Classical
Rock and Roll
Psychedelic
Punk
Romantic
Soul Blues
Blues Rock
Salsa
Chicago Blues
Celtic
Flamenco
Cool
Bebop
Swing
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado (Edition)
Fado
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.12: Strength Ratio of Two Strongest Rhythmic Pulses.
of the relative places regarding this feature. The most curious observation is
how most genres conform to the expected inversion – the early music, fado, and
hardcore rap appear now at the bottom, while the jazz-related, improvisatory and
swing genres are at the top – except for country blues. This clearly shows what
can be a defining set of features for this genre and, at the same time, a way to tell
it apart from fado.
5.8.8 Combined Strength

of Two Strongest Rhythmic Pulses
The combined strength of two strongest rhythmic pulses is defined as “the sum
of the frequencies of the two beat bins of the peaks with the highest frequencies”
(McKay, 2004, p. 70).
This feature (figure 5.13) both confirms and reinforces the trends already ob-
served in the individual strength of each of the two strongest pulses: fado is
99
Combined Strength of Two Strongest Rhythmic Pulses
1.0
Global Mean
Global Median
0.8
0.6
Frequency
0.4
0.2
Country Blues
Modern Classical
Adult Contemporary
Bebop
Swing
Cool
Salsa
Chicago Blues
Soul Blues
Celtic
Rock and Roll
Blues Rock
Punk
Flamenco
Psychedelic
Funk
Romantic
Metal
Hard Rock
Soul
Ragtime
Classical
Alternative Rock
Reggae
Hardcore Rap
Baroque
Renaissance
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Dance Pop
Pop Rap
Fado (Edition)
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.13: Combined Strength of Two Strongest Rhythmic Pulses.
placed above global mean and median, along with the early music genres, coun-
try blues and hardcore rap, clearly contrasting with the improvisatory, swing and
jazz-related ones.
5.8.9 Number of Strong Pulses
The number of strong pulses is defined as the “number of beat peaks with
normalized frequencies over 0.1” (McKay, 2004, p. 70).
The number of strong pulses ranges from 0 to 9 in the global database (figure
5.14). The original fado corpus has a median of four strong pulses, which is slightly
above global median and mean, and still on pair with country blues or hardcore
rap, and this time also accompanied by flamenco. One can notice that punk is the
style with the highest values overall, but how the improvisatory, swing and jazz-
related styles still appear at the bottom. The edited version of the corpus loses
half of its strong pulses, which might be related to the change of time signature
in most of the recordings.
100
Number of Strong Pulses
Global Mean
8 Global Median
6
Beat peaks
0
Country Blues
Modern Classical
Adult Contemporary
Salsa
Bebop
Swing
Rock and Roll
Soul Blues
Chicago Blues
Celtic
Cool
Blues Rock
Renaissance
Reggae
Soul
Funk
Classical
Psychedelic
Romantic
Flamenco
Hardcore Rap
Baroque
Hard Rock
Alternative Rock
Ragtime
Metal
Punk
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado (Edition)
Fado
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.14: Number of Strong Pulses.
5.8.10 Number of Moderate Pulses
The number of moderate pulses is defined as the “number of beat peaks with
normalized frequencies over 0.01” (McKay, 2004, p. 70).
This feature (figure 5.15) ranges from 1 up to 26 and one can see a reasonable
amount of variability among the several genres. What is noticeable is how fado,
despite having a relatively high number of strong pulses, has a remarkably relative
low number of moderate pulses (both in the raw and edited version of the cor-
pus), along with the early music genres, adult contemporary and ragtime. Usually
related styles like country blues and hardcore rap depart from this tendency for
medium positions, while flamenco appears at the opposite side of the spectrum,
which is topped by romantic music. There seems to be a tendency for the impro-
visatory, blues, swing and jazz-related genres to cluster around the high end of the
spectrum as well.
101
Number of Moderate Pulses
25 Global Mean
Global Median
20
Beat peaks
15
10
5
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Baroque
Ragtime
Soul
Alternative Rock
Salsa
Reggae
Funk
Psychedelic
Hardcore Rap
Hard Rock
Metal
Rock and Roll
Classical
Swing
Soul Blues
Chicago Blues
Blues Rock
Punk
Bebop
Celtic
Cool
Flamenco
Romantic
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.15: Number of Moderate Pulses.
5.8.11 Number of Relatively Strong Pulses
The number of relatively strong pulses is defined as the “number of beat peaks
with frequencies at least 30% as high as the magnitude of the bin with the highest
magnitude” (McKay, 2004, p. 70).
This feature (figure 5.16) varies between 1 and 16. It has some similar trends
with the number of moderate pulses: fado is again at the bottom part of the
spectrum alongside with the early music genres and adult contemporary; however,
ragtime shifted to the middle of the spectrum. At the high end, romantic and
flamenco keep their very high values; they are, however, surpassed by techno,
which has made a huge shift. This seems to be a good unique feature to help
characterize techno music. The cluster of improvisatory, blues, swing, jazz-related
genres is a bit more scattered around the mid and high part of the spectrum. It
seems remarkable how tango has almost all of its recordings concentrated around
the value 5, while also displaying distant outliers.
102
Number of Relatively Strong Pulses
Global Mean
15
Global Median
10
Beat peaks
5
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Baroque
Soul
Alternative Rock
Psychedelic
Hardcore Rap
Hard Rock
Cool
Classical
Swing
Soul Blues
Chicago Blues
Blues Rock
Ragtime
Reggae
Rock and Roll

Bebop
Salsa
Funk
Metal
Punk
Celtic
Flamenco
Romantic
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.16: Number of Relatively Strong Pulses.
5.8.12 Rhythmic Looseness
The rhythmic looseness is defined as the “average width of beat histogram

peaks (in beats per minute). Width is measured for all peaks with frequencies at
least 30% as high as the highest peak, and is defined by the distance between the
points on the peak in question that are 30% of the height of the peak” (McKay,
2004, p. 70).
This feature (figure 5.17) seems a highly contrasting one. While most genres
appear to have very few rhythmic looseness, a relevant cluster show great variety
in this matter, highly distorting the scale. This cluster, as expected, belongs to
the improvisatory, blues, swing, jazz-related genres. What also seems noticeable
is how the original fado corpus shows no rhythmic looseness at all, alongside with
the early music genres, however, the edited corpus displays a very relevant value at
the opposite side of the spectrum. We attribute this fact to encoding issues: while
the raw data has quantized values, the edited corpus has the humanizer included.
We strongly believe the edition better reflects the actual practice.
103
Rhythmic Looseness
15
Global Mean
Global Median
10
Beats per minute
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Baroque
Alternative Rock
Psychedelic
Hardcore Rap
Hard Rock
Classical
Ragtime
Reggae
Funk
Metal
Punk
Celtic
Flamenco
Romantic
Salsa
Blues Rock
Rock and Roll
Soul
Soul Blues
Cool
Swing
Chicago Blues
Bebop
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado
Pop Rap
Dance Pop
Fado (Edition)
Medieval
Traditional Country
Tango
Techno
Figure 5.17: Rhythmic Looseness.
5.8.13 Polyrhythms
Polyrhythms is defined as the “number of beat peaks with frequencies at least

30% of the highest magnitude whose bin labels are not integer multiples or factors
(using only multipliers of 1, 2, 3, 4, 6 and 8) (with an accepted error of +/- 3
bins) of the bin label of the peak with the highest magnitude. This number is
then divided by the total number of beat bins with frequencies over 30% of the
highest magnitude” (McKay, 2004, p. 70).
According to this definition one might see (figure 5.18) that fado is among the
genres in which the polyrhythms prevalence is higher, and our edition slightly
emphasizes that aspect. Traditional and contemporary country genres, adult con-
temporary and soul appears at the high end of the spectrum, which is topped by
renaissance music, certainly driven by polyphonic recordings. On the other hand,
dance pop and bossa nova, Celtic music and surprisingly funk seem relatively poor
regarding this feature.
104
Polyrhythms
1.0
Global Mean
Global Median
0.8
0.6
Frequency
0.4
0.2
Country Blues
Modern Classical
Adult Contemporary
Funk
Celtic
Romantic
Salsa
Swing
Metal
Blues Rock
Flamenco
Rock and Roll
Soul Blues
Chicago Blues
Ragtime
Reggae
Punk
Psychedelic
Hardcore Rap
Hard Rock
Classical
Bebop
Baroque
Alternative Rock
Cool
Soul
Renaissance
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Pop Rap
Fado
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.18: Polyrhythms.
5.8.14 Rhythmic Variability
Rhythmic variability is defined as the “standard deviation of the bin values

(except the first 40 empty ones)” (McKay, 2004, p. 70).
Looking at the figure 5.19, it seems that fado, alongside with renaissance music,
top the genres with more contrasting rhythmic figurations, which in a certain way
can be expected, since they can have very short and quick passages but also very
prolonged, hovering notes. Our edition, thanks to the standardization of time
signatures, and the conformation of note durations to fit into those binary time
signatures, obscures that effect and throws fado into the middle of the spectrum.
On the lower part of it one can encounter bebop, swing and jazz soul, which can be
also expected, since one often encounters very similar figurations: the variability
is, instead, obtained by small and micro-deviations.
105
Rhythmic Variability
0.05 Global Mean
Global Median
0.04
Relative Frequency
0.03
0.02
0.01
Country Blues
Modern Classical
Adult Contemporary
Bebop
Swing
Chicago Blues
Cool
Soul Blues
Salsa
Psychedelic
Rock and Roll
Celtic
Blues Rock
Funk
Romantic
Flamenco
Soul
Punk
Metal
Reggae
Classical
Hard Rock
Ragtime
Alternative Rock
Hardcore Rap
Baroque
Renaissance
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Dance Pop
Fado (Edition)
Pop Rap
Fado
Medieval
Traditional Country
Tango
Techno
Figure 5.19: Rhythmic Variability.
5.9 Note Duration
The average note duration is simply the “average duration of notes in seconds”
(McKay, 2004, p. 70). Looking at the table 5.5, the average note duration, in
seconds, is quite similar when considering both the corpus (0,36) and only the
melodies (0,38). If one considers the average tempo of fados at around 78 bpm,
then this average approximates an eight note value (0,384). When including the
average note variability (0,22 and 0,26) one obtains an average span (0,14 – 0,58
and 0,12 – 0,64) from a little less than a sixteenth note (0,192) to a bit more than
a dotted eight note (0,577) as preferred durations for the rhythms. Overall the
maximum note durations on the melodies alone are inferior (average 0,57 versus
0,71 and absolute 4,97 versus 5,00), as well as the average minimum (0,18 versus
0,23) but the absolute minimum note durations are superior (0,05 versus 0,01),
which is consistent with the range of the average note duration being inferior (0,38
versus 0,48). This fact implies more uniformity (and slight brevity, probably due
to ornamentation) and a more constrained range in the rhythm of the melodies
opposed to a wider number of possible rhythms for the accompaniments, which also
might contain extreme values. We have calculated the staccato incidence as well,
106
5.9. Note Duration
which is defined by the number of notes shorter than 0,1 seconds divided by the
total number of notes (McKay, 2004, p. 71). This parameter would capture mainly
grace notes, melismas and similar ornamentations. This ratio is negligible within
our corpus (0,02), and only slightly superior on the melodies (0,04), with a highly
skewed distribution (5,30 and 5,05) and kurtosis (35,90 and 33,61), which confirms
the existence of only a few melodies in which ornamentations are notated. In real
performances and recordings this number should be more uniform and relevant.
Table 5.5: Statistics of Note Duration
Note Duration Avg Med St Dev Var Min Max Range Skew Kurt
Average 0.36 0.36 0.09 0.01 0.23 0.71 0.48 0.81 1.37
Variability 0.22 0.20 0.10 0.01 0.08 0.59 0.51 1.42 2.78
Maximum 1.73 1.66 0.78 0.61 0.53 5.00 4.47 1.19 2.43
Minimum 0.12 0.11 0.07 0.00 0.01 0.25 0.23 0.28 -1.35
staccato inc. 0.02 0.00 0.06 0.00 0.00 0.46 0.46 5.30 35.90
Melodies-only Avg Med St Dev Var Min Max Range Skew Kurt
Average 0.38 0.39 0.08 0.01 0.18 0.57 0.38 -0.37 0.23
Variability 0.26 0.24 0.11 0.01 0.10 0.73 0.63 1.27 2.37
Maximum 1.47 1.38 0.75 0.56 0.48 4.97 4.50 2.04 5.76
Minimum 0.14 0.12 0.06 0.00 0.05 0.25 0.20 0.12 -1.72
staccato inc. 0.04 0.00 0.09 0.01 0.00 0.70 0.70 5.05 33.61
The average note duration can be also seen in comparative terms (figure 5.20).
Fado is among the taxonomies with a higher median of average note duration
along with the early music, modern classical and country genres, including country
blues. Notice the incredibly large span of modern classical music standing out of
all the rest. At the opposite side of the spectrum salsa, rock and roll and bebop
seem to be the most busy genres, which is somewhat expected.
5.9.1 Variability of Note Duration
The variability of note duration is defined as the “standard deviation of note

durations in seconds” (McKay, 2004, p. 71).
This particular feature (figure 5.21) is useful to understand the kind of rhythms
involved in the different taxonomies. Fado is positioned at the lower part of the
spectrum alongside with genres like ragtime, bluegrass, country blues, flamenco,
punk, salsa, reggae, rock and roll and bebop. On the other side, one can find
107
Seconds Seconds
108
0.0
0.5
1.0
1.5
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Ragtime Salsa
Bluegrass Rock and Roll
Country Blues Bebop
Flamenco Reggae
Global Mean
Global Mean
Punk Funk
Global Median
Global Median
Salsa Techno
Fado Punk
Reggae Metal
Rock and Roll Blues Rock
Bebop Chicago Blues
Metal Jazz Soul

Bluegrass
Funk
Dance Pop
Celtic
Ragtime
Chicago Blues
Celtic
Jazz Soul
Flamenco
Blues Rock
Hardcore Rap
Techno
Tango
Classical
Psychedelic
Hardcore Rap
Pop Rap
Alternative Rock
Cool
Cool
Hard Rock
Psychedelic
Alternative Rock
Tango
Romantic
Modern Classical
Bossa Nova
Medieval
Average Note Duration
Soul
Baroque
Classical
Variability of the note duration

Pop Rap
Soul Blues
Swing
Swing
Bossa Nova Country Blues
Figure 5.20: Average Note Duration.

Dance Pop Fado (Edition)
Hard Rock
Figure 5.21: Variability of Note Duration.

Traditional Country
Romantic Smooth Jazz
Traditional Country Fado
Soul Blues Modern Classical
Soul Contemporary Country
Renaissance Baroque
Contemporary Country Medieval
Smooth Jazz Adult Contemporary
Adult Contemporary Renaissance
5.9. Note Duration
adult contemporary, smooth jazz, contemporary country and renaissance. This

seems surprising because neither one can find relevant connections with the previ-
ous graph (average note duration) nor does one seem to be able to find a tendency
among clusters of taxonomies that have been paired so far in many features. There-
fore this seems a unique feature in that many styles actually behave independently:
fado and country blues had high average note durations but small variation; adult
contemporary had also high average note durations while maintaining high varia-
tion. Rock and roll and salsa have low average note durations and also have low
variation, while ragtime and bluegrass have remarkably low variation while having
a more generous average note durations.
One can infer based on this graph that the low variation on average note du-
ration must represent similar rhythm figurations, or at least the avoidance of
contrasting durations.
5.9.2 Maximum Note Duration
The maximum note duration is defined as the “duration of the longest note (in
seconds)” (McKay, 2004, p. 71).
Observing the figure 5.22, fado has low maximum note durations, which also
happens in the cases of ragtime, flamenco, country blues and bluegrass, all below
the four seconds mark. At the other side of the spectrum, adult contemporary,
romantic and hard rock styles present recordings which have at least one very long
note. This is also the case of some recordings among several genres, remarkably on
celtic and hardcore rap, which have very high averages compared to the median,
indicating extreme outliers. This might simply indicate the existence of some
recordings which, for some reason, have at least one very, very long note, which is
not representative of the practice, by itself. On the other hand, the total absence of
such notes (including outliers), in the genres present at the bottom of the spectrum,
seems to point out to a defining condition of these genres. The reason for that
condition might be something as simple as the instrumentation conventionally
used to actually be unable to produce such long notes.
109
Maximum Note Duration
30 Global Mean
Global Median
25
20
Seconds
15
10
Country Blues
Modern Classical
Adult Contemporary
Ragtime
Flamenco
Bebop
Celtic
Salsa
Funk
Punk
Reggae
Renaissance
Rock and Roll
Hardcore Rap
Chicago Blues
Baroque
Cool
Metal
Swing
Blues Rock
Soul Blues
Classical
Soul
Psychedelic
Alternative Rock
Hard Rock
Romantic
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.22: Maximum Note Duration.
5.9.3 Minimum Note Duration
The minimum note duration is simply the “duration of the shortest note (in
seconds)” (McKay, 2004, p. 71).
Looking at the figure 5.23, the minimum note duration in fado is among the
highest ones along with the early music genres, hardcore rap and flamenco. This
confirms the idea that the overall variability of durations in fado comprises a
rather short span when compared to other taxonomies, indicating relatively rather
uniform rhythms. On the other side of the spectrum, swing, funk, Chicago blues
and salsa have the shortest short notes.
5.9.4 Staccato Incidence
The staccato incidence refers to the “number of notes with durations of less
than a 10th of a second divided by the total number of notes in the recording”
(McKay, 2004, p. 71).
110
Relative Frequency Seconds
0.0
0.2
0.4
0.6
0.8
0.0
0.1
0.2
0.3
0.4
Fado Swing
Fado (Edition) Funk
Medieval Chicago Blues
Renaissance Salsa
Global Mean
Global Mean
Baroque Soul Blues
Global Median
Global Median
Country Blues Rock and Roll

Ragtime Contemporary Country
Modern Classical Soul
Classical Traditional Country
Flamenco Psychedelic
Bluegrass Adult Contemporary
Alternative Rock Reggae
Metal Smooth Jazz
Hardcore Rap Jazz Soul
Contemporary Country Dance Pop
Smooth Jazz Tango
Traditional Country Blues Rock
Celtic Hard Rock
Punk Techno
Adult Contemporary Metal
Romantic Bebop
Hard Rock Bossa Nova
Jazz Soul Celtic
Swing Romantic
Staccato Incidence
Cool Pop Rap
Bossa Nova Alternative Rock
Minimum Note Duration
Blues Rock Cool

Soul Modern Classical
Pop Rap Country Blues
Figure 5.24: Staccato Incidence.

Soul Blues Ragtime
Chicago Blues
Figure 5.23: Minimum Note Duration. Bluegrass
Psychedelic Punk
Tango Classical
Funk Flamenco
Rock and Roll Hardcore Rap
Reggae Baroque
Bebop Fado
Dance Pop Fado (Edition)
Salsa Medieval
Techno Renaissance
111
5.9. Note Duration
Looking at the figure 5.24,as expected, based also on the previous graphs, fado
is actually the genre with the lowest value regarding this feature, next to the early
music, country blues and ragtime. At the opposite side one finds techno, salsa,
dance pop and bebop. Given the way this feature is defined, more than the true
meaning of stacatto, which is fact an articulatory feature implying a pause of equal
value following a certain note duration, we think this feature allow us to find the
prevalence of short note durations overall. Having that in mind, one realizes that
it makes sense for genres known to have speedy rhythms and highly percussive
figurations to stand out, while slower genres to stay behind. We also think the
way the musical transcriptions were notated has influence, in this feature, reflecting
the near absence of melismatic and ornamental notation in most scores, while in
the most recent practice they could actually appear.
5.9.5 Average Time Between Attacks
The average time between attacks is defined as the “average time in seconds
between Note On events (irregardless [sic] of channel)” (McKay, 2004, p. 71).
Average Time Between Attacks
Global Mean
0.8 Global Median
0.6
Seconds
0.4
0.2
0.0
Country Blues
Modern Classical
Adult Contemporary
Bebop
Chicago Blues
Swing
Salsa
Rock and Roll
Cool
Soul Blues
Blues Rock
Soul
Psychedelic
Funk
Celtic
Reggae
Punk
Hard Rock
Romantic
Metal
Flamenco
Ragtime
Alternative Rock
Classical
Hardcore Rap
Baroque
Renaissance
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Dance Pop
Pop Rap
Fado (Edition)
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.25: Average Time Between Attacks.
112
5.9. Note Duration
This feature (figure 5.25) is especially useful to discern how “busy” a certain
genre is. One can see how fado is among the genres with highest time between
attacks next to the early music, country blues and modern classical music. This
is expected given the usual slow, dragged pace, on big part of the repertoire of
these genres and also the absence or minimal presence of percussion. On the other
hand, bebop, Chicago blues, swing, salsa and rock and roll seem rather busy,
and of course, the amount of different instruments playing at the same time, and
percussive lines, will highly contribute to this effect.
5.9.6 Variability of Time Between Attacks
The variability of time between attacks is defined as the “standard deviation

of the times, in seconds, between Note On events (irregardless [sic] of channel)”
(McKay, 2004, p. 71).
Variability of Time Between Attacks

0.7
Global Mean
0.6 Global Median
0.5
0.4
Seconds
0.3
0.2
0.1
0.0
Country Blues
Modern Classical
Adult Contemporary
Salsa
Celtic
Punk
Bebop
Rock and Roll
Ragtime
Metal
Reggae
Funk
Chicago Blues
Soul
Swing
Hard Rock
Cool
Blues Rock
Alternative Rock
Psychedelic
Soul Blues
Hardcore Rap
Flamenco
Classical
Baroque
Romantic
Renaissance
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Pop Rap
Fado
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.26: Variability of Time Between Attacks.
In this feature (figure 5.26), compared to the previous one, one notices that fado
is slightly less high in relative terms. Although the cluster of early music styles and
country blues remain close, romantic, adult contemporary and tango now show up
113
mingled in the lot. This seems to imply that these styles actually have more places
to “breathe in” and privilege the existence of rests and suspensions here and then,
which is certainly true in the case of fado. On the opposite side, bluegrass, salsa,
Celtic, funk and bebop have very low variability, implying relative busyness all
along.
5.9.7 Average Time Between Attacks For Each Voice
The average time between attacks for each voice is defined as the “average
of average time in seconds between Note On events on individual channels that
contain at least one note” (McKay, 2004, p. 71).
Average Time Between Attacks For Each Voice
12 Global Mean
Global Median
10
8
Seconds
0
Country Blues
Modern Classical
Adult Contemporary
Bebop
Celtic
Ragtime
Rock and Roll

Punk
Chicago Blues
Cool
Blues Rock
Classical
Flamenco
Salsa
Metal
Psychedelic
Baroque
Renaissance
Soul Blues
Romantic
Funk
Soul
Alternative Rock
Hard Rock
Swing
Reggae
Hardcore Rap
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado (Edition)
Fado
Dance Pop
Pop Rap
Medieval
Traditional Country
Tango
Techno
Figure 5.27: Average Time Between Attacks For Each Voice.
It is relevant to notice the discrepancies between this feature (figure 5.27) and
the previous ones. While, on the whole, fado seemed to be one of the less busy
taxonomies, when analyzed voice by voice one realizes that in fact each voice of
fado is rather busy. Fado is among the genres with less average times among with
bebop, Celtic and bluegrass musics. At the opposite side lie country blues, techno,
adult contemporary and hardcore rap, a fact that seems to dissociate fado from
their usual counterparts of story-telling music genres.
114
5.9. Note Duration
This also seems to point out the idea that some taxonomies might have a huge
contrast among their voices: some voices might be incredibly busy while others
are rather sparse, or, even if they have relatively sparse voices they might not be
aligned, thus increasing the overall business, while remaining sparse when analyzed
voice by voice. The fact that fado is relatively much busier when seen voice by
voice might mean that its voices are well aligned with each other, when played as
a group, thus increasing the blending.
5.9.8 Average Variability of Time Between Attacks For

Each Voice
The average variability of time between attacks for each voice is defined as
the “average standard deviation, in seconds, of time between Note On events on
individual channels that contain at least one note” (McKay, 2004, p. 71).
Average Variability of Time Between Attacks For Each Voice

10
Global Mean
Global Median
8
6
Seconds
0
Country Blues
Modern Classical
Adult Contemporary
Flamenco
Ragtime
Celtic
Renaissance
Punk
Bebop
Rock and Roll

Chicago Blues
Baroque
Cool
Classical
Soul
Romantic
Blues Rock
Funk
Soul Blues
Psychedelic
Swing
Reggae
Metal
Hard Rock
Salsa
Alternative Rock
Hardcore Rap
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.28: Average Variability of Time Between Attacks For Each Voice.
Looking at the figure 5.28, fado stands out at the bottom of the spectrum
alongside bluegrass, flamenco and Celtic music, while techno, adult contemporary,
dance pop and hardcore rap occupy the highest slots. This could well be a defining
115
feature for the genre, possibly meaning the existence of very similar clusters of
rhythms. When one thinks about the constant repetition of the ostinati it makes
sense, and if one thinks of the constant repetition of the melodic rhythms due to
the strophic nature of the practice (bearing in mind that the transcriptions do
not account for groove or improvisation), then it seems to make total sense. We
believe that, if this feature was measured taking into account human performances
then the relative outcomes would be much different.
5.10 Harmonic Progressions
Through the analyzes of the standardized data it was possible to study the
harmonic progressions commonly used in fado and confirm the suspicions about
the vast recurrence of the tonic-dominant and dominant-tonic patterns. Tables
were built comparing the progressions within the different sections and we have
used them to perform calculations regarding their frequency. These figures are
important since they reveal empirically what are exactly the frequencies of the
harmonic progressions through history. This provides concrete values and weights,
so the progressions can be modeled in a representative way, and also the internal
complexity and variety among different sets of sources.
Within section A it is possible to observe 33 different chord progressions (table

5.6), from which 7 are recurring and 26 are unique suggesting some variety across
the repertoire. 93% of the fados start in the tonic degree, 5% in the dominant and
2% in sixth degree. An ending perfect cadence is overwhelmingly present in 92%
of the cases (one via a pivotal parallel modulation), with only 7% of the section
A progressions ending with a half cadence, and a single 1% ending in a pivotal
relative. The tonic-dominant-dominant-tonic progression (T|D|D|T) is clearly the
archetype, as Ernesto Vieira described, accounting for 36% of cases in Major
mode and 25% in minor mode, for a grand total of 61% of total cases. The other
recurrent progressions are, as predictable, the tonic-subdominant-dominant-tonic
movements (T|S|D|T), with 9% of the cases (either via second or fourth degree
motions). Secondary dominants and diminished chords are extremely rare.
Within section B the variety increases (table 5.7): there are 55 different chord
progressions, from which 45 are unique (although some are highly derivative). Far
fewer section B progressions start in the tonic degree (54%). The ending in a
116
5.10. Harmonic Progressions
perfect cadence is still overwhelmingly present in 90% of the cases. The T|D|D|T
progression is still an important archetype, accounting for 15% of cases in Major
mode and 17% in minor mode. The cluster of T|S|D|T movements represent 10%
of the cases (either via second or fourth degree motions). Clusters of two chained
cadences emerge: a plagal cadence followed by a perfect cadence (S|T|D|T) repre-
sents 8% of the cases, while a chain of two perfect cadences (D|T|D|T), considering
the clusters of both dominant and secondary dominant movements (V/x to x), ac-
count for 20%. Such a high value is enough for them to be considered an archetype
within the B section.
The C and D sections have to be analyzed having in mind that our sample is
smaller (not all fados have these sections) and two different sets of formal possibili-
ties emerge – often C and D sections are not true C and D sections but are, in fact,
the equivalent of A and B sections of a juxtaposed fado. As discussed previously,
often a simple ABAB fado in a minor mode is followed by another ABAB fado
in the Major mode (either relative or parallel), the melodic and rhythmic material
having little to no connection to the previous material. Therefore, in these cases,
the C and D sections will display the same patterns as an A and B section would.
Other times, however, the C and D sections are indeed C and D sections, this
meaning they are related to the former material and are derivatives, variations or
evolutions on the previous A and B sections. And in these cases they will tend to
improve on the harmonic material in more interesting and complex ways.
Within section C it is possible to observe that there are 40 different chord

progressions (table 5.8), of which 28 are unique. This shows an increase in variety
compared to sections A and B. Only 50,8% of the fados start in the tonic degree,
22,5% in the subdominant, 18% in the dominant, and a few others in other degrees.
An ending perfect cadence is overwhelmingly present in 91% of the cases (3% via
secondary dominants). The T|D|D|T progression is still the archetype accounting
for 15,4% of cases in Major mode and also 15,4% in minor mode, for a grand
total of 30,8% of total cases. We believe this value reflects mostly the fados that
are indeed behaving like A section fados. The cluster of S|T|D|T represents 7,5%
of the cases, while D|T|D|T account for 15% considering the clusters of both
dominant and secondary dominant movements (V/x to x). Such a high value is
enough to consider them an archetype within the C section. It is noteworthy the
presence of a cluster of unique progressions containing movements departing from
the second degree (9%) and a considerable increase of the presence of diminished
117
chords (10,5%) and also borrowed chords from the parallel (namely the typical bVI
and bVII chords appearing in this section in 6% of the progressions). These later
clusters of more complex progressions reflect true C sections that are evolutions
on the previous sections.
Within section D it is possible to observe that there are 37 different chord pro-
gressions (table 5.9), of which 20 are unique. This represents an increase in variety
compared to sections A and B, but a decrease in relation to C. 65,8% of the fados
start in the tonic degree, 19% in the subdominant, 13,3% in the dominant, and
few others in other degrees. An ending perfect cadence is overwhelmingly present
in 86,7% of the cases (3,8% via secondary dominants). The T|D|D|T progression
is still the archetype accounting for 17% of cases in Major mode but only 9,5%
in minor mode, for a grand total of 26,5% of total cases. The cluster of T|S|D|T
movements represent 13,3% of the cases (either via second or fourth degree mo-
tions). We believe these values reflects fados that are behaving like section B fados.
The cluster of S|T|D|T represents 7,6% of the cases, while D|T|D|T accounts for
9,5%. While such value is still enough to consider them an archetype within the
D section, it clearly loses strength compared to B and C sections. The presence
of diminished chords increases (13,3%) and overall there are more solutions for
closing segments.
118
Table 5.6: Harmonic Progressions of section A.
A1 A2 A3 A4 Total
I I I V7 1%
V7 I 1%
I-V7 I-V7 V7 1%
IV V7 I 1%
ii V7 I 3%
V7 V7 I 36%
vii V7 I 1%
I-#i07 V7 ii-V7 I 1%
I-i07 V7 V7 I 1%
I-V I IV-V I 1%
IV IV IV I 2%
i i i V7 1%
i-V7 V7 1%
iv i 1%
i-iv V7 V7-i 1%
i-V7 V7 V7-i 1%
iv iv V7 1%
V7 I 1%
i 4%
V7 V7 i 25%
vii07-V7/III III 1%
V7/V V7 i 2%
i-bVII7 V7 V7 i 2%
i-V7 i ii0 i 1%
iv-i V7 1%
i-V7 i-V7 i-V7 1%
i07-ii07 i-V7/iv iv-V7 i 1%
V7 I V7 I 1%
i V7 i 2%
V7 I I 1%
I vi 1%
vi I IV I 1%
ii V7 I 1%
11 25 32 33 100%
119
Table 5.7: Harmonic Progressions of section B.
B1 B2 B3 B4 Total
I IV I-V7 I 1%
V7 I 1%
ii V7 I 5%
V7 V7 I 15%
vii0 V7 I 1%
I-V7/ii ii V7 I 1%
V7 V7 I 1%
III V7 i i 1%
V7/III V7 i 1%
V7/III-V7/V V7 i 1%
IV I V7 I 3%
IV V7 I 1%
i i V7 i 2%
ii V7 i 1%
iv i-V7 i 1%
V7 i 1%
V ii V 1%
V7 i-V7 i 4%
iv V7 1%
V7 i 13%
V7/iv-iv iv-i V7-I 1%
V7 i-V7-i 1%
i-V7 i-V7 i-V7 NC 1%
i-V7/iv ii0 V7 i 1%
iv i-V7 i 1%
ii V7 I-V7 I 1%
ii-V7 V7 V7-I I 1%
ii0 i V7 i 1%
ii7 i V7 i 1%
iv i V7 i 2%
V7 iv V7 1%
iv-V7 i i-V7 i 1%
#iv07 I ii-V7 I 1%
V V V I 1%
V7 I V7 I 1%
i i-V7 i 1%
iv V7 1%
V7 V7 1%
V7/V V 1%
iv V7 i 1%
V7 I I 1%
I-V7 I 1%
i III 1%
ii-V7 I 1%
V7-iv iv-V7 V7-i V7-i 1%
V7/III III iv-V7 i 1%
V7 i 2%
V7/ii ii V7 I 6%
V7/iii iii 1%
V7/iv iv V7 i 3%
Vi/vi vi V7 I 1%
VI VI V7 i 1%
vi V7 V7 I 1%
vi-V V7 V7 I 1%
vi-V7/vi vi-IV I-V7 I 1%
25 42 55 55 100%
120
Table 5.8: Harmonic Progressions of section C.
C1 C2 C3 C4 Total
I ii V7 I 3%
ii-V7/ii-ii ii-V7 I 1.5%
V7 I-V7 I 1.5%
V7 I 14.9%
vi-ii I-V7 I 1.5%
I-i0 V7 V7 I 1.5%
I-#iv07 V7 IV I-V7 1.5%
I-V7 I-V7 I-V7 I-V7 1.5%
I-V7/ii ii V7 I 1.5%
I/III V7/III V7/III I/III 1.5%
III III V7/III-V7/V V 1.5%
iv-V7 i V7 1.5%
IV V i-V7 i 1.5%
IV-#iv07 I I-V7 I 1.5%
IV-V7 I-V7/ii ii-V7 I 1.5%
i bVII bVII bVI 1.5%
iv V7 i 1.5%
V7 i-V7 i 1.5%
V7 i 14.9%
i-bVI V7 i-bVI V7-i 1.5%
i-V7 i-V7-i-V7/iv iv V7/III-III-V7/iv 1.5%
i-VII vi V7 i-I 1.5%
ii-i i-V7/V V7 I 1.5%
ii-ii0 I V7 I 1.5%
ii-V7/iii I V7 I 1.5%
ii-V7/vi vi V7 I 1.5%
ii0 i V7 I 1.5%
ii07-V7 i ii07-V7 i 1.5%
iv i V7 I 3%
V7 i 4.5%
V7/V-V7 V7/IV 1.5%
V7 I V7 I 4.5%
i V7 i 4.5%
V7/III III V7 I 1.5%
V7/ii ii V7 I 3.0%
V7/iv iv i-V7 i 1.5%
V7/V IV-#i07 V7/V-V7 I 1.5%
V7/vi IV V7/ii ii 1.5%
bVI I ii I 1.5%
bVI-V7 V7 V7-i i 1.5%
29 36 39 40 100%
121
Table 5.9: Harmonic Progressions of section D.
D1 D2 D3 D4 Total
I I I ii 3.8%
V7 3.8%
V7 V7 1.9%
IV i-V7 I 1.9%
ii V7 I 3.8%
V7 I-ii-V7 I 1.9%
I-V7 I 1.9%
V7 I 15.1%
I-i07 ii-V7 V7 I 1.9%
I-V7 I-V7 I-V7 I 1.9%
I-V7/ii ii V7 I 1.9%
ii-ii07 V7 I 1.9%
V7 V7 I 1.9%
IV-#iv07 I V7 I 1.9%
i ii i-V7 i 1.9%
ii07 V7 i 1.9%
iv V7 i 1.5%
V7 V7 I 1.9%
V7 i 7.5%
V7/V V7 i 1.9%
vii07 V7 i 1.9%
i-V7 i iv-V7/iv iv 1.9%
i-V7/iv iv i-V7 i 1.9%
ii I V7 I 1.9%
ii0-I I-V/V V7-I I 1.9%
iv i V7 i 5.7%
V7/V V7 1.9%
V7 iv V7 1.9%
V7 i 1.9%
iv-V7 iv-V7 i-v7 i-V7-i 1.9%
V7 I V7 I 1.9%
i V7 i 3.8%
V7/III III iv-VI V7 1.9%
V7/ii ii V7 I 1.9%
V7/iv iv V7 i 1.9%
bVI V7 V7 i 1.9%
vii07 V7 i-V7 i 1.9%
18 30 35 37 100%
122
In order to better perceive the trends analyzed, it is useful to visually ag-

gregate the distribution of Harmonic progressions in terms of functional clusters
(figure 5.29). We have highlighted the Tonic-dominant-dominant-tonic (T|D|D|T)
movements, but also the tonic-subdominant-dominant-tonic (T|S|D|T) one and
the chain of plagal-dominant cadences (S|T|D|T), as well as Dominant-dominant
(including secondary dominants – D|T|D|T) ones.
Figure 5.29: Distribution of Harmonic Progressions.
It seems clear from observing the figure that there is an enormous variety of
unique or infrequent different harmonic progressions among the data, which seems
to elude, again, Ernesto Vieira’s model. However, some patterns arise. It is over-
whelmingly frequent that all sections end with the dominant-tonic movement in
their third and fourth bars, and what really varies are the first two bars. The
archetype T|D|D|T is clear within the A section, but loses some of its strength in
the remaining ones. However, it can be noticed its preponderance overall. The
T|S|D|T movement is preferred as a “closing” movement with notorious prepon-
derance in even sections (B and D), almost vanishing in C section. The chain of
dominant cadences is an important archetype regarding the B and C sections, but
loses strength in D section in favor of T|S|D|T. The chain plagal cadence – perfect
cadence does not exist in section A and is then found on all other sections in sim-
ilar relevant percentages. The overall complexity of the progressions augments as
noticed by the increasing weight of the “others”. This implies a fragmentation and
absence of clear tendencies in favor of unique movements, which is consistent with
123
how most tonal traditions work: starting simple and then becoming increasingly
more complex and varied through the span of time. One of the criteria that most
clearly shows that is the complexity found as we depart from Ernesto Vieira’s
simple model, in relation to harmonic progression. If it is true that many fados
respect that model as it is, others add sub-dominant degrees (second or fourth de-
gree chords), while others venture into diminished seventh and augmented sixths
chords, various modulations and secondary dominants.
The notation of dynamics and articulations varies, and some fados are pre-
sented following the typical conventions and complexity of Romantic piano music
(for instance the “Canção da Noite” [Night song] or “Fado das Três Horas” [3
a.m. Fado], attributed to Reynaldo Varella (Neves, 1893, p. 208) is fairly sim-
ilar to a Chopin Nocturne). These discrepancies seem to point not only to the
documentation of several musical occurrences in different stages of its evolution,
but also to their composers’ music erudition. The transcriptions of fados coming
across the oral tradition (with unknown composers and based on intermediary
sources distant from their origins) are very different from the ones whose com-
posers were still alive, present or were schooled. This does not mean that there
was a linear evolution in the complexity of fado instrumentals, and that a fado
with only two chords is more primitive than a more complex one. It is probable
that such an assumption might be true in most cases; however, one should not
ignore the coexistence of several composers, with many different backgrounds and
erudition, and multiple cultural and social contexts occurring in the same space
and time. Even nowadays, very simple fados with two or three chords are being
composed. Harmonic complexity, in itself, is a very limited criterion to determine
the social context or date of a given fado. It seems that the model of Ernesto
Vieira evolved and, even on the date it was written, it was already not enough to
fully characterize the practice.
5.11 Tonality
Both the initial and final tonality of the fados reflect a distribution privileging
simplicity of execution (figure 5.30), since most practitioners are amateurs. 54% of
the fados start in the Major mode, almost evenly distributed among the preferred
G major (11%), D major (10%), C major (9%) and F major (8%). All other Major
options have much lesser frequencies and some, like C#/Db Major or F#/Gb
124
5.11. Tonality
Major are even non-existent. The other 46% of the fados are in the Minor mode
with strong emphasis on D minor (14%) and A minor (13%), followed closely by G
minor (9%) and C minor (5%) at distance. Six minor tonalities (C#/Db, D#/Eb,
F#/Gb, G#/Ab, A#/Bb and B) are inexistent.
Figure 5.30: Distribution of tonalities.
Ernesto Vieira’s assumption that most fados are in the minor mode is not
confirmed by the data. These tonality choices seem logical, favoring the use of
few accidentals combined with the easiness to play them on the guitarra and the
viola (favoring open string positions, knowing their tunings to be B-A-E-B-A-D
and E-B-G-D-A-E). This last factor could explain the very few fados in Bb major,
despite having fewer accidentals when compared to E or A major. There are also
fewer fados than one would expect in E minor (2%), and that is probably related
with how one modulates within the same work as explained further ahead.
As a way to complement the distribution of tonalities we have also calculated

the Dominant Spread, defined as the “largest number of consecutive pitch classes
separated by perfect 5ths that accounted for at least 9% each of the notes”, as
well as the “number of peaks in the fifths pitch histogram that each account for at
least 9% of all Note Ons” in order to perceive strong tonal centres (McKay, 2004,
p. 74). These methods identified one to three tonal centres for fados, averaging
on two, which is consistent with the results found and calculated by us manually,
also complemented with the information that on average there are four relevant
125
pitches at a perfect 5th distance – reinforcing the fact of a paired tonic-dominant

relationship on each fado.
Table 5.10: Statistics of Tonal Centres
Corpus Avg Med St Dev Var Min Max Range Skew Kurt
Dominant Spread 3.85 4.00 1.16 1.34 2.00 7.00 5.00 1.02 1.02
Strong Tonal Centres 2.04 2.00 0.62 0.38 1.00 3.00 2.00 -0.02 -0.32
5.11.1 Strong Tonal Centres
Comparative features regarding tonality could be useful so we have calculated

them and build figures (figure 5.31).
Strong Tonal Centres
4.0
3.5
3.0
Peaks
2.5
2.0
1.5
Global Mean
Global Median
1.0
Country Blues
Modern Classical
Adult Contemporary
Metal
Hardcore Rap
Flamenco
Funk
Soul Blues
Blues Rock
Rock and Roll
Chicago Blues
Romantic
Soul
Punk
Alternative Rock
Psychedelic
Salsa
Hard Rock
Reggae
Celtic
Classical
Bebop
Cool
Swing
Renaissance
Baroque
Ragtime
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Pop Rap
Fado
Fado (Edition)
Dance Pop
Medieval
Traditional Country
Techno
Tango
Figure 5.31: Strong Tonal Centres.
In a disappointing way this feature seems almost useless to define any of the
taxonomies proposed, since all of them have median and mean around two tonal
centres. Fado stands near the middle of the spectrum not standing out.
126
5.12. Modulations
Dominant Spread
7
Global Mean
Global Median
6
5
Pitch Class
Country Blues
Modern Classical
Adult Contemporary
Metal
Hardcore Rap
Flamenco
Funk
Soul Blues
Blues Rock
Rock and Roll
Chicago Blues
Romantic
Soul
Punk
Alternative Rock
Psychedelic
Salsa
Hard Rock
Reggae
Celtic
Classical
Bebop
Cool
Swing
Renaissance
Baroque
Ragtime
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Pop Rap
Fado
Fado (Edition)
Dance Pop
Medieval
Traditional Country
Techno
Tango
Figure 5.32: Dominant Spread.
5.11.2 Dominant Spread
Looking at the figure 5.32, this feature does not seem very useful to define fado,
as well, as it does not stand out in any way, aligning in the middle, along with
tango, bossa nova, jazz soul and salsa. Metal, modern classical and hardcore rap
lie at the bottom part of the spectrum, while renaissance, baroque and ragtime
occupy the top part.
5.12 Modulations
53% of the fados are monotonal, which contributes to the idea of simplicity
associated with these songs. However the remaining 47% have one or more relevant
modulations. By relevant modulations we mean the clear establishment of a new
tonal center for at least an entire section. While many fados only have one relevant
modulation, thus being purely bitonal, others return to the initial tonality, while
others present several tonalities. This aspect is also linked with the overall form
(and thus with the lyrical structure) since, for instance, in a bitonal fado an
127
AABB form will have one modulation from A to B, while an ABBA form will
carry on two modulations – from A to B and then back from B to A. 20% of the
fados modulate to the relative tonality, while 19% modulate to the parallel. These
modulations imply an alternation between the minor and the Major mode, which
seems to be archetypical to Ernesto Vieira, even if they only occur in 39% of the
overall cases. Only 4% of the fados modulate to the dominant, and 4% modulate
to other non-related tonalities.
Table 5.11: Types of Modulations
Modulations %
None 53
Relative 20
Parallel 19
Dominant 4
Others 4
Total 100
This behavior implies that often the initial tonality of a fado has to be chosen
not in isolation but taking into account the combinations of possible modulations.
Therefore, tonalities like D minor and A minor, are highly common since both the
parallel and relatives of these tonalities have few accidentals and are fairly easy to
perform, while tonalities like C major, although having a desirable relative, has a
not so friendly parallel. The case of the infrequent initial tonality of E minor can
then be explained due to having a poor parallel, therefore being more suitable as
a middle-section relative for the numerous fados starting and ending in G major.
These assumptions are also confirmed by the data in the graphical display. The
first graph presents the initial tonality versus the first relevant modulation (figure
5.33), while the second graphical representation contrasts the initial with the final
tonality (figure 5.34).
128
5.12. Modulations
Figure 5.33: Initial tonality versus first relevant modulation.
129
Figure 5.34: Initial tonality versus Final tonality.
130
5.13. Range
Observing the final tonalities one can see the relevant migrations of the fados
starting in D minor and A minor to their Major counterparts, while the movements
from G minor to G major, G major to E minor, F major to D minor and D major
to B minor seem somewhat relevant in the first graph and almost non-existent in
the second one showing how these tonalities are privileged as middle sections on
fados returning to their original key. Overall the Major fados increase from 54 to
62%, while the minor decrease from 46 to 38%. This highly suggests an overall
tendency for the politonal fados to have a first part based on a minor tonality and
a final part based on a Major one.
5.13 Range
Defining the range as the difference between the highest pitch and the lowest
one (McKay, 2004, p. 74), one can see that there is enormous variance among
the transcriptions. There are fados as concentrated in as little more than two
octaves up to fados spread along more than five octaves. On average a fado is
delimitated within the confines of three and a half octaves. Most melodies in the
sources have a range of approximately one octave-and-a-half, since they are sung
mostly by amateurs. In the transcriptions analysed the average range was of 17
semi-tones, with a minimum of 9 and maximum of 52. However, each fadista may
style the melody, the typical range changing according to their limitations and
ability. Constrained singers use narrower quasi-spoken ranges, while remarkable
divas might use more than two octaves, modifying their register. Taking into
account that the melodies alone average the range of an octave and half (which is
consistent with most vocal practices), one can assume that the accompaniment will
be placed on the two octaves below. The most common pitch is a value obtained
by dividing the most common pitch by the number of all possible pitches (127)
(McKay, 2004, p. 74). The calculated ratio tells us that on most fados the most
common pitch lies slightly below the center, indicating possible tonics ranging
from C3 to D5 averaging on A3. The primary register, which is the average MIDI
pitch (McKay, 2004, p. 74), complements and validates this information since it
ranges from G3 to C5. This minor difference is consistent with the idea of the
most common pitch being a tonic and a departure place, from which the melodies
are built arch shaped on top, and thus skewing the average pitches to a range
131
slightly above. The importance of bass register is the percentage of note-ons9

below F#3 (McKay, 2004, p. 74), and while there are notable cases of fados not
having notes in this register, up to fados having two thirds of them, it averages on
less than a third of the fados having notes on the bass register. On the other hand,
and as expected, most note ons occur in the middle register (G3-C5). The notes
happening in the high register, meaning above C5, account for an average of only
13%. The astonishing low variances regarding these values indicate us that most
fados really follow the trend and the notable exceptions seem to be clear outliers.
Table 5.12: Statistics of Range
Range 43.96 43.50 8.29 68.73 26.00 67.00 41.00 0.40 0.03
Range of Melodies-only 17.38 16.00 5.58 31.19 9.00 52.00 43.00 2.79 14.13
Most common pitch 0.45 0.45 0.04 0.00 0.38 0.58 0.20 0.70 1.05
Primary Register 60.10 60.00 2.72 7.40 55.00 72.00 17.00 1.12 3.26
Importance of Bass Register 0.28 0.27 0.12 0.01 0.00 0.66 0.66 0.70 1.67
Importance of Middle Register 0.60 0.61 0.12 0.01 0.28 0.84 0.55 -0.63 0.44
Importance of High Register 0.13 0.11 0.10 0.01 0.00 0.51 0.51 1.48 2.96
We have calculated these range features in comparative terms as well (figure

5.35).
Range seems to be a very useful feature to define fado, since fado is the fifth
lowest value along with medieval, flamenco, renaissance, country blues and punk.
This is due to being genres anchored in vocal traditions, with instrumentation
lacking extreme registers. At the opposite side of the spectrum, romantic, mod-
ern classical and the cluster of jazz-related genres appear to use a wide range of
registers. This finding somewhat confirms the empirical analysis made on fado
phonograms by Antunes et al. (Antunes et al., 2014), claiming that fado could be
easily identified because of its overall narrow range of frequencies, mainly concen-
trated in the center of the frequency spectrum.
5.13.1 Most Common Pitch
The most common pitch is defined as the “bin label of the most common pitch
divided by the number of possible pitches” (McKay, 2004, p. 74).
This feature (figure 5.36) allow us to visualize the relative place of the most
common pitch regarding to the overall range of the genre. Fado is at the bottom
9
in the MIDI protocol a note-on is a message to turn on a MIDI note.
132
Relative Frequency Semitones
0.2
0.3
0.4
0.5
0.6
20
40
60
80
Metal Medieval
Punk Flamenco
Alternative Rock Renaissance
Hard Rock Country Blues
Global Mean
Hardcore Rap Fado

Global Median
Rock and Roll Fado (Edition)

Fado Punk
Fado (Edition) Baroque
Chicago Blues Celtic
Country Blues Alternative Rock
Pop Rap Bluegrass
Renaissance Pop Rap
Traditional Country Techno
Contemporary Country Hard Rock
Psychedelic Classical
Ragtime Reggae
Soul Blues Traditional Country
Swing Contemporary Country
Techno Metal
Bluegrass Funk
Classical Psychedelic
Range
Funk Tango
Cool Salsa
Smooth Jazz Hardcore Rap
Bossa Nova Soul
Most Common Pitch

Figure 5.35: Range.
Bebop Rock and Roll
Tango Ragtime
Adult Contemporary Chicago Blues
Romantic Soul Blues
Figure 5.36: Most Common Pitch.

Medieval Cool
Flamenco Dance Pop
Reggae Adult Contemporary
Soul Blues Rock
Blues Rock Smooth Jazz
Jazz Soul Bossa Nova
Dance Pop Bebop
Modern Classical Jazz Soul
Salsa Modern Classical
Global Mean
Baroque Swing
Global Median
Celtic Romantic
5.13. Range
133
third of the spectrum, along with metal, punk, alternative rock, hard rock, and
rock and roll. This seems to indicate that the most common pitch is one that
occurs in the accompaniment and not in the voice. This also seems to point out
to reiterated pitches in the accompaniment (either a reiterated bass note, or chord
tone) and to favor more simple ostinati -like genres. Celtic, baroque, salsa and
modern classical music appear at the other side and it seems that their most
recurrent pitch occurs in the melodic or ornamental lines.
5.13.2 Primary Register
The primary register is basically the “average MIDI pitch” (McKay, 2004, p.
74).
Primary Register
80
70
Average MIDI Pitch
60
50
Global Mean
Global Median
40
Country Blues
Modern Classical
Adult Contemporary
Metal
Punk
Alternative Rock
Hard Rock
Hardcore Rap
Flamenco
Bebop
Rock and Roll
Psychedelic
Chicago Blues
Funk
Blues Rock
Soul
Cool
Reggae
Renaissance
Swing
Soul Blues
Salsa
Ragtime
Celtic
Baroque
Romantic
Classical
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Pop Rap
Dance Pop
Fado
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.37: Primary Register.
This feature (figure 5.37) complements the last one, and gives an idea about
the span of the overall range of a given style. It was previously shown that fado
possesses a relatively short range, and now one can see how its average is slightly
above global mean and median, confirming that it is a genre in which the bass
notes seem to not play a relevant role, thus relying more on the melody. At the
134
5.13. Range
lower side of the spectrum one finds taxonomies like metal, punk, alternative rock,
hard rock and techno, confirming the idea that the sound of these genres is skewed
towards the bass and accompaniment notes. Modern classical, classical, romantic
and baroque music appear at the higher side of the spectrum. This also shows
us how, historically, notated music seems to be much more contrived and valuing
higher registers and melodies.
5.13.3 Importance of Bass Register
The importance of bass register is defined as the “fraction of Note Ons between
MIDI pitches 0 and 54” (McKay, 2004, p. 74).
Importance of Bass Register

1.0
Global Mean
Global Median
0.8
Relative Frequency
0.6
0.4
0.2
0.0
Country Blues
Modern Classical
Adult Contemporary
Classical
Baroque
Celtic
Salsa
Romantic
Ragtime
Reggae
Renaissance
Soul Blues
Swing
Cool
Soul
Blues Rock
Funk
Flamenco
Psychedelic
Rock and Roll
Bebop
Chicago Blues
Hardcore Rap
Hard Rock
Alternative Rock
Punk
Metal
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Fado
Fado (Edition)
Pop Rap
Medieval
Traditional Country
Tango
Techno
Figure 5.38: Importance of Bass Register.
The pitch 54 corresponds to the F#/Gb below the central C (usually named
C4). This feature (figure 5.38) analyzes the percentage of notes in a given genre
that lay there, therefore and probably, mostly bass notes and chords, since usually
melodies do not go that low. Surprisingly, in relative terms, fado is on the average
regarding this feature. Although fado has a narrow range, and therefore one
should not expect to find many notes along the low register, it seems that their
135
percentage is indeed relevant. This is consistent with the idea that fado has a
strong component relying in bass notes and chords, even if they are not played
in the extreme low registers. It just means that the accompaniments are mostly
played rather close to the central register. On the other side, the weight of these
bass notes and chords is not something that makes fado stand out in any way,
meaning that most genres have a similar percentage of notes in the lower registers.
Medieval, classical and baroque seem to be exceptions, in the sense that they,
indeed, have fairly low percentages of notes in the lower registers, while metal,
punk, alternative rock and hard rock, confirm the relatively huge amount of notes
within this register.
5.13.4 Importance of Middle Register
The importance of middle register is the “fraction of Note Ons between MIDI
pitches 55 and 72” (McKay, 2004, p. 74).
Importance of Middle Register
1.0
Global Mean
Global Median
0.8
Relative Frequency
0.6
0.4
0.2
0.0
Country Blues
Modern Classical
Adult Contemporary
Metal
Alternative Rock
Punk
Hardcore Rap
Hard Rock
Rock and Roll
Celtic
Chicago Blues
Psychedelic
Ragtime
Romantic
Blues Rock
Funk
Soul
Bebop
Classical
Cool
Salsa
Reggae
Soul Blues
Baroque
Swing
Flamenco
Renaissance
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Dance Pop
Pop Rap
Fado
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.39: Importance of Middle Register.
The pitch 55 corresponds to the G below the central C, while pitch 72 is the C
above. This range is roughly the main range where most vocals and most melodies
136
5.13. Range
occur. Looking at the figure 5.39, fado is slightly above average, at the top third
of the spectrum, indicating the importance of this central register in the genre.
On the lower part, metal, alternative rock, punk and hardcore rap are the genres
to which this register seems less relevant, while medieval, renaissance, bluegrass
and adult contemporary highly depend on it.
5.13.5 Importance of High Register
The importance of high register is the “fraction of Note Ons between MIDI
pitches 73 and 127” (McKay, 2004, p. 74).
Importance of High Register
0.8
Global Mean
Global Median
0.6
Relative Frequency
0.4
0.2
0.0
Country Blues
Modern Classical
Adult Contemporary
Flamenco
Renaissance
Metal
Punk
Alternative Rock
Hard Rock
Bebop
Rock and Roll

Funk
Psychedelic
Cool
Swing
Soul Blues
Soul
Reggae
Blues Rock
Hardcore Rap
Chicago Blues
Salsa
Baroque
Romantic
Celtic
Classical
Ragtime
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Pop Rap
Fado
Fado (Edition)
Dance Pop
Medieval
Traditional Country
Techno
Tango
Figure 5.40: Importance of High Register.
The pitch 73 corresponds to the C#/Db, a minor ninth above the central C.
This feature (figure 5.40) is useful to understand the practices that highly value
melodic, lyrical and ornamental features, which usually transcend the uneducated
voice, since most of these notes are unsingable or only singable by sopranos.
Fado is within the average regarding this feature, meaning that although the
percentage is not relevant to define the practice, most of other genres also do
not depend on this register to be defined by it. This is extremely evident in
137
the cases of flamenco, country blues, renaissance, metal and punk, with marginal
percentages of notes within this register. On the other hand, ragtime, modern
classical, classical, Celtic and romantic music have high percentages of notes in
the higher register, reinforcing the idea of melodies and motives played by melodic
instruments, transcending the normal vocal range. Also, it is very common within
these genres the doubling of melodies one or two octaves higher, which reinforces,
even further, the percentage of notes in this register.
5.14 Pitch Prevalence
We calculated the fraction of note-ons corresponding both to the most common

pitch and pitch class. Then, we did the same for the second most common pitch
and pitch class and calculated the ratio between them in order to obtain the
relative strength of the top pitches and top pitch classes. We also calculated the
interval between these two top pitches and top pitch classes for each fado. All these
calculations were both performed in the corpus overall, and in the Melodies-only.
We have confirmed that indeed fados rely on strong tonal centres and have a fairly
skewed distribution of pitches. The most common pitch is consistently prevalent,
averaging at 16% and peaking at 33%, when compared to the average of 4,2% in
a perfectly even distribution for a two-octave range fado. This tendency is highly
reinforced if one regards pitch class, noticing a minimum of 16% and a notable
peak of 49%, compared to an average of 8,3% in a perfectly even distribution.
While the high ranges in the relative strength show that there are fados where
the second pitch is only a quarter prevalence from the top pitch (indicating a
strong tonicization, possibly huge repetition and a more fragmented distribution
of the remaining pitches), they also show that there are fados where there is an
equilibrium suggesting an alternation of tonics. However these seem to be more the
exception than the rule since the average (and median) is considerably relevant:
usually the second most common pitch represents 70% of the top’s strength, which
implies that two pitches alone average at being one third of the pitches in a given
fado, and ranging up to an average of 45% of those same pitch classes.
We have also analyzed the interval between the top pitch and second most
common pitch (figure 5.41), as well as their respective pitch classes (figure 5.42).
It is noteworthy that within the entire corpus the intervals reinforcing the main
138
5.14. Pitch Prevalence
Table 5.13: Statistics of Pitch Prevalence
Pitch Prevalence Overall Avg Med St Dev Var Min Max Range Skew Kurt
Most common pitch prevalence 0.16 0.13 0.06 0.00 0.08 0.33 0.25 1.17 0.48
Most common pitch class prevalence 0.27 0.26 0.07 0.00 0.16 0.49 0.33 0.93 0.56
Relative Strength of top pitches 0.70 0.72 0.20 0.04 0.29 1.00 0.71 -0.38 -0.81
Relative Strength of top pitch classes 0.69 0.70 0.18 0.03 0.25 1.00 0.75 -0.18 -0.70
Most common pitch prevalence 0.22 0.21 0.06 0.00 0.12 0.45 0.33 1.22 2.51
Most common pitch class prevalence 0.24 0.23 0.05 0.00 0.16 0.45 0.30 1.34 2.32
Relative Strength of top pitches 0.74 0.74 0.18 0.03 0.20 1.00 0.80 -0.46 -0.05
Relative Strength of top pitch classes 0.76 0.76 0.17 0.03 0.32 1.00 0.68 -0.54 -0.12
tonal functions take the lead – therefore the perfect fourth relationship seems to
imply fados where either the subdominant above or the dominant below the tonic
is the second most common pitch (assuming the tonic is the top pitch), while the
perfect fifth relationship seems to indicate the opposite – the subdominant below
the tonic or the dominant above. The octave and double octave relationship
imply, of course, the doubling of the tonic. When one looks only at pitch classes
then the octave and double octave intervals vanish, and the percentages relative
to perfect fourths and fifths increase naturally. Among the transcriptions, the
accompaniments rely mainly on ostinati consisting on chordal material referring
mainly to the tonic, dominant and subdominant functions. Therefore, their main
pitches are, of course, reinforced and their weight statistically increased when
compared to Melodies-only. Among the melodies, which have by its nature a
smaller range and increased repetition due to the typical stepwise movement, the
interval of a second is privileged. This seems to imply a strong emphasis on the
tonic and the dominant of the dominant, which is consistent with the idea that to
form an arch-shaped melody departing from the tonic, the most obvious step is the
second degree, immediately above and also easily harmonized with the dominant.
The tritone relationship is never to be seen in this context, either in the corpus or
in just the melodies.
5.14.1 Most Common Pitch Prevalence
The comparative statistics based on pitch do not take into account temporal
locations, and consider each recording as a whole. Furthermore “all notes occurring
on channel 10 were ignored for all of these features, as pitch values on that channel
correspond to percussion patches, not to pitches” (McKay, 2004, p. 72). The most
139
Figure 5.41: Interval between strongest pitches.
Figure 5.42: Interval between strongest pitch classes.
140
common pitch prevalence refers to the “fraction of Note Ons corresponding to the
most common pitch” (McKay, 2004, p. 72).
Most Common Pitch Prevalence
0.4
Global Mean
Global Median
0.3
Relative Frequency
0.2
0.1
Country Blues
Modern Classical
Adult Contemporary
Romantic
Ragtime
Bebop
Swing
Cool
Salsa
Classical
Baroque
Soul Blues
Psychedelic
Soul
Blues Rock
Reggae
Renaissance
Chicago Blues
Rock and Roll

Hard Rock
Celtic
Alternative Rock
Funk
Flamenco
Punk
Metal
Hardcore Rap
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado (Edition)
Fado
Dance Pop
Pop Rap
Medieval
Traditional Country
Tango
Figure 5.43: Most Common Pitch Prevalence. Techno
Looking at the figure 5.43, fado is within the average regarding this feature
not standing out in any particular way. While hardcore rap, metal and country
blues seem to be characterized by having one pitch that is substantially prevalent
(we speculate it to be either a repetitive drone or a reiterated tonal center), on
the opposite side, one can find modern classical, romantic, ragtime and bebop as
taxonomies with much more variety, no pitch particularly standing out, in relative
terms. One can also observe that the ratio between the opposite sides is superior
to four.
5.14.2 Most Common Pitch Class Prevalence
The most common pitch class prevalence is defined as the “fraction of Note Ons
corresponding to the most common pitch class” (McKay, 2004, p. 72).
141
Most Common Pitch Class Prevalence

0.7
Global Mean
0.6 Global Median
0.5
Relative Frequency
0.4
0.3
0.2
0.1
Country Blues
Modern Classical
Adult Contemporary
Swing
Bebop
Cool
Romantic
Ragtime
Baroque
Classical
Renaissance
Salsa
Soul Blues
Flamenco
Psychedelic
Reggae
Soul
Celtic
Blues Rock
Chicago Blues
Rock and Roll
Funk
Alternative Rock
Hard Rock
Punk
Hardcore Rap
Metal
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Pop Rap
Fado
Fado (Edition)
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.44: Most Common Pitch Class Prevalence.
This feature (figure 5.44) reinforces some ideas expressed in the previous one.
Fado is among the average, not standing out, while metal, country blues, hard-
core rap and punk clearly state one strong tonal center. Minding that in these
recordings the prevalence of such pitch class approaches 40%, which is five times
an evenly distribution, one has to consider that there is little room for all other
possible pitch classes. On the other hand modern classical, swing, bebop and cool
have much more even distributions, possibly indicating more variety, chromaticism
and complexity.
5.14.3 Relative Strength of Top Pitches
The relative strength of top pitches is defined as “the magnitude of the 2nd
most common pitch divided by the magnitude of the most common pitch” (McKay,
2004, p. 72).
Observing the figure 5.45, while fado seemed average regarding the prevalence
of a unique pitch and could not be told apart by such feature in itself, when
observing the relative strength of the second most common pitch, one finds that
142
Relative Strength of Top Pitches
1.0
0.9
0.8
Relative Frequency
0.7
0.6
0.5
0.4
Global Mean
Global Median
0.3
Country Blues
Modern Classical
Adult Contemporary
Metal
Celtic
Blues Rock
Flamenco
Hard Rock
Hardcore Rap
Rock and Roll
Psychedelic
Chicago Blues
Funk
Cool
Soul Blues
Soul
Reggae
Punk
Ragtime
Baroque
Renaissance
Salsa
Alternative Rock
Bebop
Swing
Classical
Romantic
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado (Edition)
Fado
Dance Pop
Pop Rap
Medieval
Traditional Country
Tango
Techno
Figure 5.45: Relative Strength of Top Pitches.
fado is clearly at the bottom of the list, along with tango, metal and Celtic music.
This indicates that fado is indeed a genre in which a tonal center is important,
regarding all other pitches, although the particular pitch of the tonal center is not
as strong as when compared with other ones. This probably means that such tonal
center, in fado, is played in different octaves while in other genres it is reiterated
as the same pitch all along. Romantic, classical, swing and jazz soul appear as
practices in which the second most prevalent pitch is almost as important as the
main one, which corroborates the idea of more variety and complexity.
5.14.4 Relative Strength of Top Pitch Classes
The relative strength of top pitch classes is defined as “the magnitude of the
2nd most common pitch class divided by the magnitude of the most common pitch
class” (McKay, 2004, p. 72).
And this feature (figure 5.46) corroborates our thesis that fado is indeed a genre
that depends on a strong tonal center, along with metal, hard rock, country blues,
and funk. While on a genre like metal it seems that such tonal center is often
143
Relative Strength of Top Pitch Classes
1.0
0.8
Relative Frequency
0.6
0.4
Global Mean
Global Median
0.2
Country Blues
Modern Classical
Adult Contemporary
Metal
Hard Rock
Funk
Blues Rock
Hardcore Rap
Alternative Rock
Chicago Blues
Soul Blues
Reggae
Rock and Roll
Flamenco
Soul
Celtic
Punk
Ragtime
Psychedelic
Salsa
Romantic
Bebop
Baroque
Classical
Cool
Renaissance
Swing
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Dance Pop
Pop Rap
Medieval
Traditional Country
Techno
Tango
Figure 5.46: Relative Strength of Top Pitch Classes.
played in the same octave, in fado the octave is not really important. Modern
classical, swing, renaissance, cool and classical are at the opposite side of the
spectrum, indicating either the absence of tonal centres, or the presence of at least
two.
5.14.5 Interval Between Strongest Pitches
The interval between strongest pitches is defined as the “absolute value of the
difference between the pitches of the two most common MIDI pitches” (McKay,
2004, p. 73).
Observing the figure 5.47, fado is included in a cluster of taxonomies in which

the most relevant interval between the strongest pitches is the perfect fifth (seven
semi tones), as denoted by the median value. Other important values seem to
be the perfect fourth and the octave among these styles. A huge cluster of other
taxonomies has instead the perfect fourth as the main value while also retaining
the perfect fifth as important. The exceptions seem to be hard rock and ragtime
with a remarkable preference for the octave. All these taxonomies seem to be
144
Interval Between Strongest Pitches
25 Global Mean
Global Median
20
Semitones
15
10
0
Country Blues
Modern Classical
Adult Contemporary
Swing
Bebop
Reggae
Soul
Celtic
Punk
Salsa
Baroque
Classical
Cool
Renaissance
Blues Rock
Alternative Rock
Chicago Blues
Soul Blues
Rock and Roll
Flamenco
Psychedelic
Hardcore Rap
Romantic
Funk
Metal
Hard Rock
Ragtime
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Pop Rap
Dance Pop
Fado
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.47: Interval Between Strongest Pitches.
tonal oriented though. At the other side of the spectrum, medieval and swing
privilege the minor third, while pop rap, bebop and modern classical prefer the
major third. These choices seem to point out to more ambiguous genres regarding
tonality, probably approaching modality or other kind of theoretical systems which
can make sense.
5.14.6 Interval Between Strongest Pitch Classes
The interval between strongest pitch classes comprises the “absolute value of
the difference between the pitches of the two most common pitch classes” (McKay,
2004, p. 73).
This feature (figure 5.48) simply reinforces the last one. Being aware that the
factor octave has vanished, one can observe how indeed the majority of the genres,
in the database, are tonal. The perfect fourth and perfect fifth, are privileged as the
main interval between the strongest pitches, indicating a clear prevalence of a tonic
and a dominant or a subdominant. Swing, medieval, pop rap, techno and hardcore
145
Interval Between Strongest Pitch Classes
Global Mean
10
Global Median
8
Semitones
Country Blues
Modern Classical
Adult Contemporary
Swing
Hardcore Rap
Bebop
Reggae
Soul
Celtic
Punk
Salsa
Baroque
Classical
Cool
Renaissance
Blues Rock
Alternative Rock
Chicago Blues
Soul Blues
Flamenco
Psychedelic
Romantic
Funk
Metal
Hard Rock
Ragtime
Rock and Roll

Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Pop Rap
Dance Pop
Fado
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.48: Interval Between Strongest Pitch Classes.
rap are the notable exceptions, with the third being the most predominant interval,
among the strongest pitches.
5.14.7 Number of Common Pitches
The number of common pitches is defined as the “number of pitches that ac-
count individually for at least 9% of all notes” (McKay, 2004, p. 73).
This feature (figure 5.49) is useful in suggesting the complexity or intricacy of a

style and fado does not stand out, being part of an enormous cluster of taxonomies
which revolve around two main pitches. Medieval seems clearly the exception, ac-
centuating probably its modal character, as well as its expected narrow range.
An interesting, and somewhat expected, cluster of genres comprising swing, be-
bop, modern classical, afro soul, salsa, classical, cool, romantic and ragtime seem
characterized by a profusion of different pitches.
146
Number of Common Pitches
7
Global Mean
Global Median
6
4
Pitches
Country Blues
Modern Classical
Adult Contemporary
Swing
Bebop
Salsa
Classical
Cool
Romantic
Ragtime
Baroque
Soul Blues
Soul
Blues Rock
Chicago Blues
Psychedelic
Funk
Metal
Hard Rock
Rock and Roll
Hardcore Rap
Reggae
Celtic
Punk
Renaissance
Alternative Rock
Flamenco
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Dance Pop
Fado
Fado (Edition)
Pop Rap
Medieval
Traditional Country
Tango
Techno
Figure 5.49: Number of Common Pitches.
5.14.8 Pitch Variety
The pitch variety is the “number of pitches used at least once” (McKay, 2004,
p. 73).
This feature (figure 5.50) is an excellent complement to the previous one, be-
cause it allows us to determine the real distribution of pitches overall. Accounting
just for the number of common pitches there might exist genres that, while having
some common ones, could still have many more scattered with very low occur-
rences, or a great number of unique pitches. This is the case of Romantic music,
being the overall style with most pitches, while still having few common ones,
which seems to be compatible with the idea of large range, while preserving some
tonality.
Fado is at the low side, in this feature, being a practice with relatively few
pitches overall. This seems to point out, not only for its simplicity, but possibly
also for a somewhat narrow range. Medieval, flamenco, hardcore rap, renaissance
and country blues are the lowest ones.
147
Pitch Variety
80 Global Mean
Global Median
60
Pitches
40
20
Country Blues
Modern Classical
Adult Contemporary
Flamenco
Hardcore Rap
Renaissance
Punk
Celtic
Reggae
Alternative Rock
Baroque
Funk
Hard Rock
Soul
Psychedelic
Metal
Rock and Roll
Blues Rock
Classical
Chicago Blues
Salsa
Soul Blues
Ragtime
Cool
Bebop
Swing
Romantic
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Pop Rap
Fado
Fado (Edition)
Dance Pop
Medieval
Traditional Country
Techno
Tango
Figure 5.50: Pitch Variety.
5.14.9 Pitch Class Variety
The pitch class variety is the “number of pitch classes used at least once”
(McKay, 2004, p. 73).
This feature (figure 5.51) complements the last one, allowing us to differentiate
between the genres that seem complex because they have many pitches, but are
indeed octave doublings. Analyzing only the pitch classes it can be seen how most
genres use all, or nearly, the full template of the twelve pitches available. So,
while fado averaging at ten pitches could seem to be a rather complex or nearly
chromatic practice, when seen in relative terms is actually on the lower third of
the spectrum. This, again, points to its simplicity, in pair with country blues,
dance pop or flamenco. Medieval, hardcore rap and punk seem to be the simplest
of them all, while romantic, swing, modern classical, and the overall cluster of
jazz-related genres confirm themselves as the most complex ones.
148
Pitch Class Variety
12
10
Pitch Classes
4 Global Mean
Global Median
Country Blues
Modern Classical
Adult Contemporary
Hardcore Rap
Punk
Celtic
Reggae
Flamenco
Renaissance
Alternative Rock
Baroque
Funk
Hard Rock
Psychedelic
Soul
Metal
Rock and Roll
Blues Rock
Classical
Chicago Blues
Salsa
Soul Blues
Ragtime
Cool
Bebop
Swing
Romantic
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Pop Rap
Fado
Fado (Edition)
Dance Pop
Medieval
Traditional Country
Techno
Tango
Figure 5.51: Pitch Class Variety.
5.14.10 Most Common Pitch Class
The most common pitch class is defined as the “bin label of the most common
pitch class” (McKay, 2004, p. 74).
The visualization of this feature as presented in the other figures makes no sense
in this case. While it is certainly important to notice what the most common pitch
class is, in a given practice, global median and mean seem to have a meaningless
value in this context, not being known exactly what they represent. On the other
hand, the mode within any given taxonomy and their overall distribution makes
sense.
Most genres have as the most common pitches the 4 (E) and the 7 (G). The
most common pitch class in Fado is 9 (A), which is also fairly common among
other genres, namely baroque, hardcore rap, reggae, psychedelic and punk. A few
number also privilege pitches 0 (C) and 5 (F), and a lesser number 2 (D) and 11
(B). No genres have as the preferred pitch any of the “black keys” of the piano.
Instrumentation is probably the main factor behind these facts.
149
Table 5.14: Most Common Pitch Class
Taxonomy Mode Taxonomy Mode Taxonomy Mode

Medieval 2 Dance Pop 7 Contemporary Country 4
Renaissance 5 Psychedelic 9 Cool 0
Baroque 9 Jazz Soul 5 Bebop 0
Hardcore Rap 9 Smooth Jazz 7 Soul 4
Classical 7 Chicago Blues 4 Romantic 4
Swing 0 Soul Blues 11 Modern Classical 5
Celtic 4 Traditional Country 7 Salsa 7
Blues Rock 4 Hard Rock 7 Tango 2
Funk 2 Bossa Nova 4 Ragtime 7
Pop Rap 11 Metal 4 Alternative Rock 4
Reggae 9 Bluegrass 7 Rock and Roll 0
Techno 5 Adult Contemporary 7 Punk 9
Fado 9 Flamenco 4 Country Blues 7
Table 5.15: Distribution of the most common pitch class
Pitch Class #
0 4
1 0
2 3
3 0
4 10
5 4
6 0
7 10
8 0
9 6
10 0
11 2
5.15 Melody
Most fado melodies are either characteristically arch shaped or undulating. The
shape of the melodies are the result of unfolding the narrative, contriving each line
into a pair of 2/4 measures, respecting both the emotional prosody and the under-
lying harmonic progression. The melodic form varies greatly within the repertoire.
The oldest fados in the sources are very simple and strophic, with little to no vari-
ation. These sources, however, present only the musical scores. Since there were
no recordings back then, one cannot be sure that the melodies were effectively
sung that way, rather than improvised and more varied. The textual descriptions
150
5.15. Melody
of the practice and the very idea that styling and improvisation are values highly
regarded and praised in fado, in the beginning of the twentieth century, lead us to
believe that the melodies would vary more than what is presented in the musical
scores, depending also on the skill of the performer. During the twentieth cen-
tury, and with the evolution of the practice, the form also evolved and complex
strophic fados with great variation appeared and even some through-composed fa-
dos resembling art songs emerged in the repertoire. All these possibilities co-exist
nowadays, with very complex forms and simple strophic fados circulating side by
side.
In fado practice, the combined ability to manipulate, not only the rhythm of
a melody, but also the pitch is called estilar (styling). Very much like other
improvisatory practices (namely Jazz), styling is a kind of improvisation. It “is
overwhelmingly spoken about by fadistas as essentially unlearnable, as the defining
characteristic of fado as genre as practice and of a fadista as an individual voice.
Instrumentalists style too.” (Gray, 2013, p. 144) If one considers again the idea of
an ideal version of each fado, an urtext, it is clear that there is a whole continuum
of possible variations on the melody, in which one can go further and further away.
Some fados are more suitable to be highly manipulated while others are not. It
is also a skill that depends on the creativity, musicality and risk taking of the
fadista and the instrumentalists (Gray, 2013, p. 146). This ideal version of a
fado is generally attributed to the first person to fix it in a given medium. In the
nineteenth century this would be the musical score, but in the twentieth century
this was mainly the phonogram. The fadista was then entitled the “creator”, as
one can see in the music scores covers and records. There was the composer (when
the fado was authored), the author (the lyricist) and this new figure – the creator.
Because, and following the reasoning of Manuela de Freitas, already discussed, it
is the fadista the one who embodies the fado, and brings it to life, in fact, the
one creating it. Without the fadista, the music is not fado, it is just music on a
score. A second creator may become a composer, when styling in such a way that
the result departs so much from the original version, and consistently within the
creator’s style or grammar, that the new version is then fixed. And it is believed
that this dynamic, dialogic and emergent process gave birth to a whole series of
fados, the ones all derived from others, during the second half of the nineteenth
century and beginning of the twentieth century.
“When performers style melodies, they draw on shared musical

151
grammars and conventions (...). Highly inventive fadistas, working

within these shared conventions and performance histories, may, through
repeat performances, develop their own melodic figures for specific fa-
dos in their repertoire, melodic formulas than can then be creatively
refigured in their future improvisations of the same fado” (Gray, 2013,
pp. 154-155).
Melodies are contrived both by the underlying harmonic progression and the
text, as described earlier. Most lines begin with unstressed syllables, thus becom-
ing pick-up notes. Hence, the archetypical situation is that most fado melodies
begin with an ascending leap of a fourth, as a consequence of those pick-up notes
being in the dominant, leading to the tonic, in the downbeat of the first bar. The
sequential melodic contours follow logically these constraints and the final tone
rests inevitably in the tonic. Sometimes, however, in very dramatic renditions of
a fado by rather skilled performers, and for emotional and climactic effect, the
peak note is chosen to be also the last. In those situations, the melody is varied
in the last strophe so that the tonic standing in a low register is transposed one or
even two octaves higher and reached via an upward movement. The archetypical
situation is shown in notation (figure 5.52) and most fados in the transcriptions
can be seen as more or less inventive variations of this archetype, depending on
the initial constraints.
j
& b 42 œ
œ ‰ œ ‰ œ
J J
œ œœ œ œœ œ œœ # œ œœ œœ # œ œœ œ œ
(V7) i i
? b 42 ‰ œ œ œ œ
V7
œ
&b ‰ j Œ
5
J œ
? œ œœ œœ œœ œœ œœ œœ œœ
i V7 i
b œ #œ #œ œ j
œ œ œ
Figure 5.52: A fado archetype.
152
5.15. Melody
5.15.0.1 Pitch Variety in melodies
The number of common pitches (table 5.16) are defined as the number of pitches
that account individually for at least 9% of all notes (McKay, 2004, p. 73) and
complement the information obtained in pitch prevalence. While it is already know
that fados are highly skewed towards tonal centres, looking at the minimum and
maximum values of this parameter one can also see fados with no common pitches,
meaning a total fragmentation of the distribution, and possibly more variety, up to
fados with five common pitches, possibly the ones where more modulations occur.
On average, though, only two pitches seem to have prevalence, and we predict them
to be the tonic and the dominant in most cases. Among the Melodies-only, since
the range is smaller (and thus there are less possibilities) the concentration around
certain pitches increases and so it can be said that on average melodies are based
around five pitches. We predict them to be the tonic, dominant, subdominant,
the second degree (dominant of dominant) and the leading tone.
The pitch variety, meaning the number of pitches that show up at least once
(McKay, 2004, p. 73), is a value presenting an enormous variance, hinting that
there are very monotonous fados based around a few number of pitches (as small as
fourteen), to more complex ones with fifty different pitches. The average is thirty.
If one translates these values into pitch classes, it can be indeed confirmed that,
although relying on strong tonal centres, most fados, mainly due to modulations,
present on average almost all of the chromatic scale averaging on ten pitch classes.
We predict the ones missing most should be the tritone and the augmented fifth.
When focusing on Melodies-only, this number is slightly less, averaging on nine
pitch classes per melody, and thirteen pitches overall.
Table 5.16: Statistics of Pitch Variety
Number of Common Pitches 1.94 2.00 1.07 1.15 0.00 5.00 5.00 0.57 0.20
Pitch Variety 29.57 28.50 9.23 85.18 14.00 50.00 36.00 0.36 -0.72
Pitch Class Variety 9.75 10.00 1.77 3.12 7.00 12.00 5.00 -0.10 -1.35
Number of Common Pitches 4.47 5.00 1.11 1.22 1.00 7.00 6.00 -0.52 0.69
Pitch Variety 12.96 13.00 3.92 15.33 7.00 25.00 18.00 0.91 0.57
Pitch Class Variety 9.02 9.00 1.61 2.61 6.00 12.00 6.00 0.20 -1.03
153
5.15.1 Pitch-Class distribution
Concerning pitch-class distribution, specifically among the melodies, it is pos-

sible to confirm (figure 5.53) that the tonic is the prevalent one by far, averaging
at 24%, followed by the second, fourth, fifth and minor seventh degrees all with
similar prevalences, above the average on a theoretical perfect evenly distribution.
This is somewhat expected since these are the pitch-classes present in most func-
tional chords, but with the surprise of the low percentage of the leading tone,
compared with, for instance, the minor seventh. As a third tier we have the major
third and sixth degrees and then all the others well below average, with the tritone
being almost inexistent.
Figure 5.53: Pitch-Class distribution in melodies.
5.16 Melodic Intervals
In order to better understand the melodies, not only pitches are relevant but
also the transitions between them. Therefore, we have performed a series of cal-
culations on the Melodies-only (table 5.17). The average melodic interval has a
rather large span among fados and by itself does not seem such a useful value to
draw conclusions from, so further calculations will be required. On the other hand,
the most common melodic interval gives us more precious information, confirming
that indeed smaller values (like stepwise motion) are highly preferred, which is ex-
pected among melodies. The distance between the most common melodic intervals
154
5.16. Melodic Intervals
reinforces this idea, since typically the second most common interval is just half
tone away, hinting at the possibility that the most common intervals on melodies
are unisons, seconds and thirds. The most common melodic interval prevalence
shows us how on average it recurs from one fifth up to half of the times in each
melody, averaging at one third, while the second most common interval averages
at 70% of this rate. Usually there are four melodic intervals in each melody that
at least account for 9% of all melodic intervals.
Table 5.17: Statistics of Melodic Intervals
Average Melodic Interval 2.48 2.42 0.84 0.71 0.90 8.42 7.52 3.73 24.76
Most Common Melodic Interval 1.72 2.00 1.38 1.90 0.00 12.00 12.00 4.06 30.75
Distance between Most Common Melodic Intervals 1.52 1.00 0.72 0.51 1.00 6.00 5.00 2.86 15.01
Most Common Melodic Interval Prevalence 0.33 0.32 0.07 0.01 0.20 0.53 0.33 0.60 0.15
Relative Strength of Most Common Melodic Intervals 0.70 0.71 0.17 0.03 0.29 1.00 0.71 -0.26 -0.71
Number of Common Melodic Intervals 3.96 4.00 0.90 0.81 2.00 6.00 4.00 0.08 -0.19
Since the average interval was an incomplete statistic, we have gone further and
explored more the type of intervals one can found on melodies (table 5.18). In the
transcriptions, the fraction of horizontal intervals that are repeated notes, minor
thirds, major thirds, perfect fifths, minor sevenths, major sevenths, octaves, minor
tenths or major tenths, on average represent 42% of all melodic material, which in-
dicates how chordal derivatives are important and expected in highly tonal tunes.
Stepwise motion is the most prominent interval accounting for half of the total,
on average, up to a maximum of 80% in some fados. Repeated notes (or unisons)
are highly relevant as they average on 17% of all intervals to a maximum of 52%
on certain fados. Chromatic motion is equally relevant with similar values, as
well as melodic thirds which also account for a fifth of the possible intervals. The
amount of bigger leaps with relevant percentages seems almost irrelevant within
the context. According to Huron, singing is usually stepwise, with larger leaps
immediately compensated by a step in contrary motion (Huron, 2008), and these
statistics seem to confirm this general trend. Fado singers seem to have a lower
maximum phonational frequency range (MPFR, an acoustic measurement repre-
sentative of the physiologic limits of the larynx) than Western Classical singers,
and this could be related with life habits, less or lack of singing training, or it
could be just a fado voice characteristic (Mendes et al., 2013).
155
Table 5.18: Statistics of Melodic Motion
Amount of Arpeggiation 0.42 0.42 0.11 0.01 0.17 0.68 0.51 -0.09 -0.06
Repeated Notes 0.17 0.16 0.10 0.01 0.00 0.52 0.52 0.63 0.36
Chromatic Motion 0.19 0.18 0.09 0.01 0.03 0.49 0.46 0.92 1.91
Stepwise Motion 0.47 0.49 0.13 0.02 0.13 0.80 0.67 0.09 0.08
Melodic Thirds 0.20 0.17 0.09 0.01 0.06 0.56 0.49 1.34 2.43
Melodic Fifths 0.03 0.02 0.04 0.00 0.00 0.24 0.24 2.94 12.48
Melodic Tritones 0.02 0.01 0.02 0.00 0.00 0.11 0.11 2.14 5.84
Melodic Octaves 0.01 0.00 0.05 0.00 0.00 0.42 0.42 7.67 64.97
5.17 Comparative features based on Melody
This set of features tries to reflect meaningful data on pitches taking into ac-
count the order in which they are played. Unfortunately it is not easy to purely
discriminate the main melodies in any given MIDI file. What Bodhidharma does
is to collect statistics regarding melodic motion, however, all notes occurring in
any given channel are treated as a melody, which might create a distortion. Since
we already had prepared a third edited corpus with just the main melodies filtered
and trimmed, according to Ernesto Vieira’s model, what we have done was to re-
place the “Edition corpus” values with those from the “Melodies-only” corpus for
this particular set of features only. Therefore in all figures one will be able to see
the relative positions of raw fado corpus within the scope of the raw database, but
the main fado Melodies-only as well. Unfortunately it was not possible to have
a version of the entire database with the main Melodies-only, which would have
been ideal to test these features.
5.17.1 Average Melodic Interval
As one can see (figure 5.54), in relative terms, fado has huge melodic intervals
if one considers all voices as a whole. This can be explained considering the
accompaniment, as there are huge leaps between adjacent pitches in the lower
voices. Those leaps correspond to the alternation of the notes played by the thumb
and the other fingers in the guitar, or the bass and chord notes on the left hand
of a piano player. Country blues, flamenco and ragtime also display huge average
melodic intervals because these files were encoded using the same principle.
156
5.17. Comparative features based on Melody
Average Melodic Interval
14 Global Mean
Global Median
12
10
Semitones
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Hardcore Rap
Baroque
Classical
Reggae
Funk
Blues Rock
Metal
Punk
Swing
Hard Rock
Soul Blues
Soul
Psychedelic
Celtic
Alternative Rock
Romantic
Chicago Blues
Rock and Roll

Cool
Salsa
Bebop
Ragtime
Flamenco
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado (Edition)
Dance Pop
Pop Rap
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.54: Average Melodic Interval.
Looking at the fado Melodies-only, then the scenario drastically changes: the in-
terval is now a little less than two semitones, along with medieval and renaissance.
This reveals a tendency to mostly conservative stepwise, undulating melodies.
5.17.2 Most Common Melodic Interval
The most common melodic interval is defined as the “melodic interval with the
highest magnitude” (McKay, 2004, p. 75).
This figure (figure 5.55) is consistent with the last interpretation. Indeed the
most common melodic interval in fado is the largest one among the entire database,
having the perfect fourth as its median point. This reflects the huge leaps in the
accompaniments, them being usually fourths, fifths and octaves. On the other
hand when looking at the Melodies-only, then fado behaves as expected and the
most common melodic interval is the whole tone. This seems to be the most
common melodic interval among most taxonomies. A generous number of genres
also privilege the unison.
157
Most Common Melodic Interval
Global Mean
Global Median
20
15
Semitones
10
Country Blues
Modern Classical
Adult Contemporary
Reggae
Metal
Punk
Hard Rock
Soul
Alternative Rock
Rock and Roll

Hardcore Rap
Swing
Romantic
Chicago Blues
Bebop
Renaissance
Baroque
Classical
Funk
Blues Rock
Soul Blues
Psychedelic
Celtic
Cool
Flamenco
Ragtime
Salsa
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Fado (Edition)
Pop Rap
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.55: Most Common Melodic Interval.
5.17.3 Distance Between Most Common Melodic Intervals
The distance between most common melodic intervals is defined as the “abso-
lute value of the difference between the most common melodic interval and the
second most common melodic interval” (McKay, 2004, p. 75).
As expected, in most genres the distance between their most common melodic
intervals is just a half or a whole step (figure 5.56). This reflects diatonic melodies
moving stepwise up or down, and the fado melodies respect that trait. When
considered as a whole, then due to the accompaniment, the leaps are bigger.
5.17.4 Most Common Melodic Interval Prevalence
The most common melodic interval prevalence is defined as the “fraction of

melodic intervals that belong to the most common interval” (McKay, 2004, p.
75).
This feature (figure 5.57) is highly interesting because it allows us to hint, re-
gardless of the how the file was encoded, the diversity of intervals used. If the most
158
Relative Frequency Semitones
0.1
0.2
0.3
0.4
0.5
0.6
5
10
15
Ragtime Swing
Country Blues Tango
Fado Bebop
Bossa Nova Medieval
Global Mean
Global Mean
Salsa Fado (Edition)
Global Median
Global Median
Cool Baroque
Contemporary Country Classical
Bluegrass Soul Blues
Tango Jazz Soul
Bebop Modern Classical
Soul Blues Cool
Modern Classical Dance Pop
Romantic Bossa Nova
Adult Contemporary Hardcore Rap
Traditional Country Romantic
Soul Renaissance
Psychedelic Funk
Flamenco Blues Rock
Swing Smooth Jazz
Jazz Soul Psychedelic
Smooth Jazz Celtic
Rock and Roll Adult Contemporary
Chicago Blues Flamenco
Classical Traditional Country
Funk Ragtime
Blues Rock Salsa
Celtic Reggae
Dance Pop Hard Rock
Reggae Soul
Hard Rock Alternative Rock
Most Common Melodic Interval Prevalence

Pop Rap Rock and Roll
Hardcore Rap Chicago Blues
Distance Between Most Common Melodic Intervals
Alternative Rock Pop Rap

Fado (Edition) Contemporary Country
Baroque Country Blues
Renaissance Bluegrass
Figure 5.57: Most Common Melodic Interval Prevalence.

Metal Fado
Techno Metal
Figure 5.56: Distance Between Most Common Melodic Intervals.
Punk Techno
Medieval Punk
159
prevalent melodic interval has a low percentage this means all possible intervals
are scattered and distributed, thus no one being highly prevalent. While if one
takes fado as a whole one sees a relatively high prevalence of one interval above all
others, the complete opposite happens when looking at the main Melodies-only.
This points out to how the accompaniments have highly repetitive structures rely-
ing on the same interval over and over, while the melodies seem to be much more
diverse. Fado melodies when compared to all other taxonomies also show signs of
being highly inventive along with bossa nova, country blues and ragtime. On the
other hand, medieval, punk and techno seem to rely too much on a given melodic
interval, although in these later cases we suspect it is because of accompaniment
figurations.
5.17.5 Relative Strength of Most Common Intervals
The relative strength of most common intervals is defined as the “fraction of

melodic intervals that belong to the second most common interval divided by the
fraction of melodic intervals belonging to the most common interval” (McKay,
2004, p. 75).
Relative Strength of Most Common Intervals
1.0
0.8
Relative Frequency
0.6
0.4
Global Mean
0.2
Global Median
Country Blues
Modern Classical
Adult Contemporary
Hard Rock
Reggae
Metal
Renaissance
Romantic
Hardcore Rap
Celtic
Punk
Baroque
Blues Rock
Chicago Blues
Classical
Alternative Rock
Rock and Roll
Salsa
Swing
Psychedelic
Flamenco
Soul
Funk
Ragtime
Cool
Bebop
Soul Blues
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Fado (Edition)
Pop Rap
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.58: Relative Strength of Most Common Intervals.
160
This feature (figure 5.58) complements the last one as it allows to better under-
stand the melodic motion distribution. In practical terms, looking at fado, it can
be seen that the accompaniment figuration not only possesses one relevant melodic
interval, but also at least two of almost equal importance in relative terms. On the
other hand, looking at melodies, it seems that the second most relevant interval
has some relative distant weight from the most common interval. It was previously
shown that the most common melodic interval on fado melodies is the whole tone,
and that the distance to the second most common melodic interval is the half tone.
Therefore it can be inferred that in the fado melodies the whole tone is relatively
more prevalent than half tones and minor thirds. Looking at the entire spectrum
of taxonomies it seems that the second most common interval in soul blues, bebop
and contemporary country is as important as the first one, while on medieval and
techno one can see that a single melodic interval clearly outperforms all others.
5.17.6 Number of Common Melodic Intervals
The number of common melodic intervals is the “number of melodic intervals

that represent at least 9% of all melodic intervals” (McKay, 2004, p. 76).
Number of Common Melodic Intervals
4
Frequency
1
Global Mean
Global Median
0
Country Blues
Modern Classical
Adult Contemporary
Metal
Renaissance
Romantic
Punk
Baroque
Alternative Rock
Flamenco
Ragtime
Hard Rock
Reggae
Hardcore Rap
Celtic
Blues Rock
Chicago Blues
Classical
Rock and Roll
Salsa
Swing
Soul
Funk
Cool
Bebop
Soul Blues
Psychedelic
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado
Dance Pop
Fado (Edition)
Pop Rap
Medieval
Traditional Country
Techno
Tango
Figure 5.59: Number of Common Melodic Intervals.
161
This feature (figure 5.59) is useful to confirm the complexity and variety within
the melodic contours. As one can see most genres present a median of four common
melodic intervals, fado melodies included. The exceptions are psychedelic and
contemporary country with their median being five, while a relatively large sample
of taxonomies have medians as three, this including fado as a whole.
5.17.7 Amount of Arpeggiation
The amount of arpeggiation is defined as the “fraction of horizontal intervals

that are repeated notes, minor thirds, major thirds, perfect fifths, minor sevenths,
major sevenths, octaves, minor tenths or major tenths” (McKay, 2004, p. 76).
Amount of Arpeggiation
Global Mean
Global Median
0.8
Relative Frequency
0.6
0.4
0.2
Country Blues
Modern Classical
Adult Contemporary
Flamenco
Baroque
Renaissance
Celtic
Cool
Bebop
Ragtime
Classical
Swing
Hardcore Rap
Romantic
Soul Blues
Salsa
Blues Rock
Funk
Chicago Blues
Psychedelic
Soul
Rock and Roll
Metal
Reggae
Hard Rock
Alternative Rock
Punk
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado (Edition)
Fado
Dance Pop
Pop Rap
Medieval
Traditional Country
Tango
Techno
Figure 5.60: Amount of Arpeggiation.
It is somewhat surprising that both fado as a whole and fado melodies have
a relatively low level of arpeggiation (figure 5.60). This can only be explained
by fado on the horizontal level being organized on diatonic steps, and fourths,
the accompaniments included. While it might seem counterintuitive at first, one
can picture an alberti bass based on the dominant, as being a frequent trait, for
instance. That is often written in the sequence fifth leading up to the tonic, low-
ering to the seventh and going back up to the tonic, which would correspond to a
162
“melody” consisting only on perfect fourths and half tones. Medieval, flamenco,
baroque, modern classical and Celtic are the other genres with low levels of arpeg-
giation, while punk, techno, alternative rock and hard rock seem to rely on these
intervals.
5.17.8 Repeated Notes
The repeated notes represent the “fraction of notes that are repeated melodi-
cally” (McKay, 2004, p. 76), in other words, unisons.
Repeated Notes
Global Mean
0.6
Global Median
0.5
Relative Frequency
0.4
0.3
0.2
0.1
0.0
Country Blues
Modern Classical
Adult Contemporary
Flamenco
Bebop
Ragtime
Cool
Celtic
Romantic
Salsa
Swing
Soul Blues
Baroque
Psychedelic
Blues Rock
Renaissance
Rock and Roll
Chicago Blues
Soul
Classical
Funk
Hardcore Rap
Reggae
Alternative Rock
Metal
Hard Rock
Punk
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.61: Repeated Notes.
Fado as a whole stands out in this feature (figure 5.61), having the second
lowest frequency of unisons of all taxonomies along with flamenco, bebop and
ragtime. This is mainly due to the accompaniments, because when analyzing only
the melodies then fado is within the average. Metal, hard rock, techno and punk
are the practices who can be defined by over emphasizing this trait.
163
5.17.9 Chromatic Motion
Chromatic Motion is the “fraction of melodic intervals corresponding to a semi-

tone” (McKay, 2004, p. 76).
Chromatic Motion
Global Mean
0.4
Global Median
0.3
Relative Frequency
0.2
0.1
0.0
Country Blues
Modern Classical
Adult Contemporary
Punk
Rock and Roll

Alternative Rock
Hard Rock
Psychedelic
Reggae
Salsa
Metal
Chicago Blues
Soul
Funk
Celtic
Soul Blues
Blues Rock
Flamenco
Ragtime
Romantic
Cool
Hardcore Rap
Baroque
Classical
Bebop
Renaissance
Swing
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado
Dance Pop
Pop Rap
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.62: Chromatic Motion.
While fado as a whole is at the lower part of the spectrum (figure 5.62), re-
garding this feature, when looking only at the melodies fado can be found at the
highest part of it, ranked third, among swing, medieval, renaissance and tango.
This clearly represents the importance of half tones in fado melodies, which, while
not being the most common interval, in relative terms, it represents a huge distinc-
tion in comparative terms. Since fado is a tonal style, and not a chromatic one,
this can only be explained as being the result of a relatively high use of leading
tones and minor seconds. This could be a defining trait.
5.17.10 Stepwise Motion
Stepwise Motion is the fraction of melodic intervals that corresponded to either

a minor or a major second.
164
Stepwise Motion
0.8
Global Mean
Global Median
0.6
Relative Frequency
0.4
0.2
0.0
Country Blues
Modern Classical
Adult Contemporary
Punk
Rock and Roll

Hard Rock
Chicago Blues
Salsa
Alternative Rock
Psychedelic
Reggae
Soul
Ragtime
Metal
Soul Blues
Funk
Romantic
Flamenco
Blues Rock
Bebop
Cool
Hardcore Rap
Celtic
Swing
Classical
Baroque
Renaissance
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Fado
Dance Pop
Pop Rap
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.63: Stepwise Motion.
This feature (figure 5.63) seems to be a highly useful one to define fado. While
fado as a whole is among the two lowest taxonomies (along with punk), the fado
melodies are among the two highest ones, only surpassed by medieval. In com-
parative and relative terms these values just show how important stepwise motion
and minor seconds are in melodic terms to define this style, and, on the other
side, how of little importance they seem to be to the accompaniment. The impor-
tance of stepwise motion, combined with the also high importance of chromatic
motion clearly seems to point out to a character of either undulating or perfectly
arch-shaped melodies, which could be modeled with Brownian noise.
5.17.11 Melodic Thirds
Melodic thirds refers to the “fraction of melodic intervals that are major or
minor thirds” (McKay, 2004, p. 76).
Melodic thirds do not seem to be a particular trait useful to define fado (figure
5.64). The main corpus lies slightly below the average, while the melodies are a
bit above the center of the spectrum, not standing out. Punk, metal and techno
165
Melodic Thirds
0.4
Global Mean
Global Median
0.3
Relative Frequency
0.2
0.1
0.0
Country Blues
Modern Classical
Adult Contemporary
Punk
Metal
Alternative Rock
Baroque
Hard Rock
Flamenco
Hardcore Rap
Renaissance
Ragtime
Reggae
Classical
Salsa
Soul
Funk
Romantic
Blues Rock
Cool
Bebop
Celtic
Swing
Chicago Blues
Rock and Roll
Psychedelic
Soul Blues
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Pop Rap
Fado
Dance Pop
Fado (Edition)
Medieval
Traditional Country
Techno
Tango
Figure 5.64: Melodic Thirds.
are the genres which least depend on this interval, while soul blues, psychedelic,
traditional country and rock and roll seem to be highly dependent on it.
5.17.12 Melodic Fifths
Melodic Fifths corresponds to the “fraction of melodic intervals that are perfect
fifths” (McKay, 2004, p,76).
Looking at fado as a whole, it stands in line with the average of other taxonomies
(figure 5.65). Minding the Melodies-only, then fado can be defined as having a
clear lack of melodic fifths, being the second lowest taxonomy along with medieval,
flamenco and punk. On the other extreme reggae, psychedelic and rock and roll
present a high prevalence on this kind of leap.
5.17.13 Melodic Tritones
Melodic Tritones is the “fraction of melodic intervals that are tritones” (McKay,
2004, p. 76).
166
Relative Frequency Relative Frequency
0.00
0.02
0.04
0.06
0.08
0.00
0.05
0.10
0.15
0.20
Medieval Medieval
Punk Fado (Edition)
Hardcore Rap Flamenco
Dance Pop Punk
Global Mean
Global Mean
Techno Modern Classical
Global Median
Global Median
Renaissance Hardcore Rap

Celtic Dance Pop
Alternative Rock Tango
Hard Rock Classical
Pop Rap Funk
Bluegrass Romantic
Reggae Blues Rock
Fado (Edition) Bebop
Classical Swing
Baroque Metal
Country Blues Techno
Smooth Jazz Baroque
Adult Contemporary Renaissance
Soul Country Blues
Traditional Country Ragtime
Contemporary Country Smooth Jazz
Rock and Roll Jazz Soul
Psychedelic Celtic
Melodic Fifths
Flamenco Soul Blues
Melodic Tritones
Tango Alternative Rock
Funk Hard Rock
Romantic Pop Rap
Figure 5.65: Melodic Fifths.

Blues Rock Fado
Figure 5.66: Melodic Tritones.

Metal Bossa Nova
Fado Adult Contemporary
Chicago Blues Soul
Salsa Cool
Modern Classical Chicago Blues
Bebop Traditional Country
Swing Bluegrass
Ragtime Salsa
Soul Blues Contemporary Country
Cool Rock and Roll
Jazz Soul Psychedelic
Bossa Nova Reggae
167
Overall the prevalence of melodic tritones is very low among all taxonomies
presented (figure 5.66), when compared to other melodic intervals, due to the fact
that most of these genres rely on tonal tradition. Still, within this low range, one
can notice how fado, overall, belongs slightly above the center of the spectrum
while the melodies appear on the bottom part of the spectrum emphasizing the
lack of prevalence of this leap in the voice. Still, other taxonomies like medieval,
punk and hardcore rap are much more extreme, with values bordering zero, while
bossa nova, cool, jazz soul and soul blues seem to tolerate this interval more.
5.17.14 Melodic Octaves
Melodic Octaves represents the “fraction of melodic intervals that are octaves”
(McKay, 2004, p. 76).
Melodic Octaves
Global Mean
Global Median
0.4
Relative Frequency
0.3
0.2
0.1
0.0
Country Blues
Modern Classical
Adult Contemporary
Hardcore Rap
Renaissance
Reggae
Celtic
Baroque
Rock and Roll
Blues Rock
Chicago Blues
Swing
Soul Blues
Classical
Soul
Psychedelic
Cool
Hard Rock
Funk
Romantic
Salsa
Bebop
Flamenco
Metal
Ragtime
Alternative Rock
Punk
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado (Edition)
Pop Rap
Dance Pop
Fado
Medieval
Traditional Country
Tango
Techno
Figure 5.67: Melodic Octaves.
This seems to be another feature useful to define fado (figure 5.67). With the
accompaniments included, fado is one of the genres with highest prevalence, along
with techno and punk. On the other hand, analyzing just the vocal melodies, fado
is the second lowest close to medieval and hardcore rap. So it seems that melodic
168
5.18. Melodic Contour
octaves only make sense in the accompaniment, being a vital part of it, while they
are absent in the vocals.
5.18 Melodic Contour
We had the intuition that fados’ melodies were mainly arch shaped in a very
smooth way. Our work with the database has shown that in terms of seeing each
melody as a whole that is true – the average pitches in the middle are higher than
the extremes, and the ending ones are below than the first ones. We have retrieved
the melodic contour of the melodies in the corpus, using Huron’s classification
system10 and compared them with the ones provided by himself, relative to the
Essen Folksong Database (Huron, 1996). This database comprises 6251 folksongs
from mostly European sources, so it is a good comparative source.
Table 5.19: Prevalence of Melodic Contours
Huron Contour Essen Folksong Database (%) Fados’ Melodies (%)

Ascending 27.3 16
Descending 16.3 25
Concave 12.9 6
Convex 41.9 53
Horizontal-Ascending 0.3 0
Horizontal-Descending 0.2 0
Ascending-Horizontal 0.4 0
Descending-Horizontal 0.3 0
Horizontal 0.4 0
Total 100 100
More than half of fados’ melodies are convex, which is a higher percentage than
the folksongs in the Essen database. Furthermore a quarter of fados’ melodies
are descending, which is also a prevalence much higher than those in the Essen
database. These two contours alone represent a total of 78% of fados’ melodies
clearly defining the archetype. On the other hand, while a quarter of the folksongs
in the Essen database are ascending, only a sixth of fados’ melodies do so. Fur-
thermore, the concave contour represents an eight of the folksongs in the Essen
10
one of the additional features available in MeloSpyGUI (http://jazzomat.hfm-weimar.
de/, accessed June, 6, 2015).
169
database, while in fado this prevalence is halved. All other contours have negligi-
ble expression in the folksongs in the Essen database and null expression in fados’
melodies.
While the overall statistics might present a general, however distorted idea,
other statistics retrieved can complement it better. Overall the ratio of melodic
intervals that rise rather than fall is very close to half, but slightly less, which gives
a tendency for fados’ melodies to slowly go down over the time. The duration of
melodic arcs, which is the average number of notes separating peaks and troughs
is incredibly small, ranging from one to three and averaging at two. The size of
those melodic arcs, which is the interval that separates each peak from each trough
is also rather small, averaging at less than a tritone. This seems to indicate rather
zigzagging melodies falling down in the long run, instead of smooth arcs.
Table 5.20: Statistics of Melodic arcs
Direction of Motion 0.46 0.46 0.06 0.00 0.27 0.68 0.41 0.01 1.22
Duration of Melodic Arcs 1.99 1.91 0.39 0.15 1.12 3.09 1.97 0.87 0.64
Size of Melodic Arcs 5.68 5.58 1.20 1.44 3.20 9.62 6.42 0.72 0.84
5.18.1 Direction of Motion
Direction of Motion is defined as the “fraction of melodic intervals that are

rising rather than falling” (McKay, 2004, p. 76).
This is another feature useful to define fado (figure 5.68): when seen as a
whole, fado is the taxonomy that most rises, and in a prevalence significantly
higher than all other taxonomies. Country blues, funk, traditional country and
ragtime are other styles which rise overall. This pattern could be explained by
a series of pitches consecutively rising and then a few huge leaps downward to
compensate. Paying attention to just the melodies the pattern changes: fado
aligns with the styles presenting downward motion, along with punk, dance pop,
flamenco, metal, pop rap and alternative rock. This, again, shows us a contrasting
tendency between the melodies and the accompaniment.
170
5.18. Melodic Contour
Direction of Motion
0.8
Relative Frequency
0.6
0.4
0.2
Global Mean
Global Median
Country Blues
Modern Classical
Adult Contemporary
Punk
Flamenco
Metal
Alternative Rock
Baroque
Classical
Renaissance
Blues Rock
Swing
Cool
Romantic
Bebop
Hardcore Rap
Reggae
Celtic
Rock and Roll
Soul Blues
Chicago Blues
Soul
Hard Rock
Salsa
Psychedelic
Ragtime
Funk
Bluegrass
Smooth Jazz
Jazz Soul
Bossa Nova
Dance Pop
Pop Rap
Fado (Edition)
Fado
Medieval
Traditional Country
Techno
Tango
Figure 5.68: Direction of Motion.
5.18.2 Duration of Melodic Arcs
The duration of melodic arcs is defined as the “average number of notes that
separate melodic peaks and troughs in any channel” (McKay, 2004, p. 76).
Overall it can be seen how fado melodies presents relatively long melodic arcs
(figure 5.69). This tendency is emphasized when fado is seen as a whole, having
the third highest median, only surpassed by punk and flamenco. On the other side
of the spectrum Celtic, bossa nova and Chicago blues present the briefest melodic
arcs.
5.18.3 Size of Melodic Arcs
The size of melodic arcs is defined as the “average melodic interval separating
the top note of melodic peaks and the bottom note of melodic troughs” (McKay,
2004, p. 76).
This seems to be another relevant feature to define fado (figure 5.70). On

the whole, fado is the taxonomy presenting the tallest melodic arcs, followed by
171
Semitones Notes
172
10
20
30
40
50
2
4
6
8
Medieval Celtic
Renaissance Bossa Nova
Fado (Edition) Chicago Blues
Baroque Bluegrass
Global Mean
Global Mean
Hardcore Rap Psychedelic
Global Median
Global Median
Classical Salsa
Swing Contemporary Country
Celtic Traditional Country
Blues Rock Rock and Roll
Funk Baroque
Pop Rap Blues Rock
Reggae Techno
Techno Bebop
Dance Pop Reggae
Psychedelic Renaissance
Jazz Soul Cool
Smooth Jazz Adult Contemporary
Chicago Blues Modern Classical
Soul Blues Soul
Traditional Country Swing

Hard Rock Ragtime
Bossa Nova Funk
Metal Hard Rock
Bluegrass Soul Blues
Adult Contemporary Jazz Soul
Size of Melodic Arcs

Contemporary Country Smooth Jazz
Duration of Melodic Arcs
Cool Classical
Bebop Fado (Edition)
Soul Medieval
Figure 5.70: Size of Melodic Arcs.

Romantic Romantic
Modern Classical Hardcore Rap
Figure 5.69: Duration of Melodic Arcs.

Salsa Metal
Tango Dance Pop
Ragtime Tango
Alternative Rock Country Blues
Rock and Roll Pop Rap
Punk Alternative Rock
Country Blues Fado
Flamenco Flamenco
Fado Punk
5.19. Position of the final tone
flamenco, country blues and punk. This is due to the accompaniment figurations
because analyzing only the melodies, fado presents one of the shortest melodic arcs,
along with medieval and renaissance. This is another clue on how fado melodies
seem to privilege stepwise motion, having fewer leaps than other taxonomies, over-
all. The values obtained also seem to indicate either undulating melodies or, in
case of arch shapes, a long arch with a short peak, within the interval of an octave.
5.19 Position of the final tone
According to the data observation, the first note of a fado is usually either
the dominant or the tonic, then followed by an arch shaped or undulating form
that invariably comes to rest on the tonic. In simple terms, this is usually the
lowest note of the song, in abstract notation. Sometimes, however, in undulating
fados, the tonic is achieved one octave higher, which breaks the previous rule. In
very dramatic performances and for emotional and climactic effects, sometimes
the peak note is chosen to be also the last. Therefore, in those situations, the
melody is varied in the last strophe in a way that the tonic standing in a low
register is transposed one or even two octaves higher and reached via an upward
movement. This can be the highest note of that particular performance of that
fado. These situations are more exceptions than rules, and usually occur only
when the performer is rather skilled and intends to show off (a diva, for instance),
and also in more complex fados that resemble the format of art songs.
5.20 Dynamics
The volume of a fado also depends greatly on the skill of the performer, the
context, performative space, and the narrative involved. In the context of street
or amateur venues, fados are sung louder and sometimes with resource to yelling
and shouting. In the context of professional venues, with a silent audience and a
mournful narrative, the performance can be extremely soft and delicate. Some-
times a fado alternates the two moments, having very soft dynamics in some
phrases and then growing and growing until it reaches a climax in very high notes
with powerful dynamics and voice projection, to immediately being finished almost
173
in a counter-climatic whisper. “In a performance of traditional fado, strophes of-

ten accrue amplified emotional significance and more marked improvisation as the
singer progresses through the poetic narrative of the fado – that is, they accrue
mais fado (more fado)” (Gray, 2013, p. 40). In empirical study, shimmer mean
values among fado singers were lower than in all other music styles (pop, country,
soul, musical theater, western classical), except Jazz (Mendes et al., 2013).
The calculation of dynamics took into account a relative loudness unit defined
as the note velocity times the channel volume divided by the overall possible
range, which will result in a value between 0 and 127 (McKay, 2004, p. 72). The
main issue is that many of the transcriptions lacked dynamics information, other
times such information was presented in a subjective way by the means of signs
or words. Even the typical indications found in conventional literature like “pp”,
“mf ” or “f ” lack an objective relation to a typical MIDI scale with a resolution
of 127, being complemented by the volume or expression CC. At first these issues
could be a problem if one thinks on how to get meaningful data having these
highly fluctuating constraints. However, since the encoding of all transcriptions
was done using the same software, having the same interpreting algorithms and
conventions, then, in a relative way all dynamics are comparable. Although we
are not aware what the exact value for “pp” is or the algorithm for a crescendo,
we know they are consistent among all the data encoded and therefore they can
be compared.
Having this in mind, we have calculated the overall dynamic range for each fado.
This range is simply the difference between the maximum and minimum loudness
found. We have also calculated the variation of Dynamics, which is the standard
deviation of loudness level of all notes, and the average change of loudness from
one note to the next note in the same channel (McKay, 2004, p. 72). All these
parameters were calculated for both the edition corpus and for the Melodies-only.
As expected, there is an enormous variance among the overall dynamic range,

reflecting both the fados totally lacking dynamic information (hence, the software
assuming default values causing very narrow and steady ranges, for the entire
performance), and the ones presenting detailed, contrasting loudnesses going from
“pp” to “ff ”. It is not linear to assume that the lack of dynamic range in some fados
is merely due to lack of information from the transcriber neglecting that parameter
– one has to consider that the devaluation of that parameter might indeed hint
at a practice where the range was somewhat flat and therefore not noteworthy
174
5.21. Comparative features based on Dynamics
enough to actually notate its fluctuation. Some of the early recordings point
to cases like these. Still, the overall dynamic range is much higher in the fado
corpus in general and much narrower considering the Melodies-only, suggesting
more and more contrasting dynamic indications within the accompaniments. That
is corroborated by the variation of dynamic also being higher within the edition
corpus than in the melodies. However, when analyzing the note to note transitions
within the same channel one notices they are very similar – in fact, slightly higher
considering the Melodies-only. This clearly suggests that the contrast in the overall
dynamic range is mainly a result of vertical discrepancies, and not horizontal.
Horizontally it seems that the voice actually fluctuates more and presents more
contrasting dynamics, while the accompaniment keep their horizontal voices more
consistent but clearly delineated among themselves.
Table 5.21: Statistics of Dynamics
Dynamics Avg Med St Dev Var Min Max Range Skew Kurt
Overall Dynamic Range 44.10 41.50 23.02 530.03 13.00 102.00 89.00 0.50 -0.83
Variation of Dynamics 9.67 7.48 4.66 21.68 4.40 26.50 22.10 1.29 1.18
Avg Note to Note 6.36 6.65 1.68 2.82 3.19 11.24 8.05 0.14 0.22
Overall Dynamic Range 26.13 18.00 15.59 242.98 8.00 72.00 64.00 1.47 1.22
Variation of Dynamics 7.52 6.04 3.71 13.80 2.07 19.18 17.11 1.81 2.44
Avg Note to Note 6.51 6.78 1.93 3.71 1.83 9.97 8.14 -0.55 -0.29
5.21 Comparative features based on Dynamics
For the purpose of comparing dynamics a unity of “loudness” was defined as

the “velocity values scaled by volume channel messages:
note loudness = note velocity × (channel volume/127)
All features based on dynamics use relative measures rather than absolute mea-
sures (such as average volume) because the default volume and velocity values set
by sequencers can vary, and many MIDI authors simply encode their files without
varying these values” (McKay, 2004, p. 72).
175
5.21.1 Overall Dynamic Range
The overall dynamic range is defined as “the maximum loudness minus the
minimum loudness value” (McKay, 2004, p. 72).
Overall Dynamic Range
120
100
80
Loudness
60
40
20
Global Mean
Global Median
0
Country Blues
Modern Classical
Adult Contemporary
Flamenco
Renaissance
Punk
Baroque
Metal
Ragtime
Celtic
Alternative Rock
Classical
Hard Rock
Hardcore Rap
Reggae
Blues Rock
Rock and Roll
Soul
Psychedelic
Romantic
Cool
Soul Blues
Funk
Chicago Blues
Swing
Salsa
Bebop
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Tango
Techno
Figure 5.71: Overall Dynamic Range.
Fado is one of the genres with lowest overall dynamic range (figure 5.71, even in
the edited version of the corpus) among the likes of early music, flamenco, country
blues and punk. At the opposite side techno, tango, bebop and salsa show very
high dynamic ranges. It is noteworthy, however, the presence of a considerable
number of genres with very large spans, occupying the entire loudness spectrum
(like metal, alternative rock, classical and blues rock), which indicates the presence
of recordings both with maximum and minimum dynamic ranges. It seems that
the density of instrumentation is a key factor influencing this feature – the more
voices present, and the presence of percussion, seems to increase the range overall,
which is also expected.
176
5.21. Comparative features based on Dynamics
5.21.2 Variation of Dynamics
The variation of dynamics is defined as the “standard deviation of loudness

levels of all notes” (McKay, 2004, p. 72).
Variation of Dynamics
35
Global Mean
Global Median
30
25
20
Loudness
15
10
0
Country Blues
Modern Classical
Adult Contemporary
Flamenco
Renaissance
Ragtime
Metal
Classical
Baroque
Punk
Alternative Rock
Psychedelic
Soul
Salsa
Reggae
Blues Rock
Celtic
Swing
Hard Rock
Romantic
Rock and Roll

Funk
Cool
Soul Blues
Chicago Blues
Hardcore Rap
Bebop
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Dance Pop
Pop Rap
Medieval
Traditional Country
Tango
Figure 5.72: Variation of Dynamics. Techno
Following the trend of a smaller relative range, fado is also among the genres
with less dynamic variation (figure 5.72), along with medieval, country blues,
flamenco, renaissance and ragtime. We highlight the fact that baroque has slightly
climbed up the scale, being a style that varies within its narrow range. At the
opposite side, bebop, techno, tango and hardcore rap are the genres with highest
variety. Also of note is the fact that salsa left the top part of this chart. Hence,
while there seems to be a strong correlation between the overall dynamic range
and the dynamic variability, in some cases that is not true.
5.21.3 Variation of Dynamics In Each Voice
The variation of dynamics in each voice is defined as “the average of the stan-
dard deviations of loudness levels within each channel that contains at least one
note” (McKay, 2004, p. 72).
177
Variation of Dynamics In Each Voice
30
Global Mean
Global Median
25
20
Loudness
15
10
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Baroque
Punk
Metal
Flamenco
Alternative Rock
Hardcore Rap
Classical
Ragtime
Celtic
Reggae
Hard Rock
Rock and Roll
Blues Rock
Funk
Soul
Bebop
Cool
Soul Blues
Psychedelic
Salsa
Swing
Chicago Blues
Romantic
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Techno
Figure 5.73: Variation of Dynamics In Each Voice. Tango
Analyzing the dynamic variation in each voice (figure 5.73), it can be seen that
more distortions occur, meaning that, overall, some styles are highly influenced
by just one or few of its instruments. It is very likely that in some styles one or
few instruments use a very large dynamic range while the majority varies much
less. In the case of fado, the general corpus is still among the less variable cases,
along with medieval, renaissance, baroque, punk and country blues; however, the
edited version of the corpus clearly approaches the global median and mean. On
the other side of the spectrum one can find romantic, Chicago blues, swing and
salsa. This fact corroborates the idea that in genres like techno and bebop one
or few instruments create contrast and influence the overall perception of their
dynamic range.
5.21.4 Average Note To Note Dynamics Change
The average note to note dynamics change is defined as the “average change of
loudness from one note to the next note in the same channel” (McKay, 2004, p.
72).
178
5.22. Pitch Bend Messages
Average Note To Note Dynamics Change
Global Mean
25
Global Median
20
Loudness
15
10
Country Blues
Modern Classical
Adult Contemporary
Renaissance
Baroque
Punk
Metal
Alternative Rock
Classical
Flamenco
Hardcore Rap
Romantic
Ragtime
Blues Rock
Celtic
Hard Rock
Swing
Reggae
Psychedelic
Bebop
Cool
Funk
Soul
Rock and Roll
Chicago Blues
Soul Blues
Salsa
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Fado
Fado (Edition)
Pop Rap
Dance Pop
Medieval
Traditional Country
Techno
Tango
Figure 5.74: Average Note To Note Dynamics Change.
Fado, along with the early music genres, punk, metal and country blues has
virtually no change of dynamics from one note to the next (figure 5.74). This
might be because it relies on smoothly gradual continuous dynamics, with no
ruptures or abrupt changes, or can also be because there are actually no changes in
many of its voices. This can indicate instruments incapable of producing dynamic
changes (like the harpsichord), or can simply indicate that the recordings were
made without this parameter taken into consideration. In the case of fado that
was already discussed as a pertinent issue, and when one tries to compensate that
fact, in the edited corpus, one realizes how fado approaches the global mean and
median. At the other side of the spectrum, tango, salsa, soul blues and Chicago
blues are genres which rely on relatively abrupt dynamic changes.
5.22 Pitch Bend Messages
The next three features rely on the measuring of pitch bend messages. However,
this is an extremely problematic issue. While, if uniformly encoded, they would
give many interesting clues, the problem lies in the encoding: it is a parameter
179
that is seldom notated in written notation, hence, most times totally absent from
the musical scores and MIDI transcriptions derived from notation programs. Even
when MIDI files are generated by live playing with controllers, many do not have
ways to encode these messages due to the lack of knobs. Other times, they are
simply neglected during performance (for instance, playing a melodic line in a
keyboard and not using the pitch bend wheel). Therefore, historically, many
musical practices in real life performances might present these features (namely
in vocal lines or continuum instruments like trombones or violins) and they are
completely missing in the MIDI files.
The fado corpus formed completely neglects this aspect, since pitch bend mes-
sages, glissandi or vibrato are not transcribed in notated music, while we often
find them in the phonograms and real life performances. We speculate that this
might happen with many other recordings in the database, and most practices in
general. Therefore, for historical and technical reasons, we do not think the data
we have is reliable enough regarding these particular features. They are presented
here just for demonstrative purposes of their potential.
5.22.1 Glissando Prevalence
The glissando prevalence is defined as the “number of Note Ons that have
at least one MIDI Pitch Bend associated with them divided by total number of
pitched Note Ons” (McKay, 2004, p. 74).
If one relies on the data we have (figure 5.75), then one has to say that fado is
among the taxonomies that do not present glissando, which are the major part of
them. At the opposite side of the spectrum, four blues-related taxonomies (country
blues, soul blues, blues rock, Chicago blues) have a significant prevalence. For the
reasons already explained we do not believe these results represent reality.
5.22.2 Average Range of Glissandos
The average range of glissandos is defined as the “average range of Pitch Bends,
where range is defined as the greatest value of the absolute difference between 64
and the second data byte of all MIDI Pitch Bend messages falling between the
Note On and Note Off messages of any note” (McKay, 2004, p. 74).
180
~ Cents Relative Frequency
0
10
20
30
40
50
60
0.00
0.05
0.10
0.15
Modern Classical Modern Classical
Hardcore Rap Hardcore Rap
Flamenco Flamenco
Techno Techno
Global Mean
Global Mean
Romantic Romantic
Global Median
Global Median
Pop Rap Pop Rap

Punk Punk
Bossa Nova Bossa Nova
Tango Tango
Fado Fado
Fado (Edition) Fado (Edition)
Salsa Salsa
Reggae Reggae
Celtic Celtic
Classical Classical
Bebop Bebop
Medieval Medieval
Cool Cool
Renaissance Renaissance
Baroque Baroque
Jazz Soul Jazz Soul
Bluegrass Bluegrass
Ragtime Ragtime
Soul Soul
Traditional Country Adult Contemporary
Glissando Prevalence
Country Blues Dance Pop

Adult Contemporary Metal
Average Range of Glissandos

Metal Traditional Country
Rock and Roll Funk
Swing
Dance Pop
Figure 5.75: Glissando Prevalence. Smooth Jazz
Alternative Rock
Chicago Blues Swing
Figure 5.76: Average Range of Glissandos.

Alternative Rock Hard Rock
Soul Blues Rock and Roll
Funk Contemporary Country
Hard Rock Psychedelic
Smooth Jazz Chicago Blues
Blues Rock Blues Rock
Psychedelic Soul Blues
Contemporary Country Country Blues
181
5.22. Pitch Bend Messages
Since we do not believe the raw data is reliable (figure 5.76) we do not find the
results credible. Since most taxonomies do not present glissandos, then their range
is next to zero, fado included in this set. It is noteworthy however to find that
among the taxonomies that present this feature encoded, contemporary country
and psychedelic are the two most prominent, even though they were not among
the ones having the highest glissando prevalence. This means that in these genres,
the few glissandos they possess are accentuated, while the set of blues-related
taxonomies seem to have more glissandos but which are weaker, overall.
5.22.3 Vibrato Prevalence
The Vibrato Prevalence is defined as the “number of notes for which Pitch
Bend messages change direction at least twice divided by total number of notes
that have Pitch Bend messages associated with them” (McKay, 2004, p. 74).
Vibrato Prevalence
0.07
Global Mean
0.06 Global Median
0.05
Relative Frequency
0.04
0.03
0.02
0.01
0.00
Country Blues
Modern Classical
Adult Contemporary
Hardcore Rap
Flamenco
Romantic
Punk
Salsa
Reggae
Celtic
Classical
Bebop
Cool
Renaissance
Baroque
Ragtime
Soul
Metal
Rock and Roll
Swing
Chicago Blues
Alternative Rock
Funk
Hard Rock
Blues Rock
Psychedelic
Soul Blues
Bluegrass
Jazz Soul
Smooth Jazz
Bossa Nova
Pop Rap
Fado
Fado (Edition)
Dance Pop
Medieval
Traditional Country
Techno
Tango
Figure 5.77: Vibrato Prevalence.
The same encoding issues remain (figure 5.77): most taxonomies do not present
vibrato at all, or just some outliers, meaning most files were not encoded taking
into account pitch bend messages, therefore the initial data is unreliable to describe
182
5.23. Embellishment, ornamentation and vocal effects
the actual practices. Among the ones that have this feature encoded soul blues
and contemporary country seem the most relevant.
5.23 Embellishment, ornamentation

and vocal effects
In the oldest written sources it is possible to detect only a minimal degree

of embellishment expressed through some grace notes and ornamentation on the
melody. It is not known exactly how they would reflect the actual practice. In the
early recordings, the singing is mainly flat, with minimum embellishment. Styling
is shaped employing numerous specific techniques grounded on timbral variations.
However, throughout the twentieth century some particular performers used and
abused this and other similar resources discussed further ahead, namely Amália
Rodrigues. Often mordentes and similar figurations were applied in certain notes
or passages, and this peculiar way of singing was described as voltinhas [twirls].
Tremolo and glottal shake is sometimes also used, especially on longer and
dramatic notes, depending much on the performers and the way they embody the
narrative. The use (sometimes overuse) of embellishment by Amália Rodrigues
was emulated in the following years, and spread through phonographic recordings.
Many new performers would follow the vocals traits found there and would try
to impersonate them. Eventually those gestures became trademarks of the genre,
and at present embellishment is standard in fado.
There are numerous specific techniques and descriptors that shape the way in
which the styling is done. Not only rhythm and pitch are slightly varied in a
random way, although within certain constraints to respect the overall structure,
but also a whole plethora of timbral variations.
This series of parameters, gestures and shapes associated with timbre and how
the voice is managed are not notated and are completely missing from the written
scores, especially the ones from the nineteenth century, hence the empirical data
obtained is not reliable regarding these features. They are, however, mentioned
in the literary descriptions and in the ethnographic studies, and can be deducted
and inferred from the discourses of the practitioners. They are absolutely crucial
183
to define the practice, because their presence or absence might lead a song to be
considered fado or not.
Many fado performers after Amália Rodrigues use her typical style, or close to
it, which is defined as “a slight portamento connecting the lower and the upper
note of the ascending intervals, and this same upper note attacked with a small
mordente leading the fall into the next lower note. The outcome is an all-time
legato singing reinforcing the meaning of each verse” (Nery, 2010c, p. 105).
The glissando (in this specific vocal context the term is used interchangeably
with portamento or slide), occurs when one goes from one tone to another in
a continuous way, passing through all intermediate levels. This effect happens
often in fado, and is notorious especially in ascending leaps, and most notably
in pickup notes or from a weak to a strong beat. Melisma is also frequent, due
to the legato. Often the same syllable is sung and prolonged through various
notes, some of them micro-tonal and out of the scale in a kind of mix between
embellishment and glissando/portamento. Inflexions are also frequent, as well as
sobs, wails and shouts. This specific way of articulating the melody is nowadays
a trademark gesture of fado, and some people dare to say that fado itself can be
inferred solely by its presence: the famous Portuguese maestro António Vitorino
d’Almeida, author of a few fados himself, on the TV show Autores, stated that
“fado is a way to launch your voice, to launch the musical phrase” or “to articulate
the melody”11 , and demonstrates his assumptions by giving an example on the
piano, based on this kind of articulation. This peculiar way of singing, so many
times repeated and made famous through time, established a model, a canon
on how to properly sing fado – also confirmed, by (Gray, 2013, pp. 210-225).
Moreover, it created the possibility of making virtually any music to be recognized
as fado. This should not be seen as a strange phenomenon, after all “the gestural
energy of a melody is phenomenologically more fundamental than the sequence
of pitches of which a melody is comprised” (Hatten, 2004, p. 114). Amália was
able to assimilate all kinds of genres, from whichever countries, and lend them her
own personal style. Therefore, she was able to produce constantly a “fado beyond
fado”, a typical unifying sound that affected all of her repertoire regardless of its
origin (Nery, 2010c, pp. 193-194). Amália was well aware of that phenomenon:
“She stresses that her fame as fadista reached such a high degree that today people
consider as fado everything she sings: ‘I may sing a malhão or a vira, or even a
11
https://www.youtube.com/watch?v=40M432Bihg0, accessed June, 6, 2015
184
simple song; everybody will consider it as a fado’ ” (Côrte-Real, 1991, p. 67). In

the same interview, quoted by Côrte-Real, she emphasizes and demonstrates how
her vocal gesture makes the style and defines the genre through a musical example:
she sings a melodic line “flat” and then she sings the same melodic line using her
typical ornaments, portamento and mannerisms, and states how the first version
would be considered one style and the second version a fado, in spite of being the
same melody with the same exact lyrics. According to Hatten,
“musical gesture is biologically and culturally grounded in commu-

nicative human movement. Gesture draws upon the close interaction
(and intermodality) of a range of human perceptual and motor sys-
tems to synthesize the energetic shaping of motion through time into
significant events with unique expressive force. The biological and cul-
tural motivations of musical gesture are further negotiated within the
conventions of a musical style, whose elements include both the dis-
crete (pitch, rhythm, meter) and the analog (dynamics, articulation,
temporal pacing). Musical gestures are emergent gestalts that convey
affective motion, emotion, and agency by fusing otherwise separate el-
ements into continuities of shape and force”12 (Bandt et al., 2007, p.
355).
Furthermore,
“Perhaps the most important function of gesture, however, comes

from its thematization as motivic idea. (. . . ) Thematic gestures, like
motives, are further marked as significant parts of the discourse of a
movement, and they play a structural as well as expressive role in the
unfolding form and expressive genre of a movement or work” (Hatten,
2006, p. 8).
Tremolo as a stylistic conscious device is only used in held notes, especially the
higher ones. Other than that it is mainly a recurrent condition in some fadistas, like
Frei Hermano da Câmara, for instance, who have a tremulous timbre by default.
Because fado is mainly sung by uneducated voices, with no formal training, the
position and projection is often made in the throat. Therefore glottal shake,
12
http://www.music.indiana.edu/html/department/theory/courses/t561-f03_
hatten.shtml, accessed June, 6, 2015
185
vibratos, shouts and speech mixed with singing is frequent. Since the melody
in fado is mainly improvised, and can be shaped to the will of the performer,
linked to the prosody, the register of fado is therefore mainly low, sometimes chest
voice, and close to the spoken register. Only very virtuosic singers risk the head
register and falsetto is rather uncommon, only happening when someone is trying
to imitate another performance and sticking to a range or key that is not tailored
to his or herself. In empirical study,
“jitter mean values were lower than Pop and Western Classical
singers and higher than those of Country, Soul, and Musical Theater
singers”, also, “for the vibrato analysis, 11/15 subjects (eight males and
three females) produced vowels with a vibrato frequency and extent
within the range described in the literature” (Mendes et al., 2013).
For Amélia Muge (singer, composer) fado is related to suffering, in the sense of
life experience, of embodying the narrative and being able to express and convey
it to the audience13 . Manuela de Freitas (lyricist) and José Luı́s Gordo (guitarra
player) shared the same thought14 . Emotion in fado is exaggerated. It is like
comparing Paulo de Carvalho and Carlos do Carmo, comparing a singer to a
fadista. Fado is more than just singing.
Amélia Muge, composer, has shown two versions of the same song, with differ-
ent arrangements, one sung by herself and another by Camané (famous fadista).
In her version she stated how she had a very light and fluid voice, and in Ca-
mané’s version, how his voice was carrying much more “weight” and how the
words acquired a much more emotional deepness compared to her original version.
According to Amélia Muge this was the main factor to inscribe fado in the sec-
ond version, that Camané had put the fado in that song. Fado was, then, a way
of living, and consequently expressed in the voice and through the body. When
asked directly about her values and if the articulation of the melody, the way to
launch one’s voice, was enough to define fado, she reinforced that it is indeed the
most important characteristic, but she had doubts about that being enough just
by itself. She believed it was more than that, and that it is an entanglement of
13
Debate “fado(s): Escritas e Autoria”, moderated by Soraia Simões, promoted by Mural
Sonoro in 15/12/2013.
14
Debate “Nova literatura para fado”, moderated by Soraia Simões, promoted by Mural Sonoro
in 26/01/2014.
186
several expressive factors that defines fado. Vital de Assunção (viola player, com-
poser) added that the metric and rhythm structure were crucial to define fado.
That fado had “rules”. Perhaps because Amélia Muge is a singer and Vital de
Assunção is a viola player, each of their own musical values reflect the importance
of their own roles in the practice. Therefore, for Vital de Assunção, the swing, the
groove and the balance of the harmonic figurations (most of the time binary, but
sometimes quaternary, never ternary), needed to be there for fado to exist. He
stated that if you do not have that layer and, even when you do, if when you play
the typical bass line and chords you do not have the urge to almost dance, or move
along, then it is not fado. In spite of these differences, the notion of movement
and the existence of an expressive layer communicated through the embodiment
of emotions in performance seems to be an important value to define the genre
among the community, in general.
One of the key characteristics of amateur fado performances, especially among

male performers, is the very narrow and tense vocal width. Often, performers es-
caping from the spoken register, instead of projecting the voice and using singing
techniques, do the opposite and embed the words in extreme emotion by narrow-
ing the vocal cavity, raising the glottis, raising the tongue and pulling it back and
tensing the muscles in the throat, while slightly tilting the head back and some-
times closing their eyes. Cumulatively, it is frequent to display a high degree of
rasp (hoarseness, harshness, grating or buzzing) when performing. It was observed
empirically that fado singers produced vibrato, but a singer’s formant was rarely
produced.
“The result of the lowering of larynx and enlargement of the phar-

ynx can be physiologically incompatible with the lateral positioning of
the head these singers adopt during the performance. The head later-
alization has a 45 angle with the shoulder and is slightly tilted back.
With this head-neck posture, it is difficult to expand the laryngeal
tube” (Mendes et al., 2013).
The idea of a strained throat is an important value, also pointed out by Ruben
de Carvalho (Carvalho, 1994, p. 95) and Ellen Gray (Gray, 2013, p. 40), because
it not only reinforces the idea that it is an unlearned skill (as opposed to lyrical
singing, for instance), but it also emphasizes the suffering, the unhappy fate, as
a husky, gritty, voice is commonly associated with drunk men in taverns and
187
with prostitutes, people who led difficult lives, smoked and drank a lot. It is a
timbre that embodies pain and suffering. A great model for this singing style
is the legendary Alfredo Marceneiro, who probably had great impact in coining
this style for men, and was largely imitated afterwards. This style of singing is
metaphorically linked with suffering, a very hard life, bad habits and tremendous
life experience, and thus, defines what it means to be a fadista. Also, “one is not
taught, nor does one learn to be a fadista; one is born a fadista” (Gray, 2013,
p. 28). In the case of women fadistas, the vocal width is usually not as narrow
as in men, and sometimes it is even absent and the voices are relaxed. However,
even in this case, rasp might be present with the precise metaphorical intention
of conveying the image of suffering and bad fate. This is widely known as “voz
de bagaço” [brandy’s voice], a timbre typical of people who drank a lot of alcohol
and never cared about the quality of their voices.
Gray also mentions how a voice marked by a “pronounced nasality, a texture

that might occasionally break into roughness, and sometimes a vocal placement
that falls between speaking and singing” might be deemed as carrying “more fado”
(Gray, 2013, p. 65), although we do not find nasalization to be that common and
relevant to define the practice. Portuguese language, by itself, has more nasal
sounds than most Indo-European languages, and this fact might have created a
perceptual bias on Gray, leading her to attribute to fado a characteristic that is
in fact from the language15 .
These differences in tempo, air pressure and placement are not easily mea-
surable or quantifiable. However, they correspond to articulatory functions like
off-beat accents (emphasizing syncopation), tenutos, slight ritardos, portamentos,
glissandi, mordentes and general ornamentations and arpeggios. There are cor-
relations between these occurrences and the shape of the melody; the strategic
places where they happen. One can imagine dynamics being correlated directly
with rhythm and pitch – the open vowels being the longest notes, thus the highest
pitches and the loudest dynamics. Therefore, if one thinks about the “statistically
ideal” heptasyllabic European Portuguese line, with pick up notes derived from the
absence of stresses in the first syllable, and one applies all the principles discussed,
a pattern will emerge – melodic contours that will contrast short rhythms, lower
registers and lower dynamics for the unstressed syllables, contrasting with larger
notes, higher pitches and louder dynamics in the stressed ones, which will need
15
As kindly pointed out by Jorge Martins Rosa.
188
5.24. Accent and enunciation of consonants
to be further compensated. According to the universal rules of entropy and iner-

tia, every step statistically privileges the least energy and movement, so singing is
usually not made in leaps but by steps. Therefore the melodies will tend to have
undulating or arch-shaped contours. The upward movements are favorable to por-
tamentos and glissandi (to approach and sustain the long, highest notes; similar
traits can be found in other improvisatory traditions, like scooping, in jazz) and
downward movement to mordentes, and melisma to compensate (like in flamenco
or Arabic songs). The long, high, open vowels are suitable for tenutos, suspen-
sions and vibratos, while also prone to fading. The intensification of these traits is
consistently observed over time in live performances and phonograms throughout
the twentieth century.
5.24 Accent and enunciation of consonants
In fado performance the accent is mostly linked to prosody. Therefore, the

stronger attacks are mainly linked with the stressed syllables containing open
vowels. Often there is a deliberate effort to place those syllables in the strong beats.
However sometimes this does not happen, hence the frequency of syncopation.
One can say that this is a very personal decision, open to the will or skill of
the fadista. Sometimes, for very dramatic effects, an accent might happen in
an unexpected place, thus breaking the prosody. A very famous case is Amália
Rodrigues’ interpretation of “Gaivota” [Seagull]. When she sings the line “Se uma
gaivota viesse” [If a seagull would come], instead of the logical stress in “vo”, she
prolongs and embellishes that syllable, in a melismatic way, but with a vanishing
dynamic. She then hits hard on the following unstressed, almost mute, syllable
“ta” and links it with the following “vie”, resulting in a cacophonous “gaivooooo....
TAVIEsse”. This particular choice was highly criticized by some peers, namely
Daniel Gouveia (Gouveia, 2010, p. 203), who points out this case (among many
others) as an example of how to not sing correctly. Diction is a value highly praised
in fado, and therefore the enunciation of consonants is usually very precise, so
that the narrative and the words are communicated in an efficient manner to the
audience.
Due to a combination of other factors, namely the timbre provoked by the nar-
row width and the raspiness associated with alcohol consumption, as well as the
189
typical legato singing, some old fadistas often present an abnormal, yet very char-
acteristic, degree of slurriness in their singing. Although this condition greatly
affects the intelligibility of the narrative, in these situations it is tolerated, since
it is the perfect metaphor for a long life being a fadista and the inevitable conse-
quence of all their extreme suffering. Therefore, paradoxically, one can say that
although diction and the precise enunciation of consonants are a goal in defin-
ing what good fado is, the practice shows that slur is characteristic of the genre,
and it is usually associated with masters and veterans (even Amália presented a
very slurred, rasp singing in her later years). When fado is mocked down and
portrayed in caricatures16 , often this exaggerated, slurred, unintelligible version is
shown. Ellen Gray points out how “the listeners perception of ‘soulfulness’ and
feeling in fado are intimately linked to the performer’s level of mastery of multiple
sonic communicative codes”, including “word placement, Lisbon-inflected pronun-
ciation, phrasing, breathing and use of rubato and vocal ornamentation” (Gray,
2013, p. 40-41).
Hence, we believe that the combination of all these timbral and articulatory
variations in styling, much more than pitch and rhythm, corresponds to a key
value expressed by most of the fadistas and practitioners, in such a way that
they often cannot describe or express in precise, technical terms, but that can be
deduced from their discourses. This is what Amélia Muge, for instance, meant
when she said that Camané sung her words with more “weight” and turned her
song into a fado; this is what Tozé Brito or José Luı́s Gordo mean when they say
that they write songs, but then fadistas make them fados. The embodiment of a
song into a fado – by fadistas, when expressed by their voices (and neglecting the
rest of the embodiment for now), corresponds to the set of timbral and articulatory
parameters described above, which are not present in an explicit technical and
detailed way in the discourses or narratives, are inconsistent with the musical
scores, but are implicitly present in the recordings, and are intuitively perceived
by any listener. They are a key to define, understand, and model the practice.
16
Caricatures are often very useful to understand a stereotype, since they portray and ex-
aggerate the most relevant traces. For notable fado caricatures, see Dulce Pontes videoclip
“Júlia Galdéria” (https://www.youtube.com/watch?v=XYT4iQ8ukWE, accessed June, 6, 2015)
and Joaquim Monchique’s impersonation of Amália Rodrigues (https://www.youtube.com/
watch?v=iC_yL8gArGw, accessed June, 6, 2015).
190
Chapter 6
Embodiment
Fado, being a total art form, relies heavily on body performance and on cues
that are beyond regular vocal and instrumental cues. The whole performance is
perceived and evaluated by the audience through the means of codes and con-
ventions that rely on visual communication, not only aural. This is the main
reason why Manuela de Freitas insists that recordings cannot replicate a live ex-
perience (sensorial information is missing). Also, a recent study by Tsay pointed
out that in piano competitions “people consistently report that sound is the most
important source of information in evaluating performance in music. However, the
findings demonstrate that people actually depend primarily on visual information
when making judgments about music performance” (Tsay, 2013), which empha-
sizes the importance of embodiment in establishing empathy with the audience for
the narrative to be successfully communicated, and thus for “fado to happen”.
According to Marc Leman, this empathy is possible because the physical con-
straints of human bodies are universal, therefore the sensorimotor basis of gestural
communication is common to most cultures. So, even within very different cul-
tures, the same gestures may have the same meaning. Furthermore,
“nonlinguistic descriptions, such as body movements in response

to music, rely on the transfer of sonic moving forms to motor moving
forms. The motor modality is then the expression of the original au-
ditory experience and, as such, it can be considered a description of
the original auditory experience and, as such, it can be considered a
description of the original music with the body” (Leman, 2008, p. 21).
191
So, in the same way that we all have a common sense of what a fast or slow
tempo is (because we all measure it up relative to the heart beat, that is a universal
trait within human species (Deutsch, 2012)), there are plenty more other gestures
and conventions we share simply because we are all humans. This can perfectly
account for why a Portuguese native can empathize with a flamenco or American
blues performance, the same way someone who does not understand Portuguese
still can be moved by fado.
“The connection between subjective experience and matter has to

do with the human brain”, however, we cannot forget the role of the
body as a mediator. “Scientific methodology has been expanding from
purely physical issues (music as sound) to more subjective issues (music
as experience). The connection between the two draws on fundamental
relationships between mind, body and matter” (Leman, 2008, p. 48).
Leman points out how a number of researchers during the twentieth century,
namely Lipps and Turlit, already saw the perception of emotions in music as a
phenomenon of physical empathy. The sonic forms are perceived in reference to
bodily movement expressions, and so we can actually perceive a music as sad,
without experiencing sadness ourselves (Leman, 2008, p. 44). This perspective
may be the solution for the subjective experience within the brain. Since much
research around music has its core on the cultural relativism and the plethora
of possible parameters through which music can be valued, without a definition
of perceived universals, the conception of the human body as mediator and its
own constraints constitute the universal by itself. Therefore, if one aims to be
objective regarding perception, emotion and value of music, one should search for
the part of the experience that is universal within every one of us – how our body
is interacting each time we experience music in context.
“Knowledge does not emerge from passive perception, but from the
need to act in an environment. In that sense, ecology is not merely
about the relationship between a subject and its environment, but
also about the knowledge which is needed to act in that environment”
(Leman, 2008, p. 43).
In this sense, then, there is something universal in fado that is recurrent and
transmitted by and through the body, and perceived by any audience. On the other
192
hand, there are also gestural and bodily conventions that only apply contextually,
because they are cultural constructions that emerge from the “ecology” of the
place. In that sense, the judgment of a fado performance by a foreigner will always
rely on different values and cues than the judgement by a native – language and
accent being the primary difference in recordings and phonograms, since the body
expression is mainly absent in that media.
Peter Kivy, one of the most reputable philosophers of music, fails to see how
emotion can be perceived in music precisely because he forgets the universality
of the mediator (the human body). His discourse is based on the experience of
music through cognition and he reaches the conclusion that the emotion must
be in the music like the redness is in the apple. He does not even conceive the
relativist assumptions of Nettl (Nettl, 2005) (and ethnomusicology in general),
because based on his own experience
“there is substantial agreement at least within broad boundaries as

to what garden-variety emotions various passages of expressive music
are expressive of, as anyone can demonstrate to himself by playing well-
chosen passages to competent listeners and asking for their responses.
So obvious is this fact that I will say no more about it” (Kivy, 2001,
p. 96).
Kivy argues that if he is emotionally moved by music, then music itself must be
the intentional object of his emotion. Also, he denies to acknowledge his emotions
as the garden-variety common names because he claims to actually feel none of
them. He also claims that music is “a thing of the mind”, and that “the musical
emotion that the music I take to beautiful or excellent in some other musical way
arouses in me is an enthusiasm, an intense musical excitement about what I am
hearing” (Kivy, 2001, p. 105). Kivy then reviews the positions of Davies and
Levinson, and how all of them agree on the point that music “is expressive of
the garden-variety emotions in virtue of the emotive qualities we recognize in it”
(Kivy, 2001, p. 106), but pointing out how for Levison, “the listener identifies with
a musical persona (. . . ) and then reacting to that, empathetically, sympathetically,
or antipathetically” (Kivy, 2001, p. 107). Kivy quickly dismisses this position by
arguing that “the analyst who tries to tease a persona and her misfortunes out
of a symphony or a string quartet is between Scylla and Charybdis. If she stays
within the bounds of sanity, she gets an abstraction that nobody can identify or
193
emphasize with. And if she tries to get Anna Karenina or Hamlet out of Opus
131, she stretches our credulity well beyond the breaking point” (Kivy, 2001, pp.
110-111). He stresses out how “the behavioral manifestations of the garden-variety
emotions are absent” (Kivy, 2001, p. 111).
What Peter Kivy seems to forget is that not only there is a whole realm of pos-
sibilities for facing even absolute music as programatic music, with an imaginary
program, based on a myriad of contexts, life situations, associations (and that in
fact, we can relate to a fictional “persona” triggered by our own subjective mind),
but also we do not necessarily need to feel those emotions, but just recognize them.
So, when we say that some piece of music is sad we are not saying that we are
sad, or feel sad, because of the music, neither that the sadness is in the music,
but instead that we recognize the physical and embodied properties of sadness in
the sonic movement we are perceiving as reference to our own bodies1 . And that
is the only reasonable explanation for any emotional universality (if one aims at
one). Our conclusion is that, then, music in itself does not carry any emotional or
expressive property in itself, but not everything is subjective/cultural experience
of the mind. Every musical experience is, then, the outcome of both the subjective
mind and the universal body, and so in every interpretation and description there
are some things that will always be very personal and unique and others that will
be common to most people, even when they do not share the same conventions
and culture. And that degree of communality is brought to us by our common
mediator, the body.
Those codes and conventions have been described extensively in the sources,
and have been changing greatly through time: in the beginning of the nineteenth
century, in Brazil, the descriptions mention a fado that was mainly a dance with
many traces in common with flamenco performances. In Lisbon the dance was
no longer the main feature, but still existed and it was an important part of
the practice. Throughout the twentieth century the dance lost importance and
nowadays fado, in professional venues, uses mainly the bodily posture conventions
of other popular ballad genres, in which the musicians sit and the singer stands
still, only allowing small steps, foot stomps, slight hip or upper body swells, and
arm and hand movement to metaphorically express and aid the narrative.
1
On this particular problem also see the distinction Noël Carroll portrays from “natural
horror” and “art-horror” (Carroll, 1990, pp. 12-42).
194
6.1. Fado as dance
“The fadista often sings the more melancholic songs with eyes
closed, with her head thrown back, her hands at times interlocked in
front of her and at others gesturing expressively, the torso in a stillness
rapt with a focus that directs all attention to the sound and expression
of her voice” (Gray, 2013, p. 29).
In amateur venues and informal contexts, in very cheerful fados, especially

the ones in march tempo, singers get looser and dance a little, clap and may re-
quest the audience to participate. Much of the posture conventions, dress codes,
lighting, decoration of the place, that are nowadays regarded as traditional val-
ues, were fixed and stylized during the dictatorship and are in fact, a “sanitized
and proper nationalist tradition”, an idealization of Portugalidade [Portugality],
created mainly as a commodity for the upper classes, tourists and foreigners.
6.1 Fado as dance
Fado appears referenced as a dance in Brazil in the beginning of the nineteenth

century. We will discuss those sources and the possible links of this practice to the
ones that reference fado in Portugal during the nineteenth century. It seems that
during a certain period of time and in some contexts the bodily expression was
heavily associated to what people perceived as fado. The original dance seemed
associated with instrumental music, however there are no clear references to voice
or lyrics in this ancient practice. Meanwhile, in nineteenth-century Lisbon, clear
documentation shows that music, voice, lyrics and dance were connected. Fado
was no longer a “dance”, but more of a performance practice where dance was
included as an important part of it. Eventually that part would gradually lose its
importance and nowadays almost nobody associates fado with dancing.
The earliest known references of fado as an artistic performance date from

the early nineteenth century. Rui Vieira Nery describes with great detail early
sources to point out how fado was seen and described as a dance in Brazil at
that time. He quotes the Italian geographer Balbi (Balbi, 1822), as well as some
European travelers like Captain Louis-Charles Desaules de Freycinet (Freycinet
et al., 1826), or the Germans Carl Schlichthorst (Schlichthorst, 1829) and Johann-
Friedrich von Weech (Weech, 1831) as they describe fado being a fascinating yet
voluptuous, immoral dance, similar to the ones performed by the black people
195
(Nery, 2004, pp. 18-22). He also uses the insights provided by local authors as
the Brazilian poet Falmeno (Cordeiro, 1827) or the novelist Manuel António de
Almeida (Almeida, 1893) to characterize the dance. These early sources provide
us with vivid observations and descriptions about the reality and daily life of
Brazil, at that time, and so we can infer valuable details about what was a current
practice.
Almeida originally published his book Memórias de um sargento de Milı́cias

[Memories of a militia sergeant] in 1852, as a part of a periodical series. Later it
was revised and published as a whole book in 1854. In Chapter VI we can read a
very detailed narrative:
“Everyone knows what fado is, that dance so varied, so voluptuous,

it seems to be the daughter of the most refined study of art. A simple
viola fits the purpose better than any other instrument. Fado has many
variants, each one more original than the other. One single person, man
or woman, dances for a while in the middle of the room, executing the
most difficult steps, adopting the most graceful positions, accompany-
ing everything with finger snaps, and then, slowly, approaches whoever
pleases him or her the most. Then the dancer performs some twists
and twirls and finally claps his or her hands, this meaning a new dancer
has been chosen. This procedure goes roundabout until everyone has
danced.
Other times a man and a woman dance together; precisely following
the tempo, either with slow or faster steps, either repelling or approach-
ing one another; sometimes the man approaches the woman with light
steps, while her, performing a slight movement with her body and
arms, slowly retreats; other times she is the one looking for the man,
who then steps back until, finally, they come together again. There
is also the circle in which many people dance; certain measures being
interrupted with clapping and loud and lasting, other times milder and
briefer, step-dancing, although always with a steady unitary time sig-
nature. Besides these, there are also other forms not discussed here.
The music is different for each one of them, but always played by a
viola. Often the player, in certain measures, sings a song, sometimes
carrying a truly poetic meaning. Once fado begins it is difficult to end,
196
6.1. Fado as dance
always lasting until dawn, sometimes going on for days and nights in
a row.2 ”
6.1.1 The hypothetical links with flamenco
Some scholars and historians of fado were too quick to dismiss the influences
of other musical practices in the shaping of fado. Clearly, in a world that is a
melting pot and where several cultures and practices cross and overlap each other,
it is impossible not to notice some characteristics of early practices making part
of rituals that would later develop on their own. Looking back at the descriptions
of fado as a dance, in Brazil, one clearly identifies many of the choreographic
characteristics of flamenco in that description. However, there are not many clues
regarding the instrumental accompaniment that the viola would have been doing
back then. Still, the clues are enough for us to believe that by then fado and
flamenco, despite having different names (or not even being recognized as such,
as autonomous musical categories), were practices co-existing in the same time
period and nearby geographical regions, and therefore could share a fair number
of common characteristics and traits and be influenced by each other. Moreover,
they probably could have common origins.
Following that reasoning, when one reads more about early flamenco, one re-
alizes that there are two distinct aspects concerning the practice: concerning the
2
“Todos sabem o que é fado, essa dança tão voluptuosa, tão variada, que parece filha do mais
apurado estudo da arte. Uma simples viola serve melhor do que instrumento algum para o efeito.
O fado tem diversas formas, cada qual mais original. Ora, uma só pessoa, homem ou mulher,
dança no meio da casa por algum tempo, fazendo passos os mais dificultosos, tomando as mais
airosas posições, acompanhando tudo isso com estalos que dá com os dedos, e vai depois pouco e
pouco aproximando-se de qualquer que lhe agrada; faz-lhe diante algumas negaças e viravoltas,
e finalmente bate palmas, o que quer dizer que a escolheu para substituir o seu lugar. Assim
corre a roda toda até que todos tenham dançado.
Outras vezes um homem e uma mulher dançam juntos; seguindo com a maior certeza o com-
passo da música, ora acompanham-se a passos lentos, ora apressados, depois repelem-se, depois
juntam-se; o homem às vezes busca a mulher com passos ligeiros, enquanto ela, fazendo um
pequeno movimento com o corpo e com os braços, recua vagarosamente, outras vezes é ela quem
procura o homem, que recua por seu turno, até que enfim acompanham-se de novo. Há também
a roda em que dançam muitas pessoas, interrompendo certos compassos com palmas e com um
sapateado às vezes estrondoso e prolongado, às vezes mais brando e mais breve, porém sempre
igual e a um só tempo. Além destas há ainda outras formas de que não falamos. A música
é diferente para cada uma, porém sempre tocada em viola. Muitas vezes o tocador canta em
certos compassos uma cantiga às vezes de pensamento verdadeiramente poético. Quando o fado
começa custa a acabar; termina sempre pela madrugada, quando não leva de enfiada dias e noites
seguidas e inteiras.”
197
vocal and choreographic tradition, one that is attributed to the Gypsies, in its
earlier forms, it
“consisted of the song or cante without any musical instruments for

accompaniment, with the exception of a long, dry stick that was held
upright and rhythmically tapped on the ground. This is known as á
palo seco, a term literally translating to ‘by means of a dry stick.’ This
terms also applies to flamenco that is accompanied solely by rhythmic
hand clapping and/or finger snapping. Some of the most primitive
flamenco songs relate tragic stories of lost love, struggle, desperation,
hardship, imprisonment, and death” (Ruiz, 2007, p. 75).
The other, concerning the music: flamenco music was influenced by Hindu and
Arabic music from the beginning, and
“the melodic song forms known as Arabesques and Melismas are

fine examples of Arabic influence upon flamenco. Reminiscent of litur-
gical chants, Melismas, or extended vocalizations of several notes to a
single syllable, are a strong component of flamenco song. Important
characteristics include the extension of vowels, expression of profound
feelings, and priority of emotions over lyrics.” Numerous ornamenta-
tions such as ay, layli, layla, lolailo, lereli, and lerele were fairly common
among traditional flamenco songs, which may have originated from the
Arabic Layali, a term for improvisation, or from the similar-sounding
Arabic expression la illaha illa’llah, translating as ‘There is no God
but Allah’ ” (Ruiz, 2007, p. 75).
These descriptions are extremely interesting because many of the characteristics

associated with fado can be recognized, and we can think of at least two different
moments of its chronology where they could have been implemented and rein-
forced. The melodic Arabic and Hindu characteristics were probably implemented
into the viola, and we can easily extrapolate that this kind of accompaniment made
sense both in the fado and flamenco of the early nineteenth century, as described
by the sources. Also, this mood and affect linked with the way the narratives are
expressed and sung (exacerbation of emotion over the lyrics) could probably be a
very good reason why one should call fado (destiny, fate) to fado and also justify
198
6.1. Fado as dance
its themes and furthermore the way it is articulated and how the stressed syllables
are constantly emphasized: they would come directly from this tradition of very
emotional, tragic songs. This thesis would be consistent with the description of
the character of the songs that would happen sometimes in the fado ritual, as
described in the sources from the early nineteenth century. Although there are
evidences that this way of singing (this almost caricatural gesture) was somewhat
lost during the nineteenth and early twentieth century, based on the descriptions
and early recordings, it is also known that this gesture still existed in performative
practices of the Iberian Peninsula, namely in Andalusia and in folk and traditional
songs. This same way of singing was then “recovered” and reinforced in fado as a
personal mark by Amália Rodrigues, who, borrowing it from the way her mother
sung local songs, unintentionally spread it as a gesture of notable recognition asso-
ciated with the practice. It seems fairly easy now to recognize how these gestures
and characteristics from the Eastern Europe and Middle-Eastern Asia are present
in fado nowadays and how we can trace them back to their origins in a logical
way. What we find notable, however, is how a kind of singing already associated
with performative practices in early flamenco, and probably in the early forms of
fado, is weakened and then reinforced later on, on a new context and geographical
space. Perhaps this just show us that Amália felt the gesture made sense and
it was “always there”, and that it was a more than adequate way to express the
moods and affects linked to those kinds of narratives. Even in a more simplis-
tic way, the mere embodiment of the emotions, affects and moods, contained in
these narratives and expressed through the means of the prosody of the European
Portuguese language naturally are more effectively conveyed through the means
of this specific kind of gesture.
What seems to be a crucial point is the categorization of fado as a dance,

something unique along its history. If it is true that fado was danced or “batido”
[hit] during several decades, it is also true that the core of the practice moved from
the choreographic value to the sung value – it has become “less of a dance” and
“more of a song”. The primordial value that defined its characterization changed,
and this is the core of Tinhorão’s thesis (Tinhorão, 1994).
It is not clear to us that the danced fado from Brazil is the same practice as the
sung fado in Lisbon. In a certain way, one could claim they are two distinct prac-
tices sharing the same name but with different origins and distinct influences that
overlapped in the same place. In Lisbon several sources point to several musical
199
and social practices. In the brothels and taverns, mainly in the poor neighbor-
hoods of Lisbon, physical fights or games where a person tries to knock out the
other person in a kind of a challenge or dance is often called “fado”. It is not
known how independent these practices might be from the lascivious Brazilian
dance. It is a certainty that challenges and social games already existed in Lisbon
before the court returned from Brazil in 1821. However, the Industrial Revolu-
tion, inner migrations, constant commerce and the interchange with sailors and
slaves had notorious impact in the social geography and for sure changed those
same practices. The main issue is to know when and how these practices started
being called “fado” and recognized as such. The thesis of the importation of the
name and some practices makes sense only for one thing: there is no performa-
tive practice called “fado” in the sources, namely in Portuguese literature, until
the nineteenth century, the Brazilian designation having precedence. However,
Pimentel, based on Ramalho Ortigão’s As Farpas, tries to explain an origin for
the label based on the character of the fadista. Being the fadista, the man of
the bad fortune, the one with a “bad fado”, and being the brothels by extension
“the fado houses”, then “fado” would be whatever one would be doing in there
(Pimentel, 1904). And this is a totally different explanation for the origin of the
name, completely independent from whatever was happening in Brazil.
It is noteworthy that what was practiced in the brothels and taverns was already
more than mere dance or physical challenges, because Pimentel already describes
fadistas singing mournful songs and telling their own sad stories, using a viola
or a guitarra to accompany themselves. This also shows how, in Lisbon, the
component “dance” is not seen as the core identity of the practice. However,
this does not necessarily exclude the possibility of the Brazilian practice occurring
simultaneously in Lisbon, and also being called fado. We find it possible and
even probable that one could hit the fado, sing mournfully the fado and dance
lasciviously the fado, all practices coexisting in the same time and place and
using the same denomination. The period lithography clearly shows all those
practices taking place in Lisbon during the nineteenth century3 . Also, one should
not exclude the possibility of some elements from a given community being aware
of the several practices and performing some of them, contextually, and opting
afterwards for the ones that made more sense. All these different descriptions and
hypotheses suggest the possibility of independent origins for the use of the label.
3
See for instance the numerous reproductions in (Nery, 2004)
200
6.1. Fado as dance
Both Sardinha and Gouveia contest the idea of a lascivious dance being the
origin of a mournful singing and challenge the ideas of Tinhorão, who was seconded
by Rui Vieira Nery. Sardinha criticizes the notion that fado was born in Brazil
and deconstructs the Brazilian dance using extensive choreographic analysis based
on the descriptions and written testimonials, concluding that even in Brazil the
dance called fado was not an isolated practice, but in fact a myriad of distinct
choreographies. Hence, their origin and evolution had to be prior to the written
sources. He also tries to establish a non-relation between those practices and
the Portuguese ones, namely invoking their character. He does not understand
how a bright, pulsating, lascivious dance might originate sad, plangent, mournful
melodies. Moreover, he accuses Nery of not explaining the transition, or how the
word “fado” applied to that dance. He accuses Ramos Tinhorão of the same flaw.
Sardinha acknowledges the contact point between the Brazilian dances and the
“fado bailado” [danced fado], but they are so general that they are also to be found
in many other national choreographic practices. Therefore, if one were to employ
the same reasoning, virtually all national dances would descend from Brazilian
dances, something that makes no sense. Finally, he claims that the Brazilian
sources of that period reveal little concrete information about the music that would
accompany those dances. While Rui Vieira Nery claims that the danced fado
from Brazil already had melodic and harmonic traces that we will then find in the
first musical transcriptions from 1852, and that will remain, in many cases, until
nowadays, Sardinha claims that is not possible to establish any relation with the
music from the fado-dance from Brazil, the lundum or even the music that would
support the Portuguese fado, and any opinion in that direction is mere speculation
or fantasy (Sardinha, 2010a, p. 450). This opinion echoes Renato de Almeida’s,
who claimed the absence of any physical document or musical transcription about
the music of the Brazilian fado that would support the connection with lundum
or the Portuguese fado (Almeida, 1926, p. 78).
The divergences between Sardinha and Nery are well documented in a news-
paper article in Jornal de Letras, in which Nery highly criticizes Sardinha’s work
by pointing out numerous methodological errors (Nery, 2010b), as well as a re-
buttal where Sardinha answers maintaining his arguments and stating that Nery
obviously missed the point and failed to interpret his work (Sardinha, 2010b).
Gouveia, in a certain way, supports Sardinha in his hypothesis. He starts

by quoting Luı́sa Guerra on how “Lundum was danced in Africa. Fado is not
201
sung. (. . . ) Lundums were sung in Brazil. Fado is not sung. (. . . ) the so-called
Brazilian danced fado lasted only about forty years. Lundum continued. Lundum
was danced and sung in Portugal. Fado is sung. It is sung because it exists a
proper context” (Guerra, 2003, pp. 96-97) (Gouveia, 2010, p. 18). The existence
of several simultaneous musical practices and influences are recognized, but not
a transference between them. Gouveia follows by analyzing Caldas de Barbosa’s
lyrics, a Brazilian native that moved to Portugal, and according to Nery and
Tinhorão one of the fado precursors. Gouveia concludes that nothing in Barbosa’s
poetry is similar to fado narratives and
“what one could listen in the salons was... salon music. There were
musical scores, evidence that schooled people wrote them for presum-
ably schooled people to enjoy. The several renditions of those genres,
performed nowadays by the meritorious efforts of Segréis de Lisboa,
Pedro Caldeira Cabral or Silvestre Fonseca, show that it was erudite
music, sung by tenors or lyrical sopranos, with trained voices, accom-
panied on the piano or the harpsichord, all following (or even depen-
dent on) a musical score. The lyrics reinforce perfectly the idea already
portrayed by the music, with elaborated metrics and poetic palatial or-
naments. “Cadê” – one might wonder, assuming a fado Afro-Brazilian
origin – the descriptive quatrain, the simplicity of the musical accom-
paniment, the melodic improvisation spontaneity?” (Gouveia, 2010, p.
21).
The situation reaches the point of comedy when in the documentary Trovas
Antigas, Saudade Louca (Ávila, 2010), broadcast by Portuguese National Televi-
sion (RTP) in September 24th, 2010, about the origins of fado, a scene where a
modinha accompanied on a harpsichord and sung with an educated voice, in a
palatial setting, is presented as a representation of a fado precursor. The disparity
between the described practices and the ones shown is so vast that one cannot
find any relation, neither lyrical, neither performative, causing perplexity.
At this point we might add that it seems that both Sardinha and Gouveia
failed to understand not only the differences, but also the relations between what
would be the practice of fado in its spontaneous context (and the contexts where
they learned what fado is) and the sanitized versions played by the upper classes
following the musical transcriptions. The existence of these two kinds of repertoire
202
6.1. Fado as dance
as reflections of the same reality is, in our opinion, a major cause for some of these
misunderstandings.
Sardinha explains that the satirical, humoristic, and non-mournful fados, derive
their musical basis from the passacalhe, an extremely popular genre in Europe and
especially in Andalusia. Obviously it would be a variant of the popular version,
before it reached the complexity of the erudite one. Several authors, namely César
das Neves, corroborate this idea: “the large number of fados, almost all variations
from each other, improvised daily, are a kind of slow passa-calle, having charac-
teristically Portuguese music” (Neves, 1893, p. 31). Sardinha also quotes Pinto
de Carvalho saying that fado was corrupted until “it became a slow passacalhe”
(Sardinha, 2010a, p. 250). Grounded on these affirmations and knowing that
“Corrido” was the term by which the Spanish passacalles (which were the basis of
their traditional ballads) were known (following an idea from Mário de Andrade),
Sardinha sustains the thesis that this was the genesis for the term “Fado Corrido”
and its vulgarization as a major subclass of canonical fados. Gouveia, on the other
hand, claims that one called “those fados ‘corridos’ because their function is to tell
a run-on sentence story [corrida] that unfolds itself through a series of quatrains
(...)” (Gouveia, 2010, p. 33), being unable to properly sustain this idea beyond
the obvious etymological play, which seems nothing more than speculation to us.
Sardinha then extrapolates that
“humoristic fados, played by the blind musicians, having the pas-

sacalhe lively and marked rhythm, and given the natural tendency for
people to dance to everything, started being danced by the audiences
and tavern costumers, alongside with other dance practices that were
already being performed there for recreation of those same costumers”
(Sardinha, 2010a, p. 267).
Based on Teófilo Braga and Ramón Menéndez Pidal, Sardinha tries to show that
the period fado practices, and others derived from the traditional ballads, needed
no external influence, and they were by themselves suitable for dancing. “Calling
fado to these other dance practices was just another step” and using, again, a
synecdoche reasoning, “this was how the label fado engulfed other practices differ-
ent from its original matrix” (Sardinha, 2010a, p. 268). Sardinha culminates his
thesis, supported on a reasoning by Carolina Michaelis de Vasconcelos, that fado
203
had an ancient, national, origin, based on the traditional ballads, sung across the
entire country.
Our own opinion is that none of these authors considered the hypothesis that all
these different theses (even if Sardinha’s one is highly speculative) have something
true in them. One cannot fully prove or dismiss any of them. Obviously the
practices derived from the traditional ballads had their evolution and continuity
alongside other contemporary practices coming from abroad. We cannot assume
that someone has ever called “fado” to the result of those practices as Sardinha
assumes, nor even assure that that fact would be decisive for the creation of an
independent musical category. However, we consider plausible that those practices
might have had influence on what constitutes the body of repertoire of the future
fado. A myriad of performative and contextual practices coexisted in the same
time and space and some of them were known as “fado” in a given period of its
history. It is likely that the same practices that were not fado started being called
by that name by the incorporation of certain elements or its inclusion in a new
context and posterior legitimation. On the other hand, some other practices could
just cease to be called fado. It seems clear, though, that the sources tell us about
several choreographic practices that were known as fado or integrated as such in
a symbolic category. However, it also seems clear that, although the dance has
been always associated with fado until very recently in its history, soon enough
that aspect was no longer the fundamental trait for its recognition as such.
6.2 Fado feeling
The body posture and the way the fadistas embody their performance is a
key feature to influence the overall sound of fado, not only because of the spatial
placement regarding the audience and the musicians and how they will interact
with each other, but also because of the recurrent tendency to place the neck in a
way that will invariably affect the breathing and the timbre of the voice. Hence,
the embodied performance will have a major influence, even in the phonograms,
when communicating emotions, moods and affects towards an audience eager to
create empathy with the singer.
In the particular case of fado, the embodiment of the narrative, mapping each
line into a pair of binary measures, is done in conjunction with a gestural device:
204
6.2. Fado feeling
the tempo of the second beat of the odd measures is slightly slowed down, while its
dynamics are increased, raising tension and creating a waddling effect that makes
one want to lift their feet. While this effect has been described in literature as a
syncopation, the data shows us that many different rhythmic figurations actually
occur, and instead, it is the gestural and articulatory characteristics of the beat
that are consistently reiterated, as shown in figure 6.1.
Tempo
œ œ œ œ- œ œ1 œ ‰2 œj
A rit. a tempo rit. a tempo
j œ œ œ œ- œ œ1 œ ‰2 œ
& 42 œ
1 2 1 2
Voice
J
œ äœ œ2 äœ
A rit. a tempo rit. a tempo
&œ œ œ œ œ ‰ œ œ œ œ œ1 œ ‰ œ
5
J
1 2 1 1
J
2 2
œ >œ
U
B rit. a tempo
œ j
rit. a tempo
œ ‰ j ‰ Œ
9
& œ œ œ œ œ œ
1 1
œ œ
1 2 2 1
œ œ
2
Figure 6.1: Fado feeling.
This gestural shape is one of the trademarks of the genre and can be described
as fado feeling. When a given line of text is repeated for emphatic purposes, often
the already emphasized open vowels tend to be over-emphasized, slightly changing
the pitch, further slowing down the tempo, and increasing even more the dynam-
ics. Sometimes, even enlarging the rhythmic value, shortening the following ones,
which would correspond roughly to something between a tenuto and a sforzando.
In fado practice rubato is often used for dramatic effects and for emphasizing
certain words from the text. Therefore, the instruments try to keep a steady tempo
(namely the viola). It is the singer who decides when to diverge, and at that point
the other instruments have to react and adapt themselves to the singer. Often in
very dramatic and long suspensions the instruments completely cease to play and
wait for the singer to finish the cadenza and let the silence resonate. “Particularly
205
in the last strophe during the finale on a key utterance, when the instruments
might pause, and in this space of silence, the voice might hover on the break of a
cry; it might become almost impossibly soft (or, for a man, impossibly high); the
voice might tremble and turn upon itself in ornamentation” (Gray, 2013, p. 39).
This silence is powerful and has a tremendous emotional effect:
“There are shifts in pitch and rhythm; suspensions of time in a

vowel drawn out or a silence prolonged; micro-timbral nuances in the
voice that inflect the feeling of a word or a phrase. There are moments
that hover on a continuum between speech and song – now a word
declaimed, then a phrase sung. There are the variations of force behind
the air as it moves out of the lungs and the throat into the air: the
wail, the whisper, the falsetto, the sometimes widely oscillating vibrato.
Then there is stillness, the strategic use of silence” (Gray, 2013, p. 139).
After that, the viola resumes the play with a steady beat, and the performance
follows. Even in the older musical scores from the nineteenth century it is possible
to infer these moments, often near the end and in a high, long note or above a key
word, marked with a fermata.
206
Chapter 7
Fado as Memory
By the end of the twentieth century, fado exists in the mind of listeners and
performers as a concept referring to certain contexts, conventions, musical gestures
and practices. It is impossible to define exactly what fado means to everybody
since, as has been shown, fado can be equated to a series of many different things,
and our definition of fado might be slightly different from others or even have a
narrower or wider scope. The meaning of fado each one brings inside themselves
can vary according to a series of factors like education, living context, informa-
tion access, local practice, etc. Even reducing fado to a particular set of musical
gestures that seem to be common to most of its occurrences nowadays might not
be enough. Another step beyond was already taken when dealing with covers
and homages. A concrete example: the Portuguese musical singer Sónia Tavares,
from the band The Gift, made a new arrangement of Amália Rodrigues’ “Gaivota”
[Seagull]. “Gaivota” in itself was already a bold innovation in the 60s, and was not
considered a fado by many of its contemporaries. However, in the 90s, everyone
would consider “Gaivota” a fado, even a canonical and respected one. It was a very
good example of a song that was assimilated into the canonical repertoire. The
new arrangement presented by Sónia Tavares was another bold innovation derived
from Amália’s one – the new song was sung completely flat with a synthesized
accompaniment. None of the musical gestures or references typically associated
with fado are present. Any common listener, outside the context, hearing this
performance would say it was a normal contemporary Portuguese pop song.
The “Gaivota” from the 60s was already very different from a typical fado –
a different structure, a complex harmonic progression, a poem with an irregular
207
form. It just did not fit the known and accepted model of its time. However,
because it was performed by Amália, sung with her peculiar vocal style, with the
guitar and viola accompaniment, and with the dressing code and bodily posture
used at the time, there were enough cues leading one to think that it was fado, and
as such it was assimilated as one. “Gaivota” was, eventually, considered a fado,
just like many other new songs throughout the 70s, 80s, 90s and even nowadays
are, by offering enough cues, namely the musical gestures, present and related to
traditional models of fado.
Sónia Tavares’ “Gaivota” cover, however, completely misses those same cues. It
lacks vocal and guitar gestures, it lacks the dressing codes and the bodily posture.
It lacks the performative context and space. Nothing in that performance has any
resemblance or connection to a traditional fado from the beginning of the twentieth
century. There are only second generation cues, one ought to say. The lyrics, and
more or less the same melody (even without the ornaments) and harmony are
preserved between the original “Gaivota” and the new one. People say the new
“Gaivota” is a fado, because it is a recreation of the original “Gaivota”, which is
already considered a fado nowadays, even if it was not back then. So, there is a
kind of chain of events, and it is this chain that is preserving the sensation of fado
along the way, even if there are not common traits shared between the first and
the last element of this chain.
Fado is then recognized through memory and association. This is a fado because
of memory and reference: because by listening to this performance one immedi-
ately equates it with a former performance that was already considered as being
a fado, even if that same performance was not considered as one in the time of its
original creation. In the twenty-first century fado can then be identified as vaguely
and imprecisely as the memory and reference of some gestures and cues present in
previous occurrences of another fado.
7.1 Desfado
The most recent album by Ana Moura, Desfado (Moura, 2012), is an interesting
case study that allows us to test several of our assumptions. First, the work is
called Desfado (literally un-fado) and the deconstruction of the fado starts by
paradoxically leading us into thinking that one is in the presence of the real thing
208
7.1. Desfado
by the use of strong cues. It then challenges them by incorporating a myriad of

other elements, usually not present in traditional fado (piano, several percussions,
strange metrics and poems, etc.), leading one to think in the opposite direction,
that one is in fact in the presence of a more folk or pop album, an album of songs.
The album was recorded and produced in Los Angeles and a web-series based on
the recording can be found in Ana Moura’s YouTube channel1 . Along with Ana
Moura, two musicians, the guitarra and the viola players, traveled with her, which
shows the importance of the local instrumentation in the definition of crucial sonic
cues. Those instruments are present throughout the entire record, the viola always
as a base support – mostly doing a rhythmic pattern of bass and chords, and the
guitarra as a counter melodic instrument when the voice is present or as a melodic
and soloist one on some instrumental sections. This use has been the most common
during the entire twentieth century in all records and live performances and so it
acts as a strong reference of an already existing tradition. Tozé Brito, a Portuguese
composer, on episode seven of the web-series, states and reinforces the idea that
he composes regular songs but then the way Ana Moura sings them (estilando),
and the arrangements “turns them into fados”. This statement reinforces the idea
that contemporary fado can be seen and recognized through some specific cues
and gestures and that its identity is linked to those. In this case, Tozé Brito seems
to identify them as being the articulation of the voice and the instrumentation.
The cherry on top of the cake is a song composed by Joni Mitchell, entitled “A
case of you”. Hearing this song we certainly hear the guitarra in its habitual role,
however the viola makes a slight smoother and fluid pattern, usually more close to
pop ballads and not fados. The strongest cue, Ana Moura’s vocal style, is totally
missing: she sings in English, and that totally deprives us of a crucial element in
the understanding of fado’s identity. Because, and on top of that, the prosody and
gesture are not present because she sings with a very discrete articulation without
legatos, portamentos and almost absent mordentes. The outcome is that it does
not sound like fado at all, it just sounds like a regular pop song. One of the most
important characteristics of fado is its relationship with language. Although some
of its themes are universal, in the sense that they are part of the daily life and
emotions of most human beings, it is the fact that they are sung in European
Portuguese language that gives them a special aura.
1
http://www.youtube.com/user/anamourafado, accessed June, 6, 2015
209
Chapter 8
The Creative Processes
It can be argued that there are as many different creative processes as there
are composers in the world. The craft of a song can seem elusive and, some-
times, arbitrary and not possible to be modeled or reduced to a correct “way of
doing it”. While we have been defending that fado emerges in a kind of linear
fashion where the melody is constrained by the text, at this point some might
argue otherwise. Often, and this is true of other song making traditions as well,
composers just whistle a textless monophonic melody, that descends on them via
“divine inspiration”. They talk about how they dreamt of it, or how some God
might have dictated it to them. Other times, they just fiddle around with a guitar
and start mumbling random melodies, derived from the chords they are playing.
Other times, they just improvise on a piano and the fingers lead their way up into
crafting beautiful melodic lines, in a largely arbitrary process, not at all related
to a text, or involving any sort of prosodic concern.
All these operative models, and a thousand more, exist, more or less related.
Yet, they all translate into one of two things: either they are generative or trans-
formative processes. They are generative when new material is being born for
the first time, out of nowhere. They are transformative when previous material is
being changed, recombined or further developed.
When composers depart from a text and derive the melody from it, for us, that
poses no issues. It is the “ideal” situation we have been defending all along, and
it is clear how one is facing a generative process in which the text constraints the
melodic creation, in the sense of the rhythms, pitches, articulations, and many
other relevant parameters involved.
211
However, sometimes, composers come up with textless melodies first. In this

case, this is the generative process: the creation of new melodic material, and
in these terms, this seems to violate the “correct ideal process” we have been
defending. We still think it is not the case. Because even when composers generate
textless melodies first, they have to deal with the text at a later moment. Either
being written by them, or provided by someone else, they have to come up with
a text whose characteristics exactly match the characteristics of their generated
melodic material. This often implies that they have to transform their initial
melody to conform to the new text. Sometimes this process is a back and forth
one, in which the text is also slightly changed and transformed to conform to some
melodic detail that composers do not wish to change. Either way, one is in the
presence of transformative processes, through which the text will indeed ending
up prevailing over and shaping the melody, and leading up to a final result that
is similar to the initial model proposed. In the cases when this does not happen,
that means, the text ends up not shaping the melody, according to its prosodic
and semantic conventions, then the final result is incongruent.
The mismatch of text and melodic material is often detected both by the audi-
ences and by the composer peers, and it is invariably seen in a very negative light
and deemed as a bad practice. This leads to a subjective positioning of what good
and bad practices are, and that is something that varies according to the different
song making traditions and their communities. In the case of fado, it is something
often pointed out, and already presented in this dissertation. Due to this fact,
sometimes subsequent transformative processes occur. Fadistas, for instance, may
receive mismatched texts and melodies and then they “correct” them in perfor-
mance, transforming the melodies so that they conform to the text in ways they
see fit. The reverse might also happen, unexperienced fadistas might ruin what
were congruent texts and melodies. In either case, these processes generally do
not overlap with the initial creative process, unless both the text and melody are
being created on the spot, which, at present, is very rare.
Our goal is to avoid any model (be it explanatory or predictive) that is modeling
what is considered a bad practice or a final result that is not adequate to the
communities involved in the practice. Hence, and having acknowledged the fact
that there are indeed a myriad of ways of composing and achieving similar or even
undesired results, we claim that, in the end, in a “good” song making process,
the final text used is indeed a crucial, foundational element in constraining the
212
melodies. Be it directly as a generative device (the melody deriving from it),
or as a transformative one (a precomposed melody being changed to conform to
it), or as a back and forth transformative one (both the melody and the initial
text are shaping one another until they conform to each other). Therefore, while
a more faithful model would encompass these three variations, we consider that
their end result sounds very similar. Both the second and third processes ending
up providing the illusion that the melody conforms to the text just as if it were
derived from it all along. Hence, one can model them like the first one in order to
avoid redundancy.
213
Chapter 9
Conclusions
Fado is a sorrowful tale, a narrative emerging from the life experience of the
people, using the vocabulary and lexical field appropriate to be shared by the
people. The text shapes the form of the music. The strophic nature of the narrative
translates into strophic musical forms repeated over and over, while more complex
texts originate more complex forms. The available musical instruments determine
the musical arrangement. Since most people have limited musical knowledge, the
most common practice is to play very simple patterns (often just arpeggios, other
times simple stock figurations and repetitive ostinati), maximizing the number of
open strings and the easiest fingering positions, using the simplest time signature.
Therefore, a binary pulse and a chord progression alternating tonic and dominant
is vastly preferred over any other more complex variation. If skilled performers
are available, then the degree of complexity and inventiveness of a fado tends to
increase accordingly. These conditions constrain the way the melody is shaped.
The rhythm of the melody is dependent both on the prosody of the text and the
time signature chosen, while the pitches of the melody are constrained by the
underlying chord progression being done by the instruments.
The tempo, dynamics and articulations of the same melody are constrained by
the emotional (semantic) content of the lyrics and by the physical characteristics
of the fadistas. Their body shape, the way they position themselves, their health
condition, will determine most of the timbral characteristics of the melody and how
they will be perceived by the audience, in a complex network of relationships and
interdependencies. The rhythms and pitches selected will also affect the dynamics
and the timbre, since they are all correlated (ascending lines will tend to increase
215
in dynamics; high-pitched longer notes will also tend to have increased dynamics
and vibrato while being slowed down; shorter and faster rhythms will tend to have
lower dynamics, while being accelerated etc.).
There are a myriad of conventions built along the nineteenth and twentieth
century that were supposed to reflect “Portugality” and convey a number of emo-
tions, affects or moods that would make us experience empathy towards them.
Portuguese natives would see themselves as individuals belonging to a community,
a culture, mirrored on those traits. Therefore, as Moore points out, “A musically
inclined Portuguese (and most Portuguese are musically inclined) can instantly
tell whether a song is a ‘fado’ or not; though he cannot successfully explain it to
any one who is not a born Portuguese.” (Moore, 1902, p. 165). Fado would, then,
be the empathic recognition or embodied experience of certain emotions, affects
or moods, learned by association to certain conventions as symbols of Portugal
(saudade, nostalgia, melancholy), in a performance practice.
However, even with such a cautious but strong definition, a very important
conclusion that we have reached is that it seems impossible to devise a definitive
definition that includes everything that we call fado, but excludes everything that
we do not call fado. This is a problem similar to the one portrayed by Wittgenstein
in his Philosophical Investigations, where he studied the problem posed by the
definition of “game” (Wittgenstein, 2010). His solution, based on the strategy of
“family resemblances”, is somewhat what we propose for the issue of fado. We
believe that fado is a name we give to a myriad of different things that share some
characteristics in common, but not all of them have to be present for us to identify
something as a fado. On the other hand, there are also a number of extraneous
traits that might be incorporated into something recognized as fado that are not
relevant characteristics to the phenomena at all. We have detailed a series of cues
and signs that may help us define it but, as stated previously, there are no clear
boundaries as to what extent those cues have to be present, neither how many,
neither in which contexts.
It seems that the stronger characteristics are the ones related to unique timbral
traits, not shared by other practices. Thus the conjunction of two gestures: the
vocal styling within a theatrical embodiment and the typical “Armandinho” gui-
tarra formulaic countermelodies, allow that virtually any musical work containing
them will be identified by most of the contemporary listeners as a fado regardless
of other factors. The use of these gestures may not be exactly in their original
216
form – one needs neither the voice of Amália, nor even any voice at all to recall her
vocal shape. One successful attempt at demonstrating this assumption are Rão
Kyao’s instrumental fado albums1 in which he recorded the melodic lines with
wind instruments (mostly saxophones and flutes) (Côrte-Real, 1991, p. 74). By
articulating the melody as the originally Amalian vocal gesture described above,
he was able to convey the feeling of fado to his listeners. Neither a voice nor
lyrics were present, however, by employing a gesture that at the time was already
canonical and present in everyone’s minds, the genre was conveyed by memory and
association with past events and contexts. A mere allusion sometimes is enough
to convey the style.
Empirically, the relatively low note density seems to be a useful feature to

define fado in relative terms. The usually sparse instrumentation and the lack of
percussion tells it apart from other musical styles.
The strength of strongest rhythmic pulse also seems very distinctive, since fado
stands out regarding this feature. Fado also presents relatively high average note
durations, with low variability (confirmed by low maximum and high minimum
note durations), and negligible very short notes, indicating mostly rhythms cen-
tered around the eight note, with absence of extreme note values. The very low
average variability of time between attacks for each voice also points out how
written fados share clusters of similar rhythms well aligned with each other.
Range is a useful feature to define fado, since fado has a relatively short range,
located mainly around the A below central C. In terms of pitches, the magnitude
of the second most common pitch in fados, divided by the most common one, is
relatively very low, which indicates a strong tonicization of the style. Its distance
is usually the perfect fifth or the perfect fourth reinforcing the idea of a tonal
tradition. The relatively low pitch class variety implies the simplicity of the style,
being A the most common pitch class overall.
Fado melodies have a high number of half and whole tones in relative terms,
contrasting with the accompaniment figurations, presenting a low level of them. In
comparative terms, fado melodies also lack perfect fifths, tritones and octaves when
compared with other styles. In terms of direction of motion, fado accompaniments
melodically rises while melodies go down. Fado melodies also present long (in
1
Fado Bailado (1983); Viva o fado (1996); Fado virado a Nascente (2001).
217
terms of duration) and short (in terms of range) arcs in relative terms, indicating
undulating melodies.
Another strong cue is a visual and social one: the presence of a fadista. The
simple presence of someone who is known or present themselves as a fadista leads
us to recognize or categorize as fado whatever songs they perform. This is a
vicious cycle: usually a fadista is someone who earns their identity because they
are professional fado singers, because they sing fados. On the other hand, after
being recognized for doing that, and one assuming their identity as being that,
then whatever they do is entitled to be part of the “fadista repertoire”, and is
therefore identified as a “fado”. This is precisely what happened with Amália
Rodrigues and many others following her, namely Carlos do Carmo.
Some musical archetypes, namely the ones based on the primary Ernesto Vieira
definition, seem also to be a strong cue, since statistically, most of the fados con-
form to that model. Although there are a large number or variations and excep-
tions, especially in more recent times (and much due to the inclusion of extraneous
repertoire by the appropriation of some popular songs by famous fadistas), the
truth is that when someone tries to compose a fado from scratch, usually they end
up following something somewhat along those lines. So it would be very unlikely
for someone to not think of a fado in binary form, without at least four melodic
arcs, each one being two bars long, in the antecedent-consequent format, usually
with a pickup note and a syncopated rhythm, and a harmonic ostinato based
around tonic and dominant. Many variations can be made, but the skeletal model
is this one.
A final and very important cue is the lyrical content and the ideological asso-
ciation with Portugal. This is a very tricky and subjective cue, leading to many
mistakes and curious phenomena. Whenever one sees a performance in which the
lyrics talk about feelings and facts associated with Portugal, and cumulatively, one
can pick up some other cues, namely the use of a guitarra or viola, for instance,
and also cues from the popular and ideological imaginary built during the dicta-
torship (the famous “Tom’s dolls”), some people tend to make an association and
recognize this as part of the fado repertoire. Recent examples of this are famous
groups like Madredeus or Deolinda, or the popular singer Dulce Pontes. Although
these Portuguese artists perform some very original Portuguese music, we would
not consider them fado. They actually have some fados in their repertoire, but in
essence they do not present themselves as fadistas, in the whole, neither do they
218
display most of the strong characteristics already mentioned. Still, somehow, some
of their songs are confused and categorized as “fado” by their listeners. This is
most visible looking at commentaries to videos of their music on YouTube (both
by Portuguese and foreigners). We can only conceive and explain this based on
those ideas of “Portugality” and lyrical content associated with their music.
The lack of a proper vocabulary and of a way to systematize this set of values
has been a recurrent problem until the present time. Also, our standard musical
notation is poor and lacks the resources to take full account of these traits. Ellen
Gray presents transcriptions with a remarkable degree of detail compared to the
Portuguese sources, which lack many of these parameters. Based on her notation
and our own findings we can provide one example (figure 9.1) of how one can
improve conventional notation to better convey a “fado feeling” when scoring a
MIDI file, for instance, for computer rendition.
A n da n te
# j j
& # 42 œ œ. œ œ œ œ œ œ ‰ œ œ. œœ œœ œ œ
œ œ œ œ œ œ œ
gra c io so
? # # 42 ‰ œ œ œ œ œœ œ œ œ œ œœ œ œ œ œ œœ œœœ
j ä ,
a te mpo
j ä
rit.
# m-
a te mpo rit.
& # œ ~~~ œ . œ œ- œ œ œ œ ‰ œ ~~~ œ ≈ œ œ œ œ ~~ œ œ

m m
œœ œœ
œ œ œ œ œ œ̆ œœœ œ œ œ œ œ̆ œœœ œ œ œ
œ
3
? ## ‰ œ œ œ̆
Figure 9.1: Alternative notation to better convey “fado feeling”.
Many more characteristics could be mentioned, essentially weaker and less rel-
evant than the ones mentioned above. Throughout the different sub-sections of
this part we have explained most of them in detail, and we think we were able
to portray an essential and global idea of what fado is, and also, what fado was
219
and can be in different times and contexts. We believe we have achieved a holis-
tic description of the phenomenon and presented solutions for its understanding,
modeling, and also possibilities for further research. Fado has been considered to
be many different things in recent history. Although one can clearly see the con-
cept evolve in time and widen, like a tree spreading into many branches, the truth
is that fado’s evolution cannot be considered as closed. All fado manifestations
exist and persist at the same time, varying their relevance or preponderance in a
completely subjective way. Fado is a dynamic and ever changing concept; hence
one might always find some new manifestation that will not be contained in the
definition one might provide. However, one can more or less agree that we recog-
nize fado when we see it, and most of the time we can present a justification for
why that is so. Therefore, we are certainly able to model some objects, based on
chosen rules and statistical data from the past to create new fados. However, we
are not creating nor exhausting the entire array of possibilities. Perhaps we might
not even be creating a representative set of what the concept means at present.
But we can for sure say that “if this object complies with the characteristics x, y,
z. . . then it is for sure a fado, in this context”.
220
Chapter 10
Future Work
Empirical Musicology is a field in current expansion and development. While

the study of fado might be considered more of a historical, anthropological or
social problem, we believe an empirical musicological approach was missing. We
also believe most approaches have merit and a syncretism of all of them might give
us the best answers to understand a phenomenon that we, now, see as a complex
and holistic one. Given the amount of data gathered and all the possibilities that
remain to gather some more, using the same or different methods, it is expected
that we have built the grounds for a fruitful study. The database and the several
corpus versions are available and there is plenty of room for more experimentation
using other parameters and statistical operations that we have not considered
given our own subjective view and limited amount of time. There are plenty more
occurrences of fado in literature waiting to be unearthed and converted and added
to the database, as well as increasing its time span, if deemed as necessary. The
comparative database should also be enlarged. Ideally, a huge set of recordings
– including not only symbolic sources but also audio recordings – sampling the
entire world musics, with much more detail and uniform encoding conditions could
prove to be worthwhile. Therefore, we consider that much future work can still be
done and built upon this data to refine and improve its effectiveness.
221
Part II
Automating the Composition

Process
223
Chapter 11
Introduction
The main objective of this dissertation is to model musics and vocal sounds
typically associated with fado practice, in a way that a computer can generate
new sound-alike instrumental music and reproduce it. There are several ways to
approach this problem. However, soon enough we have realized that the most
flexible and generic models are the strongest ones, as they can be easily manip-
ulated and adapted to represent other musical practices. In a certain way, and
following the steps of our inspiration, David Cope, a reference in the imitation of
musical styles using artificial intelligences, we see fado as just a case study and a
proof of concept of a larger problem: the modeling of the cognitive process behind
the way humans, in general, compose songs. This second problem seems a more
interesting and vast one, however, impossible to solve in the scope of a mere PhD
dissertation, so we struggled back and forth between these two axes in order to
focus on our case study without losing sight of the bigger picture.
The song composition process seems to be a series of constrained problems.

There is a task to be solved step by step, involving a series of decisions. Some
of those decisions are not consciously taken by the composer; rather, they are
the natural result of the limitations of some initial conditions. In the chain of
events, some decisions and limitations influence others. For instance, the musical
instruments available will determine the arrangement, the overall timbre, range,
even the rhythms and dynamics of the final outcome. The skill of the performers
and their health conditions will also affect them. The structure of the text will
affect the musical structure, and so on. Therefore, our model is built in a sequential
225
Part II. Automating the Composition Process
way, taking into account what we think best reflects the actual practice, based on
the observations presented in the first part of this dissertation.
It would be tempting to say that modeling the process of composition of a

song is a daunting task and a virtually impossible one, given the large number
of possibilities and different approaches the actual practitioners use in their daily
lives and performative contexts. Again, this being true, we also believe the best
models are the most elegant and general ones. Therefore, instead of focusing on
the differences and the myriads of individual approaches (plurality) we decided
to concentrate on what they could all have in common, their tendencies, their
recurrent patterns (universality). A model like this, aiming at universality, is
obviously a flawed reduction of reality. It is also far from giving any scientific
or objective explanation about the whys and hows of that same reality. Instead,
it represents an archetype, an ideal, or even a stereotype. Perhaps, a projected
vision of what an “average” of all possible ways of doing it might be, if it makes
any sense to conceive that. One might say that, if David Cope in his project tried
to teach the computer to compose like he, himself, would do (Cope, 2005), we did
virtually the same. We tried to learn how some people compose music and vocal
sounds associated with fado, how those musics and sounds were made throughout
history, how they evolved and changed, how they actually sound. Then we tried
to do it ourselves. We spent a long period learning and perfecting the craft of
composing and producing music using computers and virtual instruments. We
had to physically understand the process and be a part of it so we could be aware
of the challenges and obstacles one usually faces. We had to understand not only
more analogical tasks, such as how to create melodies from lyrics having just a
viola or a guitarra in our hands, but also the process of creating a score using a
music notation program. Moreover, we felt the need to understand how a musical
score might end up in a DAW (Digital Audio Workstation) to be recorded, mixed
and mastered, up to the point of obtaining a convincing sound result. While this is
usually a time consuming task, that in the real world might evolve several people
doing it, in several different ways, we felt we needed to understand and learn most
of them. Derived from all that experience, we have now a conception of how to
do that process from scratch. If someone asks us to compose music for a fado, we
have a way of doing it. And now we are teaching the computer the most effective
way to replicate the way we do it.
To think about an academic or scientific goal is to be less concerned with results
226
in the short-run. The pursuit of knowledge, in itself, is the main goal. However,
looking from a commercial angle, there also are a myriad of possibilities that can
be taken into account: the production of software emulating or imitating a style,
generating infinite music, is an invaluable resource from an economical point of
view. Once it is done, it consumes very few resources and can generate massive
amounts of content. Neither does one need to pay musicians or composers, nor
does one need to pay royalties. The outcome could be sold to society and the
industry. At present there are numerous software applications based in algorithmic
composition for people to try on and generate music at their will, like PG Music
Band-in-a-Box 1 or Dunn’s Artwonk 2 , as well as leisure and relaxing environments
that generate music real-time with interactive components, not only in computers,
but also in smartphones, for instance, Brian Eno’s Bloom or Trope 3 or the more
academic ANTracks (Schulz et al., 2009). In a way, by trying to imitate a style
and creating an endless repertoire of songs, several goals are being met: not only a
scientific and philosophic problem is being addressed, but also an attempt is being
made at creating something to fulfill a cultural and social need: the endless desire
for music – “Music is not a language that describes the way society seems to be,
but a metaphorical expression of feelings associated with the way society really
is” (Blacking, 1973, p. 104).
1
http://www.bandinabox.com/, accessed June, 6, 2015
2
http://algoart.com/, accessed June, 6, 2015
3
http://www.generativemusic.com/index.html, accessed June, 6, 2015
227
Chapter 12
State of the art
In the creative process of composing music one is dealing with the most intricate
processes that happen inside our brain. Composing, in a romantic view, could be
something that deals with emotions, moods, affects, senses, soul – with one’s
true being, with the self. One of the crucial aspects of composing resides in the
moment when the composer has several options on how to combine their basic
musical cells. Whether they are sounds, clusters, musical pitches, structures or
other sonic objects, human beings have two main options in hand: either they leave
the decision to chance, thus throwing dice, picking up cards, following tables, I-
Ching or several other methods human beings usually use to generate random
results, or they just deliberately pick a road at their will. When thinking about
this second option, the most common by far, one really does not know what is going
on, and one would step into an endless discussion about subjectivity, determinism
versus free-will, auto-determinacy or consciousness. In fact, the debate is wide
open and one really does not know yet the rules that govern our choices. In
many ethnographic interviews composers many times answer things like “I just
felt like it”, or “It sounded right” or “I have tried several combinations and it
was the one that sounded better”, and so on. Often, even highly specialized
musicians are unable to justify their choices, methods or decisions in a rational
way. However, it is often believed that it is a human phenomenon different from
pure chaos or chance. One tends to think that when a human makes a choice,
something personal, something belonging only to that individual is present, and
that if somebody else was put in the same situation probably the outcome would
be different. In a certain way, choices made when composing define the composers,
define their style. Bourgeois (Bourgeois et al., 2001) stated that “Wasn’t it Buffon
229
who said, ‘the style makes the man’ ? It means that the way you do something
is more important than what you do. The way you do something signals you
as an identity. You are the only one who does it that way.”, corroborating this
idea. Also, Wittgenstein (Wittgenstein et al., 1984) follows the same path when
he states about the same sentence that the style of a man is his own image1 . The
style, the way composers make their choices, is present in every one of their art
works, it is a kind of a signature, a personal mark, something that leads us to
assign a given work to a specific individual and not another.
When thinking about computers making choices, none of this happens. Com-
puters deliberate only on what they are told to do, only in predetermined instruc-
tions given by the programmer/composer. And every time a computer is called
to make a choice, by itself, out of two or more different possibilities, either a pure
algorithmic procedure or randomness (which is also based in an algorithm) are
involved. In fact, reading specifically the story of random procedures in informat-
ics (Hayes, 2001), it can be seen that in the beginning most of those procedures
were only pseudo-random, as they were based in chaotic mathematical functions
such as, for instance, f (x) = x.(1 − x).k. This and similar derivative functions
are extremely effective to solve everyday problems, since they generate apparently
random numbers. However, this is an illusion because, in fact, these functions, to
a certain extent, are predictable. Therefore, in the following years, scientists tried
to improve random generators to be more and more chaotic and less predictable,
to approach universal systems. The solutions were found using external param-
eters like the computer’s clock, the cooler speed (which is dependent on weather
conditions and the amount of work the processor is submitted to), the movements
of the mouse pointer, the frequency of the keyboard strokes, and many other al-
ternatives using physical processes on a microscopic level. When dealing with this
kind of variables one understands that the computer choices are governed by what
could be called cosmic chance, and there is no way someone could find any reliable
pattern to confine it to any recognizable style or identity. It would be extremely
hard to find any coherence and we really doubt that one day someone could claim
“This set of choices is really the style of Computer 04 in Room 302”. Unless, of
course, one limits the way computer makes choices according to some constraints
1
“Le style c’est l’homme”, “Le style c’est l’homme meme”. Der erste Ausdruck hat eine bilige
epigrammatische Kürze. Der zweite, richtige, eröffnet eine ganz andere Perspektive. Er sagt,
daß der Stil das Bild des Menschen sei.
230
and desires of the programmers themselves. But in that case, one is really as-
signing a kind of identity given by them and not one the computer would create
by itself. We have already written about this subject in more detail elsewhere
(Videira, 2007). This being said, a point has been reached in which we feel this
romantic vision of the composer being gifted and having an opaque soul might
be transcended. We have come to believe that computers might be programmed
to automate reasonably well problems that lie in the music composition process.
Algorithms can be employed, just like when a cook writes a recipe, to emulate
those processes. Several techniques were and have been developed through the
course of history and are surveyed ahead.
Systems of algorithmic composition can be found a long time before the com-
puter era. In 1026 the monk Guido D’Arezzo applied a method to derive his
compositions right from the lyrical content, assigning specific musical pitches to
vowels (Roads, 1996, p. 822), (Järveläinen, 2000, Silva, 2003). Later, during
the eighteenth century, several musical dice games were made famous: a preset of
composed measures was being recombined between each other according to chance
dictated by dice throwing (Roads, 1996, p. 823), (Alpern, 1995, Silva, 2003). How-
ever, it is with the advent of computers that these methods became increasingly
popular and efficient. A computer is able to perform calculations much faster
and more accurately than any human would do, so adopting them as an aid tool
was something quite desirable for composers seeking effectiveness. The necessity
and motivation to explore these new resources was born. There are numerous
examples of pioneers in algorithmic composition since the 50s up to today. The
first, and most relevant, was Lejaren Hiller, in 1955, when computers were still
rare and extremely big. He was one of the first to apply computers to algorithmic
composition, musical impression and physical modeling synthesis. He was also one
of the first cybernetic music historiographers contributing with his Experimental
Music (1959) and Music Composed with Computers in 1970 (Roads, 1996, p. 830).
Several composers followed his example namely Brün, Myhill, Tenney, Barbaud,
Philipot, Koenig and even Xenakis. The latter had been composing for a while us-
ing formal processes (his previous piece Metastasis, from 1955, was written based
in stochastic calculations made by hand), and quickly understood that a com-
puter would make his work a lot easier. The first processes being implemented
were deterministic procedures, these being a set of predetermined instructions that
a computer would follow and, by the end, one would obtain a certain result, sim-
ilar to following a kitchen recipe. The procedure by itself is straightforward and
231
classical composers frequently used them. Computer programs following contra-

puntal rules are a good example of this. The advantage is efficiency and speed
– the machine produces in seconds what for a conservatory student might take
hours to compose.
12.1 A survey on algorithmic composition
Since this project began many things changed. The references accumulated
and discussion around the core problems of the field grew exponentially. Thus,
this particular chapter has been revised substantially since its first version in 2010,
because it quickly became outdated, with several papers and surveys being pub-
lished recently during the development of our own research. What were ill-defined
and vague problems suddenly became well-formalized or solved issues, while many
theoretical approaches were implemented by vast teams of researchers and new
applications were tested and put in the market. So, in a way, we felt vastly over-
whelmed with a constant flux of new information while, often, also surpassed when
some of our own “original” ideas and concepts suddenly became obsolete or flimsy
when compared to recent work by other researchers around the world.
Our references for what is the field of algorithmic composition and its main
achievements also became outdated and redundant. Curtis Road’ (Roads, 1996),
Eduardo Reck Miranda’s (Miranda, 2001) and David Cope’s (Cope, 1996, 2000,
2004, 2005, 2008) books will always be historically important and full of relevant
information, taking into account their context, but Nierhaus’ (Nierhaus, 2008)
book on algorithmic composition does a major update and quickly became a ref-
erence as a state of the art survey. More recently, Jose David Fernández and
Francisco Vico (Fernández & Vico, 2013) did another massive update, their article
being a very complete and up to date survey for the field, and therefore we will
depart from these two references to introduce this discussion.
Algorithmic composition is a subset of a larger area: computational creativity.

The problem of computational creativity in itself is still difficult to define and
formalize well. It can basically be said to be “the computational analysis and/or
synthesis of works of art, in a partially or fully automated way” (Fernández & Vico,
2013, p. 513). On the other hand, the most common approaches and problems
algorithmic composition had to solve are now systematized and well-defined.
232
12.1. A survey on algorithmic composition
We will reproduce the classification system Fernández and Vico adopted (Fernández
& Vico, 2013, p. 519). We are adopting this classification system because of its
systematic and formal coherence: it focuses on the different methods implemented
from the point of view of the processes they are trying to model, which is exactly
the task we are pursuing ourselves (figure 12.1).
Figure 12.1: Fernández and Vico (2013, p. 519) classification system.
There are two big areas representing two radically different approaches. Either
one can observe how intelligences typically create and then model their behavior
or reasoning (artificial intelligence) or one can simply focus on getting good mu-
sical results even if a human would never be able to do it that way. These last
approaches usually represent processes found in nature and are assumed to be the
result of chaos or complex interactions without any intelligence behind. Among
these complex systems one can find self-similarity processes (fractals, for instance)
and cellular automata.
Among artificial intelligence, one finds processes that intelligent beings use.
Those are divided into three main areas: symbolic processes, optimization and
machine learning.
Symbolic artificial intelligence relies mainly on modeling the process of creation

by the use of simple rules. They usually represent the creation of symbols and
their subsequent manipulation and transformation. These approaches depend on
the idea that the creative processes can be described in formal terms and reduced
233
to a theory. The problems must be well-defined and their boundaries established,

as well as its constants, variables and sets of operations. In this case, the rules
must be conveyed by the programmer, who obtained them through the previous
observation and analysis of the processes involved.
Optimization processes derive mainly from the idea that one can depart from
very simple units or imperfect samples (members of a representative population)
and combine, recombine and eventually transform them through successive trial
and error operations. The strongest and fittest combinations are selected, while the
weakest and most unsuitable ones are discarded, until results start to improve and
eventually get optimal. These processes resemble the Darwinist idea of evolution.
Machine learning processes, on the other hand, focus on analyzing the outcomes
of previous creations. A representative corpus is constituted, processed and ana-
lyzed, and then new predictive or prescriptive rules are inferred by the machine.
Using those rules the computer will, then, be able to generate new works.
12.1.1 Symbolic artificial intelligence
The concept of formalizing processes through the use of symbols and their
subsequent manipulation has historical precedence in almost every scientific area of
knowledge. One of the most important historical approaches to symbolic artificial
intelligence is the use of grammars. Grammars consist of a series of abstract rules
that are provided in order to build or infer a new sequence after a given basic set
of symbols, ideas or objects. They have their origin in the study of linguistics,
namely in the model developed by Noam Chomsky in 1957 (Chomsky, 2002), and
in the parallel thought that music is a hierarchical structure (musical notes can
be seen as phonemes that can be used to build words, sentences, paragraphs,
etc.) (Bernstein, 1976), (Miranda, 2001, pp. 72-78). This kind of formalization
was very influential in the development of derived theoretical models, namely
A generative theory of tonal music developed, in the early eighties, by Lerdahl
and Jackendoff (Lerdahl et al., 1996), which was subsequently upgraded (Lerdahl,
2005, Margulis, 2005).The derivation of the grammatical rules and their subsequent
mapping seems to be the major factor determining the quality of the results.
Historically they have been mostly derived by hand, based on previous knowledge
or observation of musical corpora. However, some attempts at inferencing the rules
automatically have also been tried. The application of grammars to music systems
234
usually suffers from linearity, since they are only able to deal with one layer at a
time. Nevertheless, they are a very powerful tool and have been sometimes used
in Jazz improvisation programs and in chord progression creation (Nierhaus, 2008,
pp. 83-118), (Steedman, 1996).
A recent example concerning the use of Grammars is the work of Tojo et al., a
musical analysis system based on A generative theory of tonal music, which is a
vast elaboration on previous projects which included a music expectation method,
predicting the next notes needed to assist musical novices in playing improvisa-
tions, and a melody morphing method, generating an intermediate melody between
two melodies in a systematic order in accordance with a specific numerical mea-
sure (Tojo et al., 2013). Another recent example is the work of Quick and Hudak,
presenting a new class of generative grammars called Probabilistic Temporal Graph
Grammars, which they claim to be able to handle temporal aspects of music in a
way they retain a coherent metrical structure and also phrase repetition (Quick &
Hudak, 2013).
Grammars exist in other variants. One of them, featuring parallel rewriting, is

known as Lindenmayer Systems, or L-systems. They are mostly useful for model-
ing self-similarity patterns, especially in visual renderings using turtle graphics. A
popular use of these systems in algorithmic composition is “to pre-generate a col-
lection of short fragments and/or other musical objects, and define an algorithm to
interpret the final sequence of symbols as instructions that transform and arrange
the fragments into a composition” (Fernández & Vico, 2013, pp. 523-524).
A borderline case of grammars, Transition Networks are a graphical represen-

tation of finite automata in which the nodes represent the state of each single
automaton and where the edges indicate transitions. The structures may allow
recursiveness. In their conceptual basis they derive from grammars and are bor-
rowed from computational linguistics (Woods, 1970). In the musical applications
discussed, they can be considered more like an intelligent approach than a rule-
based system, since their application is usually based on inference and retrieval of
patterns within a provided corpus (Nierhaus, 2008, pp. 121-130).
Inside this category, the most relevant and successful case is David Cope with his
EMI and Emily Howell programs. Following a composer’s block, Cope generated
the idea for a computer program that would be able to aid him by making sense
of his overall musical style and mimicking it, composing new music that sounded
235
like him. Utopian as it may sound, searching for a machine to replicate one’s own
creativity might not be a mere “aid” but more of a kind of immortality. According
to Hayles, Minsky seemed to believe that “if we can become the information we
have constructed, we can achieve effective immortality” (Hayles, 1999, p. 13).
Just imagine the consequences of a computer producing Cope’s music after Cope
no longer exists. Moreover, “no longer about ‘life’, life science is now about the
fact that there is nothing but story, nothing but information. This information is
the sublime body” (Doyle, 1997, p. 22). Is it really feasible to extract one’s own
essence into mere code, information, and replicate it with a machine? However,
at least musically, Cope thought this could succeed.
According to Cope’s website2 and related literature (Cope, 2005, Silva, 2003),
his program functions after building a database with relevant material of a given
composer in MIDI format. Therefore, it is possible to generate a choral similar
to Bach’s, by inserting in the database as many of Bach’s chorals as one can.
The more information the computer has access to, the better the outcome will
be. Cope explains in a rough way how the computer works recognizing signatures
(to determine the overall style), patterns (to identify recurrent motifs), allusions
(to identify certain formal processes that replicate themselves). Then, using the
information retrieved, it is possible to apply a series of techniques as recombination,
association or integration to generate a new piece of work that will resemble the
ones in the original database.
Regarding these issues, there are two things to account for: the first is that
the whole process of composition relies in solving problems by making choices; the
second is that the computer does not have established rules on how to compose.
It will have to learn and deduce them from the compositions that are fed to it. In
informatics and in Cope’s case, the concept of learning is intimately related with
trial and error processing – a constant back and forth that is recursively executed
until success is achieved. Let one imagine that to accomplish a certain goal in a
composition only the note B would fit. The first time it was running, the program
would test every note one by one, and all would be rejected by the programmer
instructions until B was accepted. The computer would keep that situation in its
database for future reference, so the next time a similar situation were to occur the
machine would have already “learned” what the adequate solution would be and
would promptly answer with it. It can be seen that the more often a computer runs
2
http://artsites.ucsc.edu/faculty/cope/experiments.htm, accessed June, 6, 2015
236
these processes, the quicker it becomes, because it is “learning” and keeping the
good solutions for its problems. So, machine learning is understood as a continuous
optimization process of adequacy of the proposed solutions to presented problems
(Cope, 2005, pp. 177-219).
Many of the processes symbolic artificial intelligences model are based on rules,
thus shaping rule-based systems. Often they are not always hierarchically struc-
tured like grammars; instead, these rules are often implemented as a set of iterative
procedures. An iteration is the successive replication of a mathematical procedure,
often building up in outcome of the previous step. A value chosen to map a cer-
tain musical parameter might be constantly transformed in this way. Usually
iterative processes depend on functions that cause three types of outcome: either
one tends to achieve a stable fixed value, or oscillating between certain values,
or oscillating in an unpredictable way (chaotic behavior). Usually, the potential
in algorithmic composition of the first case is not very appealing, since listening
to music relies heavily on a good balance between repetition of musical elements
and novelty. On the other hand, chaos has an enormous potential and it is widely
used, especially because of its unpredictability and possibility of novelty genera-
tion. Associated with iterative procedures, there are also series of functions able
to generate self-similarity (fractals). These are interesting because they mirror
patterns and structures often found in nature. The main issues surrounding iter-
ative procedures lie on the control (or lack of it) over them: specifically in chaotic
functions, a very small deviation in the initial states can add up to an enormous
difference in the outcome. Moreover, once applied, they evolve by themselves and
one cannot do much to change the flow of the events. Also, the outcome is very
difficult to foresee, which is not the case with other rule-based systems applied to
music (Miranda, 2001, 83-98) (Nierhaus, 2008, pp. 131-155). For more information
on iterative procedures applied to music see Gogins (Gogins, 1991).
The evolution of rule-based systems in the 80s and 90s led to a more sophis-
ticated formalization. The process of composing, this process of making choices
could also be seen as a constraint satisfaction problem, because the choices do
not occur in an unlimited space of options. An obvious example is the choice
of tempo for any given musical work: reasonably one seldom finds tempi below
40 beats per minute or above 200 beats per measure. These two numbers es-
tablish boundaries constraining a particular musical parameter. Therefore, the
237
formalization of any given style or process, implies that every time a rule is cre-
ated or inferred, and a choice has to be made, also some constraints have to be
established to limit the available options to achieve reasonable and realistic re-
sults. The choice of constraints will, of course, enhance or impair the results:
over-constraining the problem will lead to more predictable and repetitive results,
while allowing too much fluctuation might lead to more innovative but also surre-
alistic outcomes. The recent project of Flow Machines3 , coordinated by François
Pachet, and jointly hosted by SONY CSL and UPMC – Paris 6, is an example of a
constrained based formalization. Instead of focusing on the process of creating an
individual artwork, the project targets the formalization of style, through the use
of Markov constraints. A user should be able to create not one, but several works,
all conceptually connected, assembling a coherent corpus. In their examples, they
have shown how jazz virtuosity may be modeled as a Markov sequence generation
problem with unary constraints holding on specific notes of the melody. Several
jazz lead-sheets are available and new music can be created in the style of famous
composers, or even chimeras. An example provided is the “Boulez Blues”, this
being the composition of chord sequences using the style of Charlie Parker, con-
strained by a Boulez requisite that all chords must be different (Pachet & Roy,
2014, Pachet et al., 2013).
Another frame for rule-based systems is case-based reasoning. A database of

possible cases is assembled representing problems and their solutions, generally
including structured knowledge on how the case is solved. Each time the program
runs, it goes to the database to find the more adequate solution. When accepted,
the new case and its solution might be recorded and added to the original database.
Because the musical process can be seen as the constant output and synchro-
nization of several agents at the same time (namely a live performance, or im-
provisation), another way to formally describe and model it was to make use of
concurrency models. Concurrency models provide primitives to specify the seman-
tics of distributed systems.
Papadopoulos and Wiggins (Papadopoulos & Wiggins, 1999) presented a sur-

vey pointing out many examples of symbolic artificial intelligence and describing
their scope, namely the CHORAL program by Ebcioğlu (a rule-based, constrain
satisfaction, expert system for the harmonization of chorales in the style of J.S.
3
http://www.flow-machines.com/, accessed June, 6, 2015
238
Bach) or the real-time jazz improvisations by Ramalho and Ganascia (case-based

reasoning), among many others.
They left out, however, the very famous Lexikon Sonate by Essl, one of the
best examples of a successful rule based system.
“Lexikon-Sonate is a work-in-progress which was started in 1992.

Instead of being a composition in which the structure is fixed by nota-
tion, it manifests itself as a computer program that composes the piece
– or, more precisely: an excerpt of a virtually endless piano piece – in
real time. It lacks two characteristics of a traditional piano piece: there
is no pre-composed text to be interpreted, and there is no need for a
pianist or an interpreter. Instead, the instructions for playing the pi-
ano – the indication ‘which key should be pressed how quickly and held
down for how long’ – are directly generated by a computer program
and transmitted immediately to a player piano (or a MIDI/software
synthesizer) which executes them.” (Essl, 1992).
This represents a computer that, when the program is running, is self-sufficient

and is able to compose for itself a piece in real-time and execute it without human
intervention, once the process has started. Each time the program is executed it
generates a different piece and the program can be executed as many times and
for how long as desired4 .
The piece itself is constituted by 21 possible modules. Each one of these modules
corresponds to a set of instructions leading to certain way of sounding having a
high degree of freedom. All these modules might be combined in clusters of two,
three, four, etc, a piacere of the listener, who may decide which ones to activate,
or simply lie back and rely on a randomizer to pick them. Analyzing this piece,
one acknowledges that it is a sort of dice game pieces popularized in the eighteenth
century. However, it is applied in a textural plan and not chronologically – instead
of a pre-set fragment following another one, here there is a kind of texture or way
of playing that is going to be mixed with another. In a certain way, one might
say that the piece respects a certain style. However, one’s ears may not be able to
recognize that linearly: one may imagine listening to an excerpt now having three
4
It is possible to execute an online version of Lexikon Sonate in http://www.essl.at/works/
lexson-online.html, accessed June, 6, 2015
239
modules activated, then, on another occasion, another excerpt with another three,
and then, in yet another occasion, a third excerpt with another three different
modules. It would be reasonable to think that one is listening to three different
pieces with three different styles. The only way to realize the style of the piece, or
a development of it, would be listening to it several times, many times. It must
be remembered that the Lexikon Sonate is one virtually endless piece, meaning
that whenever one hears it, one never knows which part is being played, or even
what is being listened to. There is no beginning or end. This piece transcends
traditional human listening, and moreover, has ubiquitous properties. Imagining
two or more computers running the software in the same room, one would be
listening to different instantiations of the same infinite piece. In a certain way,
Lexikon Sonate has within itself the potential of providing all possible music in
that style, and that is more than humans can handle in their limited lifetimes (the
odds of it repeating itself are very small). This kind of generated pieces of music
brings very interesting consequences, as well as many philosophical questions.
In a detailed paper on six techniques for algorithmic music composition, Langston

describes a technique called “Riffology” in which a pre-set of melodic phrases are
fed to the computer, along with rules to combine them structurally.
“The riffology algorithm makes dynamically-weighted random choices

for many parameters such as which riff from a repertoire of melody frag-
ments to play next, how fast to play the riff, how loud to play it, when
to omit or elide notes, when to insert a rhythmic break, and other
such choices. These choices are predicated on the model of a facile,
unimaginative, (and slightly lazy) guitarist” (Langston, 1989, p. 6).
He also refers in detail another technique called “Key Phrase animation”:
“Once the key frames are drawn, the ‘in-between’ frames are pro-
duced by interpolation between the key frames. (. . . ) If we let melodic
phrases take the place of the key frames (assuming that starting and
ending phrases have the same number of notes) and simply interpolate
pitches and time values between corresponding notes in the phrases, we
achieve a musical version of key frame animation.” (Langston, 1989,
p. 12).
240
One can easily imagine a fruitful combination of these two techniques in style
imitation, using the first one to provide pre-fixed structures and skeletons of each
piece and the second one to fill in the gaps creating the melodies that were missing.
12.1.2 Machine Learning
Machine learning is an approach that relies on previous sets of data. The

algorithms programmed, instead of being generative rules, are learning rules that
use the data provided to infer a model. Based on that model they can then make
decisions or predictions. In the case of algorithm composition, musical corpora are
often used as the initial data, the goal being to imitate the style of a given corpus
or to manipulate it.
A very popular approach concerning machine learning is the use of Markov mod-
els. Markov chains represent the probabilities associated with transition states and
are built in a way that future events depend on one or more past ones (Miranda,
2001, pp. 69-72), (Nierhaus, 2008, pp. 68-82). The concept was first introduced by
Russian mathematician Andrey Andreyevich Markov (1856-1922), who applied his
chains to the distribution of vowels and consonants in Pushkin’s poem “Eugeny
Onegin” (Basharin et al., 2004), (Nierhaus, 2008, p. 67). The term “Markov
chains” applied to this class of stochastic procedures was first used in 1926 in
a publication of the Russian mathematician Sergey Natanovich Bernstein (Bern-
shtein, 1926), (Nierhaus, 2008, p. 68). Harry Olson was the first to use them in
the context of algorithmic composition by analyzing eleven melodies by Stephen
Foster and producing Markov models of first and second order in regard to pitches
and rhythm (Nierhaus, 2008, p. 71), (Olson, 1967, pp. 430-433). These models,
due to their simplicity and rule-based functioning adapt very well to style imita-
tion. However, their structure only allows the description of context dependencies
in linear succession, which means that one can only analyze and generate one mu-
sical parameter at a time, when most music has a vertical dimension and not only
a horizontal one. Thus, Markov models are applied considering fractioning music
to single states in a rather theoretical approach. Furthermore, as models of higher
orders lead to very large transition tables, their use is usually generalized to the
analysis of a previous database. Cybernetic Composer (Ames & Domino, 1992) is
a program that generates pieces in different genres, such as jazz, rock or ragtime
241
using Markov chains to generate the rhythm at one stage, and the melody at a
later stage.
A step ahead, in terms of machine learning, was to try to model the way brains
process data and infer knowledge and, therefore, to assume the neuron as the main
basic unit and use this principle. While, historically, the first formalizations and
attempts date back to the early 40s (McCulloch & Pitts, 1943), mainly for image
recognition and classification purposes, only more recently they were employed in
music systems. Most artificial neural networks are designed as a type of automa-
ton and, as such, they can be represented as weighted graphs, where each node
encapsulates an individual neuron and the weighted edges represent the synaptic
links. Usually there is also an activation function associated to each node in order
to scale the output values. Training a neural network involves presenting it with
a series of samples of the problem to be solved and an example of a solution for
each sample problem (Miranda, 2001, pp. 99-118). Artificial Neural Networks
may generate satisfying results over short passages, but show weaknesses in the
creation of larger context-dependent material. On the one hand, the reason for
these difficulties lies in the application of the back-propagation algorithm used
frequently in the training, which has problems in processing an exhaustive con-
text. On the other hand, the modeling of large and context-dependent musical
sections is a general problem for all methods of algorithmic composition that are
not able to process information as a hierarchically ordered structure. For the
treatment of material which is context-dependent over long passages, generative
grammars are well suited. When knowledge about the domain to be modeled
exists, rule-based systems are generally preferred. Moreover, the generations in
a neural network often end up in stationary situations. Another disadvantage is
the fact that, for the production of longer musical segments, a great number of
training cycles is in most cases required. A great advantage of neural networks
over Markov models and generative grammars, however, becomes evident in the
production of smaller musical components. Both in generative grammars and also
in Markov models, only transitions (e.g. of tone pitches) that are also explicitly
contained in the corpus may be generated. Here, the neural network may produce
surprising movements that nevertheless meet the requirements of the underlying
corpus – this aspect of artificial neural networks may also represent an interest-
ing motivation for their application in a framework of innovative compositional
concepts (Nierhaus, 2008, pp. 205-223).
242
Papadopoulos and Wiggins (Papadopoulos & Wiggins, 1999) have done a com-
prehensible survey about programs like this, presenting the feed-forward Artificial
Neural Networks that Todd used with feedback for melody generation; the Boltz-
mann machine for harmonization by Bellgard and Tsang; Toiviainen jazz impro-
visator; Hörnel and Degenhardt baroque-style melodic improvisation; Schwanauer’s
MUSE chorale harmonizer, among many others.
12.1.3 Optimization
Optimization starts from the idea that a model can become better or more
efficient with time, thus presupposing evolution. Evolutionary systems are derived
from a Darwinist model of evolution. An initial population is expected to evolve
during the course of several iterations of successive evaluations. This implies that
at each stage members of the population suffer some kind of mutation (typically
random transformations or recombinations), which are then evaluated through
some kind of fitness function which will assert its quality. If the quality is above
a certain threshold, determined by the fitness function, then the new specimen
survives, and it is integrated in the original population; if not, it is discarded.
Overtime, the population will become more fit (its best and mean fitness will
increase), according to the established criteria.
Within the category of evolutionary systems there are the genetic algorithms.
They can be seen as an adaptive system of sounds used by a number of individu-
als in a certain community. These “musical-living-organisms” can be mapped to
evolve according to certain rules and suffer transformations, thus generating new
material or spreading and diversifying the already existing one. Domain-specific
knowledge of the problem to be solved is not necessary for the application of a
genetic algorithm. Therefore, this class of algorithms is especially suitable for
tasks that are difficult to model mathematically or for problem domains that do
not have an explicit superior rule system. The architecture of this class of algo-
rithms promotes the production of a large number of small form segments in the
generation of musical structure. An essential feature of a genetic algorithm is the
continuous generation and examination of symbolic strings – a procedure that is
highly suitable for process-like compositional concepts. However, because a genetic
algorithm usually generates new outputs by continuously arranging fragments of
chromosomes, different transformations within a chromosome are combined and
243
therefore lose their structural functions. Therefore, its application to style imita-
tion can become problematic (Miranda, 2001, pp. 119-157), (Nierhaus, 2008, pp.
157-204).
A successful case of evolutionary algorithms applied to music composition is the

project Melomics 5 . It started out with the creation of a computer cluster called
Iamus, which generates compositions based on formal constraints and aesthetic
principles (Diaz-Jerez, 2011). The compositions then evolve following a strategy
of genomic encoding, in a way resembling embryological development. A fitness
function including values on instrumentation and preferred duration evaluates the
successive steps. The whole project was transferred to the music industry and the
music it produces is sold as a commodity (Fernández & Vico, 2013, p. 550).
12.1.4 Complex Systems
Complex systems is a term used to represent all kinds of methods not considered
artificial intelligences, because they are not modeling any intelligent process. One
of them, cellular automata, fits very well modelling systems that are dynamic and
change according to time or space, in discrete quantities. Each state is represented
through cells (usually in an endless n-dimensional grid), which are elements that
will suffer the effects of predetermined transition rules. All cells are updated at the
same time and can be used to model virtually everything that changes in the course
of time-steps. A famous instantiation of this process is Conway’s Game of Life.
Eduardo Reck Miranda created CAMUS, a system using two cellular automata
to produce music: one to determine the melody; the other, the instrumentation
(Miranda, 1993). Unfortunately, the results were not as successful as he expected.
More recently, he argues that this kind of method is more suited to generate
pleasing results when applied to sound synthesis (Miranda & Biles, 2007, pp. 170-
193).
Self-similarity traits, often found in nature, are another set of tools used in
terms of algorithmic composition. It has been found that 1/f noise is able to
produce pleasing results due to this fact – “its structure is statistically similar
across several orders of magnitude” (Fernández & Vico, 2013, p. 556). Since then
it has been used numerous times either as a device to generate raw material or as
5
http://melomics.com/, accessed June, 6, 2015
244
12.2. Advantages and Disadvantages
a modifier. Although 1/f noise is preferred for that fact, other kinds of noise, like
pink or Brownian noise, are also employed. Jeon et al, use noise to both generate
and modify melodies, thus creating variants (Jeon et al., 2006).
12.2 Advantages and Disadvantages
When dealing with algorithmic systems in order to emulate a certain practice,

one must understand that the goal is somewhat different from simply using proce-
dures to empower creativity to produce a single piece of art with an aesthetic goal.
A practice may be characterized by many parameters, and so the approach must
reflect carefully as many relevant parameters as possible and provide a sufficient
margin so that an infinite number of different solutions may arise. Therefore, one
should start by considering the quality and quantity of works within the desired
practice about to be mimicked.
Thinking about symbolic rule-based systems
“first, user-provided rules limit programs only to styles with which

the user is familiar and can code. Second, user-provided rules often de-
rive from a statistical or at least generalized model of a style and rarely
reflect the idiosyncrasies which often help define musical style. Third,
user-provided rules often require exhaustive amounts of code, since
such rules can often contradict one another and/or require backtrack-
ing or other time-consuming and inelegant processes” (Cope, 2002a, p.
124).
Furthermore, the rule-based approach depends heavily on the ability of the pro-
grammer, since the computer only works with the rules assigned to it. For a
rule-based system to have a successful result the stochastic scope has to be wide
enough to provide an enormous number of valid solutions. However, by granting
this much liberty, it is not guaranteed that the outcome is appropriately gener-
ated in terms of the style to be modeled. On the other hand, highly restricting
the algorithm might end up leading to the mere reproduction of examples of the
corpus (Nierhaus, 2008, p. 271).
When working with symbolic systems one can imagine many more problems.
Grammars, for instance, are hierarchical structures. Since a lot of music is not,
245
musical grammar implementations do not make any strong claims about the se-
mantics of the pieces. Usually, a grammar can generate a large number of musical
strings of questionable quality.
Optimization methods also present the problem of subjectivity and efficiency,

since they depend on the users to evaluate and filter provisional results or inter-
mediate steps of the compositional process. This can easily lead to fatigue from
the user’s point of view, in the case of manual selection. The programming of
automatic fitness functions can be highly problematic as well and may not reflect
what the end users expect quality to be.
A corpus reflecting a musical practice that encompasses works that can be very
different between themselves might not be suitable to be modeled by a machine
learning system. The reason for this is the risk of not producing coherent rules.
That would jeopardize the quality of the future output since:
“the structure of the corpus is to a large extent jointly responsible

for the quality of the generated output. The selected examples should
be able to efficiently represent the class of possible compositions and
at the same time, depending on the strategy of the algorithm, these
musical pieces should be able to be combined with each other – this
means that the corpus must not join several classes of compositions
whose structural properties are not interchangeable” (Nierhaus, 2008,
p. 271).
In this situation, a previous human analysis could be more effective by finding

ways to standardize the corpus, or by reducing the style into smaller groups of
similar works and extracting the main rules that govern them. Moreover, when
using Artificial Neural Networks, the representation of time cannot be dealt with
efficiently, even in those cases for which it has feedback. Also, they usually solve
toy problems, with many simplifications, when compared with symbolic rule-based
approaches (Papadopoulos & Wiggins, 1999).
Complex systems present the problem that from a scientific point of view they
are not modeling any concrete process intelligent agents actually consciously use
when making music. If the challenge is to know more about the composition
process in particular and about human agency in general, then one might be
missing the point by employing them. On the other hand, it is true that many
246
12.3. Conclusion
human decisions, in particular subtasks, seem grounded on random or cosmic laws

that transcend our best tries to rationalize them. In those particular points they
might be useful.
Furthermore, there are reality constraints: with what means and within how
much time should the system be produced? Having all the time and means in the
world, one could in all seriousness be advised to use a machine learning approach,
since it deals with the most purist and abstractionist reasoning an academic can
aspire to: modeling a brain, able to learn and capture all the intricacies that even
a trained human is not able to, if one points at the big picture (for more on the
subject see (Dreyfus & Dreyfus, 1988)). However, having a strict deadline and the
need to present convincing results, then a symbolic system approach, as simple as
it might be, is always a safer one, since it will always allow the presentation of
something. It can lack depth, but that will always be a price to pay. No work can
ever be done to perfection, so one must weight very well all factors, and make a
decision.
12.3 Conclusion
Empirical experience shows that all the considered approaches to algorithmic

composition have flaws and strong points depending on the target. The main fac-
tor in consideration regards the practice to be mimicked and the kind of corpus it
presents. Practices with a large corpus, unfamiliar to the user, and similar struc-
tures seem more appropriate to machine learning systems; on the other hand,
practices that seem very intuitive to the programmer and that pose problems
regarding the consistency of the corpus are filtered more easily by a human intelli-
gence and then explained to a computer by symbolic systems. This latter approach
seems also more predictable, providing expected results very similar to the ones
in the corpus (since all the rules and constraints of the practice are read correctly
by the user), whereas the automatic machine learning systems approach is more
likely to fluctuate within the idiosyncrasies of the practice and present somewhat
unexpected results (or even gibberish), considering a wide corpus is provided.
247
Chapter 13
The model
Following the ideas of Brian Eno, it seemed to us that the most elegant model
is the one that, taking the least amount of information possible, is able to extract
the maximum of it, and yet still retain the core ingredients of the process, of the
sequential steps undertaken to compose a song. That translates into a generative
process like the one of planting a seed and then seeing how an entire tree grows out
of it (Eno, 1996). This seems to imply strong hierarchical structures and a good
degree of self-similarity. Our observations suggest that, given a lyrical text (what
would be literally a fado), then the entire musical work derives from it, using a
set of predefined conventions, shared grammars and transformations. Moreover,
we believe this same idea can be applied to most song-writing traditions, only
changing the characteristics of the text and the predefined conventions, grammars
and transformations. Furthermore, we also believe that the text in itself, can be
dispensable, sometimes, in the composition process. One can work, in fact, with
only some information about its structure, namely the number of stanzas, lines,
syllables and, eventually, accents. Having this seminal information, and the correct
rules and transformations defined, then, one is able to, just following them, very
similarly to following a kitchen recipe, obtain a suitable musical work, complying
with the genre modeled.
The model (figure 13.1) assumes the sequential order through which fados are
typically performed and simulates the various agents and decision processes in-
volved, as observed in detail in the first part of this dissertation. The point of
departure is a text (a fado, in the form of a poem) that will be sung by a per-
former on top of an instrumental accompaniment. Either the composer or the
249
Figure 13.1: A fado model.
fadista (when the fadista is the composer in real time) creates a melody out of
that text. At this stage, we are not working with actual lyrics or textual content,
although that would be ideal in the future. Therefore, the first module in only
pretending that there is a text, and, based on a seed, it creates a list of symbols in
numeric format. Those numbers represent the relevant information an original text
would carry: namely the number of stanzas, lines per stanza and syllables per line.
One could argue that in a real text all information is relevant (namely semantic
content); however, as seen in the first part of this dissertation, fado practitioners
claim that as long as the metrical structure is the same, the lyrics can be easily
interchangeable between the fados. So, while it can still be argued that accents
and vowel quality might affect the prosody (and therefore the melody to some
extent), the core values that define the importance of a text, in structural terms,
are the ones we are indeed retrieving/generating. This is because the number of
stanzas determines the number of sections the work will have (hence, the overall
form), and the number of lines determines the length of each section (number of
melodic phrases per section). The number of syllables of each line determines the
rhythms of the melody (constraints the number of notes of each melodic phrase).
A second module simulates the agents involved in the instrumental harmonic

accompaniment (either a piano, or, more commonly, a viola and a bass). This
250
module retrieves information from the “text” module (the form and general struc-
ture of the sections) and generates suitable harmonic progressions and scales to
fit into them. It also generates ostinati and bass lines based on those progressions
and scales.
A third module simulates the singer. This module retrieves the information
from the “text” module (namely the number of syllables of each line) to generate
the rhythms of the melody. It combines this information with the harmonic infor-
mation retrieved from the second module to assign suitable pitches, and therefore
simulating the creation of melodic lines on top of the previously known stock-
accompaniment.
A fourth module simulates the guitar. This module retrieves information on all
previous modules and is then able to generate suitable counter-melodies or melodic
figurations that simultaneously fit in the harmonic progressions, and at the same
time, simulate a reaction to the melodies generated by the singer.
All these four modules generate what we would call an urtext score, similar to
the ones in the database. A raw, quantized ideal, to be, then, performed. This
implies the repetition of sections and styling. A MIDI file exported at this stage
would be ideal to be printed as a music score.
A fifth module, a pseudo-humanizer, consists in a series of functions that will

transform some of the generated parameters to originate variants. The transforma-
tions, when applied to the melodies, will simulate the styling of fadistas, therefore
slightly changing rhythms, pitches and adding ornamentation. When applied to
the accompaniments they might originate pauses, and small rhythmic and melodic
variations as well. This will simulate the real practice, guaranteeing that virtually
any repetition of a section will sound slightly different each time. A MIDI file
exported at this stage reflects the performed fado, not suitable to be printed as a
score, but perfect to be imported into a digital audio workstation to be recorded
with virtual instruments.
A sixth module is idealized as an extension of the fifth module. However, we

were not able to implement it in time to produce convincing results and will be
implemented at a later stage. It will involve micro-transformations to enhance
the humanizer results, namely by controlling “groove” and syncopation in a not
random fashion, and also controlling pitch bend messages to be able to generate
portamentos, glides, slides and more convincing tremolos and mordentes.
251
Chapter 14
Methodology
The observation of the practice, the processes it encompasses, and the model
obtained determines the methodology to be followed. Since the problem of song
composition, in general, and the musics and vocal sounds of fado, in particular,
has been modeled as a hierarchical linear generative process, a symbolic artificial
intelligence is the logical choice. Each step of the process involves either the
generation of symbolic material, its retrieval from another source (a table, a list),
or its transformation, via a set of rules. It is either a chain of choices or orders,
each one having boundaries and probabilities of happening or not. Therefore, it
is mainly a consecutive series of constraint-satisfaction problems.
The main inspiration for this research project, David Cope (Cope, 1991, 1996,
2000, 2002b, 2004, 2005, 2008), uses a borderline approach based on grammars
and transition networks; however, he also relies on direct feeding from corpora for
style imitation. The strong aspect of his work are the actual results, since most
of the works displayed seem well-formed and coherent structurally. They could
perfectly be the manual labor of any aspiring composition student imitating the
style of past masters. Although Cope’s arguments and descriptions are clear, the
actual computational processes seem obscure and his algorithms are never made
explicit enough for replication. On the other hand, the more recent approach of
Pachet et al. in the Flow Machines project (Pachet & Roy, 2014, Pachet et al.,
2013) seems promising and more of an actual source of reliable inspiration. Still,
their integration with lyrics is not yet done and one is not able to discern how the
text influences or shapes the music. Also, their examples of jazz lead-sheets lack
the depth of a fully developed song in terms of form and variation. Moreover, their
253
content sometimes resembles just a collage in which their individual components

lack the semantic context from which they were originally derived, therefore the
ending result losing coherence. The work of Toivanen et al. (Toivanen et al., 2013),
however, seems a good example of what it is pursued: they aim at automatic com-
position of lyrical songs, and create a module that actually generates text and this
text directly influences the musical content. This approach, theoretically, makes
perfect sense. The mapping of the textual parameters regarding the musical ones
seem consequent regarding what are the actual processes and practices observed.
They also imply that the semantic content of the text should affect the mood of
the music (according to arbitrary Western-European conventions), by mapping it
to tonality, which seems also logical within many genres. However, their program
also lacks the formal coherence of what it would be expected a typical song to be
like. There is not a clear phrasal structure, neither a complete form defined over
time (regarding repetition and variations, the songs just seem run-on), neither
does it have a performative layer: the program merely generates quantized MIDI
scores.
These detailed examples, two of them still in progress, act as direct inspirations
for the model and all of them rely on hybrid symbolic approaches mixing several
techniques to solve the constrained-choices along the way, or to generate specific
sets of material. Therefore, it has been decided to pursue the same path, learning
from their research and trying to integrate in the model the best ideas one could
find among them, while discarding, changing or improving the ones thought to
not be adequate. Hence, a hybrid constraint-satisfaction method has been ap-
plied, in order to be possible to build the main structures of the instrumental
music desired. This method was basically a generative reverse engineering of the
analytical-decomposition done previously when we have applied the semiotic an-
alytical method to reduce each fado occurrence to its essential matrix. We were
able, after setting the model, and seeing the rules that govern the practice as re-
flected in the corpus, to build it up again. After all, “the point of analysis is to
explain what is obvious – the experience of musical unity or whatever – in terms
of structures that are not obvious and can only be deduced from analytical study”
(Cook, 1994, p. 222).
With enough rules provided, justified on previous statistical analysis and pat-
terns discussed in the first part of this thesis, we were able to generate new scores of
archetypal music associated with fado. These new sketch-scores were then “filled
254
again” (using the philological method) with the parameters we first removed (in-
troductions and codas, ornaments, and even bugs). We understand that
“each system of notation, then, involves its own patterns of empha-

sis and omission. (. . . ) Consequently when a composer writes down
music he is relying heavily on the reader’s musical ear and imagination
in supplying the precise intervallic, rhythmic and dynamic values that
the notation omits, just as he has to contribute sonorous, dramatic and
emotional values that cannot possibly be specified in score.(. . . ) None
of this happens if you make a strictly ‘scientific’ analysis of a score –
analyzing the distribution of notated intervals in terms of set theory,
say, of by means of statistical comparisons. When you do this, you
are analyzing the score without actually reading it in the sense I have
described. (. . . ) Consequently, if you analyze a given composition this
way, your analysis may be scientific in the sense of having an explicit
methodology, but it will not be at all scientific in the sense of having
any meaningful or predictable relationship to the music’s physical or
psychological reality – that is, to the noise it makes or the effect it
has on people. Indeed, the more strictly deductive your analysis of
the score is, the more directly will it be conditioned by the particular
cultural and pragmatic assumptions built into the notation and less
bearing it will have upon the music you actually wanted to find out
about” (Cook, 1994, p. 237).
So, regardless of how precise, methodic or scientific all this process might be, we
never lose sight of the fact that we are still lacking some essential components of
the performance practice we have been studying. As mentioned before, we have no
access at all to the sounds of the performance as it was in the nineteenth century,
for instance, so we are aware that our primary goal is to reconstruct musical scores
following the same conventions of the corpus, because that was the starting point in
the first place. To pretend that we would end up reconstructing the performance
practice as it was supposed to be in the nineteenth century from the musical
scores would be utopic, because the textual sources lack a variety of performative
dimensions very difficult to reproduce at this stage. We try, in fact, to add some
of those dimensions to the musical scores afterward using other methods but that
is a whole new issue related with the reproduction of a human performance of the
scores and not necessarily correlated with the procedures pointed out so far.
255
Concluding this point – we were able to reconstruct musical scores based on

the ones we began with: the instrumental music, and the vocal music reduced
to a melodic instrumental line as well, associated with fado, lacking context and
a series of human dimensions that needed to be filled in later. These are fados
that are comparable with the universe of those subjective transcriptions made by
the musicographers of that time. Having done that, we may say that we have
concluded an important part of the project, as well as the most objective and
pedagogical part of it.
14.1 Interactivity
Generating new music associated with fado practice implies to us that the user
should be able to interact within the process by choosing some parameters. We
have aimed to make the interpretation process also interactive.
“To analyze a piece of music is to weight alternatives, to judge how

it would have been if the composer had done this instead of that – it is,
in a sense, to recompose the music in a way that normal, concert-hall
listening is not” (Cook, 1994, p. 232).
We have given people (the ultimate goal of this project – music – is for people
to listen and enjoy) the opportunity to take part in the composition of “their”
fado. There were several ways and techniques to do this and we have had to
make some subjective choices regarding the parameters that could be customized
by assigning some variables and also the ways in which the user could actually
manipulate those parameters.
Interaction presupposes a dialogue in opposition to a soliloquy, and there are

many ways and degrees to do that concerning music. Robert Rowe assumes that
the participation should be occurring on the level of music making in opposition
to passive listening. The computer would be receiving input in real time by ana-
lyzing a real-life performance and reacting to it, much in the style of jazz players
interacting with each other (Rowe, 1999). This view is also shared by Chadabe
(Chadabe, 1984), and historically interactive music has often been designed this
way. A fairly recent application by Dias et al., Gimme the Blues is a Jazz impro-
viser in real time, reacting to stimuli provided by the users on virtual instruments
256
14.1. Interactivity
set up via a friendly graphic user interface using touch screen technology (Dias &
Guedes, 2012, Dias et al., 2012). All these approaches, however, start from the
notion that the users are in the role of musicians or performers and the computer
is simulating their counterparts. Our approach, on the other side, departs from the
perspective of the user as composer, and the role of the computer is to automate
as many of the composition processes as they desire. The system will interact as
little or as much in the sense that the user is able to decide how many and which
parameters they want to chose or leave them up to the software.
Although Todd Winkler wrote a reference book on interactive composition also

aiming mainly at interaction in the sense of a constant real-time performance
and feedback loops between the computer and the user in the role of performer,
one can understand how interactivity does not necessarily have to occur at this
level. He mentions in a negative light the example of television as being “not very
interactive (yet)”. Nevertheless, by doing so he is recognizing that the mere fact of
one being able to switch channels, change volume and altering controls for color,
tint and horizontal hold is, in fact, some interactivity (Winkler, 2001, p. 3). He
also mentions how a slightly interactive piece may be a predetermined score where
the user has the ability to change one parameter as tempo, and how sophistication
increases with the number of parameters one is able to change and thus shape the
final results (Winkler, 2001, p. 6). Moreover, an interactive composition is then
defined in detail by the following characteristics:
“
1. Human input, instruments – Human activity is translated into
digital information and sent to the computer.
2. Computer listening, performance analysis – The computer re-
ceives the human input and analyzes the performance information
for timing, pitch, dynamics, or other musical characteristics.
3. Interpretation – The software interprets the computer listener in-
formation, generating data that will influence the composition.
4. Computer composition – Computer processes, responsible for all
aspects of the computer generated music, are based on the results
of the computer’s interpretation of the performance.
5. Sound generation and output, performance – The computer plays
the music, using sounds created internally, or by sending musical
257
information to devices that generate sound ” (Winkler, 2001, p.

6).
As one can see, the model proposed by Winkler for an interactive piece is respected
integrally by our model, if one regards the user in the role of composer. Their per-
formance (as typical of composers) instead of corresponding to the act of playing
an instrument, corresponds to the act of deliberating and making choices and then
notating them symbolically. In this case, the results of those deliberations are fed
to the computer, which proceeds to the next steps, receiving that information and
shaping the music accordingly, and then returning it back to the composer, in the
form of a score or a rendition. One can easily understand that if the role of the
user would be the role of fadista (singer) or of instrumentalist (viola player, for
instance), then the program would necessarily be totally different. Certainly more
interactive, in the sense that it would require constant back and forth mechanisms
and feedback loops, but at the same time, it would correspond to a different inten-
tion than this one represents. Being creative and mostly interested in the process
of song composition, we tailored the design of the research according to our own
bias holding this perspective. Therefore, we constrained, to some extent, the de-
gree of interactivity the program can possibly have. Nevertheless, it can certainly
be considered interactive.
14.2 Work Flow
We have developed the initial coding work, under the mentoring of Professor
Bruce Pennycook, in Austin, regarding the modeling of the generator. At first, we
had thought of implement the program in Max/Msp, since we had taken classes
and workshops using this programing environment and we were already familiar
with it. However, considering that a musical generator of this magnitude implies a
serious amount of symbolic manipulation, and the need for a very flexible language
that allowed the definition of several modules and customized functions, we opted
for the use of a Lisp based language instead. This decision took into account prior
projects of the experts and references in the area (Nierhaus, 2008), namely David
Cope (Cope, 2004, 2008). It made sense to keep their methodologies and ideas,
in the long run, even if that implied a steeper curve, and having to learn a whole
258
14.2. Work Flow
new language from scratch. Having this in mind, it was necessary to learn to code
in Lisp, which required several months of training.
Initially we started practicing and implementing some algorithms in Common

Music, an open source software developed by Henrich Taube1 . However, after
some insipid, yet important, developments we soon reached the conclusion that
Common Music was not powerful enough for the desired purposes because it lacked
predefined functions ready to deal with a model of tonal music implementation.
It would have taken several months to create those functions from scratch. That
work, however, had already been done by Tonality Systems and their software
Symbolic Composer 2 , also Lisp based, but with a set of numerous musical functions
already predefined which would allow us to save time and resources and concentrate
on the concrete problem and work in higher-level detail, thus enhancing the final
result greatly. So, it was decided to buy this software and we started working with
it in the beginning of 2012. Although one might think that time was wasted in
previous experiences, it was not the case at all, because all coding had been done
in Lisp, the same base, so everything was relevant to the learning and mastering
of the language.
Since we began working with Symbolic Composer, the main structure of the
generator has been coded and a modular structure was conceived in order to
keep the problem the simplest and most elegant possible. The program is flexible
enough to allow several parameters to be defined by the user (like key, tempo,
ostinati), or it can be completely independent generating music at its “will”. The
experiences conducted led to the successful generation of ostinati and harmonic
progressions with controlled weightings. To determine the weight values of the
harmonic progressions and ostinati, we have used the values presented in the tables
of the previous chapter as a reference. These values were refined due to aesthetic
purposes as it will be described later. The aim was to have a final generator that
was able to represent integrally the database as it was in its final stage, and also
flexible enough to allow more variety depending on other musical factors.
It was detected that, although the “harmonizer” successfully generated func-

tional accompaniments, at some point those were not very credible due to voice
leading problems arising from the randomness of the chord inversions and the lack
of independence of the bass line in relation to chords. These problems did not
1
http://commonmusic.sourceforge.net/, accessed June, 6, 2015
2
http://www.symboliccomposer.com/page_main.shtml, accessed June, 6, 2015
259
prevent the development of the coding itself, as the final weights were only needed
in the end. In the meantime we could work with estimations or random values,
and let the database grow at its own pace and even to be completed after the
generator was completed.
When the first stage was complete, we entered a much more subjective (and
creative) area of the project, which dealt essentially with the interpretation of the
generated musical scores. As pedagogical and rewarding as it might be to have
new musical scores produced, we felt that the digital dimension of the project
would not achieve its maximum potential unless we could reproduce those scores,
at least in a moderately convincing way. Hearing the quantized MIDI files led to
some satisfaction, although clearly not enough for any casual listener. The timbre
quality of the instruments used in the actual practice (and that are crucial sonic
cues) were missing and also the groove and waddled feeling imposed by human
performance.
As the work on the fifth module progressed, with the new introductory distor-
tions provided to humanize the performance, the program was able to generate
sonic output, but it was still considered far from being acceptable by the standards
we had in mind: often the outcome created a sensation of “uncanny valley” when
heard: while it clearly seemed that we were going in the right direction, and we
were obtaining some results, they were not close enough to a human performance,
they still seemed robotic, or sometimes just odd. Sometimes we just wondered if a
minimal almost quantized version was not better all along. The poor output has
also shown that some re-writing of the module was required following empirical
and personal compositional experiences as a way to achieve more realistic results.
Although we have obtained some improvements over time, we still find the results
were not what we expected and we envision a sixth module at this stage, hopefully
to be implemented in the future.
260
Chapter 15
Implementation of the Model
During the conception of the model, we were constantly struggling with the con-
silience of the performative practice emulation with the practice reflected in the
corpus of musical scores. In the first part of this dissertation, we have presented
and discussed ideas, values and arguments in an attempt to understand fado from
the perspective of the community, practitioners and listeners. On the other hand,
we gave accounts of work with a very specific corpus of musical scores, which com-
prises a certain representation of that same practice. Having this in mind, the
initial idea was to somehow build an artificial intelligence (following mainly the
goal of David Cope) that mimicked and represented this corpus in the generation
of new scores similar to the existent ones. In order to do that, we have built a
program that mainly recombined weighted lists containing fragments, and specific
parameter values extracted from that same corpus with little to no variation. It
was a simple approach that gave us controlled and predictable results lying within
the desire scope, but that lacked inventiveness, the ability of going further. More-
over, it neither expressed algorithmically any particular new idea regarding the
methods of composing with computers, nor it was challenging to code regarding
the current state of the art. As a consequence of this fact, after the first round of
initial programming we decided to explore a second approach and build a second
artificial intelligence that, instead of just using lists with weights and data derived
from the corpus, actually generates lists from scratch using functions. This second
approach was much more challenging, because we actually had to try to formalize
the sub-processes the musical creation involves. The result was a series of ab-
stractions able to generate a whole series of different parameters that were later
combined with one another and integrated and nested in the previously designed
261
architecture. Due to the very nature of what an abstraction is, and how a func-
tion works, it allows the users either to be very strict and confined, thus providing
predictable results, or if they decide to use arguments outside the expected range,
it will provide totally unexpected and unpredictable ones. Furthermore, it can be
even more unpredictable if one decides to let a randomizer pick what the possible
range is and what parameters it may randomize. The point is that, using this
approach, we have obtained a very flexible artificial intelligence and anyone can
decide how much further one wants to go exploring it.
Each module comprises a series of constraint-satisfaction problems to be solved.

As such, having arrived at this point, we decided that the best solution would be
the combination of both approaches. Either they can be solved by resorting to
a symbolic artificial intelligence based on the statistical data retrieved from the
corpus (thus the weighted lists, the fragments) or via a second symbolic artificial
intelligence working with formalizations of the sub-process involved (translated
into grammars, rule-based algorithms or other iterative procedures). The combi-
nation of both approaches is more flexible and unpredictable regarding the results
users may have in their minds, increasing the possibilities of interactivity (by al-
lowing them to customize and manipulate more parameters). In order for this to
work in a manageable way, we have designed a coefficient of representativeness
(CR) that varies between 0 and 100, 100 being total representativeness of the cor-
pus and 0 total inventiveness. So, when the user decides this coefficient is 100,
only the artificial intelligence dealing with the corpus generates results. When it is
0, all the parameters are retrieved from the lists generated by the second artificial
intelligence. When it is any value in between, this value equates the percentage of
probability, for each calculated parameter, to be picked from either of them. Ide-
ally, the default should be something along the lines of 70 if one wants a general,
predictable score with a spice of novelty, but it can obviously be changed if the
user is willing to take some risks and decides to have a more out of boundaries
musical piece.
15.1 First Module
The first module of the model corresponds to the triggering of a seed that will
cause the information retrieval of a simulated lyrical text, in order to constrain
subsequent choices and structures.
262
15.1. First Module
15.1.1 Seed
A generative program like the one coded relies heavily on random functions and
constrained decisions that depend on seeds to run. As previously discussed there
is no true randomness inside the chaotic and random functions, as they all depend
on a seed to perform their calculations. Along with this issue comes the fact that,
if one wants each and any instance of the program to be reproducible, then they
must all be attached to the same seed. Therefore, we have decided to program
right at the beginning of the generator a function that generates a variable called
“my-random-seed”. Every time the program is called, it will generate a different
result. But, at the same time, if the user likes any of the outcomes and wants to
replicate it, then they are able to do so by choosing exactly the same seed. So,
in other words, each seed represents an instance of a given score, and it is like its
own ID number. And the user has total control over it:
(setq my-random-seed (* (float (/ (get-universal-time)

1000000000000)) (rnd-in 100) ))
(init-rnd my-random-seed)
The block of code shown basically uses the universal clock of the computer to
generate a seed. And it has a resolution up to the milliseconds. So virtually every
time the program is run it will generate a different score, because the time will
always be different. If one likes it, one just has to keep track of the inspector
window and take note of the number that will appear there and one can keep that
seed for later use. The entire score depends on this seed in order to work properly
and it is the initial trigger to derive everything that follows.
15.1.2 Text
The text is in fact just a simulation of what would be a complete lyrical text
and the relevant numerical information about its structure. It corresponds to the
parameters number of stanzas, lines per stanza and syllables per lines. These three
parameters are positive integer numbers and are stored in variables. This can be
done as merely a constrained random generation or they can be defined by the
user.
263
Those numbers, in theory, can be whatever the user envisions, but, as previously
discussed in this dissertation, they are constrained among specific values. The
number of stanzas is usually a small number between two and six, being four the
most common; the number of lines usually also varies between three and ten, four
being the most common and five and six being rather common as well, especially
since the beginning of the twentieth century; the number of syllables per line varies
between four and twelve, being seven the most common, by far.
For demonstrative purposes, in the version presented in this dissertation, we

are assuming four stanzas with four lines and seven syllables, but all other values
can be implemented with some future refining.
15.1.3 Form
The overall form will impose a hierarchical superstructure containing and con-
straining everything else that follows. As shown in previous chapters, the form
of a fado is decided by the lyrical content: namely, the number of stanzas and
lines. At this stage, this dissertation does not deal with lyrical content in con-
crete; however, one can deal with its consequences in an abstract way, by following
its trends. A text with four quatrains in heptasyllabic lines, for instance, can be
extrapolated into how one would typically sing such a text. Following Ernesto
Vieira’s model one would sing the first line over a two bar antecedent melodic
phrase (a1), and the second line over a two bar consequent melodic phrase (a2).
These two lines of text would represent a typical A section, four bars long, on top
of a certain harmonic progression. This A section could be immediately repeated
(representing the textual repetition of the same two lines) or it could be followed
by a contrasting B section representing the following two lines of the quatrain.
So, departing from this simple quatrain one could end up with several possible
combinations – AB, AAB, AABB, ABB, ABAB, all of those already explored
in previous chapters of this dissertation. At this point, and knowing that there
are still three quatrains to sing, several options arise: either they would be exact
replications of the first one, and, therefore, a typical Ernesto Vieira’s strophic fado:
ABAB ABAB ABAB ABAB or AABB AABB AABB AABB, for instance, or it
could be assumed that the first two quatrains were one fado in the minor mode
and the last two quatrains could have a different music in the major mode, and
so they would represent the doubled version of the same model, and it would be
264
15.1. First Module
something like ABAB ABAB CDCD CDCD or AABB AABB CCDD CCDD,
for instance.
Having this in mind, one then realizes that even if the concrete text is not
known (its semantic content), one can try to model the form of the music based
on its structure (number of stanzas, lines per stanza and syllables per line). The
simplest way to formalize this is to simply define a list containing the distribution
of the sections, each section represented by an alphabetic character. Going one
step further, we assume for the model different versions of each section to represent
slightly different performances of the same section whenever they are repeated. A
different version is obtained by applying transformations to some parameters in
order to obtain what can be called variations. The concrete implementation of
these variations will be discussed later. This increases variety and models what
often happens in performance. To each variation we append an integer to identify
them. There is no clear boundary on how many variations one could have. In
theory, each instance of any given section is sung uniquely, therefore it could be a
variation, resulting in at least as many variations as the total numbers of stanzas.
Therefore, the resulting variables assume the form:
Section = Xn, x ∈ {A, B, C, D . . . }, n ∈ N0
A form list is then the chronologically ordered sequence of these variables,

reflecting the idiosyncrasies of the lyrical text, completed with the uniquely titled
sections “introduction” (Intro), optional intermezzos and “coda”. Our example
of the four heptasyllabic quatrains following the simplest Ernesto Vieira’s model,
without redundancy, would result in the list:
F = ⟨Intro, A0, B0, A1, B1, A2, B2, A3, B3, coda⟩
In practice such a list might be doubled since in many fados the singers repeat
the lines. This list represents the superstructure of the piece, its overall form. It
is, in fact, the formalization of the time dimension in a linear fashion.
There are at least three ways of generating such a list. The first one is to let it
literally be defined by the user. In a fully developed app this could be a decision
mediated by a graphic user interface, in order to simplify the process, based on the
structure of the text. The user would just need to provide the number of stanzas,
the number of lines and the number of syllables, at their will.
265
The second one is using the retrieved simplified forms from the corpus and
building a weighted list of all reasonable structures. There are 53 different simpli-
fied forms in the corpus, and a table with the most common of them is presented
below. These are the simplified forms that appear at least 3 times.
Table 15.1: Most common forms in the corpus
Simplified Form #
AABB 23
AABC 7
ABAC DEDE 5
AABB CCDD 4
ABAB 4
ABCC 3
ABCD 3
These simplified forms can be implemented simply by letting the computer

randomly pick one according to its weight, and then subsequently transforming it
in order to obtain a full well-formed form. The first transformation is the decision
how many times the chosen simplified form is to be repeated, which can be done
by appending the list to itself the amount of times desired (usually no more than
four if one does not want an endless song): T1 (F ) → F ∪ F ∪ F . . .
The second transformation is to concatenate each symbol, of the resulting list,

with an integer corresponding to its variation. This integer is randomly picked
from another list. For the purpose of keeping the program within a reasonable
degree of simplicity, we are generating only 3 versions of each section which is
more than enough for demonstrative purposes:
T2 (F ) → ∀X ∈ F, X ⇒ Xn : n ∈ {0, 1, 2}.
However, technically one could have as many variations as there are unique
elements in the list to be transformed, so the general formula would be:
T2 (F ) → ∀X ∈ F, X ⇒ Xn : n ∈ {n ∈ N0 |0 ≤ n ≤ i : i = #⟨∀j, k ∈ F ⟩ : j ̸= k}.
The third transformation consists in appending the Intro section to the begin-
ning of the list, and the coda to the end, which is just a trivial list manipulation
operation:
T3 (F ) → ⟨Intro⟩ ∪ F ∪ ⟨coda⟩
266
15.1. First Module
The result could look something like this:
F = ⟨Intro, A2, A1, B1, B1, A0, A2, B1, B0, A0, A1, B1, B2, A0, A0, B2, B0, coda⟩
Notice that the variation indexes are random and do not follow any ordered
criterion; the point being that there is not a hierarchy concerning variations, since
they are all equally valid versions of a given material, therefore the users are
oblivious to which one is the presumed “original” and which one are the variations.
The third way to generate a well-formed form is formalizing an abstraction

that randomly generates a simplified form. A well-formed simplified form can be
implemented as:
1. Define the total number of sections as i, which is a positive integer: i ∈ N
2. Create the empty list F and the empty set S:F = ⟨⟩; S = {}
3. Define p as the length of the unique elements of F plus 1:

p = 1 + #⟨∀j, k ∈ F : j ̸= k⟩
4. Append the alphabetic character in position p to set S:

S = S ∪ {Xp } : X ∈ {A, B, C, D, . . . }
5. Append a randomly picked element from S to F : F = F ∪ ⟨sn ⟩ : s ∈ S
6. repeat 3 to 5 i times
This will generate a simplified form that can be subsequently transformed into
a well-formed form using the same procedures exemplified in the simplified form
derived from the corpus.
In the generator we implemented both solutions and one or the other is trig-
gered, depending on the coefficient of representativeness.
15.1.4 Section
Each section can be seen as a hierarchical class that has one relevant time-
domain property: length. This length can be one single value, but often it is a set
267
of them, each one corresponding to a bar unit, mirroring the typical divisions of
notated music. Therefore, its implementation takes the form of a list of lengths.
∀ #SectionXn = ⟨bar unit1 , bar unit2 , bar unit3 , . . . ⟩
This definition formalizes the whole-form of the song, encompassing its total
time span and constraining all the other time-dependent elements hierarchically
below this level.
The value these lengths assume depends, of course, on the song and style to be
modeled, and can be fixed or variable. They can be randomly generated or defined
by the user. Its implementation depends largely on the problem to be solved.
In the specific case of fado, and following the empirical data retrieved from the
first part of this dissertation, it is assumed that each section corresponds to four
measures, each one with a length of 2/4, and this has been implemented as such:
∀ #SectionXn = ⟨ 24 , 24 , 24 , 24 ⟩
Modifying these values could radically change the final results and the entire
perception of the practice.
In the software Symbolic Composer this can be implemented using the class
“Zone”, which is already predefined, and corresponds to the length of a section,
taking the form of a list of durations.
15.1.5 Example of a single section
At this point, and while we may be getting ahead of the formalization of the
other structural elements, we must stress that in terms of organization of the
narrative it is important to explain how the hierarchical structures were actually
implemented in code within Symbolic Composer. While the formalization of the
model allows it to be implemented in other languages and probably using other
kinds of nesting, the software Symbolic Composer that we have used to code the
generator has its own architecture of already defined objects and classes, simpli-
fying some of the low level operations. After the overall form is decided and one
has all the sections needed, it is possible to build a modular frame based on this
superstructure that can be used as a template for any other different superstruc-
ture in the future. We assumed each section as a basic module to support all other
268
15.1. First Module
parameters inside. Therefore, we will exemplify how we have programmed the gen-
erator explaining one possible way to program the first four bars of an archetypical
Ernesto Vieira’s fado as a concrete example. Then we will extrapolate to the more
general model. We will show how we have made a template defining one single
section and how it will be similar for all subsequent sections and variants.
Inside this template section we will explain the global parameters that will
apply to all instruments in this section. They are called “default” parameters and
in this case the parameters are “zone” and “tempo”.
Zone is defined as the total length of the section. In this case, we are assigning
four 2/4 bars for the A section – as extracted from the Ernesto Vieira’s model,
but it can be any length one wishes.
Tempo stands for at how many beats per minute the song will be played. One
can decide to go for a strict tempo, or let the computer decide it randomly within
a certain range acceptable for that section. Tempo is a list of values, each value
corresponding to a zone. If one has a list with four zones and a list with four
different tempos it will use a different tempo for each zone, but one can assign just
one tempo and it will play the same tempo within all the four zones.
Then we need to define the instruments. In this example we used just piano
for now, right hand, that we have called “voice” (it creates the melody) and left
hand that we have simply called “ostinato” (it creates the accompaniment), very
much like the scores from the database. One can create as many instruments as
one desires.
Inside each of these instruments we have defined our local parameters; this
means the parameters that only affect a particular instrument or voice:
Tonality, stands for the scales and chords that were used in this section. In this
work we have decided to base everything in non-accidentals scales and chords and
then use a variable as a transposition factor to contemplate the other possibilities.
Length is the rhythm of each voice. Basically it is a list of durations. Ideally

the sum of these durations should match the length of the previously defined zone.
If it is shorter, the rhythm will loop until that value is exhausted, if it is longer,
then it is truncated.
269
Symbol is a list of alphabetic characters and accidental symbols that will be

associated both with the chords and scales defined in tonality (and thus becoming
pitches) and with the durations defined in length, becoming notes. As with previ-
ous parameters the number of elements in the pitch list should match the number
of elements in the length list for ideal correspondence. In case of a mismatch, the
shorter list will be repeated and mapped onto the other until it exhausts all of its
elements.
Velocity is a list of velocities, ranging from 0 to 127, associated with each note
generated by the previous parameters. In order to obtain maximum correspon-
dence and efficiency, the number of elements of this list should also match the
number of elements both in symbol and length.
Program is the timbre of the instrument. Usually it is a value between 0-127

that corresponds to a list of the available sounds according to the synthesizers
available. In case of a generic MIDI set, the value 0 stands for a piano.
Channel is the MIDI channel associated with that particular instrument, rang-
ing from 1 to 16, and number 10 typically reserved for percussion.
270
15.1. First Module
So basically a template in Symbolic Composer, like a blank canvas for generating

one section is something that is coded like this:
;;section A
(def-section A
default
zone
tempo
voice
tonality
length
symbol
velocity
program
channel
ostinato
tonality
length
symbol
velocity
program
channel
)
And this template was the same for any section or sub-section we ever needed to
compose. We found it very reasonable, because it employs pretty much the same
parameters one would use to compose with pen and paper or with a notation
program.
After this stage, one just needs to know exactly what is desired inside each
section and generate the material for every possible parameter defined. We can
demonstrate for a hypothetical basic section A.
Inside section A, we were trying to generate our first four bars of music, meaning
we were in fact combining sub-sections a1 and a2 regarding some of the parameters.
Therefore the zone has to be a list that adds up to four bars of 2/4. Notice it
can be any length one desires, so it can be easily adapted to any other genre or
271
variation one desires in the future. Notice how a list is coded as simply a bracketed
set of numbers preceded by a ’.
’(2/4 2/4 2/4 2/4)
Then, there were many options for tempo. For the sake of simplicity we could
have simply pointed out to a single value like 78, as that is the average of the
database. Another option was to have the entire interval 72 to 84 available, there-
fore the best way was to tell the computer to generate a random number between
0 and 12 and add 72 to it. Because tempo needs to be a list of values (even when
it has only one value) we needed to tell the computer to make it a list.
(list (+ (rnd-in 12) 72))
Then we started by dealing with the instrument “voice”.
For the sake of simplicity in this example one is assuming a melody that is going
to be built on top of the archetypical T|D|D|T harmonic progression. Therefore,
the melody will be built on top of either a major or minor scale. And since it is
the right hand, to be in a comfortable realistic range it is best to center it around
octave 6 (in Symbolic Composer the first symbol of octave 6 is two octaves above
central C, midi pitch 60). So we have assigned:
(activate-tonality (0 major c 6))
Note that it can be any tonality or tonalities one wants in any range desired. It
can be even detuned tonalities. One can further explore this at one’s will or even
change the score later for any other scale one wants and see the differences.
The next step is to define the rhythm of the melody. There are at least two
approaches, the first is to assign a list of specific durations, either totally deter-
mined, or picked up from a set of different choices. The second approach is to
have a generative function that, based on algorithms, creates a list of suitable du-
rations from scratch. For this example we will just assign the Frederico de Freitas’
archetypical rhythm, however, other options will be explored further ahead:
272
15.1. First Module
’(1/8 1/8 1/16 1/8 1/16 1/8 1/8 -1/8 1/8 1/8 1/8
1/16 1/8 1/16 1/8 1/8 -1/8 1/8)
Notice how the durations fitting four entire measures are being defined, which
means the same two bar scheme repeated twice. Notice also the negative values,
which represent pauses. Finally, notice as well that the way the structure was
built the pick-up note is missing because it pertains to a previous section (an
introduction for instance), and how the last notes following the pause are actually
pick-up notes for the following sections.
The following step was to generate the pitches of the melody, so symbols were
needed. Each symbol is basically a letter that represents an element of the scale
previously activated. In this case, the scale was C major. Then the first symbol
“a” represents the first element of that scale that is the pitch class c. The second
symbol “b” represents the second element of that scale that is the pitch class d.
Negative symbols work the same way in the opposite direction.
There are countless ways to generate a list of symbols. One can type them if
specific notes are desired, or they can be generated more or less randomly. For
this example we have used a random generation around the center of the scale:
(find-minimal (vector-to-symbol -d d
(gen-noise-Brownian 4 my-random-seed 0.5)))
Basically we used a mathematical function that comes with the program “gen-
noise-Brownian” (there are many others) that generates a random walk, a singable
contour, based on Brownian noise. Using another pre-set function “vector-to-
symbol” we mapped this numeric contour into symbols within the range “-d” up
to “d” (so around the center of our scale that would be “a”), and used another pre-
set function called “find-minimal” to remove all repetitions while still preserving
the general shape of the curve. Running this set of nested functions once, we have
obtained the following result:
--> (a -c -b -d -c -b d c d b c a b a)
273
This is a perfectly singable undulating line. One can try out with different
values, seeds and nested combinations and see what happens. One can try out
some of the other mathematical functions that come with the program, like “gen-
pink-noise”, etc. More possibilities and the ones actually used in the generator
are explored further ahead.
Then we just needed to define the velocities. A velocity is a number between 0

and 127 that represents how loud a note sounds. We just wanted something very
stable and smooth. So we have opted for another kind of randomization, using
again a Brownian-noise generator:
(vector-round 64 100 (gen-noise-Brownian 5 0.75 0.75))
Again, there are countless ways of doing this. This is just one of many. One
can play with these values and see what happens. We were just happy with these
ones after trial and error. The generated values will be all between 64 and 100,
so, all in the range of mezzo-forte. A more sophisticated algorithm for dynamics
used in the generator will be explored in its own section.
Then, keeping the example simple, we assigned the default instrument piano to
the voice, which is program 1. We told it to play on channel 1 as well. One can
assign any number between 0-127 for instrument – 74, for instance is a general
MIDI flute.
program 1
channel 1
And this concludes the definition of the first four bars of melody. As one can
see it is pretty much free within some constrains: a melody with a tonal sound
around a range of one octave around c6 and with Frederico de Freitas’ archetypical
rhythm will be obtained.
Then we needed to define an ostinato to go along with it. This time we have
decided exactly what we wanted. We just defined the same zone so they will
match. The tonalities activated, since we needed the T|D|D|T movement in the
major mode, were the appropriate chords – C major and G7 major – complemented
with the appropriate voice leading inversions provided. Therefore:
274
15.1. First Module
(activate-tonality (2 ch-maj c 4) (0 ch-7 g 4)

(0 ch-7 g 4) (2 ch-maj c 4))
We have centered the ostinato around octave 4 so it does not clash with the
melody and it is in the appropriate lower register. We have provided a defined
length, since it is a stock march/fox rhythmic pattern. This list represents the
durations as we would write them on paper or a notation program. Notice how
we just need to write out two bars of the ostinato, even if the zone is four bars
long, since it is assumed that, in case where the list is shorter, it will loop until it
exhausts the time.
’(1/8 1/8 1/8 1/8 1/8 1/16 1/16 1/8 1/8)
The same is true also for the symbols, which in this case are all pre-determined.
’((-12 a) bcd -b bcd -c bcd bcd -b bcd)
Notice the logic of construction around the symbols – the “-12” attached to the
first symbol indicates to play it twelve semi-tones below, forcing it to be a bass
note, as expected in this kind of ostinato. The cluster “bcd” means that these
three symbols are played as simultaneously as possible.
We kept the same solution for the velocities as we did in the melody: let them
be somewhat random.
(vector-round 64 100 (gen-noise-Brownian 5 0.75 0.75))
For the program we kept the piano but we wanted this hand to be played on
another channel. So we assigned it to channel 2.
program 1
channel 2
275
So basically the coding of a four bar long section A is finished. It looks like
this:
(def-section A
default
zone ’(2/4 2/4 2/4 2/4)
tempo (list (+ (rnd-in 12) 72))
voice
tonality (activate-tonality (0 major c 6))
length ’(1/8 1/8 1/16 1/8 1/16 1/8 1/8 -1/8 1/8 1/8
1/8 1/16 1/8 1/16 1/8 1/8 -1/8 1/8)
symbol (find-minimal (vector-to-symbol -d d
(gen-noise-Brownian 4 my-random-seed 0.5)))
velocity (vector-round 64 100
(gen-noise-Brownian 5 0.75 0.75))
program 1
channel 1
\emph{ostinato}
tonality (activate-tonality (2 ch-maj c 4) (0 ch-7 g 4)
(0 ch-7 g 4) (2 ch-maj c 4))
length ’(1/8 1/8 1/8 1/8 1/8 1/16 1/16 1/8 1/8)
symbol ’((-12 a) bcd -b bcd -c bcd bcd -b bcd)
velocity (vector-round 64 100
(gen-noise-Brownian 5 0.75 0.75))
program 1
channel 2
)
As one can easily understand if we desired to proceed and define a section B,

all we had to do was to copy-paste the same template and change the desired
parameters, and repeat the process as many times as sections needed in order to
complete the piece.
An efficient way to automate this process is to consider this example as an

archetype and to define an abstraction that generates as many sections desired,
from scratch, just changing the relevant parameters.
Having presented a way of implementing sections, nested inside the superstruc-

ture of form, one can now understand the hierarchical dimension of the model.
276
15.2. Second Module
While it has a vertical dimension of six modules representing all the agents in-
volved, in their orders of dependency, one also understands that it has a horizontal
dimension of one section at a time, and how each section seems to be independent
from each other. And the implementation of the following modules consists in
generating the materials for these agents in a modular fashion. Generating the
material for a hypothetical section A, and then repeating the entire process for a
section B, and again for a section C and so on.
15.2 Second Module
In this section we will detail how we have formalized the harmonic layer and
which assumptions and algorithms were used.
In order to formalize any accompaniment or melodic process, the harmonic

environment has to be defined, and represented in symbolic terms. Fado musics
are tonal and, as such, tonality has to be modeled. That kind of formal exercise
has already been done several times, namely in the works of Elaine Chew, who
presented a mathematical model for tonality and its description in formal terms
(Chew, 2000a,b, 2001). We have also drawn inspiration from the work of Martin
Rohrmeier (Rohrmeier, 2011), who proposed a generative syntax of tonal harmony.
The predefined objects and classes in the software that we have used shaped how
the problem was formalized, in the sense that we were building our reasoning on top
of them. Tonality, in Symbolic Composer, is a class defined as a foundational set
of audible frequencies: T onalityx = {f req1 , f req2 , f req3 , . . . , f reqn }. Therefore, a
tonality is often equated to a scale from which the pitches will be derived. These
sounds can be virtually any, and are not constrained by any boundaries. Users
can literally build artificial sonic worlds by defining their own tonalities. While it
is possible to define a tonality in terms like
T onalityexample = {245 hz, 277 hz, 389 hz, 411 hz, 447 hz}
most of the time this is not convenient. Alternatively, by using a morphism

it is possible to map frequencies into a pair of symbols such that the first one is
an alphabetic character corresponding to a pitch and the second one is an integer
corresponding to an octave. An octave establishes a principle of equivalence such
as an increase of a unit corresponds to the doubling of the frequency. And within
277
an octave there is room for twelve equidistant different pitches. The 128 available
MIDI pitches are defined this way, with a reference frequency of A4 = 440 Hz. At
present, this convention is widely used and there is no need for further exploration
and formalization in the scope of this dissertation. It suffices to say that a tonality
can be also defined, using this convention equivalence, in terms of a scale, such as
T onalityexample = {C3, D#3, E3, F #3, G3, A#3, B3}.
In the particular case of fado, as observed in the practice and from the examples
of the corpus, there are four tonalities involved and their possible transpositions
within the twelve tone system. They correspond to the major mode (Ionian) and
the three minor modes, since fado is a part of a tonal tradition in the Western
sense of the term. Therefore, within the program, the tonality has been formalized
as such scales.
T onalityM ajor = {C3, D3, E3, F 3, G3, A3, B3}
T onalityM inor = {C3, D3, Eb3, F 3, G3, Ab3, Bb3}
T onalityHarmonic = {C3, D3, Eb3, F 3, G3, Ab3, B3}
T onalityM elodic = {C3, D3, Eb3, F 3, G3, A3, B3}
Two modifier functions were also considered to account for all possible trans-
positions, and to allow the definition of either local or global modulations.
f : ∀n ∈ T onalityx → n = (n + t + m) : {t, m ∈ Z|[−6, 6]}
In this way, the sonic domain for fado is defined and constrained in a modular
fashion. It could be argued that not all transpositions have the same weight,
and indeed statistics were presented accounting for that fact, and, at least within
the corpus there are some degrees that are never present. Therefore the value
of t, formalized above, could be further constrained to reflect these weights. At
present we do not feel that need, since in the current practice, and among skilled
performers, transpositions are often chosen based on the range of the singer and
do not reflect the original tonality the composer has chosen. Moreover, the lack
of use among some tonalities often reflected the mere inability of amateurs to
perform them, it did not reflect any intrinsic aesthetic decision regarding the
musical content. Since these works are to be performed by the computer, and the
computer has no difficulty in performing them in whatever transposition assigned,
then it makes no sense to artificially constrain them. The modifier m, however, is
278
15.2. Second Module
constrained, and only makes sense in very specific cases. Not all modulations occur
in the practice, and often this modifier is contextually attached to the harmonic
progressions used, namely to generate secondary dominants, which is something
to be developed further ahead.
While it is now possible to derive all possible pitches for the harmonies and
melodies from this foundational base, it is clear that this structure, by itself, is
a huge determinant in shaping the final sounds and in the characterization of
the practice. It seems also obvious that a simple change in these definitions can
radically alter the final result. After some experimentations we dare to say that
entirely different practices and genres might be achieved simply by changing this
parameter alone.
The problem of automatic harmonization is now considered a well-defined prob-

lem, especially when seen from the typical viewpoint of generating logical, yet
inventive, chords to accompany a given melody. One of the most recent papers
by Pachet and Roy has shown that “standard Markov models, using various ver-
tical viewpoints, are not adapted for such a task, because the problem is basi-
cally over constrained” (Pachet & Roy, 2014). Their proposed solution consists
in implementing their technique of Markov constraints to fioriture (melodic orna-
mentation, defined as random walks with unary constraints). The end results are
claimed to be more effective and musically interesting.
While these approaches seem promising, they all rely on the idea that there is
a base melody per se to be harmonized. However, in the case of fado, what one
usually sees in performance is the reverse: there is a prior harmonic structure,
on top of which the text is sung. So, while the rhythm of the melody and its
general contour derive from constraints emerging from the lyrical structure, the
pitches are mostly derived from the scalar structure of the underlying harmonic
progression. Hence, one needs another kind of approach to solve this problem.
If the overall form is a foundation to structure the span of time and the internal
recurrences of patterns inside a fado, the harmonic foundation reflected in the
scalar structures, is the foundation of each section of fado regarding the way the
pitches behave. By choosing a scale, derived from the tonalities assigned, one is
able to derive pitches that will be coherent with the melody, accompaniment and
counter-melodies, within context, and will make the existence of the piece possible.
279
15.2.1 Harmonic Progressions
Based on the tables previously obtained, and detailed in the first part of this
dissertation, we made a list containing all the relevant harmonic progressions found
in the musical scores.
A chord is formalized as an object corresponding to a set of symbols that is

defined using four attributes (or modifiers): quality, root, octave and inversion.
Chx = ⟨Q, r, o, i⟩
The quality is a set that specifies the amount and relation between the symbols
that compose the chord. A major chord is predefined as a triad in which the
interval between the root and the second symbol is of four semi-tones, while the
interval between the root and third symbol is of seven semi-tones, for instance:
Qmaj = {0, 4, 7}; Qmin = {0, 3, 7}; Q7 = {0, 4, 7, 10}, etc. . .
There are numerous possible qualities for chords (diminished, augmented, ex-
tensions. . . ) and all may be formalized in the same fashion.
The root of a chord is a modifier that assumes the form of an alphabetic symbol
relative to the pitch of the defined tonality that is active at the time the chord is
called. Hence, the root “a” corresponds to the first pitch of the scale defined (in
case the tonality was defined as a scale) or to the first frequency defined (in case
the tonality was defined as a list of frequencies). The root “b” corresponds to the
second, “c” to the third and so on. Negative values work in the opposite direction.
The octave of a chord is also another modifier, which is an integer corresponding

to the octave in which the root is to be placed. Finally, the inversion is yet another
modifier, corresponding to an integer, which will determine which element from
the ones available in the quality set will correspond to the first note.
A harmonic progression can be thus formalized as a list of chords.
HPx = ⟨Ch1 , Ch2 , Ch3 , Ch4 , . . . ⟩
As an example, the formalization of a typical T|D|D|T progression can be done

as:
280
15.2. Second Module
HPT |D|D|T = ⟨⟨{0, 4, 7}, c, 3, 2⟩, ⟨{0, 4, 7, 10}, g, 3, 0⟩,

⟨{0, 4, 7, 10}, g, 3, 0⟩, ⟨{0, 4, 7}, c, 3, 2⟩⟩
This specific case reflects the formalization of a progression of I6/4 |V7|V7|I6/4 ,

in the third octave. In this example, the tonic chord is presented in the second
inversion, so that its bass note is a common note with the following dominant.
Most of these modifiers, instead of being fixed, can be variables. The modifier
“inversion” can be further modified by an offset variable, to increase variability.
As an example attaching a fixed offset to every inversion in each chord, could be
easily done as such:
HPT |D|D|T = ⟨⟨{0, 4, 7}, c, 3, (2 + i)⟩, ⟨{0, 4, 7, 10}, g, 3, i⟩,

⟨{0, 4, 7, 10}, g, 3, i⟩, ⟨{0, 4, 7}, c, 3, (2 + i)⟩⟩ : i ∈ Z
Defining the variable i as “+1”, would cause all chords to be offset by one
inversion, so that the tonic chord would be presented in “third inversion”. In this
case, that would be the fundamental state again (one octave higher though), while
the dominant chord would go for the first inversion, which would still create a very
acceptable voice leading context with minimum finger movement.
The implementation of the harmonic progressions, as seen in the corpus, can be

done as weighted lists. To optimize resources, all possible progressions are defined
as a sequence, in which the position in the list identifies the progression.
P ossible HP s = {HP1 , HP2 , HP3 , HP4 , . . . }
Each defined progression has scales associated with it, reflecting both the tonal-
ity that is active and a modifier offset for local modulations (concrete values for
the parameter m). For instance, if in a given progression there is a secondary
dominant, this means that this progression involves two different versions of a
given scale, therefore the modifier has to change accordingly. If a progression has
borrowed chords from the relative, then it involves two different scales. As such,
parallel to the sequence of possible harmonic progressions a sequence of lists of
scales is also defined, respecting the same order, so that their position matches.
Scalex = ⟨T onalityx , t, m⟩
281
Sx = ⟨Scale1 , Scale2 , Scale3 , . . . ⟩
P ossible Scales = {S1 , S2 , S3 , S4 , . . . }
Then, further sets for each section are built containing the weights, in the same
order the progressions and scales were defined.
W eights f or Section Xn = {w1 , w2 , w3 , w4 , . . . }
As such, when one wants to assign a progression and a scale for a section Xn
there is a probability wn , constraining both the choice of HPn and Sn simultane-
ously.
∀ wn ∃ HPn , Sn : (HPn , Sn ) ⇒ Xn
Implementing the progressions in this modular fashion means one only has to
define them once, and can alter them at any stage. It is also easy to change the
weights among each section, alter them, nullify some of its elements (meaning
that in any given sections, some progressions will never be picked up, even when
they are defined) or force them. It also means one can actually make them vary
with time, if one decides to define the list of weights as variables instead of fixed
values. The same can be said about the progressions and scales themselves, since
they are defined as symbolic data. In this way, instead of needing to create a
complex generative function to act as a second artificial intelligence to deal with
the coefficient of representativeness, we just had to make a second set of weights
with a more unpredictable behavior. This way, the program might either pick up
progressions and scales from the set of weights that fully mirror the corpus (as
shown in the tables, in the first part of this dissertation), or pick up progressions
and scales drawing values from the unpredictable set of weights, hence creating
more adventurous musical results.
The option to use weighted lists, instead of other methods, relies on the fact
that we believe this best represents the process that actually happens in the prac-
tice. People do not generate progressions out of thin air, neither do they seem
to generate them in a purely sequential form, Markov chain or random-walk like
fashion. Instead, they seem to rely on previous implicit knowledge, in a kind
of mental scheme of already known progressions that work within the tradition.
This is done either by copying the records, other musicians or by resorting to other
songs they already know. It is not by chance that harmonic progressions in the
musics of fado also depend on the education of the musicians. Schooled ones like
282
15.2. Second Module
Reynaldo Varela, Alain Oulman or Jorge Fernando often employ more complex
progressions, since they are borrowing them from other traditions, namely from
the erudite music or jazz universes. Therefore, by using weighted lists this previ-
ous knowledge of which progressions work best and which not (also according to
their temporal context within the repertoire), its distribution among the overall
population, and the choice to pick them, is modeled.
Symbolic Composer already has objects defined to predict all the possible chords
used in tonal music and it is possible to define custom chords as well, therefore this
template can be modified or extended at one’s will in case one wants to model other
practices or refine this one. Each chord of the progression was defined according
to a specific inversion in order to respect voice leading rules and logical transitions
between chords. In order to not constrain the progressions to just one possible
inversion, each chord was also assigned to an off-set variable to be picked up at
random at each generation. This offset variable allows that, while keeping the
transitions consistent (and thus the voice leading), one can still obtain all possible
inversions in each progression.
Using the formalization described above we are presenting a bit of code as an

example of the implementation of three different harmonic progressions. Notice
that the syntax of Symbolic Composer uses the modifier inversion in the first place,
and only then, quality, root and octave.
;;0-I-V7-V7-I
((,(+ i 2) ch-maj c 3) (,(+ i 0) ch-7 g 3)
(,(+ i 0) ch-7 g 3) (,(+ i 2) ch-maj c 3))
;;1-I-ii-V7-I
((,(+ i 2) ch-maj c 3) (,(+ i 1) ch-m d 3)
(,(+ i 0) ch-7 g 3) (,(+ i 2) ch-maj c 3))
;;2-I-IV-V7-I
((,(+ i 2) ch-maj c 3) (,(+ i 1) ch-maj f 3)
(,(+ i 0) ch-7 g 3) (,(+ i 2) ch-maj c 3))
Alongside with this list of all possible progressions, we coded a way to call them
in a weighted form according to the section where they will be used. In this case,
as an example:
283
;;---WEIGHTS FOR PROGRESSIONS OF SECTION "A"

(setq iProgA
(first
(gn-weight 1 ’(
(0 36) (1 3) (2 1) (3 0) (4 1) (5 0) (6 0) (7 0) (8 25)
(9 1) (10 5) (11 0) (12 2) (13 0) (14 0) (15 0)
))))
;;---WEIGHTS FOR PROGRESSIONS OF SECTION "B"

(setq iProgB
(first
(gn-weight 1 ’(
(0 15) (1 5) (2 2) (3 3) (4 4) (5 6) (6 3) (7 1) (8 17)
(9 2) (10 2) (11 1) (12 1) (13 2) (14 3) (15 1)
))))
The variable “iProgA” represents the element of the list of progressions to be

picked up, to be later used as a harmonic progression in section A. In this case,
one can see that the element 0 has a 36 factor associated with it, followed by the
element 8 with a factor 25, and they are the elements most likely to be picked
up. In the progressions list, the element 0, as shown before, corresponds to the
progression T|D|D|T in the major mode, while the element 8 corresponds to the
progression t|D|D|t in the minor mode. These are the factors obtained in the
tables, and, as one can see, they are very easy to change. The way we have
built the functions separate means that we can have just one large list to have all
possible progressions and we have smaller lists to carry the weights for all different
sections. This allows us to also refine each of these modules and, for instance, to
code a function to generate new progressions from scratch and add them to the
list of another abstraction able to generate weights from scratch.
284
15.2. Second Module
15.2.2 Ostinati
The analysis of the corpus has shown that the accompaniment layer in fados is
mainly comprised of ostinati: a stock rhythmic pattern that repeats itself over and
over during a certain period. Sometimes it goes along the entire fado, but other
times it just spans through some sections, having other contrasting ostinati in the
other sections. In order to mimic this layer, we have studied the kind of ostinati
that appear in the corpus. Most of these ostinati are recurrent figurations vastly
used in other performative traditions and already have designations and categories
by which they are known and there are already some established rules regarding
the way players should approach them. One can find marches, alberti bass like
figurations, habaneras, popular marches, and hybrid versions between them, as
already detailed in the first part of this dissertation.
The first approach regarding this problem was to make a weighted list of all the
relevant patterns that appear in the corpus and then let the computer pick one,
because this models what happens in the actual practice: players just chose one
figuration from a previously known group. The formalization of these stock pat-
terns was done by literally expressing the archetypical stock figurations as music
fragments, as displayed in the first part of this dissertation. In order to facilitate
the process each ostinato was broken in two parts, representing what happens
in the real practice: a bass layer and a harmonic chordal layer. Often, the bass
and chordal layers are played by the same instrument and they do not overlap
each other, but other times they can be separated, especially when there is a bass
instrument present. Moreover, the chordal layer usually has notes deriving from
the chords while the bass layer may present additional scalar material. Both lay-
ers were expressed as lists containing the explicit durations and symbols of the
patterns in their atomic form. The atomic form is the minimum fragment that
needs to be provided so it can be looped in order to generate the complete ac-
companiment. In order to be looped correctly, the total duration of this fragment
should match the total duration of a section or at least be divisible by it without a
remainder. In the case of fado, since the sections are four binary bars long, equiv-
alent to eight quarter notes, this implies fragments with the length equivalent to
two, four, or eight quarter notes. Both the durations and symbols are represented
in lists in which the position matters, since it is this linearity and correspondence
that allows each value to be matched with the others at a later stage.
285
Ostinato Chord Durationsx = ⟨dur1 , dur2 , dur3 , dur4 , . . . ⟩,

m(#Ostinato Chord Durationsx ) = #Xn : m ∈ N
Ostinato Chord Symbolsx = ⟨Ch1 , Ch2 , Ch3 , Ch4 , . . . ⟩ : Chn ∈ HPn ⇐ Xn
Ostinato Bass Durationsx = ⟨dur1 , dur2 , dur3 , dur4 , . . . ⟩,

m(#Ostinato Bass Durationsx ) = #Xn : m ∈ N
Ostinato Bass Symbolsx = ⟨s1 , s2 , s3 , s4 , . . . ⟩ : sn ∈ Sn ⇐ Xn
These lists comprise the rhythms and symbols, abstracted from these patterns.
They have to be mapped into any given section Xn, so they will retrieve the
information regarding which harmonic progressions and scales are active at that
point, and that will allow the symbols to be relatively converted into suitable
pitches.
As suggested by Professor Bruce Pennycook, we have also programmed a func-

tion that would, somewhat randomly, generate pauses within the ostinati, and
so approaching a live performance and the freedom the players have when they
actually perform them. One of the main concerns in increasing the realism of this
function was to make sure the pauses would neither happen in the strong beats of
each section, nor be so many that they would prevent the listener from actually
recognizing the ostinati as such. This approach concerning the accompaniment
has the advantage of being totally representative of the corpus and coherent within
the tradition. It also handles very well the problems of voice leading, since by hav-
ing concrete lists of symbols to pitches, usually one automatically assigns the most
commonly used chord inversions at each stage, which actually are more natural
to the way a keyboard player would approach them. The model of Toivanen et
al, creates accompaniments in many styles, like alberti bass and many other chord
patterns, which, we believe, can be similar to our own. They have used a model
of voice leading that chooses chord inversions based on a minimal total movement
(smallest sum of intervals of simultaneous voices) (Toivanen et al., 2013).
Our second approach to this issue was to actually build a second artificial
intelligence, triggered according to the coefficient of representativeness, consist-
ing of some functions that would generate ostinati and rhythmic patterns from
scratch. The symbols derived from the harmonic progression were then assigned
286
15.3. Third Module
to those durations and new and inventive ostinati were obtained. This approach
has the advantage of generating out-of-the-box rhythmic patterns and expands
the possibilities for diversity and for unexpected results making the outcome more
interesting. However, this approach has also the problem that it might generate
patterns that lack coherence regarding voice leading, the way a keyboard player
usually performs them. It also presents the problem that, even within the con-
straints we have designed, it may generate ostinati that might not be recognizable
within the tradition, therefore jeopardizing the common expectations of what a
fado should sound like.
We believe the weighted combination of these two approaches is the best solu-
tion in order to generate convincing ostinati for the fados, but one may keep the
door open for having some sections with original accompaniment, something that
makes sense when one thinks of a live performance with inventive players.
15.3 Third Module
In this section we will detail how we have coded the melody and which assump-
tions and algorithms were used.
There are several ways to approach the problem of generating a melody. Ac-
cording to previous research on the topic, in most cases the authors adopt a mod-
ular strategy in which the main focus is on two parameters: rhythm and pitch.
Usually the first task is to generate a rhythm, a pattern, and then assign pitches
to that same rhythm. Since the melody in fado is intimately connected with lyrics
and prosody, our main references were the state of the art references dealing with
the generation of melodies in the context of songs. Therefore, we studied the ap-
proach proposed by a team of Finnish researchers who developed the prototype
M.U. Sicus-Apparatus (Toivanen et al., 2013). This program generates art songs,
writing lyrics first and then composing music to match the lyrics. The idea of mu-
sic composition in order to match a previous lyrical content is of crucial interest
to our own project, for obvious reasons. The program consists of two modules:
one that writes the lyrics based on user input of a theme and a second one that
receives an emotional valence and composes the music. In order to generate the
melody, the program first generates a rhythm based on the prosody of the already
written lyrics. It breaks down each word into syllables and then assigns a rhythmic
287
element with as many notes as there are syllables. Those rhythmic elements come
randomly from a set of commonly found patterns in art songs. Then, the program
randomly chooses a scale coherent with the emotional valence previously provided
and generates a harmonic progression based on a second-order Markov chain with
common progressions found in diatonic Western classical music. It also assigns
time values, in a probabilistic manner, for each chord generated so that the overall
length of the accompaniment structure will be consistent with the rhythm of the
melody. After this stage, pitches are assigned to the rhythm of the melody. The
underlying harmony defines a probability distribution for pitches which can be
used, and the subsequent note is found by the means of a random-walk favoring
small intervals (Toivanen et al., 2013, p. 89).
Another approach to melody generation and edition was the one presented by
Young-Woo Jeon et al. using a noise function, specifically Perlin noise, because
it “is bounded, band-limited, non-periodic, stationary, and isotropic” (Jeon et al.,
2006, p. 164). These traits would make the noise function ideal because, while
it preserves a decent amount of randomness, it still is controllable and presents
somewhat predictable results to the user. Mainly, Jeon et al. present the idea
that Perlin noise can generate a melody from scratch (if mapped into pitches)
or be mapped into an already existing melody to change some of its elements
while still preserving its original shape. This concept can also be extended to any
other parameter, like timing information, tempo or dynamics. The smoothness
of the curve that noise can produce, under certain values, is ideal for altering
any parameter originating what one would call a “humanizer”, because in the
final outcome it produces small deviations similar to the ones a human would
introduce when performing a certain task, since human behavior is slightly erratic
and imperfect. Jeon et al, also present the idea that after the noise is applied to
anything it can then be further processed to revise some things that might get
off-place or that the user desires to be constrained, for instance, certain melody
notes on key points to be determined instead of purely random. They conclude
that
“noise function makes the modification of an existing melody easy

and fast. The modified melody shares the basic form of the original
melody, but the details are randomly different. The extreme changes
that can be caused by white noise are not found in the modification
because of the ideal band-limited feature of the [perlin] noise function.
288
15.3. Third Module
By noise editing, it is possible to control the composition or modifica-

tion process precisely to reflect a composer’s intent in the work” (Jeon
et al., 2006, pp. 167-168).
In our own explorations with the software Symbolic Composer we have noticed
how several functions based on noise already came predefined and ready to use,
namely generators of white, pink, red or Brownian noise. Also, another set of
generators based on other types of curves and waves like triangular, for instance,
or even real time readings of astronomic patterns and walks like solar wind or
flux particles. We consider that the same ideas Jeon et al. present in their paper
are suitable for our own use; it is just a matter of, again, picking up the suitable
generator and deciding its seed and steepness and one is able to achieve desirable
results. We have found that Brownian noise and its graphic curves are suitable
for melodic generation the way they, themselves, generate contours similar to the
ones a human voice can do. A number of combinations of more than one noise are
able to generate virtually all the melodic contour archetypes described by Huron
and already surveyed in the first part of this dissertation (Huron, 1996, 2008),
and thus be successfully applied in algorithms with the intent of either melody
generation or creating melody variation on previous melodies.
Also, we have pushed further into noise explorations to also make them suitable
for creating variations or simple humanizations on several parameters like rhythms,
tempo, dynamics, etc.
Our approach to the problem of melody generation was to model what we have
found in the data analyzed and to try to replicate those results. Basically, when
one is composing a melody for a fado, one is really setting up a lyrical line on
top of a harmonic progression set over the span of two measures. The first task
was to decide how to divide and break the total span into smaller figures, thus
creating a rhythm. And, according to the data, nearly almost any combination
is suitable, there is really no clear pattern on what might happen. Surely, the
semiographic schemes tend to favor regular and simple divisions, like beats with
quarter notes, a group of two eight notes, or the syncopated figure of two sixteenth
notes with an eight note in between them. However, as already seen and discussed
in the “rhythm of the melody” section, triplets are frequent and dotted figures
and other combinations also arise. Furthermore, even when they are not notated,
they eventually show up later in the performance and can effectively be heard.
289
The conclusion is that one can expect virtually any subdivision of a measure into
smaller figures as long as it consists of a reasonable number corresponding roughly
to the number of syllables of the lyrical line. And even that is just a rough
estimate, because sometimes a singer can prolong the same syllable over two or
more different notes. Therefore, taking into account the typical heptasyllabic
verse, it would make sense to hear any subdivision consisting of seven to ten
notes. Since there is no concrete textual line to which one is referring, there is no
prosodic reference to be coherent with in terms of accents, for instance. Therefore,
the issue is not so much the apparent chaotic randomness of such freedom, but
rather the consistency of the recurrence of this same pattern once generated. This
is so because what the lines have in common is their metric regularity. As such,
even if one does not know exactly how a particular rhythm was generated, because
of the lack of a concrete text, one knows that the quantized version of such rhythm
has to be repeated over and over, simulating a consistent stanza.
Having this in mind, our first thought was to initially have actual weighted list
of the most common rhythms of the melodies used in the corpus. Having these
lists and weights, the artificial intelligence would simply pick up a list comprising
a rhythm that we knew made sense because of it being used successfully in the
past. This approach, however, relies on previous material and, of course, lacks
inventiveness. Moreover, contrary to what happens with harmonic progressions
and ostinati, such an approach actually does not reflect the practice. In a good
practice fadistas style their melodies on the spot, this being one of the most
defining traits, therefore using quantized versions of past performances is not an
elegant solution. This solution would eventually model what some recent amateur
performers do when they literally imitate the records. However, this way of doing
things, regarding this specific parameter, is not considered a good practice among
the members of the community. Hence, we took another approach and decided to
code an abstraction that would generate an acceptable list from scratch, trying to
model the behavior that the rhythm of the melodies is generated on the spot.
This particular list is important because its quantized version represents a base
pattern that will give consistency to the entire work in a cross-sectional fashion. As
discussed before, the rhythm of the melody is important because of its constant
recurrence with small variation (the non-quantized versions), within the same
work. Therefore, after we have generated it once, we have a structural foundation
290
15.3. Third Module
that we can use to build upon or further manipulate in order to create variations,
and in this particular case, extending beyond one simple section.
In order to efficiently generate the rhythm of one line, we broke down each line
into its parts. As one can observe from the semiographic scheme from Frederico
de Freitas (for instance), a line comprises pick-up notes, followed by what we will
call “rhythm head” (the first measure), and then a “rhythm-tail”, corresponding
to the second measure. Typically there are less figures and larger durations in the
“rhythm-tail” in order to end the line. The second beat of the second bar is, in
fact, the pick-up of the following bar:
Rhythmline = ⟨⟨Rhythmpick−up ⟩, ⟨Rhythmhead ⟩, ⟨Rhythmtail ⟩⟩.
Each one of those rhythms can be generated by slicing the available length by
a random integer. Usually, smaller numbers are preferred, since one is usually
dealing with already short durations. Two, three, four and eight will yield more
realistic results, while five, six and seven generate complex rhythms. Above nine
things can become surreal. The second step is picking up those slices and con-
catenating some of them into larger units in order to obtain durations of unequal
length. Obviously, the sum of the durations of all resulting slices remains equal to
the original available length.
! !
Rhythmn = {[#(Rhythmn )/j] × ki } : ( ki ) = j; j, k ∈ N
This process of slicing lengths and redistributing the sliced durations can be
refined if applied to smaller hierarchical units when desired. For instance, since
“rhythm-head” represents an entire bar, which in the case of fado comprises two
beats, it can make sense to actually apply the process separately to each beat,
instead of the whole bar. Also, no matter which integer j has been used to slice
the “rhythm-head”, it makes sense for it to be a smaller value when applied to
“rhythm-tail” so it generates relatively longer durations, and one actually obtains
a tail effect.
Once all those sub-rhythms have been generated and joined into a single list
comprising what would be the rhythm of a line, one might notice that all durations
are positive values and represent notes. This is not realistic, since every line has
to contain a few pauses. One of them, in the case of fado, is always determined
and occurs in the beginning of the second beat of the second bar, corresponding
to the end of the tail and the beginning of the pick-up notes. Therefore, for the
291
sake of simplicity, the pick-up rhythm is implemented as a quarter-note duration

slot, including an eight-note pause, followed by the actual pick-up note or notes.
Experience taught us that in most of the cases it is just a single eight-note, and
seldom other durations. A small weighted list suffices in this particular case. All
other rhythms might include additional pauses that can be implemented using a
random function to switch any of the available durations to a negative value.
In order to code such implementation in Symbolic Composer, we started by

coding the “rhythm head”. We have simply coded a small function that slices each
beat into smaller units (even or unevenly for more realism), and assemble those
durations onto a list. The function has weighted optional parameters that can lead
to favor even divisions or larger durations on the odd beats, for instance. Since
the resulting list is always a list of positive durations, a second function “length-
syncopate” filters its content having the possibility of creating some pauses.
;;rhythm for head in section A
(setq rhythm-A-head
(mapcar ’get-ratio
(length-syncopate my-random-seed 5 1
(flatten (rhythm-ornament 4 8 ’(1/4 1/4)
:seed my-random-seed :pcnt 100 :to-pcnt 70)
)
)
))
Basically, we are asking the computer to divide each 1/4 duration 4 up to

8 times. We are also telling it to only divide 70 to 100 percent of these dura-
tions. Note the variable “my-random-seed” there. Everything is customizable
and changeable in order to be extremely flexible and to easily allow modifications.
Then, we coded the “rhythm tail”, a second function similar to the previous one,
but with changed seed and parameters in order to slice only one beat into smaller
units (but not so small) appended to the outcome of a weighted list representing
possible pick-up notes. These two lists are then appended and flattened. Every
time one runs this function one obtains a list comprising the durations of the
rhythm of one line.
292
15.3. Third Module
Once obtained the rhythm of the melody of one line (a list of durations and
pauses adding up to two 2/4 bars, in fact): RhythmXn = ⟨d1 , d2 , d3 , . . . , dn ⟩, the
next step is to assign pitches to each one of those durations. Again, there are
several ways to approach this problem. The most obvious and consistent way of
assigning pitches to durations is to rely on the underlying harmonic progression.
The harmonic progressions for the ostinati are already defined in the second mod-
ule, so the most logical step would be to assign pitches derived from the chords
used in each beat by means of a weighted function. By doing this one obtains, for
sure, a melody that will sound consonant, however, lacking fluidity and contour.
The second approach is to apply only this principle to hierarchical relevant dura-
tions (those which stand in the first tick of each beat, for instance) and then fill
in the gaps with other notes from the scale, but not necessarily from the chord.
These notes will behave as passing tones. The problem with this approach is that
it might simply generate non-realistic intervals and strange contours between the
notes, so one has to refine this reasoning through the means of that constraint:
use tones from the scale but that are stepwise apart from the previous chord tone
assigned in a random-walk fashion.
Another totally different approach is to simply rely on templates and formulas.

Based on the data collected from the corpus, there are already a series of melodies
known to work, that sound good and realistic and that, in fact, represent the
corpus. So it is possible to obtain a supply of reliable melodies simply by having an
archive with sequences of pitches that work together and represent different types
of melodic contours and then using a weighted function to randomly pick some.
Then, they can be transformed using a second weighted function to add slight
variety to it. The outcome of this method has the enormous advantage of being
somewhat predictable, aesthetically pleasing and making sense in the context, and
has the extreme disadvantage of being a small remix of already existing material,
thus not increasing the span of possibilities by that much.
However, we believe the best approach is to try to mimic what happens in

real performances with real singers. What they do is to improvise a melody on
top of a harmonic texture. By doing that they are relying on the knowledge
of the underlying harmonic progression, but mainly the scale associated with it,
respecting the mode in which it pertains. The precise way in which the pitches
are chosen among those constraints is incredibly complex and derives from an
entanglement not yet fully understood. There are obvious relations between the
293
semantic and syntactic content of the lyrical text, since the speech contours should
influence the melodic contour. Not only that but the language in which the text
is written also influences the pitch. It is known that the pitch of speech is a
cultural construct based on the pitches people hear around them when they are
infants. It is a function of the linguistic community (Deutsch et al., 2009, Dolson,
1994). Those dimensions are impossible to model in our current state. Moreover,
the physical constraints of the singers also constrain the available choices, since
some people have wider ranges and theatrical performances while others prefer to
sing in an almost spoken form thus, narrowing the overall range of their melodies.
Therefore acknowledging these difficulties and limitations we have opted, in this
case, to try to simulate this complex behavior using the ideas proposed by Jeon et
al. (Jeon et al., 2006) and using noise as a central concept in shaping the melodic
contour.
For each of the chordal harmonic progressions listed previously, and used to
generate the ostinati, we have built a second list of all the possible associated har-
monic scalar progressions, indexed in the same position. Therefore, it is possible
to use them to retrieve the pitches for the melody. These scales, however, should
also be modified and further constrained by an offset variable in order to build
coherent contours and avoid strange intervals, influencing the voice leading.
M elodyx = ⟨⟨Scale1 , o1 ⟩, ⟨Scale2 , o2 ⟩, ⟨Scale3 , o3 ⟩, . . . ⟩ :

∀ sn ∈ Scalex , sn = sn+o : o ∈ Z
The offset variables oi are local transformational devices in the form of a rota-
tion implying that all elements on a scale change their positions back or forward
accordingly. This transformation is used to force the first symbol in any given
scale to be the one desired, in order to condition the future mapping into pitches.
As an example, the progression T|D|D|T in the minor mode, “i|V7|V7|i”, the

first chord “i” is best associated with the “natural-minor” scale, starting in the
first symbol, with an offset 0. The second chord “V7”, however, is best associated
with the “harmonic-minor” scale. This is so because the second note of the chord
“V7” does not exist in the “natural-minor” scale; it is part of the harmonic or
borrowed from the major mode. Also, it would be better starting with an offset
of either 1 or -1, influencing the voice leading so that the first symbol of the scale
is the nearest one matching a chordal tone. Following this reasoning, the third
294
15.3. Third Module
chord, again “V7”, and thinking in how a melody usually evolves, arch shaped
up and not down, could be best associated with the same “harmonic-minor” scale
but with an offset of 3 or 4, the nearest notes of the scale that are chordal tones.
M elodyT |D|D|T = ⟨⟨⟨T onalityminor , t, m⟩, 0⟩, ⟨⟨T onalityharmonic , t, m⟩, 1⟩,
⟨⟨T onalityharmonic , t, m⟩, 4⟩, ⟨⟨T onalityminor , t, m⟩, 0⟩⟩ :
∀ sn ∈ Scalex , sn = sn+o : o ∈ Z
In terms of Symbolic Composer coding this can be done as follows:
;;0-I-V7-V7-I
((0 major c 5) (1 major c 5) (4 major c 5) (0 major c 5)
)
;;1-I-ii-V7-I
)
;;2-I-IV-V7-I
)
Of course, some of these are subjective decisions and in a way influence the
voice-leading and the general melodic contour, but none of these decisions is final.
They all can be refined at any moment, and any of these offsets (or all of them)
can be set to variables.
The point being made is that, having a harmonic progression based on scales,
constrained by the chords, one is now fully apt to apply a contour to it to generate
suitable pitches.
In order to generate such a contour, a function that generates a vector based

on Brownian noise (P SD ∝ 1/f 2 ) with 257 samples was used, in order to have
a considerable shape while avoiding repetition. Then, the vector is scaled to the
length of the actual “rhythm-head” list, and converted to symbols within the
desired range. This allows having the precise length needed, while retaining the
overall shape of the contour. The operation is repeated to generate a second vector
that is applied to the “rhythm-tail” list. Both vectors are appended and the result
295
is a list of symbols, whose length is the same as the list of durations, and that has
a realistic contour: M elodyXn = ⟨s1 , s2 , s3 , . . . , sn ⟩ : n = #(RhythmXn )
When that contour is applied on top of the corresponding harmonic scalar

progression the symbols are converted to pitches that actually make sense within
the context of the chords being played in the ostinato. Therefore, the melodies
will always sound consonant, respecting the tradition involved. The coding in
Symbolic Composer corresponds to:
;;Melody ‘‘a1’’ generator

;head
(setq melody-a1-head
(vector-to-symbol a f (vector-quantize 257
(length rhythm-A-head) (gen-noise-Brownian 8 my-random-seed 0.5)))
)
Modifying the range of the symbols or the samples of the vectors can greatly
influence the result and the quality and realism of the melodies. Using other noise
functions and hybrid functions that generate graphical contours can have similar
effects. The limit is one’s imagination, and the code is flexible enough to allow
very easily the refinement of the abstraction.
15.4 Fourth Module
In this section we will detail how we have coded the countermelody and bass
layers and which assumptions and algorithms were used.
The countermelody is a layer that often appears in the practice. However, it is

many times omitted in the earliest scores. The more complex scores include parts
of it and it roughly corresponds to what a modern guitarra does, by filling in the
spaces in between the melody. In order to mimic this layer, we have prepared a set
of stock figurations based on the figures of Salwa Castelo-Branco already presented
in the first part of this dissertation, but also on a set of observations we did on our
own, observing players and listening to records. Basically, these stock figurations
are based on arpeggiations, small scalar figures or even motifs drawn out of the
296
15.5. Fifth Module
melody. If one uses Ernesto Vieira’s model as a base, then they usually occur
in the middle of the second bar, right when the voice is beginning to end a line,
and end in the downbeat of the following bar, when a new line has just started.
In terms of coding, since they apply mostly scalar figures, we had to prepare a
weighted list of possible stock figurations, much alike the one we have done for
the ostinati, just with the difference that, instead of associating them with the
chordal harmonic progressions, we have associated them with the scalar harmonic
progressions used for the melody, to allow passing tones and half-scales. In a way,
they are fragments of pre-composed music selected more or less at random and
having variations. A way to refine this whole process was to, once again, code
an abstraction that generates little motives based on small vectors with Brownian
(and similar) noises. Of course, this later approach originates more unexpected
results and sometimes results that do not conform with the tradition. The best
approach, in our opinion, is a weighted combination of both solutions.
The bass line is covered in this same section because the principle it is the same.
The bass line in fado practice is usually redundant, since most of the times one does
not have a bass present and the bass is contained in the ostinati performed by the
viola (already described in the second module). However, the solution found for
ostinati only allows chordal material and, although this is enough for most of the
earliest scores, and probably reflects the simpler forms of the practice, the truth
is that often the bass also performs scalar material, sometimes performing like a
walking bass when connecting sections or even in the same ostinati patterns like
marches and foxes. In these cases, one can really say that the bass presents counter-
melodic material. So we have decided to take this into account and prepare a layer
that most of the time performs the same bass notes as the viola, but it is linked
to the scalar harmonic progressions and also performs stock scalar figurations,
namely alternating ascending and descending motives.
15.5 Fifth Module
As we have been presenting in this text, every musical score is omissive in sev-
eral parameters that are to be completed by the performer within the cultural
practice, context, society and audience expectations. Fado is no exception, and
for a musical score to be a convincing display of this practice, that human factor
ought to be present, so the musical score generated should not be played neither
297
too randomly, neither strictly and mechanically, as any MIDI file can be. The
automatic performance of quantized MIDI scores has been studied in the Royal
Institute of Technology (KTH) and models for human performance are presented
in the studies of Friberg and colleagues (Friberg et al., 2000, 2006). These models
present and implement a series of rules (hereby referred as the KTH performance
rules) inferred from the empirical study of real performances, namely how the per-
former shapes all audible parameters such as tempo, sound level, or articulations.
It has been found that each note suffers micro-deviations relatively to the notated
value, not only due to human imprecision, but also as a systematic variations,
implying things like long notes tend to be lengthened and louder and short notes
to be shortened and softened; the higher the pitch the louder the note; notes in
uphill motion have a decreasing duration, etc. . .
We have thought to include in the generated musical scores something that

would model the major characteristics inherent to human performance, such as
crucial articulations that our analysis has shown to define the genre, and that
we have defined and discussed as “gestures” – some particular legatos, mordentes,
tenutos, suspensions, portamentos, and so on. This can be done with an extra layer
of transformative functions built on top of the quantized musical score generated.
We see them as a “humanizer” that adds some kind of swing and deflection on the
exact values provided by the MIDI score. We will explain with examples how the
humanizer was coded and implemented and how it can be refined at any time.
As detailed in previous sections most of what the generator does is to generate

fragments of music, semiographic schemes, much like a normal score generated
by pen and pencil would present. The problem with this approach is that each
instance of the scheme would be presented exactly as the previous one, thus be-
coming dull and repetitive. It is known for a fact that in human performance
the same does not happen, as each instance of a semiographic scheme is varied.
Therefore, and in order to have the best of all worlds, the way to have consistency
and coherence among the score is to effectively have these semiographic schemes
repeated over and over again in each section, instead of constantly generating new
bits of music. However, each time they are repeated something about them has
also to change. Therefore we formalized a series of abstractions, inspired in the
KTH performance rules, that generate a little variety in each scheme and thus
allow us to create variations regarding the same scheme.
298
15.5. Fifth Module
For any given section X0 in the form, transformations are applied originating
slightly varied sections Xn. Further transformations built on top of Xn might be
applied, originating more varied sections:
f : Tvariation (X0) → Xn; f : Tvariation (Xn) → Xm : n ̸= m
The best way to take advantage of the implementation of these functions is to

use them as transformations to create alternative sections for each base section
created. Therefore when a section A0 is generated, it has corresponding lists of
rhythms and lists of symbols. The section A1 would have the same lists affected
by some of these gestural functions. The section A2 could have the same base lists
affected again by the same or other gestural functions to randomly distort them in
a different way. The ornamental functions can also be used cumulatively in order
to create more variety.
15.5.1 Rhythmic variations
If one observes the text, the prosody and the defining base of the sections,
one understands that the rhythm of the melody is one of the anchors present
in every section of every work. Each rhythm is a list of values that has a fixed
total length. Therefore, the only way to increase variety within this parameter is
to slightly skew the values within the constraint of this fixed length by stealing
some duration to some notes and adding it to others. We have created three
functions to so do. The first one, “length-emphasize”, emphasizes a rhythm by a
factor, keeping the length intact within resolution, implying that 0% – almost no
distortion, 100% – maximum practical distortion to avoid smaller values becoming
void. This function is implemented as follows: the absolute value of each duration
is powered by a factor (a variable between 0 and 100) divided by 33.4 plus 1.
This will mean that each duration is powered by a number varying between 1 and
4. Not only that but each duration is also added up to a random micro-number
varying between 0 and one tenth of itself:
f : ∀ r ∈ Rhythmmelody , ri → qi = r[1+(n/33.4)] + x : x ∈ [0, r/10], n ∈ [0, 100]
Then all values will be divided by a common factor that is the result between
the sum of all original durations divided by the sum of all the new durations. This
will guarantee that the length of the list of the transformed durations will be the
same as the original durations:
299
!
f : ∀ q, q → q × ( ! rqii )
Depending on the factor and original durations, sometimes micro-values might

be originated in this process. Therefore it is suggested that the final values might
be quantized to a minimum of a 32nd note, for instance. This algorithm will create
distortions in each duration in such a way that in practical terms the longer values
will become even longer and the shorter values will become shorter. The coding
of such a function in Symbolic Composer becomes rather complex and nested:
300
15.5. Fifth Module
(defun length-emphasize (factor list &optional seed)

(diagnostic2 "length-emphasize" $cr$)
(do-quietly
(init-rnd seed)
;;quantizes the list to realistic resolution values
(let* ((a (length-quantize-grid ’1/32
;;resizes list to adjust the differences
(length-resize (make-zone list)
(let ((result nil))
(dotimes (i (length list))
;;emphasizes list by factor and
adds a bit distortion to each element
(setq f (expt (abs (nth i list))
(+ (/ factor 33.4) 1)))
(push (+ f (rnd2 0 (/ f 10)))
result))
(reverse result)))))
(b (append (list (+ (first a)
(- (make-zone list) (make-zone a)))) (rest a)))
(result nil))
(dotimes (i (length b))
(push (if (minusp (nth i list))
(- (abs (nth i b))) (nth i b)) result))
(reverse result))))
\#\textbar
(length-emphasize 30 ’(1/1 1/2 1/2 1/4 1/4 1/4 1/4))
--> (53/32 13/32 7/16 1/8 1/8 1/8 1/8)
Secondly, we have created the “length-homogenize” function that does exactly

the reverse, meaning that the larger values become smaller and the smaller ones
become longer, thus making the lists more “uniform”. Finally, in a similar fash-
ion, “length-distort”, randomly steals durations from some values adding them
to others, distorting a rhythm by a certain factor, in a non-uniform way, keeping
the length intact within the desired resolution. With a wise combination of these
three functions, we have a useful resource to obtain variations of the semiographic
rhythmic schemes.
301
15.5.2 Melodic variations
Another way to obtain variation is to slightly change some of the symbols from
a list that pertains to a melody. Symbolic Composer already has such built-in
functions, so there was no need to build them. Mainly the idea is to use noise
to modulate any list, by slightly increasing or decreasing some of its elements at
random:
∃ sn ∈ M elodyXn : sn → (s + t)n : t ∈ [−2, 2]
Thinking idiomatically about the piano, and how some fados might be played on
a piano, one of the most common resources in the literature is the repetition of the
same melodic motif played in octaves as a reinforcement. Such transformational
function can be implemented as well, simply by concatenating each symbol, from
a list of melodies, with itself displaced by a variable:
∀ sn ∈ M elodyXn , sn → sn ∗ (s + t)n : t ∈ Z
In this case, if the variable is twelve, then this will generate a list of dyads in
which each element is an octave apart from the original. Modifying the value of
the variable originates different kinds of dyads resulting in other interesting effects.
302
15.5. Fifth Module
In Symbolic Composer, we created the function “Octavate” to allow the id-

iomatic piano effect.
(defun octavate (melody)

(let ((result nil))
(dotimes (i (length melody))
(push (merge-symbols (append ’(+12)
(list ‘,(nth (- (- (length melody) i) 1) melody)))
‘,(nth (- (- (length melody) i) 1) melody)) result))
result)
)
And then we have decided to create a similar and even more powerful abstrac-
tion when we realized that one could decide to have the same effect by whichever
interval desired. The reasoning is the same, and one can append whichever dis-
tance one wants. One can even randomize the interval and create diverse melodies.
All this just shows how having a modular structure allows us infinite variations
and effects with relative ease.
15.5.3 Mixed Variations
Another class of variations can be obtained by simultaneously changing pitches

and durations. Although trickier, they are also doable. We decided to do so
by creating an abstraction for one of the most common gestures heard in fado,
namely performed by Amália Rodrigues: the mordente. We define the mordente
as the quick alternation between a note and the diatonic note immediately above.
Sometimes, in highly melismatic singing, a chain of mordentes is performed. The
algorithm consists in dividing the duration of the note in four parts, knowing that
the first quarter is the original note, the second quarter is assigned to the diatonic
note above, and the third and fourth quarters are again the original note. This
specific function is already implemented in Symbolic composer as “mord”, but to
be assigned to a specific note. It has been improved in a way to randomly assign
it to a whole phrase. Since the rhythm of the melodies and the symbols of the
melodies have been defined as paired lists with the same length, such task can
be performed by randomly pick specific positions and applying the algorithm to
303
those positions in both lists.
f : M ordenter = ∃ dn ∈ RhythmXn , sn ∈ M elodyXn :

dn dn dn
dn → ⟨ , , ⟩, sn → ⟨sn , (s + 1)n , sn ⟩
4 4 2
The abstraction coded, “mordenter”, receives as an input a list of rhythms and

a list of symbols and outputs a mixed list. Then the auxiliary functions “rhymord”
and “melmord” allow the rhythm and symbols lists to be retrieved. Both lists are
equal to the original ones, except they now have random mordentes assigned to
them.
This example shows yet another way of bringing color to the performances in
a flexible modular way.
304
15.5. Fifth Module
;;Function "Mordenter": Randomly assigns mordentes to a rhythm+melody

(defun mordenter (rhythm melody)
(prog (out)
(dotimes (i (length rhythm))
(setq a (random 10))
(if (> 7 a)
(append (push (nth i rhythm) out)
(push (nth i melody) out))
(append (push (reverse
(nth 0 (mord (nth i rhythm) (nth i melody)))) out)
(push (reverse (nth 1 (mord
(nth i rhythm) (nth i melody)))) out))
))
(return (reverse (flatten out)))
))
;;returns the rhythm of mordenter

(defun rhymord (list)
(prog (out)
(if (numberp (nth i list))
( push (nth i list) out)))
(return (reverse out))))
;;returns the melody of mordenter

(defun melmord (list)
(prog (out)
(if (numberp (nth i list)) nil
( push (nth i list) out)))
(return (reverse out))))
305
15.6 Global Parameters
15.6.1 Dynamics
Dynamics in MIDI files are usually implemented as a combination of velocities

and volume and they are associated with a given note (therefore a pitch and a
duration). Since the durations and pitches have been implemented as lists of equal
length, the velocities follow the same principle and therefore are implemented as
such: V elocityXn = ⟨v1 , v2 , v3 , v4 , . . . ⟩
The model of Toivanen et al. creates dynamic markings in the music based on
a coefficient of arousal retrieved from the lyrical content. This coefficient is calcu-
lated according to the proportion of substituted words in a given line (Toivanen
et al., 2013, p. 90). We have decided that, although one does not have a text,
one has the conventions from the practice associated with emotional prosody, from
which it is possible to extrapolate the dynamics as a function of durations and
symbols, following the reasoning presented in the KTH performance rules (Friberg
et al., 2000, 2006). Therefore, the velocities are defined as a function of the product
of durations and symbols.
The conversion is not direct, since some mapping or scaling has to occur first.
Symbols are expressed as alphabetic characters, therefore one has to guarantee that
all symbols are prior mapped into an integer between 0 and 100. This operation
is trivial and in Symbolic Composer can be done applying the predefined function
“symbol-to-vector” and then specifying the desired range:
f : ∀ sn ∈ M elodyXn → NM elodyXn : N ∈ [1, 2, 3, . . . , 100]
Furthermore all durations have to be positive, so their absolute value is taken

instead.
f : ∀ dn ∈ RhythmXn , ∃ vn ∈ V elocityXn : vn = |dn |×NM elodyXn
And finally, one can also constrain the range of results to realistic velocities by
scaling the possible range to values between 64 and 100 instead of 0 to 127, for
instance.
f : ∀ vn ∈ V elocityXn → V elV elocityXn : V el ∈ [64, 100]
306
15.6. Global Parameters
This last step largely depends on taste and on the genre to be modeled and can
be modified at will. In terms of Symbolic Composer the code looks as follows:
(setq velocity-section-A (vector-round 64 100

(mapcar ’* (symbol-to-vector 0 100 symbol-section-A)
(mapcar ’abs rhythm-section-A))))
The formula basically obtains the absolute ratios of the list of durations of a
certain section and multiplies them by the symbols, which were previously con-
verted into a vector mapped within a range of 0 to 100 values. The outcome
is then mapped to a vector within the range of 64 to 100. These values we de-
cided for voice (all within the mezzo-forte range) can be, of course, modified at
any time. Notice how the formula applied is exactly the inverse of the used in
tempo. We consider that dynamics and tempo have an inverse relationship in this
practice. The exact coefficients are yet to be determined with more study and
experimentation and are not the scope of this work.
15.6.2 Tempo
Opposing all other parameters discussed so far, which are generally local pa-
rameters and have to be generated and assigned for each melodic line, in each
possible instrument, tempo is a global parameter. This means that in each section
it affects all instruments and lines being performed. The first and most obvious
solution is to simply assign one simple value for tempo that affects the entire piece
or section. This is what often happens in most quantized scores. A more realistic
approach, also based on the KTH performance rules, however, is to have tempo
dealt in a local way.
In the program, tempo is not fixed and its fluctuation is defined as a function of
other factors, trying to emulate what happens in a real performance. According to
the conventions of emotional prosodic styles, relying on text, and as a general rule
of thumb, the higher the pitch is, the longer the duration, consequently the louder
the velocity and the slower the tempo. Conversely, the shortest rhythms and the
lower pitches are usually sung quieter and also slightly rushed. So, having this
general idea, the tempo fluctuation was formalized as the inversion of the product
of durations and pitches: Tmax − Tmin ⇐ (DurationXn × P itchXn )−1 .
307
For the implementation of this idea a base tempo is set up in the initial settings.
Since 72-84 bpm was referred as the tempo range acceptable by the standards of
Ernesto Vieira’s dictionary (Vieira, 1890, p. 55), we decided to remove one third
of the top value to find a minimum tempo of 56 as a base. Using the same 12 bpm
range, our minimum tempo will vary between 56 up to 68 bpm:Tmin = 62 ± 6. The
maximum tempo is defined as four-thirds times this value (reverse-engineering the
one-third fluctuation ratio):Tmax = 4/3 × Tmin .
Having these references as a fluctuation threshold, then inside each section

the tempo is mapped according to the inversion of symbols and durations. Each
symbol is converted and mapped into an integer between 100 and 1. This can be
done in Symbolic Composer using the “symbol-to-vector” function.
f : ∀ SymbolXn → NSymbolXn : N ∈ [100, 99, 98, . . . , 1]
Each duration is converted to an absolute ratio (to avoid negative values),

and the product of these two values is then proportionally mapped into a value
between our previously defined minimum and maximum tempos. This mapping
can be coded in Symbolic Composer using the “vector-round” function.
f : ∀ (|DurationXn |×NSymbolXn ) → TXn : T ∈ [Tmin , . . . , Tmax ]
The outcome is a list of tempo values, one for each note. Having such a detailed
tempo, on the note level, and reflecting the direct effect of the other parameters,
makes the performance fluid and organic.
15.6.3 Groove
Cumulatively with the tempo and velocities that are derived from the durations
and symbols (and symbols in a certain way are already derived from the rhythms),
we also consider that groove can be defined as a function of inverse of velocities,
much similar to tempo.
The principle is similar to the one applied earlier – the list with the velocities is
mapped onto values, in this case a range between -1 and 2, which are divided by
an extremely high value, in this case 256, in order to obtain really small durations.
[−1,2]
f : ∀ vn ∈ V elocityXn → gXn : g ∈ 256
308
15.6. Global Parameters
This resulting list of really small durations can in fact be considered a pattern
of small deviations, to be applied within the section: GrooveXn = ⟨g1 , g2 , g3 , . . . ⟩.
We have experimented with several values and, for instance, the use of divisions
for 64 or 128 gave a sense of hiccup or uncanny valley to the performance. The
coding can be done as:
(setq groove-section-A
(mapcar \#’(lambda (x) (/ x 256))
(vector-round -1 2 velocity-section-A )))
The formula on groove can be refined, as all others, given more time, since
studies have been made, namely (Naveda et al., 2011, Sioros et al., 2012).
309
Chapter 16
Templates for rendering the

MIDI files
In order to listen to and record the generated music, several options are avail-
able. The most recent version of Symbolic Composer has a system that integrates
the software MIDI Trail, which allows following the score in almost real time with
an interesting graphic user interface, and using internal soft synths. It has also a
connection to LilyPond, so it generates a score for immediate visualization1 . The
score poses problems since the gestural parameters rendered by the humanizer
originate many strange symbols due to lack of quantization and constant tempo
changes that render the score unreadable. It must be noticed that a score is pre-
pared to be read as a semiographic simple scheme, therefore, it is only suitable
if the gestural parameters are not applied to it, which is only possible with a
dumbed-down version of the generator (only the first four modules active).
Another option was to rewire the software to a sequencer and create a template
with the number of MIDI tracks needed. We find extremely important the pa-
rameter timbre, crucial in defining the practice of fado, as shown in the first part
of this dissertation. Unfortunately MIDI scores are very poor in dealing with this
particular parameter. Therefore, taking advantage of recent sample libraries re-
leased, namely Alfama 2 by Adamastor Virtual Instruments, we have rewired both
the melody and countermelody tracks to be played by a sampled guitarra, with the
harmony and bass tracks to be played by a sampled viola. This sampled guitarra
1
The latest version of the generator as well all the necessary auxiliary functions can be
retrieved in http://tiagovideira.com/2015/06/07/instrumental-fado/.
2
http://adamastorvi.com/, accessed April, 6, 2015
311
allows us to recreate the iconic timbre that most of the time is a valuable cue
in identifying the practice. We have also pre-mixed and pre-mastered the tracks,
setting up the equalizers, compressors, reverbs and the additional plug-ins we de-
sired. We saved that file as a template. Now all we have to do is run Symbolic
Composer and generate new MIDI files, and then switch to the template and drag
the file into the sequencer. The tracks will fit into the right place and they are
ready to go. We can listen to them immediately or bounce them as high-quality
mp3s. If we are not happy with this MIDI file, we can obviously make changes,
tweaks, or whatever we want. Or we can just select the tracks and delete them,
switch to Symbolic Composer and evaluate the program again. It will generate a
new MIDI file that replaces the previous one. So, we just have to drag it again into
the same setup and we are ready to go again. We can virtually listen in almost
real time or record several high-quality Mp3’s in a flash as soon as we have built
our templates.
312
Chapter 17
Conclusions
In this section we will detail the outcome of the model and why we think it is
successful. We will highlight its strengths and limitations and how and why the
generated outcome can be perceived as instrumental interactive music that is usu-
ally associated with fado practice. We will also detail the pedagogical framework
of the project and how this research was an ever flowing path of self-discovery and
personal growth. We will also emphasize all the relevant data and material that
has been produced and it is now available for the general public to use in future
work and research.
We consider we have accomplished our goals in the sense that we were able
to deal with the problems posed by David Cope and at the same time respect
the ideas of Brian Eno1 , two important precursors in this kind of systems. First,
we have built a generator that, with the click of a button, even without further
human interaction, generates a new piece from scratch. In this sense it works like
an organic unique seed from which a whole plant grows. By having all abstractions
indexed to the same seed, the outcomes are reproducible to anyone who desires
to do so. The work is also interactive enough in the sense that a user that wishes
to provide some constraints or make some decisions on their own is able to do
so – the user can decide several of the parameters and by doing so we believe
the process becomes interactive. The architecture of the program, being modular
and highly flexible and customizable, acts as a case study for fado but can be
easily adapted to other contexts. This workflow shows how the automatization of
the composition process is very useful in case one just desires to produce a large
1
http://www.inmotionmagazine.com/eno1.html, accessed June, 6, 2015
313
amount of instrumental songs in a short span of time. It obviously works better

if they are all within the same style, maintaining the same instrumentation, in
order to generate a thematic album. Still, the modular structure of the model
allows for relatively easy modeling of other styles and ensembles by changing the
relevant variables. Basically, we believe we have fulfilled our own expectations in
the sense that we have not only built a proof of concept generator of music usually
associated with fado practice, but also developed a set of skills and templates that
allow us and any user that follows our steps and algorithms to develop their own
generators for virtually any kind of music. We have built a template for style
imitation, a virtual composer that is able to compose polkas, marches, rock music
or even music in the style of Satie just by changing a few of the values, parameters
and constraints of the generator.
We think the project falls short in terms of the melodic component mainly
because of the lack of a real text. The more we worked the more we understood
the importance of the semantic content and the patterns of stresses and lengths of
the vowels in shaping the actual rhythms and pitches governing a melody. More-
over, when coding the gestural performance and paying attention to the KTH
performance rules we also realized that these parameters were also constrained
by the same factor. So, while we still acknowledge the hierarchical importance
of this parameter, in the whole process, the lack of its ideal implementation, in
fact, impacts the results. Therefore in order to refine this or similar projects, we
cannot stress enough the importance of the text implementation to its full extent,
building a database of a given linguistic community and both a lyrical generator
and extensive rules on how the words shape the subsequent parameters in a more
realistic way.
We have also developed a set of tools useful for humanizing the scores generated
and explained ways and solutions to easily render and record them. This makes
this set of tools a highly appreciated device for the music industry, namely for
the composer who wishes to work for the music libraries industry, allowing the
composition, production and recording of large sets of music in much reduced
times.
The project successfully identified a series of traits in fado music that were
never explored up to date. In this sense, we believe we have identified a series of
correlations and useful ways with pedagogical value. One is now able to discern
how to build a fado from scratch using pen and paper and understand how the
314
text is a valuable seed and how most parameters, including tempo, dynamics and
groove might have a relation to the text. The project also helps in understanding
the importance of the social and sound features, in the sense that now there is a
comparative point between the musical materials, the symbolic content and the
practice in real life. And each one is now able to be confronted with both and
decide which ones are relevant to their own sense and classification of the genre.
For instance, if one listens to the outcome of the generator in a MIDI piano sound,
does that make a difference from listening to the same output rendered using the
Alfama virtual library? Or does it make a difference to listen to the same score
played live by real musicians? Does a member of the fado community practice
consider fado the scores and MIDI files we provided in our database?
By building a digital database of fado, we have also unearthed repertoire that

was long forgotten. We have studied it, we have found ways to recreate it and,
in a sense, we want to be involved with the community and give it back to the
community so that this knowledge is integrated in the proper context, and the
debate can go on with new sets of information. We will realize better what the
values involved in all these issues are and what is really important to make the
practice evolve. We are confident we have contributed in an original way for the
creation of more tools to understand fado and help it to evolve.
We believe all this can be extrapolated to other musical practices, namely vocal
practices relying on text. We believe our main original contribution to the field
was basically to demonstrate that indeed it is possible to do so by identifying a
series of simple steps and a series of algorithms. We demonstrated how one can
make music, can understand music, can understand how some humans compose
music, in the same way someone can deconstruct a culinary practice by writing
a recipe book. In a certain way, this dissertation is a proposition for a holistic
method that shows it is possible to derive, deduce and build musical recipes in
order to generate new musical works that are actually aesthetically pleasing. It
also shows that even when the composer does not want to fully automatize the
whole process, they can do so by shortening just some aspects of it.
315
17.1 Evaluation
An important issue when considering the production of any system is the pro-
cess of evaluation of the success or failure of that system. When thinking of an
automatic music generator to mimic a practice, one might wonder how its efficiency
can be measured. Considering that music produced in this manner, especially in
academic context, aims to achieve a somewhat utopian status of scientific valida-
tion of the rules that govern a certain style, there has to be a way to test the out-
come produced. The most suitable way, and since music is a human phenomenon,
seems to be presenting it to human listeners. Cope (Cope, 2005, pp. 345-359),
however, discusses the problems that arise when listeners know, a priori, that it
is computer generated music, being influenced by this fact on their judgment. To
avoid that, then, a kind of Turing test could be a suitable solution. The music
produced could be presented to random listeners who should not be aware that
the music they are hearing is automatically generated and take into account their
feedback. If the listeners recognize what they are listening as genuine examples of
tunes within the practice pretended, then one may be able to say that the system
accomplished its goal.
We acknowledge the limitations of any aesthetic evaluation system, because

each one has different values. And even within the same culture tastes differ
greatly, and so we aspired to also have a more objective goal of evaluation than
just to rely on human ears.
It occurred to us that a simple test could be devised to perform an automatic

evaluation task using the same software used to retrieve musical information. One
can simply generate a fair number of new fado recordings and then supply them
to the trained classifier, just as we did when we tested the coherence of the corpus.
If most of them are classified as fado and not as any of the 38 other taxonomies
then we have a guarantee that the generator has produced music that is closer to
the music of the corpus than to the other music from the database. Of course,
this still does not tell us anything about its closeness to many other genres not
present in the database, but it is a promising start. Using these procedures, one
would be able to have a better evaluation of the system based on objective and
reproducible standards. Regarding this specific hypothesis, only the first four
modules of the generator should be used to generate scores, since in the trained
corpus all exemplars are based on quantized short-form piano versions and not
316
17.1. Evaluation
full fado instrumentals, with melodies and counter-melodies. These last versions
(the ones generated with all five modules and rendered with Alfama libraries), we
speculate, could only be evaluated correctly by human hears, since they aim to
model the actual practice and not just the scores.
In order to test these hypotheses 200 fados recordings were generated, in the
MIDI file format2 . 100 were generated using only the four modules version, what
we can call “Piano fados” resembling the ones in the database. But we have also
generated 100 fados using all modules, what we can call ensemble fados “for viola
and guitarra”. In order to control bias, it was decided to look up the internet for
recorded MIDI files based on piano reductions. We have found the site of Doug
McKenzie3 , from which a random sample of 25 recordings were downloaded, all
supposedly from “jazz related” genres.
We trained Cory McKay’s database using 38 taxonomies with 25 recordings

each, plus the 100 fado recordings from the corpus. Then, the 100 piano fado
recordings, plus the 100 ensemble fado recordings and the 25 piano jazz recordings
were supplied. We asked Bodhidharma to classify all these recordings. The results
give a very positive evaluation of the generator. 99 of the 100 piano fado recordings
were indeed classified as fado. 77 of the ensemble fado recordings were classified
as fado, as well, while 16 were classified as “unknown”. Three gave error and were
impossible to classify at all. Only 2 of the jazz recordings from the control group
were classified as fado, most of them being classified into jazz related genres (cool,
bebop).
These results show that the output of the generator is closer to the corpus than
to any other of the recordings available to sample the world musics. Moreover,
they show that even the generated fados in the form of ensemble still retain more
characteristics of fado than any other genres available in the other taxonomies.
However, the relevant number of 16 “unknown” and 3 impossible to classify at
all also show that refinement is possible. On the other hand, we are seeing this
as a good thing, since we also intend for the system to innovate in some way.
Hence, some fados not being recognized as such, and at the same time not being
recognized as anything else, can be seen as a possible pathway for innovation.
The control group results have also shown that there seems to be very little bias
induced due to both the overrepresentation of fados in the training process and
2
They can be retrieved in http://tiagovideira.com/2015/06/07/instrumental-fado/.
3
http://www.bushgrafts.com/jazz/midi.htm, accessed 22 May 2015
317
the fact that they are all in the form of piano reductions. The fact that the vast
majority of recordings classified as jazz piano solos, by an external source, were
indeed classified as jazz related genres reinforces the idea that fado piano solos
in particular and fado in general carry enough relevant harmonic, melodic and
rhythmic characteristics, per se, to be possible to be identified as a genre.
Regarding the pleasantness of the output, the most suitable way, and since
music is a human phenomenon, seems to be presenting it to human listeners.
Cope (Cope, 2005, pp. 345-359), however, discusses the problems that arise when
listeners know, a priori, that it is computer generated music, being influenced by
this fact on their judgment. To avoid that, then, a kind of Turing test was a
suitable solution. Five random samples were drawn from the 200 generated MIDI
files (C1, C2, C3, C4 and C5). Another five random samples were drawn from the
corpus (H1, H2, H3, H4 and H5). Excerpts with duration of roughly 30 seconds
(corresponding to coherent sections, no cuts in between thematic material) were
recorded using the same template to prevent bias from different sound quality or
production issues. These excerpts were mixed randomly. We obtained a final track
with 10 coherent excerpts, with the same sound quality, in which the only variable
was the fact that five were generated by the program and five were composed by
humans. The listeners were oblivious to this fact. An online quiz was created
in which we asked the subjects to listen to the track, in its entirety, for them to
have an idea of the overall sounding environment they were judging. Then, they
were asked to attribute a note ranging from 0 to 10 to each excerpt, according to
their personal, subjective, aesthetic criteria. There was also an open text box for
users to leave their general commentaries, if they wished to. There was no further
leading and no personal data was collected, apart from their relationship with
music (listener, amateur or professional performers) and level of music education.
Although we recognize the sample seems small, there was no other practical way
to do this in a timely manner, since most people have no time or disposition to
answer to long or complex enquires. 65 subjects answered the enquiry and their
answers were then analyzed. We realized that most excerpts had all kinds of
possible classifications, showing the enormous variety of opinions among different
people, although some tendencies emerged. Overall, the distributions were rather
normal and there were neither exceptionally unpleasant nor particularly pleasant
excerpts, with global means ranging between 4 and 7.
318
17.1. Evaluation
A one-way ANOVA test revealed that there were differences among the classi-
fications not explained by chance (figure 17.1). A Tukeys HSD (honest significant
difference) test revealed that C1 was significantly better classified that the other
four computer generated excerpts. It also revealed that H3, H4 and H5 were bet-
ter classified than H1 and H2. C1 was better classified than the two lowest (H1,
H2) and in pair with the other three human composed (H3, H4, H5) excerpts.
However, the other four computer generated (C2, C3, C4, C5) excerpts were just
comparable with the two lowest and significantly worse than the three highest
human composed excerpts.
Evaluation of fados
10
6
Pleasantness
0
C1
C2
C3
C4
C5
H1
H2
H3
H4
H5
Figure 17.1: Model Evaluation with human subjects.
Overall (figure 17.2), the cluster of computer generated excerpts had a lower
classification than the cluster of human composed excerpts in a significative dif-
ference of 1.06 points of pleasantness in the scale of 0 to 10.
These results indicate that the generator succeeded in creating output that has
relatively relevant pleasantness (global mean of 5) and the ability of producing
some high quality samples, once in a while, while avoiding garbage. However, it
has also shown that there is room for improvement, since most of the time its
results are perceived as slightly worse than human-composed music (global mean
319
Evaluation of fados
10
6
Pleasantness
0
C
Figure 17.2: Overall Human versus Computer composed samples.
of 6). The difference of only 1,06 points of pleasantness, although significative,

gives us hope, since it could be much worse.
We can also say, based on our knowledge of the samples, that the most out-of-
the-box, innovative fados generated by the program polarized the subjects. While
some of them applauded the results and preferred the difference and innovation,
and thus gave better grades, others, greatly penalized these excerpts in favor of
more traditional ones (both computer generated and human composed). There-
fore, the subjective taste of the audience and the general level of music education
of the subjects has and will always play a role in their preferences. An output
that may be seen as garbage by many will actually be seen as pleasant for others.
320
17.1. Evaluation
Therefore, the fact that there was no negative consensus, and even the worst clas-
sified samples were also able to obtain a relevant number of high classifications,
leads us to believe that the generator has good perspectives.
321
Chapter 18
Future Work
18.1 Regarding the model
We believe the database should be concluded. There are several more fados to
be transcribed into digital versions, as well as a series of other pieces of Portuguese
repertoire waiting to be unveiled. It could be of extreme patrimonial importance
to recover and analyze all these pieces and maybe compare them contextually and
find convergence points and rules. Automatic music classification and generation
of this repertoire could also be a future project (namely habanerass, descantes,
lundums, work songs...). The critical editions of all these scores, also in physical
format, is an open project that we think should be pursued. Having the database
completed to its full extent would also allow for the refining of the weights and all
the statistical work around them.
The integration of sound records and data in the form of actual sound could
be an ambitious step to complement these. Having more accurate and complete
data can never be a bad thing, in order to build better models. Also the statistical
work could be deepened with more tests, different features, and more inferential
instead of merely descriptive or comparative statistics. In order to pursue a real
explicative theory much more data (especially the data in sound format) and work
are necessary.
As already stated and stressed, the implementation of the lyrical module to its
full extent should be mandatory in order to optimize this model. Such work would
require the coordination with other researchers already working with the lyrics of
323
fado, in order to have more empirical studies regarding their data. The work of
Toivanen et al. is a good inspiration in order to accomplish this task (Toivanen
et al., 2013).
At this point we will briefly discuss the concept of “performer profile” as an idea
that could, and certainly will, lead to another project too complex to detail here,
but that will be open for future research connected to this one. The main idea
behind the “performer profile” is that a score has a human dimension associated
with the interpretation of the instructions it carries, much like the projects that
lead to the KTH performance rules (Friberg et al., 2000, 2006). In our project
this was done by the means of the fifth module, which was inspired in the work of
Friberg et al. and implemented in a sort of intuitive way. However, it would be
much more interesting if these gestures, instead of being experimental, intuitive
and vaguely based on our perception of what the records are, were based on
real empirical data based on the typical behaviors of well-known fadistas. So,
if one could engage in a performance practice study based on recordings of real
performances made by, say, Amália Rodrigues or Alfredo Marceneiro, one could
build up a profile of those performances. Then it would be possible to program
the computer so that it would apply those profiles to the newly generated fados,
leading to a much more curious and interesting rendition of them. That, however,
is an idea that still requires some maturation, resources and a whole new set of
methods to be carried out. About performance practice methodology, however,
we can leave as a reference the chapter by Eric Clarke (Clarke & Cook, 2004).
18.2 Regarding Industry
A proof-of-concept modular generator that can recreate other styles was pro-
grammed. We have also demonstrated how a setup can be used to quickly generate
content suitable to be used in the music library industry in general, or allow the
individual composer to produce general abstractions, instead of individual com-
positions, thus reducing the overall time. We believe the next step would be to
test the market and develop applications, namely generative standalone applica-
tions, such as the ones devised by Brian Eno, already on the market, but this
time devoted to other markets. We believe the association of the generator with
the sample engine Alfama and a friendly graphic user interface can easily be used
in the market of background music in many Portuguese contexts where it makes
324
18.2. Regarding Industry
sense: monuments, restaurants and public places. We also believe the code and
architecture in itself is suitable to be replicated in open source software, like Com-
mon Music, to be taught in schools. We believe a new paradigm for the teaching
of music composition and music applications is now open.
325
Bibliography
Adams, C. R. 1976. Melodic contour typology. Ethnomusicology, 20(2),

179–215.
Almeida, Manuel António de. 1893. Memórias de um Sargento de Milı́cias.

Romance de Costumes Brasileiros. Rio de Janeiro: H. Garnier.
Almeida, Renato. 1926. História da música brasileira. Rio de Janeiro: F.

Briguiet et Comp. Eds.
Alpern, A. 1995. Techniques for algorithmic composition of music. Hampshire

College.
Alves, Oriana. 2003. Mais uma voz que nos visita: práticas e discursos do
fado - a Tasca do Chico como ponto de partida e de chegada. M.Phil. thesis,
FCSH/NOVA, Lisboa.
Ames, Charles, & Domino, Michael. 1992. Cybernetic composer: an

overview. Pages 186–205 of: Understanding music with AI: perspectives on
music cognition. MIT Press.
Anderson, Benedict. 2006. Imagined Communities: Reflections on the Origin

and Spread of Nationalism. New edition edn. Verso.
Antunes, Pedro Girão, de Matos, David Martins, Ribeiro, Ricardo,

& Trancoso, Isabel. 2014. Automatic Fado Music Classification. arXiv
preprint arXiv:1406.4447.
Ávila, Fernando. 2010. Trovas antigas, Saudade Louca: TV Series. Portugal:

RTP.
Balbi, Adriano. 1822. Essai statistique sur le royaume de Portugal et

d’Algarve: comparé aux autres états de l’Europe, et suivi d’un coup d’oeil sur
327
l’état actuel des sciences, des lettres et des beaux-arts parmi les Portugais des
deux hémisphères. Rey et Gravier.
Bandt, Ros, MacKinnon, Michelle Duffy, & Dolly. 2007. Hearing

Places: Sound, Place, Time and Culture. Cambridge Scholars Publishing.
Bardin, Laurence. 2009. Análise de Conteúdo. Edições 70.
Barreto, Mascarenhas. 1959. Fado: A canção Portuguêsa. Lisboa: Oficina

Gráfica Boa Nova.
Barreto, Mascarenhas. 1994. Fado: origens lı́ricas e motivação poética.

TFM.
Basharin, Gely P., Langville, Amy N., & Naumov, Valeriy A. 2004.
The life and work of A.A. Markov. Linear Algebra and its Applications,
386(Jul), 3–26.
Beckford, William, & Alexander, Boyd. 2005. The journal of William

Beckford in Portugal & Spain, 1787-1788. Nonsuch Publishing.
Beran, Jan. 2004. Statistics in musicology. Chapman & Hall/CRC.
Bernshtein, Sergei Natanovich. 1926. Sur les sommes de quantités

dépendantes. Izvestiya Rossiiskoi Akademii Nauk. Seriya Matematicheskaya,
20(15-17), 1459–1478.
Bernstein, Leonard. 1976. The Unanswered Question: Six Talks at Harvard.

Harvard University Press.
Blacking, John. 1973. How musical is man? University of Washington Press.
Bourgeois, Louise, Bernadac, Marie-Laure, & Obrist, Hans-Ulrich.

2001. Destruction of the father, reconstruction of the father: Schriften und
Interviews 1923-2000. Ammann.
Branco, Carlos, & Barreto, Mascarenhas. 1960. Portugal do fado.

Guimarães.
Brito, Joaquim Pais de (ed). 1994. Fado: vozes e sombras. Lisboa: Electa.
Brown, S., & Jordania, J. 2011. Universals in the world’s musics.

Psychology of Music, 41(2), 229–248.
328
BIBLIOGRAPHY
Cabral, Pedro Caldeira. 1982. Guitarra Portuguêsa. In: Instrumentos

Musicais Populares Portuguêses, Veiga de Oliveira, Ernesto edn. Lisboa:
Fundação Calouste Gulbenkian.
Carroll, Noel. 1990. The Philosophy of Horror: Or, Paradoxes of the Heart.
New York: Routledge.
Carvalho, Pinto de. 1903. História do fado [por] Pinto de Carvalho (Tinop).
Empreza da História de Portugal.
Carvalho, Rúben de. 1994. As músicas do fado. Campo das Letras.
Castelo-Branco, Salwa El-Shawan. 1994. Vozes e guitarras na prática

interpretativa do fado. Pages 125–141 of: Fado: vozes e sombras. Electa.
Castelo-Branco, Salwa El-Shawan. 1998. Voix du Portugal. Cité de la

Musique.
Castro, São Luı́s, & Lima, César F. 2010. Recognizing emotions in spoken
language: A validated set of Portuguese sentences and pseudosentences for
research on emotional prosody. Behavior Research Methods, 42(1), 74–81.
Chadabe, Joel. 1984. Interactive Composing: An Overview. Computer Music

Journal, 8(1), 22–27.
Chew, Elaine. 2000a. The Spiral Array: Description and Calibration of a

Model for Tonality.
Chew, Elaine. 2000b. Towards a Mathematical Model of Tonality. Ph.D.

thesis, MIT, USA.
Chew, Elaine. 2001. Modeling tonality: Applications to music cognition. 4–9.
Chomsky, Noam. 2002. Syntactic Structures. 2nd edition edn. New York: De
Gruyter Mouton.
Clarke, Eric, & Cook, Nicholas. 2004. Empirical Musicology: Aims,

Methods, Prospects. Oxford University Press, USA.
Colvin, Michael. 2008. The Reconstruction of Lisbon: Severa’s Legacy and

the Fado’s Rewriting of Urban History. Lewisburg: Bucknell University Press.
329
Conklin, Darrell, & Anagnostopoulou, Christina. 2011. Comparative

Pattern Analysis of Cretan Folk Songs. Journal of New Music Research, 40(2),
119–125.
Cook, Nicholas. 1994. A guide to musical analysis. Oxford University Press.
Cook, Nicholas (ed). 2004. Computational and comparative Musicology.

Pages 103–126 of: Cook, Nicholas (ed), Empirical musicology: aims,
methods, prospects.
Cope, David. 1991. Computers and musical style. A-R Editions.
Cope, David. 1996. Experiments in musical intelligence. A-R Editions.
Cope, David. 2000. The algorithmic composer. A-R Editions, Inc.
Cope, David. 2002a. Computer analysis and composition using atonal

voice-leading techniques. Perspectives of New Music, 121–146.
Cope, David. 2002b. Computer analysis and composition using atonal

voice-leading techniques. Perspectives of New Music, 121–146.
Cope, David. 2004. Virtual Music: Computer Synthesis of Musical Style. MIT
Press.
Cope, David. 2005. Computer models of musical creativity. MIT Press.
Cope, David. 2008. Hidden structure: music analysis using computers. A-R
Editions.
Cordeiro, Felisberto Ignacio Januario. 1827. Obras poeticas de Falmeno.

Typographia Imperial e Nacional.
Côrte-Real, Maria de São José. 2000. Cultural Policy and Musical

Expression in Lisbon in the Transition from Dictatorship to Democracy
(1960s-1980s). Ph.D. thesis, Columbia University, USA.
Côrte-Real, Maria de São José. 2002. Musical Priorities in the Cultural

Policy of Estado Novo. Revista Portuguesa de Musicologia, 227–252.
Côrte-Real, Maria de São José. 2010. Revising citizenship: migration and

fado in the play of identities in the United States. Migrações, 73–96.
330
BIBLIOGRAPHY
Côrte-Real, Maria de São José. 1991. Retention of Musical Models: Fado

Performance among Portuguese Migrants in New York. M.Phil. thesis,
Columbia University, USA.
Costa, António Firmino da, & Guerreiro, Maria das Dores. 1984. O
trágico e o contraste: o fado no bairro de Alfama. Publicações Dom Quixote.
Cristo, Nuno Jose Dos Santos Anaia. 2014 (Jan). The Portuguese Guitar:
History and Transformation of an Instrument Associated with Fado. M.Phil.
thesis, York University, Toronto, Canada.
Deutsch, Diana (ed). 2012. The Psychology of Music. 3rd edn. Academic
Press.
Deutsch, Diana, Le, Jinghong, Shen, Jing, & Henthorn, Trevor.

2009. The pitch levels of female speech in two Chinese villages. The Journal of
the Acoustical Society of America, 125(5), EL208.
Dias, Rui, & Guedes, Carlos. 2012. GimmeDaBlues app for iOS: overview
and ongoing developments. Pages 2:1–2:1 of: SIGGRAPH Asia 2012
Symposium on Apps. SA ’12. New York, NY, USA: ACM.
Dias, Rui, Marques, Telmo, Sioros, George, & Guedes, Carlos. 2012.
GimmeDaBlues: An Intelligent Jazz/Blues Player And Comping Generator for
iOS devices.
Diaz-Jerez, Gustavo. 2011. Composing with Melomics: Delving into the

Computational World for Musical Inspiration. Leonardo Music Journal, -(21),
13–14.
Dolson, Mark. 1994. The Pitch of Speech as a Function of Linguistic

Community. Music Perception: An Interdisciplinary Journal, 11(3), 321–331.
Doyle, Richard. 1997. On beyond living: rhetorical transformations of the life

sciences. Stanford University Press.
Dreyfus, Hubert L., & Dreyfus, Stuart E. 1988. Making a Mind versus
Modeling the Brain: Artificial Intelligence Back at a Branchpoint. Daedalus,
117(1), 15–43.
Eerola, Tuomas, & Toiviainen, Petri. 2004. MIR In Matlab: The MIDI
Toolbox. In: ISMIR. Citeseer.
331
Elliott, Richard. 2010. Fado and the Place of Longing: Loss, Memory and
the City. Ashgate Publishing, Ltd.
Eno, Brian. 1996. Generative Music. In: Transcript of a talk.
Essl, Karlheinz. 1992. Lexikon-Sonate - algorithmic music generator.
Félix, James. 2015. Folk or Fake: The Notion of Authenticity in Portuguese

Fado. Ph.D. thesis, University of Leeds, UK.
Félix, Pedro. 2013. O Espectáculo Músico Teatral - Ensaio em dois actos com
um prólogo. Pages 25–79 of: O Fado e o Teatro. Museu do Fado/Museu
Nacional do Teatro.
Fernández, Jose David, & Vico, Francisco. 2013. AI methods in

algorithmic composition: a comprehensive survey. Journal of Artificial
Intelligence Research, 513–582.
Fonseca, Ricardo Xavier Machado Silva. 2011. O novo fado: uma leitura
transcultural. M.Phil. thesis, Universidade do Minho, Portugal.
Freitas, Frederico de. 1973. O Fado, Canção da cidade de Lisboa. Lı́ngua e

Cultura, 225–237.
Freitas, Manuela de. 2013. Ofı́cios Inquietos. Pages 133–137 of: O Fado e o
Teatro. Museu do Fado/Museu Nacional do Teatro.
Freycinet, Louis-Claude Desaulces de, Persoon,

Christiaan Hendrik, Agardh, Carl Adolf, Freycinet, Louis-René
Desaulces de, Schewaerichen, & Lamothe, Félix. 1826. Voyage
autour du monde: entrepris par ordre du Roi [...], exécuté sur les corvettes de
S.M. l’Uranie et la Physicienne, pendant les années 1817, 1818, 1819 et 1820.
Friberg, Anders, Colombo, Vittorio, Frydén, Lars, & Sundberg,

Johan. 2000. Generating musical performances with Director Musices.
Computer Music Journal, 24(3), 23–29.
Friberg, Anders, Bresin, Roberto, & Sundberg, Johan. 2006.

Overview of the KTH rule system for musical performance. Advances in
Cognitive Psychology, 2(2-3), 145–161.
Frith, Simon. 1998. Performing Rites: On the Value of Popular Music.

Harvard University Press.
332
BIBLIOGRAPHY
Frota, Sónia. 2000. Prosody and Focus in European Portuguese: Phonological

Phrasing and Intonation. Psychology Press.
Geertz, Clifford. 1973. The Interpretation of Cultures: Selected Essays.

Basic Books.
Geertz, Clifford. 1983. Local Knowledge: Further Essays in Interpretive

Anthropology. Basic Books.
Gogins, M. 1991. Iterated functions systems music. Computer Music Journal,

15(1), 40–48.
Gouveia, Daniel. 2010. Ao Fado tudo se canta? Linda-a-Velha: DG Edições.
Gray, Lila Ellen. 2007. Memories of Empire, Mythologies of the soul: Fado
performance and the shaping of Saudade. Ethnomusicology, 51(1), 106–130.
Gray, Lila Ellen. 2011. Fado’s City. Anthropology and Humanism, 36(2),
141–163.
Gray, Lila Ellen. 2013. Fado Resounding: Affective Politics and Urban Life.
Duke University Press.
Grier, James. 1996. The Critical Editing of Music: History, Method, and
Practice. Cambridge University Press.
Guedes, Carlos, Pennycook, Bruce, & Henriques, Tomás. 2009.

Kinetic controller driven music systems.
Guerra, Maria Luı́sa. 2003. Fado: alma de um povo?; (origem histórica).

Imprensa nacional-Casa da Moeda.
Guimarães, Pinto (ed.). 1938. Fado, canção vencedora. História, propaganda

e defesa do fado.
Guinot, Maria, Carvalho, Ruben de, & Osório, José Manuel. 1999.
Um século de fado. Ediclube.
Hartin, Patrick J., & Petzer, J. H. 1991. Text and interpretation: new
approaches in the criticism of the New Testament. Brill.
Hatten, Robert S. 2004. Interpreting Musical Gestures, Topics, And Tropes:

Mozart, Beethoven, Schubert. Indiana University Press.
333
Hatten, Robert S. 2006. A theory of Musical Gesture and its Application to

Beethoven and Schubert. Pages 1–23 of: Music And Gesture, Gritten,
Anthony and King, Elaine edn. UK: Ashgate Publishing, Ltd.
Hayes, Brian. 2001. Randomness as a Resource. American Scientist, 89(4),

300.
Hayles, N. Katherine. 1999. How we became posthuman: virtual bodies in

cybernetics, literature, and informatics. University of Chicago Press.
Holt, Fabian. 2007. Genre in Popular Music. University of Chicago Press.
Holton, Kimberly da Costa. 2006. Fado Historiography: Old Myths and

New Frontiers. P: Journal of Portuguese Cultural Studies (Winter), 1–17.
Hood, Mantle. 1960. The Challenge of “Bi-Musicality”. Ethnomusicology,

4(2), 55.
Huron, David. 1996. The Melodic Arch in Western Folksongs. Computing in

Musicology, 10, 3–23.
Huron, David. 2008. Sweet Anticipation: Music and the Psychology of

Expectation. Cambridge, Mass.; London: A Bradford Book.
Järveläinen, H. 2000. Algorithmic musical composition. Seminar on content

creation, Telecommunications software and multimedia laboratory, Helsinki
University of Technology.
Jeon, Yong-Woo, Lee, In-Kwon, & Yoon, Jong-Chul. 2006. Generating

and modifying melody using editable noise function. Pages 164–168 of:
Computer Music Modeling and Retrieval. Springer.
Kivy, Peter. 2001. New Essays on Musical Understanding: The Arousal

Theory of Musical Expression: Rethinking the Unthinkable. Oxford University
Press.
Kolinski, Mieczyslaw. 1965. The general direction of melodic movement.

Ethnomusicology, 9(3), 240–264.
Kolinski, Mieczyslaw. 1973. A cross-cultural approach to metro-rhythmic

patterns. Ethnomusicology, 17(3), 494–506.
334
BIBLIOGRAPHY
Kolinski, Mieczyslaw. 1982. Reiteration Quotients: A Cross-Cultural

Comparison. Ethnomusicology, 26(1), 85–90.
Langston, P. S. 1989. Six techniques for algorithmic music composition.

Proceedings of the International Computer Music Conference (ICMC).
Leitão da Silva, João. 2011 (Sep). Music, theatre and the nation: The
entertainment market in Lisbon (1865-1908). Ph.D. thesis, Newcastle
University, UK.
Leman, Marc. 2008. Embodied Music: Cognition and Mediation Technology.

MIT Press.
Leman, Marc. 2012. Musical gestures and embodied cognition. Pages 5–7 of:
Journées d’informatique musicale (JIM-2012).
Lerdahl, Fred. 2005. Tonal Pitch Space. Oxford University Press US.
Lerdahl, Fred, Jackendoff, Ray, & Jackendoff, Ray S. 1996. A

generative theory of tonal music. MIT Press.
Lomax, Alan. 1978. Folk song style and culture. New Brunswick, N.J.:
Transaction Books.
Losa, Leonor. 2013. Machinas Fallantes, A Música Gravada em Portugal no

Inı́cio do Século XX. Lisboa: Tinta da China.
Machado, Alberto Victor. 1937. Os ı́dolos do Fado: Biografias,

Comentários, Antologia. Lisboa: Tipografia Gonçalves.
Margulis, Elizabeth Hellmuth. 2005. A Model of Melodic Expectation.

Music Perception: An Interdisciplinary Journal, 22(4), 663–714. ArticleType:
review-article / Full publication date: Summer 2005 / Copyright 2005
University of California Press.
McCulloch, Warren S., & Pitts, Walter. 1943. A logical calculus of the
ideas immanent in nervous activity. The bulletin of mathematical biophysics,
5(4), 115–133.
McKay, Cory. 2004. Automatic Genre Classification of MIDI Recordings.

M.Phil. thesis, McGill University, Montréal, Canada.
335
McKay, Cory. 2010. Automatic music classification with jMIR. Ph.D. thesis,
McGill University, Montréal, Canada.
Melucci, M., & Orio, N. 2002. Evaluating automatic melody segmentation

aimed at music information retrieval. Proceedings of the 2nd ACM/IEEE-CS
joint conference on Digital libraries, 311.
Mendes, Ana P., Rodrigues, Aira F., & Guerreiro, David Michael.
2013. Acoustic and Phonatory Characterization of the Fado Voice. Journal of
Voice, 27(5), 655.e9–655.e15.
Merriam, Alan P. 1964. The Anthropology of Music. Northwestern University

Press.
Miranda, Eduardo Reck. 1993. Cellular Automata Music: An

Interdisciplinary Project. Interface, 22(1), 3–21.
Miranda, Eduardo Reck. 2001. Composing music with computers. Focal

Press.
Miranda, Eduardo Reck, & Biles, John (eds). 2007. Evolutionary

Computer Music. 2007 edition edn. London: Springer.
Moita, Luiz. 1936. O fado, canção de vencidos ...: Oito palestras na Emissora
nacional. Oficinas Gráficas da Emprésa do Annuário Comercial.
Moore, Isabel. 1902. Portuguese Folk-Songs. The Journal of American

Folklore, 15(58), 165–169.
Moreira, Pedro Russo. 2012 (Dec). Cantando espalharei por toda a parte:
programação, produção musical e o “aportuguesamento” da “música ligeira”
na Emissora nacional da Radiodifusão (1934-1949). Ph.D. thesis,
FCSH/NOVA, Lisboa.
Morris, R. D. 1993. New directions in the theory and analysis of musical

contour. Music Theory Spectrum, 15(2), 205–228.
Moura, Ana. 2012. Desfado (CD). Universal Portugal.
Naveda, Luiz, Gouyon, Fabien, Guedes, Carlos, & Leman, Marc.

2011. Microtiming Patterns and Interactions with Musical Properties in
Samba Music. Journal of New Music Research, 40(3), 225–238.
336
BIBLIOGRAPHY
Negus, Keith. 1997. Popular Music in Theory: An Introduction. Wesleyan

University Press.
Nery, Rui Vieira. 2004. Para uma história do Fado. Público.
Nery, Rui Vieira. 2010a. Fado. Pages 433–453 of: Branco,

Salwa Castelo (ed), Enciclopédia da Música em Portugal no Século XX.
Temas e Debates.
Nery, Rui Vieira. 2010b. Fado Maior e Menor. Jornal de Letras, Aug.
Nery, Rui Vieira. 2010c. Pensar Amália. Tugaland.
Nery, Rui Vieira, & Mateus, Paula. 2012. Fados para a república.
Imprensa Nacional-Casa da Moeda.
Nettl, Bruno. 2005. The study of ethnomusicology: thirty-one issues and

concepts. University of Illinois Press.
Neves, César das. 1893. Cancioneiro de musicas populares: contendo letra e

musica de canções, serenatas, chulas, danças, descantes ... e cançonetas
estrangeiras vulgarisadas em Portugal. Typographia occidental.
Nierhaus, Gerhard. 2008. Algorithmic composition. Springer.
Nketia, J. H. Kwabena. 2002. Musicology and Linguistics: Integrating the

Phraseology of Text and Tune in the Creative Process. Black Music Research
Journal, 22(2), 143.
Olson, Harry F. 1967. Music, Physics and Engineering. Revised edition edn.
New York: Dover Publications.
Ortigão, Ramalho. 1898. O Fadista. As Farpas, V(7).
Osório, António. 1974. A mitologia fadista. Livros Horizonte.
Pachet, François, & Roy, Pierre. 2014. Non-Conformant Harmonization:

the Real Book in the Style of Take 6. In: Proceedings of ICCC 2014.
Pachet, François, Roy, Pierre, & Ghedini, Fiammetta. 2013.

Creativity through Style Manipulation: the Flow Machines project. In: 2013
Marconi Institute for Creativity Conference, Proc.(Bologna, Italy, vol. 80.
337
Papadopoulos, G., & Wiggins, G. 1999. AI methods for algorithmic

composition: A survey, a critical view and future prospects. In: AISB
Symposium on Musical Creativity.
Parreira, António. 2014. O livro dos Fados - 180 fados tradicionais em

partituras. Lisboa: Museu do Fado.
Patel, Aniruddh D. 2008. Music, Language, and the Brain. Oxford

University Press.
Patel, Aniruddh D. 2014. The Evolutionary Biology of Musical Rhythm:

Was Darwin Wrong? PLoS Biology, 12(3), e1001821.
Pattison, Pat. 2010. Writing Better Lyrics. Writer’s Digest Books.
Paulus, J., & Klapuri, A. 2006. Music structure analysis by finding repeated
parts. Page 68 of: Proceedings of the 1st ACM workshop on Audio and music
computing multimedia.
Peixoto, Rocha. 1896. O cruel e triste fado. Figueira: Imprensa Lusitana.
Pereira, Sara de Melo Barata. 2015. Ecos do silêncio?: para um estudo

iconológico do fado. Ph.D. thesis, Faculdade de Letras/Universidade de
Lisboa, Portugal.
Pestana, Maria do Rosário. 2012. Armando Leça e a Música Portuguesa,

1910-1940. Tinta-da-China.
Pimentel, Alberto. 1904. A triste canção do sul (subsidios para a historia do

fado). Lisboa: Gomes de Carvalho.
Pinto, João Ricardo. 2004. Marchas Populares de Lisboa: a performação

musical na identidade bairrista. M.Phil. thesis, FCSH/NOVA, Lisboa.
Quick, Donya, & Hudak, Paul. 2013. Grammar-based Automated Music

Composition in Haskell. Pages 59–70 of: Proceedings of the First ACM
SIGPLAN Workshop on Functional Art, Music, Modeling & Design.
FARM ’13. New York, NY, USA: ACM.
Rice, Timothy. 1987. Toward the Remodeling of Ethnomusicology.

Ethnomusicology, 31(3), 469–488.
338
BIBLIOGRAPHY
Rice, Timothy. 2003. Time, Place, and Metaphor in Musical Experience and
Ethnography. Ethnomusicology, 47(2), 151–179.
Rice, Timothy. 2013. Ethnomusicology: A Very Short Introduction. 3rd edn.

New York: Oxford University Press.
Roads, Curtis. 1996. The computer music tutorial. MIT Press.
Robinson, Jenefer, & Hatten, Robert S. 2012. Emotions in Music. Music

Theory Spectrum, 34(2), 71–106.
Rodrigues da Costa, Liliana. 2012 (Oct). Fado - Matriz para uma (nova)
Polı́tica Cultural Externa. Uma estratégia Cultural ao serviço de Portugal.
Ph.D. thesis, Insituto Superior de Ciências Sociais e Polı́ticas, Lisboa.
Rohrmeier, Martin. 2011. Towards a generative syntax of tonal harmony.

Journal of Mathematics and Music, 5(1), 35–53.
Rowe, Robert. 1999. The aesthetics of interactive music systems.

Contemporary music review, 18(3), 83–87.
Ruiz, Ana. 2007. Vibrant Andalusia: The Spice of Life in Southern Spain.
Algora Publishing.
Sá de Oliveira, Múcio. 2014. O Papel da Viola no Fado de Lisboa. M.Phil.

thesis, FCSH/NOVA, Lisboa.
Santos, Pedro Rafael Pavão dos. 2015 (Jan). O fado e as artes?: um

século de cumplicidades e ambiguidades. Ph.D. thesis, FCSH/NOVA, Lisboa.
Sardinha, José Alberto. 2010a. A Origem do Fado. Vila Verde: Tradisom.
Sardinha, José Alberto. 2010b. Polémica As origens do Fado. Jornal de

Letras.
Schlichthorst, Carl. 1829. Rio de Janeiro wie es ist: Beitrge zur tagesund
sitten-geschichte der hauptstadt von Brasilien mit vorzglicher vcksicht auf die
lage des dortigen deutschen militairs. Hahn.
Schulz, Florian, Geiger, Chris, & Reckter, Holger. 2009. ANTracks:

generative mobile music composition. Pages 302–305 of: Proceedings of the
International Conference on Advances in Computer Enterntainment
Technology. Athens, Greece: ACM.
339
Sebastião, Liliana. 2009. O fado no bairro de Alfama: um estudo de caso.

M.Phil. thesis, FCSH/NOVA, Lisboa.
Seeger, Anthony. 2002. A Tropical Meditation on Comparison in

Ethnomusicology: A Metaphoric Knife, a Real Banana and an Edible
Demonstration. Yearbook for Traditional Music, 34, 187.
Silva, Patrı́cio da. 2003. David Cope and Experiments in Musical Intelligence.
Spectrumpress.com.
Sioros, George, Holzapfel, André, & Guedes, Carlos. 2012. On

Measuring Syncopation to drive an interactive music system.
Stammen, D., & Pennycook, Bruce. 1993. Real-time Recognition of

Melodic Fragments Using the Dynamic Timewarp Algorithm. In: Proceedings
of the International Computer Music Conference.
Stammen, D., & Pennycook, Bruce. 1994. Real time segmentation of music
using an adaptation of Lerdahl and Jackendoff’s grouping principles. In: Proc.
of the 3rd. Intl. Conf. on Music Perception and Cognition.
Steedman, M. 1996. The blues and the abstract truth: Music and mental
models. Mental Models in Cognitive Science: Essays in Honour of Phil
Johnson-Laird.
Tinhorão, José Ramos. 1994. Fado, dança do Brasil, cantar de Lisboa: o fim
de um mito. Editorial Caminho.
Toivanen, Jukka M., Toivonen, Hannu, & Valitutti, Alessandro.

2013. Automatical Composition of Lyrical Songs. Page 87 of: Proceedings of
the Fourth International Conference on Computational Creativity.
Tojo, Satoshi, Hirata, Keiji, & Hamanaka, Masatoshi. 2013.

Computational Reconstruction of Cognitive Music Theory. New Generation
Computing, 31(2), 89–113.
Tsay, Chia-Jung. 2013. Sight over sound in the judgment of music

performance. Proceedings of the National Academy of Sciences, 110(36),
14580–14585. PMID: 23959902.
Tschmuck, Peter. 2010. Creativity and Innovation in the Music Industry.

Springer.
340
BIBLIOGRAPHY
Vernon, Paul. 1998. A history of the Portuguese fado. Ashgate.
Videira, Tiago. 2007. O Artista Cibernético - O estatuto do artista na música

de vanguarda - Análise de dois estudos de caso: a Lexikon-sonate e o EMI.
M.Phil. thesis, FCSH/NOVA, Lisboa.
Videira, Tiago. 2011. Describing Music: Perception and Metaphor. The

international Journal of the Arts in Society, 6(4), 255–262.
Vieira, Ernesto. 1890. Diccionario musical contendo todos os termos

technicos, com a etymologia da maior parte d’elles ... Gazeta Musical de
Lisboa.
Weech, Joseph F. von. 1831. Reise über England und Portugal nach
Brasilien und den vereinigten Staaten des La-Plata-Stromes während den
Jahren 1823 bis 1827. Auer.
Wiering, Frans, de Nooijer, Justin, Volk, Anja, &

Tabachneck-Schijf, Hermi J.M. 2009. Cognition-based Segmentation for
Music Information Retrieval Systems. Journal of New Music Research, 38(2),
139–154.
Winkler, Todd. 2001. Composing interactive music: techniques and ideas

using Max. MIT Press.
Wittgenstein, Ludwig. 2010. Philosophical Investigations. John Wiley &

Sons.
Wittgenstein, Ludwig, Wright, Georg Henrik, & Nyman, Heikki.

1984. Culture and value. University of Chicago Press.
Woods, W. A. 1970. Transition Network Grammars for Natural Language

Analysis. Commun. ACM, 13(10), 591–606.
Zagorski-Thomas, Simon. 2005. Shouting Quietly: Changing musical

meaning by changing timbre with recording technology. In: Actes du Colloque
interdisciplinaire de musicologie (CIM05).
Zagorski-Thomas, Simon. 2012. Musical Meaning and the Musicology of

Record Production. Beitraege zur Popularmusikforschung, 38.
341
Appendix A
Supplemental Material
Included in this dissertation is:
1. The References for all the 100 fados transcribed
2. The musical scores of the Urtext corpus
3. The musical scores of the Edition corpus
4. Statistics on the 100 fados
Further Supplemental Material can be retrieved on:
http://tiagovideira.com/2015/06/07/instrumental-fado
343
ID Fonte Data Compositor Designação Comentário da Fonte Comentário Pessoal Comentário Filológico Cp. Andamento Tom Compassos Forma Progressão
1 Pimentel, A., A Triste Canção 1840 Fado do Marinheiro Parece ser dos primeiros que se vulgarizaram em 2/4 Ré menor 16 compassos AA'BB' || (Dm) i || V7 || V7 || i ||
do Sul , Livraria Central de Lisboa, segundo informa o velho guitarrista Ambrósio || i || V7 || V7 || i ||
Gomes de Carvalho, Lisboa, Fernandes Maia. Ele não tem ideia de outro qualquer || i || V7 || i-V7 || i ||
1904, p. 35 fado mais antigo.
2 1. Pimentel, A., A Triste 1845 Fado Corrido O Cancioneiro de músicas Populares diz que este fado Transcrição descritiva. 2/4 Andante Fá menor – Fá 44 Compassos I1I2 AA B CC' D O || (F) I || V7 || V7 || I ||
Canção do Sul , Livraria já era popularíssimo em 1870, e dá-lhe a letra de maior || ii || V7 || I-V7 || I ||
Central de Gomes de algumas quadras que andam na tradição oral. Mas o || I || V7 || V7 || I ||
Carvalho, Lisboa, 1904, p. Corrido não é mais que o simples acompanhamento do || I || V7 || V7 || I ||
862. Neves, César das, canto. Sobre este tipo melódico têm sido bordadas
Cancioneiro de Músicas muitas variantes por Alexandre Rey Colaço, Reynaldo
Populares , vol.3, Empresa Varella, Militão e outros, incluindo um compositor
Editora, Porto, 1898, pp. 186- estrangeiro, Munier.
188 O Fado Corrido anda em todas as colecções.
3 1. Pimentel, A., A Triste 1845 Fado Choradinho O Cancioneiro de musicas populares (fase. 18) traz Transcrição descritiva.Na 2/4 Andante Ré Menor 32 Compassos AABB CC CC || (Dm) i || V7 || V7 || i || (2x)
Canção do Sul , Livraria este popularissimo Fado com a seguinte annotação: fonte não aparece a (4+4+4+4 || i || iv || V7 || i || (2x)
Central de Gomes de «Recolhido em Lisboa, em 1850. Este é um dos fados repetição de CCNa fonte +4+4+4+4) || i || V7 || V7 || i || (4x)
Carvalho, Lisboa, 1904, p. 139 propriamente ditos, e dos mais antigos, por onde se apenas A aparece à oitava
2. Neves, César das, moldaram outros muitos que posteriomente na repetiçãoNa fonte o 2º
Cancioneiro de Músicas appareceram.» tempo do 4º compasso
Populares , vol.1, Typographia aparece com pausa no
Occidental, Porto, 1893, p. Não ha duvida que é, chronologicamente, dos acompanhamento.
217 primeiros. Mas o primeiro decerto não foi. (Vide Fado
do marinheiro). Bem podia ele ter estabelecido o tipo
deste género de canções a julgar pela sua grande
espontaneidade de sentimento, singela e profunda.
Todavia a designação de Fado choradinho parece
indicar que já havia outros, e que este se distinguia por
um tom ainda mais plangente, donde lhe viera o nome.
Em verdade, dir-se-ia um rosário musical feito de
gemidos e suspiros.
O Cancioneiro dâ-lhe a lettra de algumas quadras

populares, tais como esta:
Quem tiver filhas no mundo

Não fale das malfadadas:
Porque as filhas da desgraça
Também nasceram honradas.
É claro que se lhe pôde applicar qualquer outra

letra.Este Fado está vulgarisadissimo.
Em Lisboa anda nas collecções das casas Sassetti,
Engestrom, Lambertini, etc.
No Porto, Moreira de Sá escreveu sobre ele variações

para piano, para rabeca, bandolim ou flauta com
acompanhamento de piano ou de violão e guitarra.
Em Lisboa, Rey Colaço também glosou o Fado

choradinho; dizia o jornal Novidades, de 31 de janeiro
de 1903:
«Foi agora posto á venda mais um trabalho do illustre

pianista Rey Colaço, uma das mais poderosas
organisações artisticas do nosso paiz. É um fado
(choradinho), e, como todas as producções inspiradas
na melopeia que é o característico da nossa raça, esta
de
Rey Colaço é cheia d'uma melancolia de poente do

outono, terna e triste como uma despedida.
«O delicioso fado, que é dedicado ao sr. Raul Lino e

estampa no frontispicio a reproducção da casa
portugueza que aquelle architecto anda a construir
para Rey Colaço, está á venda em todos os armazéns
de musica.»
4 1. Pimentel, A.,"A Triste 1848 Sousa do Casacão Fado da Severa A letra do Fado refere "Hoje Transcrição descritiva. 2/4 Allegretto Lá menor 3 + 32 Compassos Intro ABAB(3 + || (Am) i || V7 || i ||
Canção do Sul", Livraria mesmo faz um ano que a 8+8+8+8) || i || V7 || V7 || i ||
Central de Gomes de Severa faleceu". Sendo a || i || V7 || i-V7 || i || (2x)
Carvalho, Lisboa, 1904, p. data da sua morte apontada
1582. Neves, Céssar das, como 1846, estimo que este
Cancioneiro de Músicas fado não seja muito
Populares , vol.3, Empresa posterior.
Editora, Porto, 1898, p. 127
5 1. Pimentel, A., A Triste 1864 Sales Patuscão Fado do Conde de O Fado do Vimioso foi publicado no fascículo 61 do 2/4 Andantino Lá menor 20 Compassos Intro AABB' || (C) I || V || V || (Am) i ||
Canção do Sul , Livraria Vimioso Cancioneiro de musicas populares portuguesas. (4+4+4+4+4) || i || V7 || V7 || i || (4x)
Central de Gomes de Acompanham-no algumas quadras, de origem
Carvalho, Lisboa, 1904, p. manifestamente popular, porque são incorrectas e
1832. Neves, César das, banais.
Cancioneiro de Músicas
Populares , vol.3, Empresa O conde é ahi tratado pelo anagramma de Moisivo.
A lettra é muito posterior á morte de D. Francisco de
Paula, como se vè da seguinte quadra, penúltima do
Fado:
Aqui ponho agora ponto.
Na lenda que finda está :
Foram casos d'oulra era,
SSo voltas que o mundo dá.
O auctor da lettra explora a lenda (elle mesmo

emprega esta palavra) da «paixão» do conde pela
Severa, cujos encantos avulta com poética phantasia:
Quem lhe vê a lace morena,
Quem vê seus olhos tyrannos,
Nada vê que mais captive
lnda que viva mil annos.
Quem uma vez lhe ouviu
Sua voz enternecida,
Ainda depois da morte
Aos seus ais recobra a vida.
Quem a vê dançar o fado
Com rigor desconhecido,
Ao vel-a batendo forte
liça um doido pt-rdido.
E, continuando a lenda, falia do conde como de um
cego amante que tivesse morrido de amor:
Assim Moisivo carpia
No auge da desventura.
E ao outro dia, já cadáver,
Foi levado á sepultura.
Chorae, fadistas, chorae,
Ah! chorai a mais não ser,
Que d'outro tão fino amante
Não torna o fado a dizer
O que é certo é que a lenda da Severa e do conde de

Vimioso, tal como a musa popular a foi cantando ao
sabor das multidões, estimulou a corrente que
vulgarisou o Fado, especialmente no sul do paiz, e que
lhe reforçou um caracter de vaga saudade, de tristeza
plangente, em que parece pairar a longínqua memoria
de uma supposta allucinação amorosa que um fidalgo
bohemio experimentou por uma pobre moça fadista, de
chinella de polimento ponteada a retroz vermelho.
Todas as mulheres dos bairros iufamados, todas as

criadas de servir, todas as camareras de botequim
cantam de preferencia o Fado da Severa e o Fado do
conde de Vimioso, dando-lhe uma intenção de
aristocracia rehabilitadora pela esperança de que um
novo conde, seguindo o exemplo de D. Francisco de
Paula, venha enamorado, dedilhar a banza, em honra
de uma segunda Severa plebea.
«Ainda hoje, diz Pinheiro Chagas, (1) se ouve cantar a

deshoras, com acompanhamento de guitarras, por
vozes nem sempre da primeira frescura, uma melodia
melancólica e plangente, que se denomina o
fado do conde de Vimioso..»
(1) Dicc. Pop., vocab. Vimioso.

6 1 .Pimentel, A., A Triste 1890 Augusto Hylario Fado Serenata Cancioneiro, fase. 23. Fado serenata. Traz a Fado de Coimbra. Transcrição descritiva. 2/4 Andante Fá maior 3 + 32 compassos Intro AABC AABC || (F) I || V7 || I ||
Canção do Sul , Livraria designação de: Musica de Augusto Hylario. Progressão harmónica mais (3 + 4+4+4+4 || I || V7 || V7 || I || (2x)
Central de Gomes de complexa e divergente do +4+4+4+4) || (Fm) i || V7 || ii || V7 ||
Carvalho, Lisboa, 1904, p. 226 Foi este que ouvimos ao próprio auctor em Espinho, e tipo de Ernesto Vieira. || ii-ii0 || i || (F) V7 || I ||
o que, de todos que elle indubitavelmente compoz, se || I || V7 || V7 || I || (2x)
2. Neves, César das, tornou mais popular. || (Fm) i || V7 || ii || V7 ||
Cancioneiro de Músicas || ii-ii0 || i || (F) V7 || I ||
Populares , vol.1, Typographia
Occidental, Porto, 1893, p.
280
7 1. Pimentel, A., A Triste 1895 Augusto Hylario O último Fado Classificado como Serenata em sub-título. Fado de Coimbra. Na fonte, é omisso o bemol 2/4 Andante Ré menor - Fá 2 + 32 compassos Intro AA BB CC DD || (Dm) i-V7 || i ||
Canção do Sul , Livraria Progressão harmónica mais no Ré ao 28º compasso, e Expressivo Maior - Sib menor (2+ 4+4 +4+4 || i || iv || V7 || i || (2x)
Central de Gomes de Cancioneiro, fase. 34, «O Ultimo Fado», com a complexa que o tipo de no Mi ao 29º compasso. +4+4+4+4) || (F) #iv07 || I || ii-V7 || I || (2x)
Carvalho, Lisboa, 1904, p. designação de Musica de Augusto Hylario e a seguinte Ernesto Vieira. || (Bbm) i || V7 || V7 || i || (2x)
2282. Neves, César das, nota: «Quando nas férias de 1895, Hylario se || i || vii07 || V7 || i || (2x)
Cancioneiro de Músicas hospedou em uma dependência do escriptorio da
Populares , vol.2, Empresa
Editora, Porto, 1895, pp. 102- nossa Empreza, offereceu-nos esta composição
103 dizendo-nos que era o seu ultimo fado, mas que
tencionava addicionar-lhe algumas variações, e que
reservássemos a publicação para quando elle as
tivesse composto definitivamente.»
8 Neves, César das, 1868 João Maria dos Fado das Salas A música do fado já não é hoje, como foi outr ora, O termo passa-calle aqui 2/4 Andante Sol menor 32 + 2 compassos. A A A B || i || V7 || V7 || i ||
Cancioneiro de Músicas Anjos considerada musica torpe e obscena, propria só das utilizado pelo autor para Coda(8+8+8+8+ 2) || i || V7 || i-V7 || i || (3x)
Populares , vol.1, Typographia viellas e dos antros de vício, onde a maruja e descrever o fado refere-se || i || V7 || V7 || i ||
Occidental, Porto, 1893, soldadesca embriagada, tangiam brutalmente em provavelmente à acepção || i || ii0 || i-V7 || i ||
pp.30-31 banzas immundas, acompanhando-a com indecorosos de passacaglia no século || i || i ||
versos e batiam com danças lascivas. XIX: isto é, uma harmonia
no baixo e voz grave em
Ha quarenta annos já se faziam fados especiaes, ou ostinato que se repete
para narrar crimes ou algum escandalo amoroso, indefinidamente durante a
satyrizar homens celebres ou políticos importantes, ou peça inteira,
para rebaixar homens altamente collocados, ou para complementada com uma
ridicularizar corporações respeitaveis, ou para melodia na voz aguda, mais
descompor qualquer sujeito. Eis um exemplo: o ou menos livre, em secções
fallecido jornalista e poeta satyrico Urbano Loureiro, regulares de quatro, seis ou
comparando um dia no jornal a Lucta os versos d um oito compassos.
escriptor nosso contemporaneo, aos fados do
Marcolino (um pobre musico ambulante, improvisador
de fados), este sabendo da comparação, procurou o
jornalista n um estabelecimento da rua de Santo
Antonio, que elle costumava frequentar e deixou-lhe o
seguinte recado: - «Diga a esse snr. Urbanes Loureiro
que se me torna, na sua gazeta, a comparar a esse
outro snr. poeta, eu faço-lhe um fado que o... arrazo.»
Publicada na agua lustral da civilisação, a monotona

musica dos fados, alegre ou sentimental, ingenua ou
luxuriante, já não ofende os castos ouvidos do bello
sexo, que a perfuma sob seus dedos sem se roborisar
de pudor, porque essas simples melodias não maculam
as candidas rosas da sua alma pura. É que a ideia do
seu ponto de partida tem sido substituida pela
assimilaçao do serio e pelo sentimento artístico que
despreza a materialidade da sua primitiva aplicação e a
propria poesia sensualista.
O grande numero de fados, quasi todos variantes uns

dos outros, que se improvisam todos os dias, não são
mais que uma espécie de passa-calle, lento, de musica
caracteristicamente portugueza: a muitos d elles nunca
os seus auctores applicaram lettra.
9 Neves, César das, 1860 José Maria dos Melodia Popular D Esta musica é vulgarmente conhecida pelo nome de Primeiro fado recolhido 2/2 Andante Fá menor 28 compassos A A B(8+12+8) || (Bb) i || V7 || V7 || i ||
Cancioneiro de Músicas Cavalinhos Anadia Fado de Anadia; é uma das musicas no estylo moderno escrito em 2/2.Variante || i || V7 || i-V7 || i ||
Populares , vol.1, Typographia do genero mais distincta e nao monotona. A poesia que irregular de 12 compassos || i || V7 || V7 || i || (2x)
Occidental, Porto, 1893, lhe applicamos, por nao conhecermos lettra propria, é por repetição de uma célula || i || V7 || i-V7 || i || (3x)
pp.58-59 antiga mas merece aceitação, por ser conceituisissima. de 4 compassos do tema
A.É curioso o autor
classificá-la como mais
distinta e não monótona por
comparação já que
harmonicamente o fado
apenas se sustém com 2
acordes sempre iguais, sem
cambiantes rítmicas ou
melódicas.
10 Neves, César das, 1885 Hija del Gudalquivir Recolhida em Alcobaça em 1885, por F. Pinto Não soa nada a As ligaduras no primeiro, 4/4 Andante Sib maior 28 compassos AAB(8+8+12) || (Bb) I || ii || V7 || I || (4x)
Cancioneiro de Músicas Nogueira. fado.Atentar no desenho da terceiro e sétimo || I || V7 || V7 || I || (3x)
Populares , vol.1, Typographia figura harmónica: arpejo compassos nos agudos
Occidental, Porto, 1893, Os fados novos teem actualmente a designação de ascendente em vez de são omissas na fonte. No
pp.152-153 Nocturnos. quebrado. entanto, tal não é
consentâneo com o que se
Esta melodia tem a suavidade germanica. Parece que passa ao quinto compasso
a musica teutonica inspirara o melodista, que não nem com a lógica da
temos o gosto de saber quem é. Tambem não escrita melódia e optei por
conhecemos o author da poesia que, sobre um antigo colocá-las.
mote portuguez, glosou esse gracioso mixto das
linguas hespanhola e portugueza, em doce união
ibérica.
Esta melodia está muito divulgada em todo o paiz

porém com innumeras variantes.
11 Neves, César das, 1852 O Atroador Esta cantiga, assim como quasi todas as suas - Ocorrência de ritmo Não se compreende a 2/4 Andante Ré Maior 30 compassos AB Int AAA BB || (D) I || V7 || V7 || I || (partes A)
Cancioneiro de Músicas congeneres, da nova musa popular dos fados, não tem arpejado e ritmo marchado indicação FIM na fonte, um Coda || I || ii || V7 || I || (partes B)
Populares , vol.1, Typographia lettra propria. Applicam-lhe diversas poesias; nós em simultâneo. compasso antes do final, (4+4 + 1 + 4+4+4 +
Occidental, Porto, 1893, dámos preferencia á presente, não só por estar mais pelo que optei por omiti-la. 4+4 + 1)
pp.120-121 em relação com o titulo e rythmo da musica, mas ainda - Intermezzo
por ser uma engraçada hyperbole, excellentemente desestruturante. Não se
glosada. É vagamente conhecida, apesar de ser antiga. percebe o porquê do
compasso extra.
12 Neves, César das, 1890 Augusto Hylario Às Estrellas Recolhida em Coimbra em 1890, pelo Exmo. Snr. Dr. Fado de Ao 8º compasso a fonte 2/4 Andante Sol Maior 32 Compassos A B C D(8+8+8+8) || (G) I || V7 || V7 || I ||(2x)
Cancioneiro de Músicas Manuel Maria de Castro Corte Real. Coimbra.Progressão tem sustenido ao invés de || V7/ii || ii || V7 || I || (2x)
Populares , vol.1, Typographia harmónica complexa nas bequadro, o que é ilógico || I || V7 || V7 || I || (2x)
Occidental, Porto, 1893, Primeiro Fado de Augusto Hylario. partes B e D, com uma visto o sustenido já estar || V7/ii || ii || V7 || I || (2x)
pp.192-193 Dominante secundária. na armação de clave e na
repetição do gesto, ao 12º
compasso, se verificar um
bequadro.
13 Neves, César das, 1887 Reynaldo Varella Canção da Noite Esta serenata é vulgarmente conhecida pela Soa mais às serenatas do É omisso o sustenido no 2/4 Andante Mib Maior 2 + 16 compassos Intro A B || (Eb) I || V7-I ||
Cancioneiro de Músicas denominação de Fado das tres horas nome com que Hylario. Fado de Coimbra? Fa, no compasso décimo. (2 + 8 + 8) || V7 || V7 || I || (2x)
Populares , vol.1, Typographia seu auctor primeiro a baptisara por ser àquella hora da || V7/vi || vi || V7/V || V ||
Occidental, Porto, 1893, p. noite que elle a improvisou. É a musica d este genero, Progressão harmónica || V7 || I || V7 || I ||
208 de mais actualidade. complexa com inflexão a
dominantes secundárias.
14 Neves, César das, 1892 O Engeitado Fado como subtítulo. A transcrição parece-me 2/4 Andante Ré menor - Fá 5 + 16 + 2 Intro A B Coda || (Dm) i || i || V7 || V7 || i ||
Cancioneiro de Músicas duvidosa e com erros de maior compassos (5 + 8 + 8 + 2) || i || V7 || V7 || i || (2x)
Populares , vol.1, Typographia Recolhida em Vizella em 1892, por F. P. Nogueira estrutura (na forma de || (F) I || V7 || V7 || I ||
Occidental, Porto, 1893, p. abordar a anacruza) e de || I || vi-ii || I-V7 || I ||
254 harmonia (no 2º tempo do
compasso 18).
15 Neves, César das, 1870 João Maria dos O Lisbonense Recolhida em Lisboa em 1870. Tom marcial. Centro tonal Na fonte, no penúltimo 2/2 Andante Lá Maior - Mi Maior 28 compassos AA B AA B || (E) IV || IV || IV || I ||
Cancioneiro de Músicas Anjos oscilante. compasso, na mão B(4+4+4+4+4+4+4) || IV || IV || V || I ||
Populares , vol.1, Typographia esquerda no 2º acorde || I || V || V || I || (2x)
Occidental, Porto, 1893, pp. aparece uma semínima, || I || V || I-V7 || I ||
266-267 que não cabe no
compasso. Por coerência
coloquei uma colcheia.
16 Neves, César das, 1894 Saudades da Recolhida em Alijó por F. P. Nogueira, em 1894. Opção de 4/4 ao dobro do Na fonte, na segunda 4/4 Allegro vivo Ré Maior 4 + 16 compassos Intro AA BB || (D) I || ii || V7 || I ||
Cancioneiro de Músicas Aldeia (Descante) andamento habitual. metade do 4º compasso (4 + 4+4 + 4+4) || I || V7 || V7 || I || (2x)
Populares , vol.1, Typographia Este descante pode ser classificado na ordem dos Poderia perfeitamente ter aparece uma pausa de || I || ii || V7 || I || (2x)
Occidental, Porto, 1893, pp. fados modernos, tanto pelo seu estylo musical, como sido escrito como 2/4 em semínima que não cabe no
288-289 pela ideia poetica. Andante. mesmo. Optei por omiti-la.
17 Neves, César das, 1895 F. P. Nogueira Fado Madrugada Não se percebe muito bem 2/4 Andantino Si Maior 12+2+32 Intro A A B B || (B) I || I || ii || ii || V7 || V7 || I ||
Cancioneiro de Músicas a existência dos 2 compassos (12+2+8+8+8+8) || I || V7 || V7 || I || (2x)
Populares , vol.2, Empresa compassos extra entre a || I || ii || V7 || I || (2x)
Editora, Porto, 1895, pp.20-21 introdução e o início do
fado.
18 Neves, César das, 1852 Fado da Figueira O compasso na fonte 4/4 Moderato Mi menor 2+24 compassos Intro A B C || (Em) i || i ||
Cancioneiro de Músicas da Foz aparece 2/4 mas depois as (2+8+8+8) || i || V7 || V7 || i || (2x)
Populares , vol.2, Empresa figuras todas, bem como o || i || i || V || i || (2x)
Editora, Porto, 1895, pp.34-35 andamento, estão || i || V || V || i || (2x)
adequados para um
compasso de 4/4.Não se
entende a semínima ao
compasso 16, 1º tempo. A
reparar pela lógica da
melodia deveria ser
colcheia e pausa de
colcheia.
19 Neves, César das, 1875 Ambrósio Fado de Coimbra Primeiro fado no compasso 3/4 Andantino Mib Maior 4 + 16 compassos Intro AABB || (Eb) I || V7 || V7 || I || (5x)
Cancioneiro de Músicas Fernandes Maia 3/4, algo bizarro. (4+ 4+4+4+4)
20 Neves, César das, 1895 A Cantadeira Descante. Classificado como Descante 2/4 Gracioso Ré Maior 12 compassos ABB4+4+4 || I || V7 || V7 || I || (3x)
Cancioneiro de Músicas mas com todas as
Populares , vol.2, Empresa características do tipo
Editora, Porto, 1895, p. 49 definido por Ernesto Vieira e
muito semelhante a outras
ocorrências classificadas
como Fado.
21 Neves, César das, 1895 A Menina dos 4/4 Moderato Ré Maior 1+16 compassos Intro AABB(1+ || I || V7 || V7 || I || (2x)
Cancioneiro de Músicas Olhos Negros 4+4+4+4) || IV || I || V7 || I || (2x)
22 Neves, César das, 1870 Fado Campestre Recolhida no Porto em 1870. Na fonte ao 2º,4º,6º e 8º 2/4 Andante Dó Maior 16 compassos AABB(4+4+4+4) || I || V7 || V7 || I || (4x)
Cancioneiro de Músicas compassos, o primeiro
Populares , vol.2, Empresa tempo aparece com duas
Editora, Porto, 1895, p. 96 semi-colcheias ao invés de
colcheias, o que não bate
certo com o espaçamento
utilizado nem com a
coerência da frase musical,
como se comprova depois
pelo seguimento da obra.
23 Neves, César das, 1850 A Nau Affonso - O andamento é 2/4 Moderato Lá menor 32 compassos AABC(8+8+8+8) || i || V7 || V7 || i || (4x)
Cancioneiro de Músicas demasiado rápido. O midi || i || i || V7 || i || (2x)
Populares , vol.2, Empresa foi gravado em Andante à || i || V7 || V7 || i || (2x)
Editora, Porto, 1895, pp. 126- semelhança dos outros
127 fados do tipo definido por
Ernesto Vieira.- Não se
compreende a frase
musical do penúltimo
compasso ter Fa# e a sua
imitação quatro compassos
antes não ter.
24 Neves, César das, 1850 Canção do Fado em subtítulo. - Não se compreende o 2/4 Andantino Sol menor 32 compassos AAAB(8+8+8+8) || i || V7 || V7 || i || (8x)
Cancioneiro de Músicas Marítimo baixo em Fá ao terceiro
Populares , vol.2, Empresa compasso. Atendendo ao
Editora, Porto, 1895, pp. 142- resto da canção deveria
143 ser Ré.
- Na fonte, no penúltimo
compasso, no segundo
tempo a colcheia de Ré
aparece pontuada, o que
não cabe no compasso.
- Na fonte, no último
compasso não aparece a
pausa de semi-colheia
antes da barra cheia.
25 Neves, César das, 1894 O que é amor Fado em subtítulo. - Na fonte, na introdução, 2/4 Andante Sol Maior 4 + 16 compassos Intro AABB || I || ii || V7 || I ||
Cancioneiro de Músicas apenas as notas de cima (4+ 4+4+4+4) || I || V7 || V7 || I || (2x)
Populares , vol.2, Empresa Esta musica é antiga e commum a muitas poesias do baixo aparecem ligadas. || I || ii || V7 || I || (2x)
Editora, Porto, 1895, p. 190 sentimentaes; com a presente lettra data de 1894, a
qual é do apreciado escriptor e jornalista o Exmo Snr. - Na fonte, ao quarto
Sousa Rocha. compasso, na fonte,
aparece uma semínima ao
invés de uma colcheia.
Não cabe no compasso.
- Na fonte, na repetição da
primeira frase do fado, ao
décimo compasso, o
bequadro de precaução
aparece apenas ao
segundo dó.
- Na fonte, ao oitavo e ao
décimo segundo
compassos aparecem
ligados os primeiros sis ao
invés do segundo para o
terceiro.
- Na fonte, no último
compasso falta o sinal de
marcação que orienta a
repetição.
26 Neves, César das, 1849 Fado Amphiguri Este amphiguri cantava-se em Lisboa ha cincoenta Variante das partes A e B do 2/4 Andantino Ré menor 8 compassos AB(4+4) || i || V7 || V7 || i || (2x)
Cancioneiro de Músicas annos, aproximadamente, mas é mais antigo. fado 003.
27 Neves, César das, 1870 Maggiolli Fado Maggiolli A musica deste fado foi recolhida em 1870. O nome Harmonia mais complexa, 2/4 Moderato Sol Maior - Mi 2 + 32 compassos Intro AA BB || (G) I ||
Cancioneiro de Músicas provém-lhe do appellido do author, que era, segundo como a de Hylario ou menor (2 + 8+8+8+8) || I || V7 || V7 || i ||
Populares , vol.2, Empresa nos consta, um bohemio da fina roda de Lisboa. Varella. || i || V7 || V7-#iv07 || (Em) V7 ||
Editora, Porto, 1895, pp. 292- || i || iv || V7 || i-V7 ||
293 Andamento Moderato, || i || vii07 || V7 || i-V7 || i ||
parece um pouco rápido.
28 Neves, César das, 1875 Serenata à Morena Esta musica que recolhemos na Praia da Granja em Andamento de Moderato, 2/4 Moderato Sol Maior 1+16 compassos Intro ABCD || I ||
Cancioneiro de Músicas 1875, quando era simplesmente executada em parece um pouco rápido. (1+ 4+4+4+4) || I || V7 || V7 || I ||
Populares , vol.3, Empresa guitarras (instrumento predominante naquella epocha), || V7/ii || ii || V7 || I ||
Editora, Porto, 1898, p. 26 por um grupo de academicos acaba de reapparecer || I || V7 || V7 || I ||
mas agora cantada com a bellissima poesia do Exmo. || I || V7 || I-ii-V7 || I ||
Snr. Conde de Monssarraz.
29 Neves, César das, 1870 Fado Nacional Este fado foi recolhido em 1870 n uma edição da casa Fado em 2/4 mas com Na penúltima repetição do 2/4 Andante Sol menor 16+8+41+17 ABCD A B C C C C || i || V7 || vii0-V7/bIII || bIII ||
Cancioneiro de Músicas Sassetti & C.ª, de Lisboa, porém sem lettra, que a não duração das figuras rítmicas
tema recorrente (13º compassos C D A B C D EDED || V7 || i-V7 || i ||
Populares , vol.3, Empresa tem propria, addiccionando-lhe os cantadores, trovas dobradas em relação ao que compasso a contar do fim), || i || V7 || V7 || i ||
Editora, Porto, 1898, pp. 43-45 populares. tem aparecido. O ao primeiro compasso do || i || i-V7 || V7 || V7-i || i || (2x)
Andamento andante , pois, mesmo, a fonte tem o || i || i-V7 || V7-i || V7-i ||
não faz sentido, acorde errado (Gm), || i || i-V7 || V7 || V7-i ||
entendendo-se aqui como conforme se atesta por || i || i-V7 || V7 || i-V7 || i ||
algo na ordem das 108-120 todas as outras || i || V7 || V7 || i || (2x)
bpm. ocorrências do tema, onde
este compasso aparece
A estrutura aparece algo com D7. No segundo
confusa, parecendo um compasso dessa mesma
rondó com um tema repetição a primeira nota
recorrente, mas depois há da fonte também aparece
secções coxas e outras com trocada - lá ao invés de sol.
compassos a mais, dando
números não-múltiplos de 4
como espectável. A ligação
entre secções parece ser o
principal problema do
transcritor.
30 Neves, César das, 1898 Henrique Carneiro Fado do Sofrimento Este fado que tende a popularisar-se vem publicado no Progressão harmónica mais 2/4 Andante Ré menor 24 compassos ABC(8+8+8) || V7 || i || V7 || i ||
Cancioneiro de Músicas Missal d um crente , poema lyrico de Mariano Gracías. acelerada, Síncopas no || V7 || i || ii07-V7 || i ||
Populares , vol.3, Empresa acompanhamento, mais || i || ii07 || V7 || i ||
Editora, Porto, 1898, p. 54 expressividade. || i || ii07 || i-V7 || i ||
|| V7 || i || iv-V7/V || V7 ||
A suspensão perto da || i-V7/iv || iv-ii07 || i || V7 || i ||
cadência final.
Resquícios de contra-canto
no acompanhamento.
31 Neves, César das, 1870 Ambrósio Fado de Cascaes Ao 3º e 7º compassos a 2/4 Andante Ré Maior 2 + 20 compassos Intro AA BBB || I || I ||
Cancioneiro de Músicas Fernandes da Maia fonte apresenta pausa de (2+ 4+4 +4+4+4) || I || V7 || V7 || I ||(5x)
Populares , vol.3, Empresa semicolcheia que não cabe
Editora, Porto, 1898, p. 63 no compasso.
32 Neves, César das, 1871 Fado dos Este fado é dos estudantes açorianos; foi recolhido em 2/4 Andantino Sol Maior 16 compassos ABAB(4+4+4+4) || I || V7 || V7 || I ||
Cancioneiro de Músicas Estudantes 1871. || I || ii || V7 || I || (2x)
33 Neves, César das, 1879 Robles Fado Robles Recolhido no Porto em 1895. A designação de Robles Ao 13 2/4 Andante Dó Maior - Dó 5 + 32 compassos Intro AA BB || (C) I || iv || I || V || I ||
Cancioneiro de Músicas provém-lhe do nome do author da musica. menor (5 + 8+8 +8+8) || I || V7 || V7 || I || (2x)
Populares , vol.3, Empresa || (Cm ) iv || i || V || i || (2x)
Editora, Porto, 1898, pp. 76-77
34 Neves, César das, 1898 Ora Toma 2/4 Allegretto Sib maior 16 compassos AABB(4+4+4+4) || (Bb) I || V7 || V7 || I || (2x)
Cancioneiro de Músicas Mariquinhas || I-V7/ii || ii || V7 || I || (2x)
Editora, Porto, 1898, pp. 88
35 Neves, César das, 1872 Eduardo Pereira Fado do Leça A musica d este fado é antiga e não tem lettra propria: Depois da introdução, ao 2/4 Moderato Mi menor 8 + 24 compassos Intro AABCCC || i || iv-vii07 || iv || V7 ||
Cancioneiro de Músicas Baptista Leça a que lhe addicionamos é do Ex.mo Snr. J. Nunes quarto compasso do fado (8+ 4+4+4+4+4+4) || i-V7/iv || iv || V7 || i ||
Populares , vol.3, Empresa Ponte. propriamente dito, a fonte || i || iv || iv || V7 ||
Editora, Porto, 1898, pp. 106- não tem os sustenidos no || i || iv-vii07 || iv || V7 ||
107 ré, no acompanhamento. O || i-V7/iv || iv || V7 || i ||
que não é coerente com a || V7 || i || V7 || i || (3x)
harmonia, nem com a
repetição do mesmo tema,
quatro compassos à frente,
onde eles aparecem.
36 Neves, César das, 1898 Ao Hylario Recolhido pelo Ex.mo Snr. Eduardo da Fonseca. Este 2/4 Allegretto Dó menor 16 compassos AABB(4+4+4+4) || i || iv || V7 || i || (2x)
Cancioneiro de Músicas fado acha-se vulgarisado por todo o paiz com diversa || iv || i || V7 || i || (2x)
Populares , vol.3, Empresa lettra.
37 Neves, César das, 1872 Fado Carmona Recolhida em Lisboa em 1874. Fado com ritmo de 2/4 Andante Lá menor 16 compassos ABCC(4+4+4+4) || i || V7 || V7 || i || (4x)
Cancioneiro de Músicas Tango.Possível parelha com
Populares , vol.3, Empresa o 039.
38 Neves, César das, 1898 Fado Visconti Este fado apparecey no Porto na presente decada, 2/4 Allegretto Sol Maior - Ré 16 Compassos AABB(4+4+4+4) || (G) I || V7 || V7 || I || (2x)
Cancioneiro de Músicas trazido por um palhaço portuguez, d uma companhia Maior || (D) I-V7/ii || V7 || V7 || I || (2x)
Populares , vol.3, Empresa equestre, por appelido Visconti.
39 Neves, César das, 1898 Fado do Gato Vulgo do Taborda, em subtítulo. Fado com ritmo de 2/4 Andante Lá menor 12 compassos ABB(4+4+4) || i || V7 || V7 || i || (4x)
Cancioneiro de Músicas Tango.Possível parelha com
Populares , vol.3, Empresa o 037.
40 Neves, César das, 1898 Canção das Este fado foi recolhido em Avanca pelo Ex.mo Snr. Dr. 2/4 Andante Mib Maior - Sib 4 + 16 Compassos Intr AABB || (Eb) I || ii || V7 || I ||
Cancioneiro de Músicas Morenas M. M. de Castro Corte Real. Maior (4 + 4+4+4+4) || I || V7 || V7 || I || (2x)
Populares , vol.3, Empresa || (Bb) V7 || I || V7 || I || (2x)
41 Neves, César das, 1898 Augusto Hylario Fado Posthumo do Este fado foi recolhido em Sinfães pelo Ex.mo Snr. Dr. 2/4 Andante Fá maior 12 compassos ABA(4+4+4) || (F) I || V7 || V7 || I (2x)
Cancioneiro de Músicas Hylario M. M. Castro Côrte Real, que nol-o enviou com a || V7/vi || vi || 6G || V/vi ||
Populares , vol.3, Empresa seguinte nota: «Fado do Hylario (ultimo). O fado que || I || V7 || V7 || I ||
Editora, Porto, 1898, p. 208 vem no Cancioneiro com a designação de ultimo é
anterior a este. Este é que é geralmente conhecido
pelo ultimo; sempre assim o ouvi designar oas
estudantes coevos do grande bohemio.» A lettra é do
Ex.mo Snr. Luiz Osorio. A primeira estrophe que vae na
musica canta-se tambem no fim.
42 Neves, César das, 1850 O Marinheiro 2/4 Andante Dó Maior 20 compassos ABA B || I || V7 || V7 || I || (partes A)
Cancioneiro de Músicas B(4+4+4+4+4) || V7/ii || ii || V7 || I || (partes B)
43 Neves, César das, 1896 Leandro Fado Leandro Recolhido em Sinfães em 1896, pelo Ex.mo Snr. Dr. M. 2/4 Andante Dó maior 4 + 16 compassos Intro AABB || I || ii || V7 || I ||
Cancioneiro de Músicas M. Castro Corte Real. O author chama-se Leandro, (4+ 4+4+4+4) || I || V7 || V7 || I || (2x)
Populares , vol.3, Empresa musico ambulante que acompanha um cego. || I || ii || V7 || I || (2x)
44 Neves, César das, 1866 Fado de Tancos Este fado data da installação do campo de manobras Segundo pesquisa, o 2/4 Andante Láb Maior - Mib 1 + 16 Compassos Intro AABB || (Eb) IV ||
Cancioneiro de Músicas em Tancos. Campo de Manobras de Maior (1 + 4+4+4+4) || IV || IV || IV || I || (2x)
Populares , vol.3, Empresa Tancos foi instalado em || V7 || V7 || I-V7 || I ||(2x)
Editora, Porto, 1898, p. 234 1866.
45 Neves, César das, 1850 Canto do Suicida A musica d este fado foi recolhida em 1874 e não tinha Ao 14º compasso a fonte 2/4 Andante Láb maior - Fá 8 + 16 compassos Intro AABB || I || V || vi || V7/vi ||
Cancioneiro de Músicas lettra propria; a presente poesia foi-lhe posteriormente tem pausa de colcheia que menor (8+ 4+4+4+4) || ii || I || V7 || I ||
Populares , vol.3, Empresa applicada. não cabe no compasso. || I || V7 || V7 || I ||
Editora, Porto, 1898, p. 246 || I || V7 || (Fm) V7 || i ||
|| iv || i || V7 || i || (2x)
46 Neves, César das, 1880 Francisco Fado do Zé A lettra que ouvimos cantar com este fado pertence a 2/4 Andante Sol maior 16 compassos AABB(4+4+4+4) || I || V7 || V7 || I || (2x)
Cancioneiro de Músicas Alvarenga Povinho um poemasinho Auto da Lavradora de Ayré publicado || V7/ii || ii || V7 || I || (2x)
Populares , vol.3, Empresa em 1678; vem nas Poesias de Antonio de Villaboas e
Editora, Porto, 1898, p. 256 Sampaio, impressas em Coimbra em 1841.
47 Neves, César das, 1898 Fado do Celta Esta canção deve ser cantada a duo por uma senhora Na fonte, ao quinto 2/4 Andante Dó maior - Lá 12 compassos ABB(4+4+4) || (C) I || V7 || V7 || I || (2x)
Cancioneiro de Músicas e um homem (fazendo de mãe e filho como indica a compasso, no segundo lá menor || (Am) i || V7 || V7 || i || (2x)
Populares , vol.3, Empresa poesia) do segundo tempo aparece
Editora, Porto, 1898, p. 260 um ponto de aumentação
que não cabe no
compasso.
48 Neves, César das, 1858 Fado Lazarista À amabilidade do Ex.mo Snr. L. Sollari e Allegro, Graus harmónicos 2/4 Allegro Sol menor 6 + 32 compassos Intro ABCB || (Bb) V7 || I || ii-6G ||
Cancioneiro de Músicas distincto paleographo da Camara Municipal do Porto, complexos, como a Sexta (6+ 8+8+8+8) || (Gm) V7 || V7 || V7 ||
Populares , vol.3, Empresa devemos o presente fado Lazarista que foi publicado Alemã ou a substituição || i || i || i || V7 ||
Editora, Porto, 1898, pp. 268- em Lisboa no Asmodeu de Setembro de 1858. Este Tritónica. || V7 || i || i-V7 || i ||
269 fado conservou uma excessiva popularidade durante || bII7-V7 || i || bII7-V7 || i || (2x)
uma dezena d annos e para isso concorreram os factos || (Bb) V7 || I || ii-6G ||
que lhe deram origem. || (Gm) V7 || i || bII7 || i-V7 || i ||
|| bII7-V7 || i || bII7-V7 || 6G/III ||
Em 1856 foram introduzidas em Portugal as Irmãs da || V7 || bVI || i ||
Caridade vindas de França, para o serviço dos
hospitaes e das escholas. Esta novidade no nosso paiz
excitou em muitas senhoras de todas as classes
sociaes, ou por sugestão ou por capricho, o desejo de
se filiarem n aquella instituição. Isto provocou uma
corrente de opinião opposta que se levantou por todo o
paiz. A imprensa liberal e os comicios, combatendo
com as armas desdo o mais serio ao mais ridiculo,
fizeram com que um dia as Irmãs da Caridade
abandonassem precepitadamente os hospitaes e os
azylos. Foi devido a este sucesso que o grande tribuno
José Estevão Coelho de Magalhães instituiu em Lisboa
o primeiro asylo de S. João; achando na rua as
creanças que as irmãs da caridade haviam
abandonado, foi com ellas mendigar donativos, para
crear aquelle asylo.
Este fado resume as manifestações populares d

aquella epocha.
49 Neves, César das, 1898 João de Deus Fado João de Deus O author da musica inspirando-se na poesia do grande 2/4 Andante Sol menor 12 compassos Intro AB || i || ii || i-V7 || i ||
Cancioneiro de Músicas lyrico, baptisou-a com o seu nome. Lettra d este fado (4+4+4) || i || V7 || V7 || i || (2x)
Populares , vol.3, Empresa João de Deus. || V//iv || iv || i-V7 || i || (2x)
50 MNTPRT008 - A. Moraes Lda. 1918 H. Rubiales Fadinho da Nita Os grandes sucessos do Salão Foz 4/4 Moderato Lá menor/Lá maior 9 + 16 + 1 + 8 Intro ABAC interm. || (Am) i-V7/iv || iv || V7 || i || (2x)
Editores Compassos DE || i || ii0 || V7 || i ||
Creação da notável e gentil Cancionista Nita Ibañez (4+4+1 +4+4+4+4 || V7/iv || iv || V7 || i ||
+1 +4+4) || i || V7 || V7 || i ||
Letra de Pedro Bandeira || V7/iv || iv || V7 || i || i ||
|| (A) I || V7 || V7 || I ||
|| V7/ii || ii-ii07 || V7 || I ||
51 MNTPRT014 - Sassetti 1927 António Menano O meu menino Segundo Octávio Sérgio, o 4/4 S/indica Láb Maior 3 + 16 + 9 + 16 Intro AABC || I || V7 || I ||
Editores original é de Alexandre compassos Intermezzo AABC || I || V7 || V7 || I || (10x)
Augusto de Rezende (3 + 4+4+4+4
Mendes de 1915, sob a +1+4+4 +4+4+4+4)
égide o meu menino d oiro .
Menano ter-se-á apropriado
da música com outra letra
em 1927.
52 MNTPRT016 - Sassetti 1927 António Menano Phases da Lua Octávio Sérgio sugere a 2/4 Muito Fá maior 5 + 16 + 4 Intro AABC Coda || I || V7 || I || V7 || I ||
Editores data de 1927 para este Vagamente compassos (5 + 4+4+4+4 + 4) || I || V7 || V7 || I || (2x)
fado, dizendo que é de autor || V7/ii || ii || V7/iii || iii ||
desconhecido, apesar de || ii-#iv07 || I-bvi0 || V7 || I ||
gravado por António || V7 || I || V7 || I ||
Menano.
53 MNTPRT027 - Canções 1940 Armando Machado Fado Araceli Letra de: João de Freitas Na partitura original falta 4/4 S/indica Dó Maior 2 + 16 compassos Intro ABAB || ii || V7 || I || I ||
Populares Portuguesas , nº18, uma pausa de colcheia ao (2 + 4+4+4+4) || I || I-#i07 || V7 || V7 ||
Tip. A. Pinto de Campos 6º, 10º e 14º compassos e || ii || V7 || I || I ||
há uma pausa de colcheia || vi || V || V7 || V7 ||
a mais no penúltimo || V7 || V7 || I || I || (2x)
compasso.
54 MNTPRT034 - Sassetti 1927 Alexandre de Fado da Mentira Oct 4/4 Moderato Sol Maior 9 + 16 + 4 Intro AABB Coda || I || I || V7 || V7 || I ||
Editores Resende compassos (9 + 4+4+4+4 + 4) || I || V7 || V7 || I ||
|| I || V7 || V7 || I || (2x)
|| V7/vi || vi || V7 || I || (2x)
|| I || V7 || V7 || I ||
55 MNTPRT038 - Irmãos Vitale 1953 Artur Ribeiro Maria da Graça A partitura da fonte era 2/4 S/indica Ré menor 4 + 32 compassos A BB B C AA DA(4+ || i || V7 || i-V7 || i ||
Editores, São Paulo , Brasil para acordeão, só se 4+4+4+4+ || i || i || i || iv ||
escrevendo a tónica do 4+4+4+4) || V7 || V7 || V7 || i ||
acorde e a cifra, assim || i || i || V7/iv || iv ||
sendo eu completei os || iv || iv || V7 || V7-i ||
acordes com as notas em || i || V7 || i-V7 || i-V7 ||
falta. || iv || iv || V7/III || III-V7/iv ||
|| iv || V7 || iv || V7 ||
|| i || V7 || i-V7 || i ||
56 MNTPRT040 - Valentim de 1920 Alves Coelho Fado de Portugal Versos de Acácio de Paiva 4/4 S/indica Lá Maior/Lá menor 4 + 16 + 12 + 4 Intro ABAC DD Intro || I || V7 || V7 || I ||
Carvalho Editor compassos (4 + 4+4+4+4 +6+6 || i-V7 || i || ii0 || i ||
+4) || V7 || i || V7 || V7 || (2x)
|| I || I || V7 || V7 || V7 || I ||
57 MNTPRT063 - Sassetti 1925 César Magliano Fado das Quatro Letra de Salema Vaz.Repertório da notável cantora Uniformizaram-se as 4/4 Allegretto Lá Maior/Lá menor 8 + 32 compassos Intro ABAC DED F || I || V7 || V7 || I || (2x)
Editores Estações Manuela Pinto Basto. articulações: amiúde (8 + 4+4+4+4 || (Am) i || i || iv || i ||
faltavam staccatos ou +4+4+4+4) || V7 || i || VI-iv-6G || V7 ||
arpeggios em repetições || i || i || iv || i ||
onde se previa que || V7/III || III || V7 || I ||
estivessem. || (A) I || I || I || V7 ||
|| V7 || V7 || V7 || I ||
|| I || I || V7/ii || ii ||
|| ii-iv || I || V7 || I ||
58 MNTPRT069 - Sassetti 1960 Fernando de O Fado foi a Paris Versos de Amadeu do Vale Uso de acordes estendidos Falta um bemol ao mi do 2/2 Fox Medium Fá maior – Fá 2 + 32 compassos Intro ABAC DEDF || V9/V || V7 || i ||
Editores Carvalho nada usuais, bem como segundo compasso na menor (2 + 4+4+4+4 || i || V7 || V7 || i || (3x)
linha melódica bastante fonte. +4+4+4+4) || V7/iv || iv || V7 || I ||
cromática. Não o enquadro || (F)I-V9/V || V7 || V7-#ii07 || I ||
nos fados castiços, || I-i || V7 || ii || I ||
claramente já uma inovação. || I-V9/V || V7 || V7-#ii07 || I ||
|| ii07 || I-V9/V || V7 || i ||
59 MNTPRT071 - Lith. Universal, 1900 Guido Fábio Não chores, que 2/4 Moderato Ré menor- Ré 1+8+16+8 Intro AB CD Intro || (Dm) i || V7 || i || V7 || i ||
E. do Carmo, 17 também vais Maior compassos (1+ 4+4 +4+4 +4+4 || V7 || i || V7 || i ||
+4+4) || i || V7 || V7 || i ||
|| i-bVII || iv || V7 || i-I ||
|| (D) I || I || I || V7 ||
|| V7 || V7 || V7 || I ||
|| (Dm) V7 || i || V7 || i || (2x)
60 MNTPRT092 - Sassetti 1946 José de Oliveira Dedicação Criação de Lídia Nunes nas emissões do Rádio Club 2/2 S/Indica Ré menor/Ré Maior 8 + 32 compassos Intro AA BC DD EF || i || i || V7 || V7 ||
Editores Cosme Português (4+4 || V7 || V7 || i || i ||
+4+4+4+4+4+4+4+ || i || i || (F) ii || ii ||
4) || V7 || V7 || I || I ||
|| (Dm) i || i || iv || iv ||
|| i || V7 || i || i ||
|| (D) I || #iv07 || V7 || V7 ||
|| V7 || V7 || I || I ||
|| (Dm) i || i || V7 || V7 ||
|| V7 || V7 || i || i ||
61 MNTPRT112 - Edições 1947 João Nobre Canção Portuguesa Fado Slow. Estranha aparição de um 2/2 s/indica Dó Maior 4 + 24 compassos Intro ABAC DE || V7/ii || ii-vii07 || I-V7 || I ||
Sassetti acorde aumentado ao (4 +4+4+4+4 +4+4) || I || V7 || V7 || I || (2x)
Um grande êxito de Maria Clara ao microfone da sétimo compasso. || I || V7 || V7 || V7/(V/V) ||
Emissora Nacional. || V7/V || I-#i07 || V7/V-V7 || I ||
|| I-V7/ii || ii || V7 || V7-I ||
|| V7/ii || ii-vii07 || I-V7 || I ||
62 MNTPRT116 - Edições 1920 Luiz de La Cruz Despedida Versos de José Coelho da Cunha Bizarra secção central com Ao 19º compasso a última 4/4 Lento Ré menor/Ré maior 8 + 16 + 3 + 14 + 7 Intro ABCD || (Dm) i || i-iv || iv-V7 || V7-i ||
Sassetti Quesada dísticos encaixados em sete nota da fonte aparece + 4 compassos Intermezzo EF G || V7/III || i || V7 || i ||
compassos. como semínima ao invés Outro || i-V7/III || V7 || V7 || i ||
decolcheia. (4+4 +4+4+4+4 + || V7/III || III || iv-V7 || V7 ||
3+ 7+7 +7 +4) || iv-V7 || i || V7/III || III ||
Foram acrescentados dois || V7 || i || V7 || i ||
piano na dinâmica por || i || V7 || V7 || I ||
motivos de audível || (D) V7 || V7 || I || V7 ||
rendição do ficheiro midi. || V7 || I || V7 || V7 || I ||
|| I || V7 || V7 || i ||
|| (Dm) i || V7 || V7 || V7 ||
|| V7 || V7 || i-iv || iv-V7 || i-V7 || i ||
63 MNTPRT121 - Casa Moreira 1914 Nicolino Milano Lu Lu Fado Versos de Guedes d Oliveira Datação pouco segura, Figuração errada ao 43º 2/4 Moderato Dó Maior 5 + 15 + 16 + 16 + Intro AA BCBC || I || I || V7 || V7 || I ||
de Sá Editora, Porto proveniente de uma compasso da fonte. 5 compassos DEDF Coda || (Cm) i || 6F || V7 || i || iv-V7 || i ||
listagem num cache de um Depreende-se que seja (5 + 6+9 +4+4+4+4 || (Eb) V7 || V7 || I || (Cm) vii0-6F || V7-i ||
sítio da internet. uma síncopa e não três +4+4+4+4 +5) || V7-i || V7 || V7 || (C) I ||
colcheias. || V7 || I || I || I ||
|| V7/V || V7 || I || V7 || I ||
|| I-V7/(V/V) || V7/V || V7 || I ||
|| I || V7 || V7 || I || (3x)
|| V7/(V/V) || V7/V || I-V7 || I ||
|| I || I || V7 || V7 || I ||
64 MNTPRT130 - Sassetti 1937 Raúl Portela Lisboa Antiga Fado Marcha. Letra de José Galhardo e Amadeu do Estranha divisão das frases Foi decidido por questão 2/4 s/indica Ré menor/Ré maior 8 + 22 + 29 Intro ABC DEFG || (D) #iv07 || V || ii-#iv0 || V7 ||
Editores Vale. melódicas em números de de coerência transformar o compassos (8 + 8+8+6 + || V7 || I || I || I ||
compassos ímpares. primeiro compasso que 8+7+7+7) || (Dm) i || V7 || V7 || V7 ||
estava escrito completo na || V7 || iv-V7 || i || i-V7 || i ||
fonte, em anacruze. || i || bVII || bVII || bVI ||
|| bVI || V7 || V7 || V7 ||
Ao 22º compasso da fonte || i || iv || iv || i || V7 || I ||
aparecem semi-colheias de || (D) V7/ii || ii || ii || V7 ||
tercina quando são || V7 || I || I || I ||
colcheias. Ao 10º || #iv07 || V7 || V7 || V7 ||
compasso a fonte || V7 || I || I || IV || ii || ii || V7 || I ||
apresenta o sustenido no || I || I || #iv07 || V7 ||
lá ao invés de dó, no || ii-#iv07 || V7 || V7 || I || I ||
acorde de la7. Ao 29º
compasso o sustenido do
dó é omisso na mão
esquerda, no mesmo
acorde de la7.
65 MNTPRT131 - Livraria Neves 1916 Raul de Campos Serenata ao Luar Letra de Americo S. d Oliveira 4/4 Allegro Mi Maior/Mi menor 16 + 32 Intro ABCD EFE G || (E) I || V7 || V7 || I || (3x)
Editora, Coimbra Animato Compassos (16 +4+4+4+4 || V7/ii || ii || V7 || I ||
+4+4+4+4) || (Em) i-V7 || i || iv || iv ||
|| V7 || V7 || V7 || V7 ||
|| i-#ii0-iv || iv || V7 || i ||
|| iv || i || V7 || i ||
|| (E) I || I || i || V7 ||
|| V7 || V7 || V7 || I ||
|| I || V7/ii || V7/ii || ii ||
|| iv || I || V7 || I ||
66 MNTPRT135 - A. Moraes, 1940 Raúl Ferrão Fado da Malva Ao compasso 23, na fonte, 4/4 s/Indicação Si Bemol/Sol Maior 5 + 24 + 4 Intro AABB CC || (Bb) I || I || V7 || V7 || I ||
Lda. Editores falta o sustenido no fá, no compassos Intermezzo (5 + || I || V7 || V7 || I || (2x)
acorde da mão esquerda. 4+4+4+4 +4+4 +4) || V7/ii || ii || V7 || I || (2x)
|| (Gm) i || V7 || V7 || i ||
|| iv || i || V7 || i || V7 || V7 || i ||
67 MNTPRT138 - Tipografia da 1927 Ruy Coelho Fado Miss Portugal Versos de Rocha Júnior Na fonte faltam os pontos 4/4 Muito Sol Maior 1+12+10+1+12 Intro AABCD inter || (G) I || I || V7 || V7 || || I ||
Empresa Diário de Notícias, de aumentação na mão devagar compassos AAB || I || V7 || V7 || I ||
Lisboa direita aos compassos 10 e (1+ 4+4+4 +6+4 || vi || V7 || V7 || I ||
19. +1+4+4+4) || V7/vi || vi || V7 || I ||
|| V7 || I ||
|| (Em) V7/iv || iv || V7 || i ||
68 MNTPRT144 – Raul 1925 Vasco Rocha Fado Ballada 2/4 S/indicação Mi Maior 5 + 16 + 8 + 8 + 4 Intro AA' B A Coda || (E) I || I-V7 || I || I-V7 || I ||
Venancio, Editor de Música, compassos || I-V7 || I || IV-V7 || I ||
Lisboa || V7 || V7 || V7 || I ||
|| I-V7 || I || IV-V7/vi || i/vi ||
|| ii-#iv07 || I || I-V7 || I ||
|| (C) I || V7 || V7 || I ||
|| V7-I || iii-I || V7/III ||
|| (E) I-V7 || I || IV-V7/vi || i/vi ||
|| ii-#iv07 || I || I-V7 || I ||
|| IV-V7 || I || IV-V7 || I ||
69 MNTPRT165 – Cruz e Sousa, 1934 Cruz e Sousa Terras de Portugal Ao célebre cantor ligeiro Guilherme Kjölner 2/4 Tempo de Fá Maior 16 +32 + 32 Intro ABAC DEDF || (F) V7 || V7 || I || I || (4x)
Ponta Delgada Fado Marcha. Marcha compassos || (Fm) i || i || V7 || V7 ||
|| V7 || V7 || i || i ||
|| IV/V || ii0/V || V7 || IV ||
|| IV/V || V7 || i ||
|| i || V7 || V7 || V7 || V7 || i ||
|| i || vii07/IV || V7/IV || I/IV ||
|| I/IV || i || V7 || I || I ||
|| (F) I || I || V7 || V7-I ||
|| IV || V7 || V7 || I ||
|| #ii07 || V7 || V7 || V7 || V7 ||
70 MNTPRT166 – Sassetti & Cª. 1925 Filipe Duarte Mulher de Portugal A Aldina de Souza, distincta actriz cantora Não se compreende a 4/4 Lento Ré Maior 4 + 16 + 8 + 8 Intro AA'BB' CC' || (D) I || V7 || I || V7 ||
Editores Fado Canção função harmónica ao compassos Coda || I || I || I || V7 ||
Letra de D. José Paulo da Câmara. compasso 14, dentro do || V7 || V7 || ii7-V7 || I ||
contexto. Talvez um Mi || (Em) V7/V || i7 || V7 || i (D) ii ||
menor em lugar de um Sol. || I || V7 || I || (D)I ||
|| I || V7/ii || ii || IV ||
|| I || ii-V7 || I || bVI ||
|| I || bVI || I || I || I || I || I ||
71 MNTPRT167 – Edições 1936 Marques Dias (arr.) Torre de Santa Do repertório da orquestra Almeida CruzArranjo de 2/4 S/indicação Ré Maior/Mi Maior 32 + 32 compassos ABAB CDCD ABAB || (D) I || I || ii || ii ||
Sassetti Lisboa Cruz Marques Dias CDCD || V7 || V7 || I || I || (2x)
|| (Bm) i || i || V7 || V7 ||
|| iv || iv || V7 || V7 ||
|| i || i || V7 || V7 ||
|| iv || V7 || I (E) i07-V7 ||
|| (E) I || I-V/ii || ii || ii ||
|| V7 || V7 || I || I ||
|| I || I-V/ii || ii || ii ||
|| V7 || V7 || I || V7/vi ||
|| (C#m) i || i || V7 || V7 ||
|| V7 || V7 || i || i || (2x)
72 MNTPRT168 - A. Moraes Lda. 1918 L. Serra e Moura Amor Eterno Do repertório do barítono Alberto Reis 4/4 S/indicação Sol menor/Sol 4 + 16 + 16 + 4 Intro AABB' CCDD' || (Gm) i || ii0 || V7 || i ||
Editores Dedicado à Exma. Sra. Dna. Elvira Adelaide Duarte Maior compassos Coda || i || iv || V7 || i ||
Versos de L. P. Luz || i || iv || V7 || i ||
|| ii0 || i || V7 || i ||
|| ii0 || i || V7 || I ||
|| (G) I || V7 || V7 || I ||
|| I-i || V7 || V7 || I ||
|| I || V7 || V7 || I-V7/IV ||
|| IV || V7 || I ||
|| (Gm) i || ii0 || V7 || i ||
73 MNTPRT182 – Manuscrito 1909 Thomaz António Fado da Severa Extraído da revista “Ó da Guarda” 2/4 Tempo de Lá menor 8 + 16 + 8 AB A'CDD' AB (Am) i || V7 || V7 || i || i-V7/ii || ii || i-V7 || i
original, Alter do Chã Camoezas (arr.) Fado compassos (A) I || V7 || V7 || I || I || V7 || (F#m) V7 || i || iv ||
i || V7 || i || iv || i || V7 || i
(Am) i || V7 || V7 || i || i-V7/ii || ii || i-V7 || i
74 MNTPRT183 – Manuscrito 1916 Raffaello Segre Fado do Ganga A fonte é uma partitura 4/4 S/indicação Lá menor 4 + 8 + 4 + 24 Intro AB intermezzo || (Am) i || V7 || V7 || i ||
original, Belém do Pará, Brazil (arr.) manuscrita com arranjo compassos CCDEED || i || V7 || V7 || i ||
orquestral e respectivas || i || V7 || V7 || i ||
partes. || (A) I || V7 || V7 || I ||
O original deste fado foi || I || V7 || V7 || I ||
interpretado por Estêvão || I || V7 || V7 || I ||
Amarante na revista "O || I || V7 || V7 || I ||
Novo Mundo" (1916), sendo || ii || I || V7 || I ||
um original de Ernesto || ii || I || V7 || I ||
Rodrigues, Felix Bermudes || I || V7 || V7 || I ||
e João Bastos.
75 MNTPRT191 – Custódio 1920 João F. Victoria Fado dos Cabelos Versos do livro “Rimas” de Emílio Ernesto. Na fonte o acorde do V7 4/4 Allegretto Lá Maior 8 + 16 + 12 Intro AABB CC'D (A) I || V7 || V7 || I || I || IV || I-V7 || II || V7 || V7
Cardoso Pereira e C.ª Lindos aparece muitas vezes com compassos || I || I || V7 || V7 || I || I || IV || I-V7 || I || I || IV || I-
a fundamental omitida. V7 || I(Fm) V7 || i || V7 || i (E) V7 || I || V7 ||
I(A)V7 || I || V7 || I
76 MNTPRT192 – Proprietaria- 1915 Manuel Benjamin Fado dos Homenagem ao jornal “O Mundo”. O novo ritmo refere-se 3/4 Lento Dó Maior 1 + 16 compassos Intro AABC || (C) I || I || V7 || V7 || I ||
Editora Musica e Theatros, Adeantamentos Versos de Pedro Bandeira. certamente a este arranjo || I || V7 || V7 || I || (2x)
Typ. A Publicidade Ultima cantiga popular (com um novo rytmo). em ternário, completamente
incomum nos fados.
Especulamos que o ritmo
originário seria um binário
convencional, com divisão
ternária.
77 MNTPRT193 – J. Heliodoro 1910 Luiz Pinto Meu Lindo Amor Distinto artista dramático e societário do Teatro A primeira secção do fado 4/4 Andante Sol menor 4 + 9 + 16 +4 Intro AA' BCBC || (Gm) i || V7 || V7 || i ||
d'Oliveira Nacional “Almeida Garrett”. tem uma transcrição compassos Coda || i || iv || V7 || i ||
errónea porque apresenta || i || iv || iv || V7 || i ||
nove compassos em vez || (G) I || V7 || V7 || I ||
de oito. Tal acontece || I || V7 || V7 || I ||
devido ao terceiro verso || I || V7 || V7 || I ||
ser estendido sobre uma || I || IV || I-V7 || I ||
nota longa e uma || (Gm) i || V7 || V7 || i ||
ornamentação do piano
(cp. 10-11): em vez de o
autor ter escrito uma
suspensão e uma cadenza
instrumental preferiu uma
transcrição literal e logo,
isso traduz-se num enxerto
de um compasso extra que
pouco sentido faz na
estrutura. A evidência
interna mostra-nos também
a incongruência, já que na
respectiva segunda secção
maior, no penúltimo verso
(cp. 27), a sílaba longa é
prolongada
“correctamente” com uma
suspensão e sem adenda
de compasso extra.
78 MNTPRT194 – Sassetti & Cª. 1934 Duarte Machado Fado Maria Creação da actriz cantora Maria Albertina Percurso harmónico pouco 2/4 S/indicação Sol Maior 8 + 16 + 16 Intro ABA'B' || (G) ii07/i || I || I || ii ||
Editores Albertina Letra de Luiz da Silva Gouveia convencional e com a compassos A''B''CD || V7 || I || I || I ||
progressão a metade da || I || I || I || V7 ||
velocidade. || I || V7 || V7 || V7 ||
|| ii || V7 || V7 || V7 ||
|| V7 || I || I || I ||
|| I-(Am) VII7 || V7 || V7 || V7 ||
|| V7 || i || i || i ||
|| (G)ii07/i || I || I-(A)V7 || I ||
|| (G)V7 || I || I || I ||
79 MNTPRT195 – Valentim de 1924 Manuel de Toada Singela Último fado de Manuel de Figueiredo. Fado com secções bem O primeiro compasso é na 4/4 S/indicação Lá menor/Lá Maior 8 + 16 + 8 + 16 + 4 Intro ABAB || (Am) i-V7-i || V7 || V7 || VI ||
Carvalho Figueiredo Creação de Aldina Sousa. delineadas, em que os realidade uma anacruza + 8 + 8 compassos Intermezzo CDCD || VI-iv-6F || i || V7 || I ||
contracantos da guitarra com três tempos. Intermezzo EE' || (A) I || V7 || V7 || I ||
estão escritos, o que não é Coda || I || V7 || V7 || I ||
habitual.Também estão || I || V7 || V7 || vi ||
escritas secções || V7/V || iv/V-6F/V || V7 || I ||
intermédias, puramente || I || V7 || V7 || i ||
instrumentais. || (Am ) i || V7 || V7 || i ||
|| i || V7 || V7 || i ||
|| i || iv || V7 || i || (2x)
|| i-V7-i || V7 || V7 || i ||
|| i || V7 || V7 || i ||
|| i || V7 || V7 || i ||
|| i-V7-i || V7 || V7 || VI ||
|| VI-iv-6F || i || V7 || i ||
80 MNTPRT196 – Custódio 1920 João F. Victoria Fado dos Aviadores Música e versos de João Victoria 4/4 Vagaroso Ré menor/Ré Maior 8 + 16 compassos Intro AA BB || (Dm) i || V7 || V7 || i ||
Cardoso Pereira & Cª. || i || iv || i-V7 || i ||
|| (D) I || IV || V7 || I ||
|| I || IV || V7 || I ||
|| IV || I || V7 || I ||
|| IV || I || V7 || I ||
81 MNTPRT204 – Edições 1942 Frederico de Freitas O páteo das Versos de António Lopes RibeiroCuriosa ambiguidade Curiosa ambiguidade entre 4/4 Vagaroso Ré menor 4 + 16 + 4 + 4 Intro AABB C D || (Dm) iv || i || V7 || i ||
Sassetti Cantigas – Primeiro entre o modo maior e menor na secção B e D em que a o modo maior e menor na compassos || i || V7 || V7 || i ||
Fado própria tonalidade notada não parece ser coerente. secção B e D em que a || i || V7 || V7 || i ||
própria tonalidade notada || I || iv || V7 || i ||
não parece ser coerente. || I || iv || i-V7 || I ||
|| (D) I || V7 || V7 || I ||
|| (Dm) iv || i || V7 || i ||
82 MNTPRT206 – Edições 1933 Frederico de Freitas Fado da Espera de Versos do Dr Júlio DantasCriação de Dina TeresaDo A “Coda” é na realidade a 4/4 S/indicação Sol Maior 2 + 36 compassos Intro ABAB CD ABB || (G) I || I ||
Sassetti e Cª. Toiros Fonofilme “A Severa” repetição reforçada do || I || V7 || V7 || I ||
último verso sobre o tema || I || V7 || V7 || I || (2x)
“B”. || (Em) i || V7 || V7 || i ||
|| iv || V7 || V7 || i-(G) V7 ||
|| I || V7 || V7 || I ||
|| I || V7 || V7 || I ||
|| I || V7 || V7 || I ||
83 MNTPRT207 – Edição de 1905 Reynaldo Varella Fado da Rosa Versos de Reynaldo Varella Na fonte, ao compasso 2/4 Allegretto Fá menor 2 + 16 compassos Intro AA' BC || (Fm) i || V7 ||
Eduardo Roza À distinta actriz cantora Maria Vitória dois, a mão direita Moderato || i || V7 || V7 || i ||
apresenta uma semínima || i || V7 || V7 || i ||
em vez de uma colcheia. || (Ab) V7 || I || (Fm) V7 || i ||
Na fonte, ao compasso || iv || i || V7 || i ||
treze, a mão direita
apresenta um fá a mais no
baixo, que não faz sentido,
em conjugação com o mi
bemol.
84 MNTPRT208 – Cruz e Sousa, 1950 Cruz e Sousa Saudades sem fim Letra de José PatrícioDedicada pelos autores a Amália A fonte ao compasso Na fonte ao compasso três, 4/4 S/indicação Ré Maior/Ré menor 4 + 16 + 16 Intro ABAC DEDE || (D) I-vi07 || ii || V7 || I ||
Ponta Delgada Rodrigues dezassete apresenta fá no falta a tercina na mão compassos || (Dm) i || V7 || V7 || i ||
baixo. Dado o contexto direita.Na fonte ao || I || V7 || V7 || I-i ||
harmónico de quarto grau, compasso dezanove faltam || i || V7 || V7 || i ||
por decisão pessoal e de as tercinas na || iv || i || V7 || i ||
coerência decidi substituir ornamentação da mão || (D) I || V7 || V7 || I ||
por sol, sendo a nota que direita.Na fonte ao || I-vi07 || ii || V7 || I || (2x)
faz mais sentido.Constante compasso vinte e sete falta
alternância entre o modo um tempo de pausa na
maior e menor. A fonte mão esquerda.
apresenta uma detalhe rico
em termos de transcrição de
ornamentos vocais e dos
contracantos, bem como da
linha do baixo, sendo
claramente escrita com
quatro vozes em mente e
não na típica estrutura de
melodia e
acompanhamento.
85 MNTPRT230 – Tip. A. Pinto 1946 Fernando de Guitarras da Sé Letra de Arnaldo Leite e Campos Monteiro Na fonte antes dos últimos Na fonte, ao compasso 2/4 Fado Fá Maior/Dó menor 4 + 32 + 16 Intro ABAC DE || (F) I || V7 || I || V7/v ||
de Campos Carvalho Grande sucesso da opereta “A bonequinha do Porto” dois compassos a armação décimo falta o bequadro no Marcha compassos || (Cm) i || i || V7/V || V7/V ||
Criação dos simpáticos artistas Alberto Reis e Saluquia de clave é alterada para dó si, na mão esquerda. || V7 || V7 || i || i ||
Rentini. menor. Optei por manter a || (Eb) I || I || V7 || V7 ||
armação em Fá maior || V7 || V7 || I || V7/iii ||
porque o final faz sentido || (Cm) i || i || V7/V || V7/V ||
com a cadência perfeita e || V7 || V7 || i || i ||
não com a cadência ao || iv || iv || i || i ||
modo menor de fá que não || V7/V || V7 || I || I ||
pertence ao contexto da || (F) I || I || V7 || V7 ||
peça e da tradição em || V7 || V7 || I || I ||
causa.O ritmo harmónico || V7/ii || V7/ii || ii || ii ||
neste fado está dobrado, || I || V7 || I-V7 || I ||
devido à escolha de
compasso 2/4.
86 MNTPRT233 – Edições 1941 Raúl Ferrão O Colete Letra de José Galhardo Estranha repartição de Na fonte ao compasso 2/4 Moderato Ré menor/Ré Maior 9 + 15 + 18 + 16 + Intro AA' BB' CD || (Dm) i || iv || V7 || i ||
Sassetti Encarnado Criação de Mirita Casimiro compassos na forma. dezoito é omisso um dos 2 compassos Coda || i || iv || V7 || i ||
Da Opereta “O colete Encarnado” fás do acorde na mão || (D) I || I-i0 || V7 || V7 ||
esquerda. || V7 || V7 || I || I || (2x)
Na fonte ao compasso || (F) V7 || V7 || I (D) ||
décimo é omisso o “a || I || V7 || V7 || I ||
tempo” para contrariar o || (F) V7 || V7 || I ||
ralentando da introdução. || I || V7 || V7 || I ||
|| (Dm)V7 || i || i || i ||
|| i || V7 || V7 || V7 ||
|| V7 || i || i || (A) bII ||
|| bII || V || i#07 ||
|| I (D) || V7 || I-V7 || I ||
87 MNTPRT234 – Edições 1943 Fernando de Um adeus que me Letra de Arnaldo Leite e Campos Monteiro Curioso uso de dominantes 2/4 Fado Sol menor/Sol 8 + 32 + 16 Intro ABAC DE || (G) (V7/V)/V || (V7/V)/V || V7/V || V7/V ||
Valentim de Carvalho Carvalho esqueceu Criação da novel actriz Maria Clara terciárias. Marcha Maior compassos || V7 || V7 || I-V7 || I ||
Da Opereta “A costureirinha da Sé” Fado sem entrada em || (Gm) i || i || iv || iv ||
anacruze no canto, isto é || V7 || V7 || i || i ||
altamente invulgar. || i || i || V7 || V7 ||
|| V7 || V7 || i || i-V7 ||
|| i || i || iv || iv ||
|| V7 || V7 || i || i ||
|| iv || iv || i || i ||
|| V7 || V7 || I || I ||
|| (G) I || I || ii || ii-vi ||
|| V7 || V7 || I || I ||
|| (V7/V)V || (V7/V)/V || V7/V || V7/V |
||| V7 || V7 || I || I ||
88 MNTPRT236 – Lit. Monteiro T. 1930 Alfredo Mantua A canção do Sul Letra de Acácio Antunes e Machado CorreiaFado Fado com arranjo 4/4 Allegretto Ré menor/Ré Maior 9 + 8 + 12 Intro AB CDE || (Dm) i || i || V7 || i-V7 ||
Pedras Negras canção da revista “De ponta a ponta”Cantado pelo extremamente completo e Moderato compassos || i || i || ii || i ||
tenor Alves da Silva complexo. Várias vozes e || V7 || i || i-#iv07-ii07 || i-V7/iv ||
camadas detalhadas.Muitas || iv-V7 || i || iv6 || i || V7 || I ||
indicações de dinâmicas, || (D) I || V7 || V7 || I ||
ornamentos e || (Dm) iv6 || i || (D)V7/V || V7-I ||
expressões.Grande || (Dm) V7/iv || iv6 || i-V7 || i ||
ambiguidade entre o modo
maior e menor. Inclusão de
acordes complexos, como
sendo o quarto grau com a
sexta agregada.
89 MNTPRT237 – Depósito 1900 Fado Amoroso Canções Populares do Norte nº2 Separata sem qualquer 2/4 S/indicação Lá menor/Dó Maior 32 compassos ABAB' CDCD' || (Am) i || V7 || V7 || i ||
Praça de D Pedro 57 (Figueira da Foz) indicação de autor, editora || i || V7 || V7 || i ||
ou data. || (C) I || V7 || V7 || I ||
|| I || V7 || V7 || I ||
|| (Am) i || iv || V7 || i ||
|| i || iv || V7 || i ||
|| (C) I || V7 || V7 || I ||
|| I || V7 || V7 || I ||
90 MNTPRT238 1915 António Marques da Fado da Elegância Da revista “À última hora” Separata sem qualquer A fonte não contém 4/4 Allegretto Lá menor/Lá Maior 4 + 16 + 8 + 4 Intro ABCD EF || (A) I || V7 || V7 || I ||
Silva e da rua indicação de editora ou data indicação de compasso.Na Moderato compassos Coda || (Am) i-V7 || i || iv || i-V7 ||
Segundo o Arquivo distrital fonte, ao compasso vinte, || i-V7 || i-(C) IV || I-V7 || I ||
do Porto a revista foi falta o ponto de || V7 || i || V7 || i ||
performada em 1915 ou aumentação e uma || iv || i || V7 || I ||
1916 no Teatro Apolo no ligadura na 2ª voz da mão || (A) I || V7 || V7 || I ||
Porto, em Lisboa, com direita.Na fonte, ao || I-V7/ii || ii || I-V7 || I ||
libreto de Augusto Veras e compasso vinte e cinco, a || I || V7 || V7 || I ||
Simões de Castro música de mão direita apresenta uma
Manuel de Figueiredo e semínima ao invés de uma
Vasco de Macedo. mínima.
91 MNTPRT239 – Valentim de 1967 Carlos Dias Fado do Cacilheiro Letra de Paulo da Fonseca.Criação de José A fonte apresenta uma Na transcrição foi 2/4 Fado Sol menor/Sol 4 + 31 + 16 + 2 Intro ABA'C DD' || (G) V || I || I || I ||
Carvalho Viana.Gravação Columbia. primeira secção com 31 eliminada uma Marcha, Não Maior compassos Coda || (Gm) n.c. || i || i || i ||
compassos o que não faz redundância de dois Depressa || i-#vi07 || V7 || V7 || V7 ||
sentido. Quanto a mim a compassos na fonte (cp. || n.c. || iv || iv || V7 ||
última frase é deficitária de 54-55 no original), quanto à || V7 || i || V7 || i ||
um compasso e nota-se a última repetição. || n.c. || i || i || i ||
sua falta.A estrutura da || i || iv || iv || iv ||
fonte aparece algo confusa || iv || i || i || V/V || V || I || I ||
e a progressão harmónica || (G) I || I || I || I || I ||
aparece com ritmo dobrado. || I-i07 || ii || ii || ii || iv ||
|| I || I-i07 || ii || V || I-V || I ||
92 MNTPRT246 – Salão 1905 Filipe Duarte Canção de Da Revista “O Cabo da Caçarola Mágica”F. A. Oliveira Este fado tem uma forma Sistematicamente são 2/4 Allegro Fá Maior 7 + 12 + 14 + 12 + Intro AcA'c BcB'cc || (F) I || V7 || I || V7 ||
Neuparth – Neuparth e Belphogor e S. Marques pouco vulgar na medida em omissas ligaduras de Moderato 12 + 2 compassos CcC'c DD'cc Coda || I || V7 || II || I || V7 ||
Carneiro que apresenta um pequeno expressão na mão *c = pequeno côro || I || V7 || I || I (Am) i || V7 || i ||
côro de dois compassos no esquerda que foram de 2 compassos. || V7 || i (F) V7 || I || V7 ||
final de cada frase. Por acrescentadas de acordo || I || V7 || I || V7 ||
vezes esse próprio côro com o evidência interna. || I || V7 || I || V7 ||
repete-se a si mesmo, || I || V7 || I (Dm) V7 || i || V7 ||
causando um apêndice de || i || V7 || i || (C) V7 ||
quatro compassos aos || I-V7 || I-V7 || I || V7 ||
versos. || I-(F) V7 || I || V7 || I || V7 ||
|| I || V7 || I || V7 ||
|| I || V7 || I || I ||
93 MNTPRT247 – Sassetti & Cª 1927 Venceslau Pinto, Fado das Iscas Da Opereta “Bairro Alto”Letra de Avelino de Na fonte, o compasso 31, 4/4 Moderato Fá Maior 5 + 20 + 12 Intro ABABB' CC'C'' || (F) I || V7 || V7 || I ||
Editores Alves Coelho, Raúl SousaCriação dos distintos actores Vasco Santana e na mão esquerda compassos || I || V7 || V7 || I ||
Portela Aurélio Ribeiro apresenta o acorde errado, || (Dm) i || V7 || V7 || i || (2x)
tal como é comprovado por || (Dm) iv || i || V7 || i ||
evidência interna ao || (D) I || V7 || V7 || I ||
compasso 35. Em lugar do || V7/ii || ii || I-V7 || I ||
segundo grau é a tónica. || V7/ii || ii || I-V7 || I ||
94 MNTPRT248 – s/indicação 1927 Venceslau Pinto, Fado do Bairro Alto Da Opereta “Bairro Alto” Este fado apresenta uma A fonte é omissa no 4/4 S/indicação Lá Maior/Lá menor 4 + 8 + 4 + 16 + 9* Intro AA' Intermezzo || (A) I || V7 || V7 || I ||
Alves Coelho, Raúl dinâmica de ff no início e acompanhamento compassos BCB'D EE' || I || I || V7 || V7 ||
Portela uma dinâmica de pp ao limitando-se a apresentar || V7 || V7 || V7 || i ||
quarto compasso. Isto melodia e posições. Optei || (Am) i || V7 || V7 || i ||
causa um desequilíbrio por escrever as posições || i || V7 || V7 || i ||
notável pelo que me parece indicadas sob a forma de || III || iv-V7 || i || V ||
uma decisão pouco feliz. acordes em semínimas.Na || i || V7 || V7 || i ||
Optei por homogeneizar o fonte ao compasso catorze || iv || V7 || iv || V7 ||
fado como meio-forte. falta o sustenido no sol da || (A) I || V7 || V7 || I ||
A transcrição parece-me mão direita.A fonte ao || bVI || I || V7 || V7 || I ||
duvidosa na estrutura na compasso trinta e três
forma de abordar a nota apresenta pausa de semi-
longa na última secção, colcheia ao invés de pausa
acrescentando-lhe um de colcheia.
compasso e tornando-a
assimétrica.
Este fado apresenta alguma
curiosidade harmónica na
introdução do III e do bVI
em compassos esporádicos
no seu interior.
95 MNTPRT249 – E. Bevilacqua 1911 Filipe Duarte Fado do Colete Da Opereta “O Fado”Letra de Bento Faria 2/4 Moderato Lá menor/Lá Maior 5 + 16 + 16 + 2 Intro ABAB' CDCD' || (Am) i || V7 || i || V7 || i ||
& C., Rio de Janeiro Encarnado compassos Coda || i || V/III || IV/III || I/III ||
MNTPRT252 – Neuparth & || iv || V7 || iv || V7 ||
Carneiro, Lisboa || i || V/III || IV/III || I/III ||
|| iv || V7 || V7 || I ||
|| (A) I || V7 || V7 || I ||
|| I || V7 || V7 || I || (2x)
|| V7 || I ||
96 MNTPRT250 – Sassetti & Cª. 1920 Coutinho de Fado Triste Da opereta “O casal da Margarida”Ao João da Câmara 4/4 S/indicação Ré Maior/Ré menor 8 + 16 compassos Intro AABB' || (D) I || V7 || V7 || I ||
Editores Oliveira || I || ii || I-V7 || I ||
|| (Dm) i || V7 || V7 || i ||
|| i || V7 || V7 || i ||
|| (D) IV || I || V7 || I ||
|| IV || I || V7 || I ||
97 MNTPRT251 – Lit. Salles, 1920 Filipe Duarte Fado Da Opereta “A leiteira de Entre-Arroios” Harmonia mais complexa, Na fonte ao 9º compasso é 4/4 Allegretto Ré Maior/Si menor 6 + 40 + 6 Intro AA BB || (D) V7 || I || ii || I || V7 || I-V7-I ||
Lda. Versos de Penha Coutinho com encadeamentos omissa a indicação de compassos || I || ii || V7 || I ||
Cantado pela actriz Ausenda de Oliveira e Sales sucessivos de dominantes tercina na mão esquerda. || I || ii || V7 || I ||
Ribeiro sobre dominantes. Na fonte ao 14º e 18º || I || vi07 || V7 || I ||
compassos é omisso o || bVI || I || ii-V7 || I ||
ponto de aumentação na || (Bm) i || vii07 || V7 || i ||
mão esquerda. || (D) vii07 || vi || IV || [(V7/V)/V]/V ||
Na fonte ao 20º compasso || (V7/V)/V || V7/V || V7 || I ||
a mão direita apresenta || (Bm) ii07 || i || V7 || i ||
semi-colcheias ao invés de || (D) i || ii || V7 || i ||
colcheias. || bVI || I || ii-V7 || I ||
Na fonte ao 34º compasso || V7 || I || ii || I || V7 || I-V7-I ||
é omisso o ponto de
aumentação na mão direita
Na fonte ao 34º compasso
a mão esquerda apresenta
colcheias e não semi-
colcheias no ostinato.
Na fonte ao 48º compasso
é omisso o bequadro no dó
da mão direita.
98 MNTPRT253 – Tip. A. Pinto 1945 Raúl Ferrão Lisboa não sejas Letra de José Galhardo A única referência a fado 2/2 Tempo de Lá menor/Lá Maior 4 + 16 + 16 Intro AA'BA' CDCD' || (Am) V7 || i || V7 || i ||
de Campos Francesa Criação da actriz Mirita Casimiro nesta obra é a indicação de Fado Slow compassos || i || V7 || V7 || i ||
Da Opereta “A Invasão” tempo. || i || V7 || V7 || i ||
|| i-(C) V7 || I-V7/ii || ii-V7 || I-(Am) V7 ||
|| I || V7 || V7 || I ||
|| (A) I || I || I || ii ||
|| ii || V7 || V7 || I ||
|| I || I || I || ii ||
|| ii-ii0 || I || V7/V-V7 || I ||
99 MNTPRT259 – A Editora, 1911 Filipe Duarte Fado das Da Opereta “O Fado” Não há indicação do autor Na fonte, ao 1º compasso 4/4 Allegretto Dó menor/Dó maior 2 + 14 + 16 + 4 Intro ABAB CDCD || (Cm) i-V7 || i-V7i-V7 || i-V7/bIII || bIII-V7 ||
Lisboa Educandas da letra. é omisso o bequadro no si, Agitato compassos Coda || i-V7 || i-V7 || i-V7 || i-V7 || (2x)
A estrutura formal parece- do acorde do V7. || (C) I-V7 || I-V7 || I-V7 || I-V7 ||
nos muito peculiar. O autor || I-V7 || I-V7 || I-V7 || I-V7 ||
ao invés de entrar em || I-V7 || I-V7 || I-V || I-V7 ||
anacruze, inicia o verso no || I-V7 || I-V7 || i-V7 || i-V7 ||
tempo forte do compasso || I || I || I || I ||
prolongando-o.
Conservando esta estranha
opção, obriga-o a prolongar
também a última sílaba. Tal
escolha traduz-se numa
frase de 3 compassos,
seguida de uma de quatro.
Quanto a mim este fado
poderia ser rescrito num
molde mais tradicional com
frases de 2+2 compassos
em anacruze, e sem
prolongamentos internos.
100 MNTPRT260 – Edições 1945 Fernando de Fado das Letra de Arnaldo Leite e Campos Monteiroda Opereta 2/4 Tempo de Dó Maior/Dó menor 8 + 16 + 1 +16 Intro ABA'C DEFG || (C) IV || IV-#iv07 || V7 || I ||
Populares Portuguesas, Tip. Carvalho Caravelas “O Gaiato da Rua”Criação da notável artista Irene Isidro Fado compassos || V7 || V7 || I || i ||
A. Pinto de Campos Marcha || (Cm) i || i || i || V7 ||
|| V7 || V7 || V7 || i ||
|| i || i || V/iv || iv ||
|| iv || i || V7 || I || I ||
|| (C) I || I || ii || ii ||
|| V7 || V7 || I || I ||
|| IV || IV-#iv07 || I || I ||
|| V7 || V7 || I-V7 || I ||
001
2 œ œ œœœ œ œ œ #œ œ. œ œ œ œ œ œ. #œ œ œ œ œ
&b 4 J œ œ œ
Piano
? 2     
b 4
œ
œ œ œ œ œ œ œ #œ œ. œ œ œ œ œ œ. #œ œ œœ
&b
5
œ œ œ
?    
b
œ œ œ œ œ œ #œ œ nœ œ œ œ œ œ œ œ œ
œ œ œ  œ
&b
9
?b    
œ œ œ œ œ œ œ œœœ œ
œ œ œ œ œ œ œ œœ j
&b #œ
13
?    
b
002
Andante
b bbb 2 œ œ œ œ œ œ œ œ œ œ nœ œ œ œ œ œ œ nœ œ œ œ œ œ œ œ œ œ œ
& 4 J œ œ œ
? b b 2  œ œœ œœ œœ œœ œœ œœ
nœ œ œ œ
œœ œœ œœ œ œœ œœ œœ
Piano sentimental
bb 4 œ œ
n œ œ
b œ œœœœ nœ œ œ œ œ œ œ œ œœœœœœœ
& b bb œ œ œ œ œ œ œ œ œ œ nœ œ
5
œœ œœ œœ œ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ

? bb b œ n œ œœ œ œ
b
bb b b œ œ œr œ œ œ œ n œ œ œ œ œ œ œ n œ œ œr œ œ œ œ œ œ œ œ œ œ œ œ œ
9 K K
& œ œ œ œ
œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ
? bb b œ œ œ œ
b
bb b œ œ œKr œ œ œ nn nn b
& b œ œ œ œ nœ œ œ œ œ œ œ œ œ œ œ œ nœ œ œ ‰ œJ
13
? bb œ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œ œœ œœJ ‰

bb œ œ nn nn b
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
&b ‰ ‰
17
J J
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
?b œ œ œ œ œ œ
&b ‰ ‰ J
21
J
œ œ
2 002
œ œ œ œ œ œ œ ‰ œJ
&b œ œ œ
25
œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœœœ
? œ œ
b œ œ œ J ‰
b œ œ n œ œ œ œ œ œ œ œ œ  œ œ œ nœ œ œ œ œ œ œ œ œ  œ
29
&
? œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œœœ œœœ œœœ

b œ œ œ œ œ œ œ œ
œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
&b nœ œ  nœ œ
33
?b œ œœœ œœœ œœœ œœ œœ œœ œœœ œœœ œœœ œ œœ œœ œœ

œ œ œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰ œ
37
& J
œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œ œ
?b œ œ œ œ œ œJ ‰
œ œ œ œ œ œ œ œ œ œ j
œ œ œ œ
&b œ œ œ œ œ
41
œ œ œ œ œ œ œ œ œ œ
? œ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œ œœœ œœœ

b œ œ œ J
003
œ œ. œ œ œ œ
Andante
2 j œ. œ œ œ œ œ #œ œ . œ œ œ œ œ œ
&b 4 œ ‰ œJ ‰ J
p sentimental
? 2  œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
b 4 œ
œ #œ œ œ. œ œ œ œ œ œ
.
&b ‰ J ‰ œJ œ œ œ œ œ œ œ ‰ œ œ œ œ œ
6
J œ
? b œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ
œ œ œ œ œ
œ œ. œœ œ œ œ œ
œ œ œœœ œ œ œ ‰ j
&b œ œ ‰ J ‰ J œ
12
?b œœœœœœœœ œœœœœœœœ œœœ œœœ œ œ œ œ œ œœœœœ ‰

œ œ œ #œ œ J
j œ œ œ j
&b œ œœœ œ œ œ ‰ œ œ œ œ œ ‰ œ œ œœœ œ
17
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ
œ œ œ
œ œ œœœ œ œ œ œ œ œ
& b œ œ ‰ œJ œ œœœ œ ‰J ‰ J
22
? b œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ
œ J œ
œ œ œ œ œœœ œ œ œ œ œ œœœ œ
b ‰ J ‰ J œ ‰
28
&
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ
œ J
004
Allegretto
2   Œ ‰ œj œ . œ œJ œ œ œ œ
&4 J J
? 2 œ œœ œ œ # œœœœ œœ œ œ #œ
Piano
4 œ œ
œœ Œ œ œ œ œ œ œ œ œ œ œ œ œ
j œ. œ œ œ œ ˙ œ ‰ œJ œ. œ
& œ ‰
7
œ J J J J
? œ œ # œœ œ #œ
œ œ œ œ
#œ
œ œ œ œ œ œœ œ œ œœ œ œ œœ œ
œ œ œ
œ œ œ ˙ œ ‰ œJ œ. œ œ œ #œ ˙
13
& J
? œ œœ œ #œ
œ œ œ œ
#œ
œ œ œ œ œ œœ œ œ # œœ œ œ œœ œ
œ œ œ œ
& œ ‰ œj œ . œ œ œ œ œ œ œ ‰ œj œ . œ
19
J J J J
? œ œ œ œ œ œ #œ #œ #œ
œ Œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
˙ œ œ œ
œ
&J œ
œ ‰ œJ œ. œ œ ˙
25
J J
? œ œ # œœ œ œ œœ œ œ œœ œ œ œœ œ œ œœ œ #œ
œ œ œ œ
œ œ œ œ
œ ‰ œJ œ. œ œ œ #œ Œ
˙ œ
31
& J
? œ œ # œœ œ œ œœ œ œ # œœ œ œ œœ œ Œ
œ œ œ œ
005
2 œ œ œ œ œ œ œ œ œJ œ
Andantino
œ œœ œœ œ œ œ
&4 J J œ J J
?2  œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
4 œ œ œ œ œ œ
œ œ
œ œ œ œ #œ œ œ œ #œ œ œ œj œ œ
5
& J J J
? œ œ œ œ œ œ
œ œ œ œ œ œJ # œ œ œ œ œ œ œ œ ‰ œJ
9
& J #œ
? j
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœœ œ ‰
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
&œ œ œ œ
13
#œ
?
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œj
&œ œ œ œ œ œ
17
#œ
? j
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœœœ
006
Andante
& b 42  œ œ  œ œ  œ œ œœœ
j
œœ ‰ ‰ œj œ œ œ œ œ œ œœœ œ œ œ œ
œ œ
pœ œ œœ œ œ
j j j
? b 42 œ ‰ œ ‰ œj ‰ ‰ j œ ‰ Œ œœœœœœœœ œœœœœœœœ œœœœœœœœ
Piano
œ œ œ œ
œ œ
&b œ ‰ j œ œœœ œ œ œ œ œ œ  œ bœ œ œ œ œ
7
œ œ œ œœ œ œ œ
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ bœ œ œ œ œ œ
dolce
U
&b œ œ œ œ œ œ œœ œ œ œ œ œ œ œ œœ œ œ œ œ œ œ œ œœ œ
13
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ b œ œ œ œ œ œ œœ  Œ œ œ œ
U
b œ œ œ œ œ œœ
&b œ œ œ œ œ œ œœ œ œ œœœ œ œ œ œ œ ‰ j œ œ œ œ
19
œ œ œ œ
? œœœœœ ‰ œœœœœœœœ œœœœœœœœ œœœœœœœœ œœœœœœœœ œœœœœœœœ

b J
&b œ œœœ œ œ œ œ œ œ  œ bœ œ œ œ œ œ œ œ œ œ œ œœ œ
25
œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
dolce
? b œ œ œ œ œ œ œ œ œ œ œ œ œ bœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
U j
&b œ  œ œ œ œ œ œ œ œ œ œ œ œ œ œ
31
œ œ œ œ œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ b œ œ œ œ œ œ œœ  Œ œ œ œ œ œ œ œœœœœ
U
b œ œ J
007
Augusto Hylario
U % j
Andante expressivo
2 œ œ œœ œ œ œ œ œ œ œ œ œ œ
&b 4 œ
f p
œ gg œœœ œœ œœ œœ œœœ œœœ œœœ
? 2 ggg œœœ g# œœœœ # œœœ œ
œ œœ
Piano
b 4 gg œ ggg œ ‰ œ œ œ œ œ œ
œ œ
œ œ  œ #œ
& b œ œ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
7
œ œ œ œœ œœ œœ œœœ œœœ œœœ b n œœ œœ

? b œ œœ œœ œœ œ œ œ œ œœœ # œ œœœ œ œ
œ œ œJ ‰ œ œ nœ œ
œ œ
œ œ œœ œ
& b œ œ œ  œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
13
? b œ œœ œœ œœ œœ œœ œœ œœ ‰ b n œœœ œœœ œœœ œ œ œœ œœœ

œ œ œ œ œJ nœ œ œ œJ ‰ œ œ
œ  b œœ b œœ b œœ œ œœ œ œœ œ œ b œœ b œœ b œ œ œ œ b œ b œ œ œœ b œœ b œœ œ œœ œ
œ œ œ œ œ
œœ
&b œ œ œ œ  œ œœ œ œ œ œ  œ œ
19
œ œ œ
? ggg œœœ U œœ œœ œœ b œœœ œœœ œœœ b œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ

dolce
b gJ ‰ Œ œ œœ œ œ
b œ œ œ œ
b œ œœ œ
b
œ
œ bœ bœ bœ œ bœ bœ œ œ bœ œ œ œ œ bœ j bœ
& b œ œœ œœ  œ œ œ œœ œ œœ œ œ œ  œ œ œ œ œ œ œ œœ ‰ b œœ b œœ œ œœ b œœ œœ
25
? b œ b œœœ œœœ œœœ b œœœ œœœ œœœ œ œœ œ œ œ b b œœ œœ œœ b œœœ œœœ

œ œ b œœ œ ‰
J œ b œœ œœ œœ œ œ œ œ œ œ
bœ j j %
& b b œœ œœ œœ  b œœ b œœ œœ œœ œœ œœ œ œœ ‰ b œœ b œœ b œœ œœ b œœ œ b œœ œœ ‰
31
? b œ b œœœ œœœ œœœ œ œ

œ b œœ œ œœ b b œœ œœ œœ b œœœ œœœ œ œ
œ b œœ œœ ‰
œ œ œ œ œ œ
008
Andante
b 2 œ œ. œ #œ œ œ œ œ n œ ‰. œœœ œ œ œ
œ œ. œ œ œ œ œ
œ œ. ‰.
&b 4 J R R
œ œ œ œ œ œ œœœ œ œ œœ œ œœœ œ œ œœœ œœœ œœœ œœœ
express.
? b 2  œ œ œ œ œ œ œ œ œ
Piano
b 4
b
&b œ œ ‰ œ œ œ œ œ #œ ‰ .. œJ œ . œ #œ œ œ œ œ nœ ‰. œ
6
J œ œ R
œ œ œ œ œ œ œ œ œ œœœ œœœ œ œ œ œ œ œ œ œ
? b œ œ œ œ œ œ .‰ œ œ œ œ œ œ œ œ
b J .
b œ œ œ œ œ œ œ ‰.
œ œ . œ œ œ œ œ œ œ ‰ œ œ œ œ œ # œ œ œ ‰ œJ
& b œ.
12
R J œ
œ œ œœœ œ œ œœœ œœœ œœœ œœœ œœ œ œœœ œ œ œ œ œœœ œœœ œœœ
? bb œ œ œ œ œ œ œ œ œ œ
b œ. œ#œ œ œ œ œ nœ ‰. œ œ. œœœ œ œ œ
‰. œ œ. œ œ œ œ œ œ œ ‰ œ
&b
18
R R J
œ œ œ œ œ œ œœœ œ œ œœ œ œœœ œ œ œœœ œœœ œœœ œœœ œœ œ œœœ œ œ
? bb œ œ œ œ œ œ œ œ œ
b œ œ œ œ œ œ
& b œ œ œ œ œ #œ œ œ ‰ J œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ
24
œ # œ œ œ
œ œ œ œ œ
? b œœœ œ œ œ œ œ ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b J œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ
%
b œœœœœœ
b œœ œ œœœ œ œ œ œ #œ ‰ ‰ œ œ
30 D.C. ao
& œœ œ œœ œœ ˙˙
œ ˙
œ œ œ
? b œ œ œœœ œœœœœœœœ œ œœ œœœœ œ œ œ ‰ œœ œœ ˙˙
b œ œ œ J œ œ ˙
009
Andante
b b b b C œ œ œ . œj œ œr œ œ œ œ œ n œ œ . j œ œr œ n œ œ . œj œ œ
& œ œ œ nœ œ œ œ nœ œœ
Piano
p
? bb b C 
b œœœœ œœ œœ œœœ œœœœ œ œœœœ œœ œ œ œœœœ œœœœ
b œ œ œ œ nœ. œœ ‰ j % j œ œr œ œ
& b bb J œ œœœ œ Œ .. ‰ œ œ œ œ . œœ
5
œ
œ œ nœ œ
? b b b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ .. œ œ œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
bb j œ œr œ n œ œ . j j r
& b b œ œ œ nœ œ œ œ nœ œ. œ œ œ œ œ œ. œ œ œ œ œ œ œ œ œ nœ œ œ nœ
11
œ nœ œ
? bb b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b j j œ œ œ œ nœ. œ œ. œ nœ œ œ.
& b bb œ . œ nœ œ œ œ nœ œ. œ œœœœ
r
J J œœœ
16
? bb b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
bb œ œ
r
j œ œ œ
r
& b b œ œ œ œ œ œ œ œ œ œ nœ œ œ œ. œ
œ œ œ. œ œ œ n œ œ œ œ ‰ œ
21
n œ œ
nœ œ œ .
? bb b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b j %
& b bb œ œ œ œ n œ œ œ œ. œ œ œ. œ œ œ œ n œ œ œ .. 
26
œ
D.C.
n œ œ
nœ œ œ .
? bb b œ œ œ œ .
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ . œ Œ
010
œ œ œœœœ œ ˙ œ œ œ œœœœ œ
œ œ œ œ œ œœ ˙
Andante
b œ œ œ œœœœ œ œ Œ œ
&b c œ œ œ œ œ œ
œ œ œœœœœœ œœœ œœœ œœœ œœœ
Piano
? bb c  œ œ œ œ œ œ œ œ œ
œ œ œ
œ œ œœœ œ ˙
œ œœœ œœ˙ œ œ œ œœ
œ œ œ œ %U
b
&b œ œ œ œ œ œ œœ
œœ œ œ Œ œ œ œ œœœœ
5
œ
p
œ
? bb œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œœœ œœ œœœœ œ Œ œ
œœœ œœœ
œ
œ œ œ
b
&b œœ˙ œ œ œ œœœœ œ œ Œ œ œ œ œœœœ œœ˙ œ
11
œ œ œ œ œ
? bb œœ œœœ œœœ œœ œœœœœœœœ œœœœœœœœ œœ œœœ
œ œ œ œ
b
&b œ œ œœœœ œ œ Œ œ œœœœ œ
con 8ª
œœœœ œ œ œ œœ œ œ œœœœ
16
f
? bb œ œ œ œ œ œ œ œ œ œ œ Œ œ œ œ œ œœœ œœœ œœœ
œœœœœ œ œ
b œ œ œœ
&b œ œ Œ œ œœœœœ œ œ œ œœœœ œ œ Œ œ
21
œœœ œ
? bb œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ œœœ œœœœœœœœ

œœ œœ œ œ
œœœœœ œ
%
b
&b œ œ œ œ œ œ œ œ œ œ œ œ œ 
26
œ œ œ œ œ œ
œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ Œ
œ œ œ œ
œ œ œ
011
œ œ œ œ œ œ ‰ œj œ .
Andante
## 2 j œ . j œœ œœ
& 4 œ œ œ œ œ œ œ ‰ œ œ.
œ œ œ œ œ œ œ œœ œ œœ
gracioso
? # # 2 ‰ œ œ œ œ œœ œ œ œ œ œœ œ œ œ œ œœ œ œ
Piano
œ œ œ œ œ œ œ œ
4
## œ . œ œ œ œ . œ j j %
œ œ
œ œ œ œj ‰ œœœ ‰ œœ ‰ .. ‰ œj œ . œ œ œ j
œ œ ‰ œ œ. œ œ œ œ
6
& œœ œ
? ## œ œ œœ œ œ œ œœœ œœ œœ œœ . œ œ œ œœœ œ œ œ œ
œ œ œ œ œœ œ œ œ œ œœ
œ
œ œ œ œ œ ‰œ ‰ œ ‰ .  œ œ œ œ
J J J
# œ œ ‰ j œ. j œ œ œ œ œ œ ‰ œj œ .
& # œ œ œ œ œ œ œ ‰ œ œ. œœ œœ
14
œ œ œ œ œ œ œ œ œ œœ œ œœ
? # # œ œ œ œ œœ œ œ œ œ œœ œ œ œ œ œœ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ
# j
& # œ œ ‰ œ œ. œ œ œ œ œ œ ‰ œ œ. œ œ œ œ œ. œ œ œ œ. œ œ
œ œœ
20
? # # œ œ œ œ œ œœœ œ œ œ œ œ œœœ œ œ œœœ œ œ œ œœœ œ œ œœœ œ œœœ

œ œ œœ œ œ œ œ œ œ œ œ œ œ
## œ œ. œ œ œ œ œ. œ œ œ œ œ. %
‰ œ œ œ œ j ‰ .. ggg œœœ ggg œœœ Œ
26
& œ œ J œ gg œ gg œ
œ œ œ œ
? # # œ œ œ œ œœ œ œ œ œ œœ œ œ œœœ œ œ œ œœœ œ œœœ .. œ œ Œ
œ œ œ œ œ
œ
012
Andante
# r
& 42 œr œ œ œ œ œ œ œ œ ‰ . œ œ œ œ œ œ œ œ œ ‰ . œr
œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ œœœ œœ
? # 42  œ œ œ œ œ œ œ
Piano
œ œ 
# r œ œœ
œ œ œ œ œ œ œ œ œ ‰ œœ .. n œœ # œœ œ œ œ œœ
œœ œ œœ œ œ ‰.
5
& œ
œ œ œ œ œœœœœœœ œœœ œœœ œœœ œœœ œœœ œœœ
?# œ œ œ œ œ œ œ œ œ œ œ
# œ œ œ œ œ œ œ œ ‰ œ .. œ œ œœ œ œœ
œ œ ‰
œœ .. œœ œœ œ n œ # œœ œ œ œ œœ œ ‰ œœ .. œœ
10
& œ œœ œ œ œ
œœœœœœœ œœœ œœœ œœœœœœœœ œœœ œœœ œœœœœœœœ
?# œ œ œ œ œ
œœ œœ  œ œ œœ œœ œ œ
# j
& œœ œœ œœ œœ œ œ ‰ ‰ . œ œ œ œœ œ œ
15
R
œ œ œ œ œ œ œ œœ œœœœœœœœ œœœœœœœœ œœœœ œœœ
?# œ œ œ œ œ œ œ œ
# œ œœ œœ œ œ œœ œ œ œœ œœ  œ œ œœ œœ œ œ œ œ œ
 œ œ œ œ #œ
20
&
œ œ œ œ œ œ œ œ œœœœœœœ œœœ œœœ œœœ œœ œ
?# œ œ œ œ œ œ œ œ œ œ œ œ œ
# œ nœ œ œ œ #œ œ œ œ œ œ œ œ ‰ œ #œ œ nœ œ œ œ
œ œ œ ‰
25
&
œœœ œœœ œœ œ œœœ œ œ œœœ œœœ œœœ œœ œ œœœ œœœ
? # #œ œ œ œ œ œ # œ œ
# œ œ œ #œ œ œ œ œ j œœ œœ
D.C.
‰ œ œ ‰
30
& œ œ J J
?# œ œ œ œ œ œ œ œ œ œ œ œ œ ‰
J
013
Reynaldo Varella
ggg œœœœ œ œ œ œ œ œ œ œœœ %U r

Andante
b œ œ.
2 œ œ  œ œ œ
&bb 4 gg J œ œ. œ œ œ œ œ œ
f p
2 ˙˙˙ œœ .. U œ œ œ
œ œ œ œ œœœ œœœ œœœ
Piano
? b ‰
bb 4 œ. œ
b œ œ.
& b b œ. nœ œ bœ œ #œ œ  œ œ. œ œ œ œ œ œ œ œ œ œ. nœ œ bœ œ
6
? b œ œœœ œœœ œœœ œ œ œ

œ œ œ œ
œ œ œ
œ œ œ œ œœœ œœœ œœœ œ œ œ
œ œœ œœ œœ
bb œ
b œ œ œ œ
& b b #œ œ  œ œ. nœ nœ œ œ 
œ œ. œ nœ œ. œ œ œ
r
11
? b œ œœ œœ œœ œœ œœ œ œ œ
œ œ œ œ œœ œœ
œœ œœ œœ
bb nœ œ œ œ nœ œ œ œ œ œ œ œ
U
œ. œ œ.
r
bb œ œ œ œ. œ bœ. j‰
œ
& b œ.
16
œ œ œ œ œ œ
U
? b œ œœœ œœœ œ œ œ
œ œ œ œ œœ
œ œ
œœœ œ œ
œ œ œJ ‰
bb œ œ
014
Andante
2
&b 4 Œ  œ #œ œ œ  œ #œ œ
#œ
 œ #œ œ œ œ
œ œ
dolce
? 2 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ
Piano
b 4 œ
j œ j
&b œ ‰ œ œ œ œ œ # œ œ ‰ œj œ œ œœ œ œ œ ‰ œ
5
? œ œœœ œœœ ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b J œ
œ œ œ œ œ #œ œ ‰ j j
œ œ œœ œ œ œ ‰ œ œ œ œ œ
œ
&b
10
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ
œ œ œ œœ j œ œ
&b œ œ ‰ J œ œ œ ‰ œ œ œ œœ œ ‰ œJ
15
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ
b œ
œ œ œ œ œ
&b œ œ J ‰ Œ
20
œœ œœ œœ
D.C.
œ œ œ
œ œ œ œ œ œ
?b œ œ œ œ œ œ œ ‰ œ œ œ Œ
J œ œ œ
015
##
Andante
& # C œœœ œ œ œœœœ œ œ œœœœ œœœœœœœœ œ œ œ œœ

r
œ
œ
f
? # ## C Œ ‰ œœ œœ œœ œ œ œœ œœ œœ œœ œœ œœ
Piano
œ œ œ œ œ œœ œœ œ œ œ œ œ
œ
œ œ œ
## œ œ œ
& # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ #˙ œœœœ
5
? # ## œœ œœ œœ œœ œœ œœ œ œ œ œœœ œœœ œœœ œœœ œœœ œœœ œ œ œ

œ œ œ œ œ œ œ œ œ # œœ œœ œœ œ œ œ # œœ œœ œœ
# ## œ œ œ œ œ œ œ
œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
11
&
? # ## œœœ œœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œœ

œ œ
# Œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
## r œ œ
& # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰ œ œ #œ
16
? # ## œœœ œœœ œœœ œ œ œœ œ œ œœ œœ œœ œœ œœœ œœœ

œ œ œœ œœ œ œ œœ œœ œ œ œ œ
# œ
œ
Œ
# ##
œ œ ‰œ œ #œ œ ‰ #˙ œœœœ
œ œ œ œ œ œ #œ œ œ œ ‰ œ œ #œ œ œ ‰
œ œ œ #œ œ
‰
œ
21
&
? # ## œœœ œœœ œœœ œ œ œœœ œ œœœ œœœ œœœ œœœ œœœ

œ # œœ œœ Œ œ # œœ Œ
œ œ œ œ
# ## # ˙ %
œ œ œ œ œ œ œ œ œ œ #œ œ j Œ ‰
26
& œ œ œ
? # ## œ œ œœ œœ œœ œœœ œœ œœ
# œœ œ œ œ œ œ œ œ
J Œ ‰
œ
016

## œ œ œ
c ‰ œ œ œj ‰ œ œ œ œ œj ‰ œ œ œ œ œ ‰ œ œ œ œ
Allegro vivo
& J œ ˙ œ
p
c  œ œœœœœœœ œœœ œœœ œœœœœœœœ œœœœœ
Piano
? ##
œ œ
## . j
& .œ œ œ œ œ œj œ œ œ œ œ ‰ œ œ œ œJ œ œ œ œ Œ œ
5
œ J
? # # ..  œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
# j j œœ
& # œ œ œœ œ œœœœœœ‰
œ œ œ œ œ Œ œ
10
J J œ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
## œ œ œœ œ œœ œœœ‰œ œ j j
œ œœ Œ œ
14
& J J œ œ œ œ
? ## œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ

œ œ œ œ
œ œ œ œ
## œ œ œ œ œ œ œ j
œœœœ‰œ œ œ œœ j Œ .. Œ
18
& J J œ œ œ
? ## œœœ œœœ œœœ œœœ œœœ œœœ œœœ .

œ œ œ œ . Œ
œ œ œ œ
017
# ### 2 œ
Andantino
œ œœœœ
& # 4 ‰ œ œœœœ Œ ‰ œ Œ
œ œœœ œ œœœ
p
? # ### 2 œ œ œ
Piano
# 4 ˙ ˙ ˙ œ Œ ˙œ ˙ ˙ œ Œ
˙ ˙ ˙ œ ˙ ˙ ˙ œ
### # ‰ œ Œ Œ ‰ œJ œ œ j j
& # œœœœ  œœ œ œ œ
9
œ œœœœ
? ### # œ œ œ œ œ œ œ
# ˙˙ ˙˙ ˙˙ œœ Œ œ œ œ œœœ œœœ œœœ œœœ
### # œ œ
# œj ‰ ‰ J œ œ j ‰ ‰ J œ œ œj œ œj œ œ
j j
18
& œ œœ œ œ œ
? ### # œ œ œ œ œ œ
œ œ œ œœœœ œœ œ œ œ œœœœ
œœ œœ œ œ œ œ
# œ œ
### # œ j œ œ
# œj ‰ ‰ J œ œ j j j j œ œ œJ œ œ
26
& œœ œ œ œ œœ œ œ œ
œ œœ œœœ œœœ œœœ œœœ œœœ
? ### # œ œœ œœ œ œ œ œ œ
# œ œ œ œ
### # œ œ œ œ j œ œ œj œ œ œœ œœ
j
& # ˙ œœ œœ j œœ œ œ œj œ œ
34
œœ œ Œ œ J œ œ
œ œ œ œ œ œ
? ### # œœ œœœ œœ œ œœœ œ œ œ œœœ œ œœ œœ
# œ œ œ œ
### # œ œ œ œ j
& # ˙ œœ j j ‰ Œ
42
œœ œœ œœ œœ œ œ œ
? ### # œ œ
œ œ œ œ
# œ œ œ œ œ œ œ œ j ‰ Œ
œ œ œ
018
Moderato
# 4 ‰ U j j
. œ œ œ œ œ œ.
& 4 œ‰ œ ‰ œ‰œœœ œ œ œ œ œ #œ
p
? # 44 œ œ œ œ œ œ œ œ Uj ‰ Œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ
Piano
œ‰Œ œ‰ Œ œ‰Œ œ
J J J
# j j j j
& œ. œ œ œ œ #œ œ. œ œ œ œ œ œ. œ œ œ œ œ œ. œ œ œ œ #œ
5
? # œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ
# j j œ œ œ j
& œ. œ œ œ œ #œ ˙ œ # œœ œœ œœ œ œ œ œ œ œ ‰ œœ œœ œœ œœ œœ
œ ‰ œ œ œ œ
9
œ œ œ œ œ œ œ œ œœœ œœœ œœœ

?# œ #œ œ œ œ ‰ Œ œ œ œ œ‰Œ
J J
# œœ œœ œœ œœ j j œj ‰ œ œ œ œœ œ œ œ œ
& œœ œœ œ˙ .
œ œœ # œœ œœ œ œ œ œ œ œ œ œ œ œœ # œœ œœ
œ
13
J
œœœ œœœ œœœ
?# œ œ œ œ œ œ œ œ œ œ œ œ œ ‰ Œ
3
œ œ œ œ‰ Œ
J J
# œ. j j . j
œ œœ œ œ # œ ˙ œ‰œ œ œ œ œ œ œ
17
& #˙ œ œœœ œ #œ œ œ œ œ #œ
J
?# œœœœ œœœ œ œœœœœ ‰Œ œ œœ œœ œœ œ # œœœ œœœ Œ
œœ œœ œœ
œ œ œ œ
#
J
# ‰ j
j œ œ œ œ . j
œ œ œ‰Œ
22
& œ˙œ œ œ œ œ œœœ œ #œ œ œ œ œ #œ ˙

J œ
? # œ œœ œœ Œ œ œœ œœ œœ œœ œœ
œ œ œ
# Œ
œœ œœ œœ
œ œ œ œ
# œ œœ œœœ ‰ Œ
J
019
Andantino
b 3 Œ Œ ‰ œ œ œ œ œ. œ
&bb 4  
J

p
œ œ œ œ œœœœœ
? b 3 œœœœ œ œœœœ œ œœœ œ œœœœ
Piano
bb 4 œ œ˙ .
œ Œ Œ œ Œ Œ œ Œ Œ œ Œ Œ
b œ. œ œ j œœ œ. œ œ
&bb œ j
œ‰œ
œœœ œ‰œ j
œ‰œ
6
œ œ œœ œœœœœ œ œ œœ œ œœœ œ
? b œ˙ . œ œ˙ . œ˙ . œ œ˙ . œ œ˙ œ. œ œ œ
bb
b œ œ. œ œ j œ œ œ œ. œ œ
r
&bb œœ œ‰œ œ j‰ œ œ œœ
11
œ œœ
œ œ œœ œ œ œœ œ œ œœ œ. œ œ œ œ œ˙ œ. œ œ œ œ
? b œ˙ . œ œ˙ . œ œ˙ . œ œ˙
bb
b j‰ œ œ œ œ œ œ . œ œ j‰
r
j‰ Œ
&bb œ œ œœ
16
œ œ œ œœ œ œ
œ œ œ œ œ
? b œ˙ œ. œ œ œ œ˙ .œ œ œ œ œ˙ œ. œ œ œ œ˙ œ. œ œ œ œ˙ œ. œ œ
bb
020
Gracioso
# 2 j
& # 4 œ #œ œ œ œ œ œ nœ œ
p
Piano
œ œ œ œ œ œ œ œ œ œ œ œ
? ## 2 ‰ œ œ œ
4 œ
## r
#œ œ œ œ nœ œ ‰. œ œ. œ œ œ #œ
3
&
? ## œ œ œ œ œ
œ
## r
œ œ œ
œ. œœ œœ ‰. œ œ. œœ œ #œ
6
& œ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ
## œ. œ œ ‰ ‰
œ œ
10
& œ œ
D.C.
œ œ
œ œ œ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ ‰
œ J
021
j j
Moderato
# œœ œœ œ œœ˙ œ œ œœ œœ œœ œœ œœ
& # c Ó Œ œ J J œœ œ
œ
œ œ œ œœ œœ œœ
? # # c œœœ œœœ œœœ Œ œœœ œœœ œœœ œœœ œœœ œœœ
Piano
œ œ œ œ œ œ
œ œ œ
## œ
& œ œœ Œ œ œœ œœ œ œœ˙ œ œ œœj œœ œœj œœ œ œœ œ œ œ Œ ‰ j
5
J J œ œ œ œ œ
œ œ œ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ
? # # œ œœ œœ œœ œ œ œ œ œ œ œ œ
œ œ
œ œ œ œ œ
## j j œ œj Œ j
& œ œ œJ œ œ ‰ œj œj œ œj œ œ œ œ œ Œ ‰ œj
10
œ J
? ## œœ œœ œœ œœ œœ œœ œœœ œœœ œœœ œœ œœ œœ

œ
# j j j ‰ œj j œj œ
& # œ œ œJ œ œ œ œ œ Œ Œ
14
œ œ œ ˙ œ
? ## œœ œœ œœ œœ œœ œœ œœœ œœœ œœœ œ œ

œ œ œ œ œ œ œ œ œ œ œœ œœ Œ
œ
022
Andante
2 j œ œ œ œ œ œ œ œ ‰ œ œ œ œ œ
&4 œ J œ œ
dolce
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
?2 ‰
4
j œ
&œ œ ‰ œœœ œ œœ œ œ ‰ J œœœ œ œœ
4
?œœœœœœœœ œœœœœœœœ œœœ œœœ œœœ œœœ

œ œ œ œ
j œ œ
&œ œ ‰ œ œ œ œ œ œ
8
œ œ
?œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
j œ œ œ œ œ œ
&œ œ œ ‰
11
œ œ œ œ
œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ œ
&œ œ œ œ œ ‰ ‰
14
œ œ
D.C.
œ œ œ œ
?œ œ œ œ œ œ œ œ œ œ œ œ œ œ j
œ œ œ œ œ œ œ ‰
023
Moderato
2 œœ œ œ œ ‰ œ œ œ œ œ œ œœ ‰ j œ œœ œ
& 4 œj œ œ J
œ œ œ
Piano f
? 2 ‰ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ
4 œ œ #œ #œ œ œ œ œ
œ œ
œ œ ‰ œ %
œ œ œ œ œ œœœ ‰ j œ œ œœ œ œ œ ‰ œ
6
& J œ J
p
? œœœ œœœ œ œ œ œ œ œ œ œ œ œ œœœœ  œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ
#œ œ #œ œ œ
J
œ œ œ ‰ œ
& œ œ œ œ œ œ œ œ ‰ œj œ œ œ œ œ œ œœ œ œœœ
12
? œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ
#œ œ œ œ œ œ #œ œ #œ œ œ
œ œœ
& ‰ œJ œ œ œ œ œ œ œ ‰ # œ œ œ # œ œ œ ‰ œJ œ œ œ œ œ
18
F
? œœœ œœœ œœœ œœœ œœœ œœ œ œœ œ œœœ œœœ œœœ œœœ
œ œ œ œ œ œ œ œ œ œ œ
& œ œ ‰ #œ œ œ œ œ œ #œ œ œ œ œ œ œœ ‰ j
24
#œ œ œ œ œ œ œ #œ
? œœœ œœœ œ œ œ œ œœœ œœœ œœœ œœœ œ œ œ œ œ œ œ œ

œ œ #œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ
%
& œ œ ‰ œj œ œ œ œ # œ œ ‰ j #œ œ œ ‰ ‰
30
œ œ œ œ #œ
? œœœ œœœ œœœ œœœ œ œ œ œ œ œ œ œ œ œ œœ

œ œ œ œ #œ œ œ œ œ œ œ œ œœ ‰
œ œ J
024
Andantino
b b 2 œ œ œ # œ
r
œ . œ œ œ œ œr œ œ œ œ œ . œ œr œ œ # œ œ . œ œ œ
& 4 J œ œ œœ œ œ
œ
p
? b 2 ‰ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
Piano
b 4 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ
œ œ œ
b %
&b œ œœœœ œ ‰ œJ œ œ œ œ œ # œ œ . œ œ œ œ œ œ œ œ œ œ . œ œ œ
7 r r r
œ
F
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? b œ œ œ œœ ‰ œ œ œ œ œ œ œ œ œ œ œœ
b œ œ œ
œ œ œ
bb œ . œ œ œ œœ œ œ œ œ #œ œ. œ
r
œ # œ œ œœ
r r r
œ œœ œ ‰
œ œ J œ œ œœ œ œ œ
14
& œ œ
œ
? b œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ

b œ œ œœ œ œ œ œœœœ œœœ œ œ œœ œœ œ
œ œ œ œ
b œ . œ œr œ œ # œ œ . œ
œœ œ œ œ œ œ œ ‰ j œ. œ
r
b
21
œ
& œœ œ œ œ œ œ œ œ. œ œ #œ
œœ œœ œœ œœ œœ œœ œœ œœ œœ # œœ œœ œœ
? b œœœ œœœ œœœ œœœ œœœ œœœ œ œœ œœ œ
œœ œœ
œœ‰ œ œ œ œ œ œ œ
b œ œ œ œ
œ œ
b %
&b œ
œ. œ œ œ #œ œ. œ œ œ. œ œ œ œ œ. œ œ #œ œ. œ œ œ #œ œ
‰ ‰
28
œ œ œ œœ œœ œœ œœ œœ œœ # œœ œœ œœ # œœ œœ œœ œœ œœ
? bb # œœ œœ œœ œ œœ œ œ œ œ œ œ œ œ œ œ œ œœ ‰
œ œ œ œ
025
U %
Andante
# 2 œ ‰ gg œœ gg œœœ ‰ ggœœœ ggg œœœ ‰ ‰ œj  œR
& 4 ‰ gg œœœ ggg œœ ‰ gg œ g
‰
gg g gg œœ
‰
g gg
p
j j j
?# 2 œ œ ‰ j œ œ
Piano
4 œ œ œ œ ‰ œ œ œ œ ‰
œ œ œ œ
# œ œ œ œ #œ œ nœ œ  œ œ œ œ œ œ œ œ œ  œ
6
&
œœ œœ œœ œœ œœ œœ œœ œœ
?# œ œ ‰ œœ œœ ‰ œœ œœ ‰ œ œ ‰
œ œ œ
œ
10
# œ œ œ œ #œ œ nœ œ  œ œ œ œ œ œ œ œ œ  œ
& œ
?# œœ œœ œœœ œœœ œœœ œœœ œœ œœ

œ œ œ ‰ œ œ œ ‰
œ
œ œ ‰ œ œ œ ‰
# r
œ œ œ œ œ œ  œ œ œ œ œ œ œ œ ‰. œ
14
& œ œ œ œ œ œ œ œ œ œ
œœ œœ œœ œœ ‰ œœ œœ œœ œœ
?# œ œ ‰ œœ œœ ‰ œ œ ‰
œ œ œ
œ
# %
œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰ ‰
18
& œ œ œ œ œ œ œ œ œ
?# œ œ ‰ ‰ œœ œœ ‰ œ œ ‰
œ œ œ
œ
026
Andantino
2 œ œ. œ œ œ œ œ #œ ‰ œ œ. œ œ œ œ œ œ ‰ œ
&b 4 J J J
f
? 2  œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
b 4 œ
œ. œœœ œ œ
&b œ ‰ œJ j ‰
D.C.
œ.
5
œœœ
#œ œ œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ ‰
b œ J
027
Moderato
# 2 jU‰ ‰ j
& 4 œ
œœœ œœœ œ œœœ œ œ œ œ œ œœœœœ œœœ œ œ œ œ #œ œ œ ‰ œ œJ
f
j œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
œ œ œ œ‰ 
Piano
?# 2 œ œ œ œ œ œ œœ œœ œœ œœ œœ œœ
4 œ œ œ œ œ œ œ œ œ
# œœœ œœœ j j œ
œ œ œœœœ œ œ œ œ #œ œ
8
& œœœ œ œ œ œ œ œœœœ œœœ
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
?# œ œœ œœ œœ œœ œœ œ œœ œœ œœ
œœ œ œ
œ œ œ
œœœ
œ
œ
œ
œ
œ
œœœ
œ œ
# œ j œ
‰ ‰ # œ œ œœœ œ œ œ œ œ œ œj ‰ ‰ œ œ œ œ œ œ œ œ œ œœœ œ
J #œ
15
& J œœœ
J #œ
? # œ œœœ œ # # œœœ # œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œ œ œ œ œ œœ œœ

œ œœ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ
œ
#
œ œ œ nœ œ œ œ œ œ œ j‰ Œ ‰ œ œ œ œ œ œ œ œ œ #œ
23
& œ œ #œ œœœ
œ #œ œ œ J J
œœ œœ œœ œ œœ œœ œœ œœœ œœœ œœœ œ œœ œœ œœ œ œœ œœ œœ œœ œœ œ œ œ
?# œ # œ œ œ œ # œ œœ œ # œ œœ œœ # œœ
#
œ œ œ œ œ nœ œ œ œ œ œ œ j‰ 
30
& œ œ#œ œ œ #œ
œ #œ œ œ
?# œœ œœ œœœ œœœ œœœ œ œœ œœ œœ œ œœœ œœœ œœœ œ œœ œœ œœ œ œœ œœœ ‰

œ œ œ
028
r U % œ œ # œ œ œ œ
Moderato
# 2 nœ œ œœœ œ
& 4 œœ
œ
œœ
œ
œœ
œ J œ œ œ
? # 2 œ œ œ r U ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
4 œ œ œ œ œ œ œ œ
# œ
& œ œ œ œ œ #œ œ œ œ œœ œ œœ œœ œœ œœ œœ œ œœ œ œ œ œ ‰ j
7
œœ œ œ œ
? # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ bœ œ œ œ œ œ œ œ œ œ œ
œ œ #œ œ œ œ œ œ œ œ œ œ
œ œ
#
& œ œ œ œ œ œ œ ‰ œj œ œ œ œ œ œ œ ‰ œJ
12
œ œ œ œ
œ
?# œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ
# %
œ œ œ œ œ œ œ ‰ ‰
bœ œ œ
17
& œ  œ œ œ
?# œ œ œ œ œ œ œ œ œ œœ œœœ œœ œœœ œœœ

œ œ œ œ œ J ‰
029
Andante
b
& b 42 œ œ ˙
j
œ œœœ ˙ j œ œ œ œ œ œ œ ‰ œ ‰ ˙˙ œœ œ œ œ œ œ œ ‰ œ ‰
#œ J J J J J
F
œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ œœ œ œ œœœœ œœœœ œœœœ
? b 2 Œ œ œ œ œ œ
Piano
#œ nœ
b 4
j j
b œ œ œ œ œ œ œ œ œ # œ œ œ œ œj ‰ Œ
Œ œ œ œ ‰ œœ ‰ œœ ‰ œœ ‰ # œœ ‰ Œ ‰ œ œ œ œ ‰ œœ ‰
&b ˙
11
J J J J J J
œœœ œ œ œ œ œœ œœœœ œœœ œœœ œœœ œœœ œ œ œ œ œœœœ œœœœ
? b #œ #œ œ œ œ œ œ
b

j
b œ œ œ œœœ ˙ œ œ #œ œ œ œ
& b œJ ‰ # œ ‰ œ ‰ Œ ‰ œ n œœ œœ ˙œ œ œ œ # œ œ œ œ œ œ œ
22
˙ ˙ œ œ œ œ œ
J
œ œ œ œœœ œœœ œœœ œ œ œ œ œ œ œœœ œœœ œœœ
? b œ œ œ œ œ œ œ œ #œ œ #œ #œ #œ
b
b b () œ œ
œ œœ œ œ œ œ œ œœ
32
& ˙˙œ œ œ œ #œ
œ œ œ œ # ˙˙œ œ œ #œ nœ œ œœ œ œ œœ # œœ ˙˙œ œ œ œ #œ
œ œ œ œ œ œ œ œ œœœœ œœœœ œœœœ œœœœ œœœœ œ œ

() œ œ œ œ œ #œ
bb œ œ œ œ œ œ œ œ œ œ œœœ œ j j œ œ œ œ œ #œ
œ œ #œ œ œ œ œ ‰ œ ‰ Œ
42
& œ œ œ œ
œ œ œ œ œœœ œœœœ œ œœ œœœœ œœœ œ œ œ œ œœœœ
? b œ œ œ œ œ œ œ œ Œ œ œ œ œ
b
œ œœ œ œ œ œ œ œ œ œ œ œ œ œ œœœœ
bb Œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ J
Œ Œ ‰Œ Œ ‰J
53
&
œœœ œœœœ œœœœ œœœ œœœ œœœ œœœ œœ œœ œ œ œœ
? b œ œ œ œ œ
J‰Œ  œ
b
b #œ œ œ œ œ j
&b Œ ? Œ j ‰ j ‰ œj ‰ œj ‰ œ ‰ œ ‰ Œ & ‰ œ œ œ œ œ œ œ œ # œ œ œ œ œ Œ ? Œ j ‰
64
œ œ J œ
œœœ œœœ œ œ œ œ œ œ œ œ œ œ œœœœ œœœ œœ œ
? b œ œ œ Œ œ œ œ œ œ œ œ œ œ œ œ œ
b
j j
? bb j ‰ j ‰ œ ‰ œ ‰ œ ‰ Œ ‰ œ œ œ œ œœœ œ #œ œ œ œ œ Œ 
75
œ œ J &
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ Œ
? b œ œ
b
030
%j œ œ œ œœ œœ œ œ œ œœ
2 # œ œ œ # œ
Andante
b œ œ œ œ œ œ œ œ. œ
r
& 4 œ œ œ œ œ
p
# œœ œ œ œ #œ œ œ œ œ
œ # œ œ œœ œ œ œ œ
? 2 ‰ œ œ œ œ œ œ œ œ œ œ œ œ œJ
Piano
b 4 œ J
j
œ œ- .  œ .  Œ œœ œ œ œ œœ
rit. a tempo
b ‰ . œR œ œ œ œ
r
œ . œ œ œ
7
& J - œ œ œœ œ œ
p
œ. #œ. œ œ œœ œœ œ œ # œœ œœ œ œœ œœ œœ
? b œœ ..  œ .  œ œ œ œ œ œ œ œ œJ œ œ œ œœ œ œ œ
J œ. J œ œ J J
J
œ œ œ . œ œ œ œ œ # œ œ . œJ
& b œ J  œr œ œ œ œ œ œ ‰ . œr œ
r
œ
14 cresc.
œ œ œ
œ
œ œ œœ # œœ œ œ n œ bœ
œ œœ œ œœ œœ œ œœ # œœœ œœj
? œœ œ œ œ œœ ‰ J œJ œ œ
œ #œ #œ nœ
b J J œ œ
r
Jœœ œ
œ
œ %
œ œ
&b œ œ œ.  œ œ œ œ œ œ ‰ ‰
21
#œ œ
p Uu
œœ œ œ œ œœ # œœœ
? œ œ n œ # œœ j œœ œœ  œ œœ
œ œ ‰
b œ R œ œ J œ
031
% j
Andante
## 2 ‰ œ œ œ œ ®œ œ œ œ œ œ  œ
& 4 œœ œœ œœ
œ œ œ
p
œ œœ œœ œœ œ œœ œœ
Piano
? ## 2 œ ‰ ‰ ‰
4 œ œ œ
## œ œ œ œ œ œ œ j œ œ œ œ œ œ œ œœ œœ œ œœœ
‰œ ® ‰ œJ œ œ œ ‰J œœ
6
&
? # # œ œœ œœ ˙œ œœ œœ œœ œ œœ œœ œœ œ œœ œœ ‰ œ œœ œœ œ œœ œœ ‰ œ œœ œœ œœ
œ œ
## œ œ œ œœ
 œ œ œ œ ‰ ‰ j œ œœ œ ‰œ
13
& œ œ œ œœ œ J œ œ œ œ œœ J
? # # ˙œ œœ œœ œœ œœ œœ œ œœ œœ ‰ œ œœ œœ œœ œ œœ œœ œœ œœ œ œœ œ œœ œœ ‰
œ œ œ
# œ œ œ %
& # œ œ œ ‰ j œ œ œ ‰ ‰
20
œ œ œ œ œ
? # # œ œœ œœ œœ œ œœ œœ œœ œœ
œ
œœ œ œ
œ œ œ ‰
œ
032
Andantino
# 2 j œ œ œ œ œ œ. œœ œ œ œ œ œœ œ œ œ
& 4 œ
p
œœ œœ œœœ œœœ ‰ œœœ œ œœ œœ
Piano
?# 2 ‰ œ œ ‰ œœ œ œ ‰
4 œ œ œ
œ œ
# œ. #œ œ œ œ œ. œ j
œ œ œ œ œ œ œ œ  œr
5
&
f
œ œœ œœ œ œœ œœ œœ œœ œœ œœ œœ
?# ˙œ œ ˙œ œ œ œ
˙œ œ œ
œ œ ‰
# œ œ œœ œ œ
& œœ œœ œœ œœ œœ œœ .. œœ œœ  œœ œœ œ œ œ œ œœ  œœ œœ .. # # œœ œœ n œœ œœ
9
œœ œœ œ œ œ œ œœ œœ œ œ œ
?# œœ œœ ‰ œœ œœ ‰
œ œ œ ‰ œ œ œ œ œ ‰ ˙œ œ œ œ
# j
& œœ .. œœ œœ œœ œ œœ ‰
14
œœ œœ œ
D.C.
œœ œ
œ
? # ˙œ ˙œ œ œ œ œ ‰
œ œ
033
Andante
2 Œ œ œ‰œ œœœœ
&4 ‰ ‰
œœ œœ œ œœ b œœ
‰
œ œœ ‰ œ
œœ œ J
œ bœ œœ œ œœ
p œ F
œ œ œ bœ œ œœ œ œœ œœœ œœœ œœœ
Piano
? 2 ˙œ œ ˙œ ˙œ œ œ œ œ
4 ˙œ
j j
& œJ ‰ œ œ‰œ œ ‰œ œœœœ œ œ œ ‰œ œ‰œ œœœœ œ‰œ
8
J J J
œ œ œ œ
? œœœœ œœœœ œœœ œœœ œœœ œ œœ œœœ œœœ œœœœ
œ œ
œ j j b œ œ b œj
& œJ ‰ œ‰œ œœœœ œ œ œ‰œ bœ œ œ
17
J œ
œ œ bœ œ bœ œ œ œ
? œœœœ œœœ œœœ œœœ œ œœ œ œ œ œ œ bœ
œ œ
j bœ œ j
& œJ ‰ b œ b œ œ # œj œ . j j‰ œ bœ œ J bœ œ œ
25
œ œ bœ œ
œ œ œ œ œ œ œ œ bœ œ œ bœ œ œ œ œ
? œ œ bœ œ œ œ œ bœ œ bœ œ œ œ bœ
œ œ
j
& œJ ‰ b œ bœ œ Œ
33 D.C.
œ œ j ˙ œ
œ
œ œ œ œ œ œ œ œ bœ œ
? œ œ bœ œ œ Œ
œ œ
œ
034
Allegretto
b 2 j œ. œ œ œ œ œ œ ‰ j œ. œœ œœ œ œ ‰ j
&b 4 œ œ œ
Piano
? b 2 ‰ œ œ œœ œœ œœ œœ œœ œ œ œ
b 4 œ œ œ œ œ œ œ œ œ œ œ œ œ
b .
&b œ œ œ œœ œ œ ‰ j œ. œ œ œ œ œ œ ‰ œ
5
œ J
? b b œ œœ œœ œœœ œœœ œœœ œœœ œœœ œ œœ œœœ ‰

œ œ œ œ
b ‰ œJ
& b œ. œ œ œ œ ‰ œJ œ.
10
? b œœ œœ œ œ œ
œ œ œ œ œœœ œœœ œœœ œ œ œ
b œ nœ œ œ œ œ œ œ
b
& b œ. œ œ œ œ ‰ œJ œ. œ œ œ ‰ ‰
14
œ œ
D.C.
œ œ
? b œœ œœ œ œ œ
œ œ œ œ œœœ œœœ œœœ œ œ
b œ nœ œ œ œ œ œ ‰
035
# 2 œ œ .. œ œ œ œ œ. œ œ œ œ œ œœœ œ œ œ œ œ œœ œ œœ
Moderato
& 4 œJ œ œ œ œ œ œ. œ œ œ œ œ œœœ œ œ œ œ œ œœ œ œœ
J J
œ œ œ œ œ œ
? # 42 ‰ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ # œ œ œj ‰ œ œ œ œ œ œ œ
Piano
#œ
œ
# œœ œ œœ œ œ œ œ œ œ œj œ
œ œ œœ œ œ œ. œ œ œ œ œ. œœ œœ œ œœœ œ œ œ œ
œ
6
& œ œ J J J
?# œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ
œ œ œ œ J œ œ
14
# œ. œ œ œ œ œ. œ œ œœ œ œœœ œ œ œ œ œ. œ œ œœ œ œ œ œ
& J J
? # œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ
#œ œ œ
# œ. œ œ.
œ œœœ œ œ œ œœœ œ œ œ j j œ œœœ œ
20
& #œ œ œ œ œ œ œ œ œ
? # œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ
# j j
œœ œœ œ œ œœœ œ œœ œœ œ j ‰
27
& œ #œ œ œ œ œ œ #œ œ œ œ
? # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰
œ œ œ œ J
036
Allegretto
b 2 j
&bb 4 œ œ œ œœ œ œ j j
œ. œ œ œ œ œ œ œ œ œ
œ œœ œ œ œœ œ œ n œœ œ œ œœ œ
Piano
? b 2 ‰ œ œ œ
bb 4 œ
b b œ œ œ œ œ œ j j
& b œ œ œ
5
œ. œ œ œ œ œ œ
œ œ œœ œ œ n œœ œ œœ œ
? bb œ œ œ œ
œ
œ
œ
b œ
b j j
& b b œ œ œ œ œ œj œ j œ œ
9
œ œ œ œ œœ œ œ n œœ œ œ œœ œ
? b œ œ œ œ
bb œ
b
& b b œ œ œ œ œ œj œ j j ‰
13
? b œ œœ œ œ œœ œ œ n œœ œ œ œ
bb œ œ œ ‰
œ œ
037
Andante
2 œ œ œ œ œ. œ œ œ ‰ œJ œ œ œ œ. œ œ œ ‰ œ
&4 J J
œ
œœ œ #œ œ  œ #œ œ œœ œ
Piano
?2 ‰
4 œ  œ œ œ 
œ
& œ œ œ œ. œ œ œ ‰ œ œ œ ‰ œJ
5
J œ œ #œ #œ œ
œœ
? œ œ  œ #œ œ  œ #œ œ œœ œ
œ  #œ œ œ 
œ œ œ œ ‰ œ œ œ œ œ #œ ‰ œJ
œ œ
9
& #œ œ J
œ œ œ œ œ
?
œ  œ œ  œ #œ œ  œ œ œ  œ
œ
œ œ œ œ ‰ œ œ œ œ œ #œ Œ
œ œ
13
& #œ œ J
œ
?  œ œ œ œ  œ #œ œ  œ œ Œ
œ œ œ
038
Allegretto
# 2 œœ œ œ œ
& 4 JR œ œ œ œ œœ œ œ œ œœ œ œ œ œœ œ
œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œœœ œœœ
?# 2 ‰ œ œ œ œ œ œ œ œ œ
Piano
4 œ
# œ
œ œœ œ œ œ œ œ
œœ œ œ œ œœ œ
5
& œ œ œ
œœœ œ œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ
?# œ œ œ œ œ œ œ ‰
J
#
& œ œ œœ œ œ #œ œ œ #œ œ œ œ œ œ œ œ œ œ
9
œ
? # œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œœœœœœœœ œœœœœœœœ
œ
#
œœ œ œ œ œ #œ œ œ œ r ‰
D.C.
& œ œ
13
#œ œ #œ œ œ
? # œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰
œ œ R
039
Andante
2 œ œ. œ œ œ œ œ œ œ. œ œ œ œ œ œ ‰ .. œj
r
‰ œJ
œ r
œ
&4 J
œœ œ #œ  œ #œ œ œ œ .. 
Piano
?2 ‰
4 œ  œ œ œ œ 
œ
œœ œ j j
&œ œ œ œ ‰ œ #œ. œ œ œ œ œ œ ‰ œ
6
?  œœ œ  œ #œ œ  œ #œ œ  œ œ œ
œ œ œ œ
œœ œ j
&œ œ œ œ ‰ œ #œ. œ œ œ œ œ œ ‰ ‰
10
?  œœ œ  œ #œ œ  œ #œ œ œ œœœ ‰
œ œ œ œ. œ
J
040
%
Andante
b 2 œ œ œ œ œ œ
&bb 4 ‰ œ œ ‰ ‰ œ œ ‰ ‰ œœœ œœœ ‰ ‰
œœ œœ R
œœ œœ œœ œœ œ œ œœ
j œœ œœ œœ
j œ œ œ œ
Piano
? b 2 œ. j . œ œ Œ
bb 4 . œ œ. œ œ
œ œ œ . œ œ. œ œ
b œ œ œ œ œ œ œ œ œ œ
& b b œJ œ œœ œ œ œ œ J œ œœ
7
œ
œ œ œ œœ œœ œœ œœ œœ œœ
? b b œ œœ œœ œœ œœœ œœœ œœœ œ œ œ œ œ œ œ œ
œ œ œ
œ œœ œœ œœ
b œ
b b œ œ œ œ œ œ  œr œ n œ œ œ œ œJ
œ œ œ œ œ œ œ nœ
& b œ
12
J J
œ œ œ œœ œœ œœ œœ œœ œœ œœ œœ n œœœœ œœœ œœœ
? b œœ œœ œœ œ œ œ ‰ œ nœ œ œ œ œ œ œ œ
bb œ œ œ
b j œ œ œ œ œ œ œ œ œ nœ œ %
& b b œ œ ‰ œ œ nœ œ œ J œ ‰
17
J J
œœ œœ œœ n œœœ œœœ œœœ œœ œœ œœ œ œ œ
n œœœ œœœ œœœ œœ œœ
? b œ œ œ œ œ œ œ œ œ ‰
bb œ œ œ œ
041
Andante
2 nœ œ œ œ bœ j
& b 4 œJ œ œ œ œ œ œ œ . ‰ œJ ‰ .
œ œ . œ
œ œ œ œ œœ œœ œœ œœ œœ œ œ
? 2  œ œ œ œ œ œ œ œ œ œ œ .. ‰
Piano
b 4
œ œ œ œ œ œ œ œ œ œ #œ œ ‰ nœ
& b œ #œ #œ
6
J J J
#œ œ œœ œœ œœ
? œ œœ # œ œœ œ œ œ œ # œœ œœ œœ œœ œœ
b œ œ œ œ œ œ œ ‰
#
& b œ œ œ œ œ œj œ œ nœ œ œ bœ ‰ ‰
10
J œ œ œ
œ œ œ œ œœ œœ œœ œœœ œœœ œ œ
? œ œ œ œ œ œ œ œ œ ‰
b
042
Andante
2 j .. ‰ œj
&4 œ œœ œœ œ œ œ ‰ œ œ œ œœœœ œ œ
r
œ
J
?2 ‰ œœœœœœœœ œœœœœœœœ œœœ œœœ œœœ . œœœ

Piano
4 œ œ œ .œ
& œ œ œ œ œ œ œ œ ‰ œj œ œ œ œ œ .. ‰ œj œ œ œ œ œ
6
œ œ
? # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ .. œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
U
& œœœœœœœœ œ œ œ œ œ œ œ œ œ œ ‰ œj
j
œœœœœœœœ
r
œœœœœœœœ
r r
12
œ œ œ œ
? œœœœœœœœ œœœ œœœ œœœ œœœ œ œ œ œ œ œ œ œ Uœ œ œ œ

œ œ #œ œ œ œ
œ œ
U
j
‰ œ œ œ œ œ œ œ ‰ œj œ ‰ ‰
17
& œœ œœœ œ œ œ œ œœœ œ œ

œ
U
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰
œ œ œ œ J
043
œœ œ
Andante
2 j œ
& 4 ‰ œœœ œœœ ‰ ‰ œœœ œœœ ‰ ‰ œœœ œœœ ‰ œ œ
‰ œ œ œ
œ œ
œ
œ. j
œ œ. j
œ œ .. j œœœ œœœ œœœ
Piano
? 2 œ. j ‰ ‰
4 œ œ. œ œ œ œ œ
œ œ œ
œ œ œœ œ œ œ
&J œ œ œ œœ œ ‰ œj œ œ J œ
7
J œ J
œœ œœ œœ œœ œœ œœ
? œ œœœ œœœ œœœ œœœ œœœ œœœ œ œ œ œ œ œ œ œ
œ œ œ
œ œœ œœ œœ
œ
&œ œ œœ œ œ ‰ œj œ œ œ œ œ # œ œ. œ œœ œ
12
œ œ
œœ œœ œœ œœ œœ œ œ œ
? œœœ œœœ œœœ œ œ œ ‰ œ œ œ œ œ œ œ œ œœœ œœœ œœœ
œ œ
œ œœ
&œ œ. œ œ œ. œ œ œ œ œ j ‰
17
œ #œ
œ œ œ
œ œ œ
? œ œœœ œœœ œœœ œœ œœ œœ
œ œ œ œ œ œ œ œ œœœ œœœ œœœ œ œœ œœ ‰
œ
044
Andante
b b b 2 Œ ‰ œr œ n œ œ œr œ n œ œ œr œ n œ œ œ œ œ œ œr œ œ œ œr œ œ  œr œ n œ
& b 4 œ œ nœ
R
f
? b b 2 œœœ œœœ ‰ œœ œœ œœ œœ œœ œœ œœ œœ œœ
Piano
bb 4 œ œ œ ‰
œ œ œ œ œ œ
b j
& b bb œ œ nœ œ œ nœ œ œ œ œ œ œœœ œ
r r r
œ œ œ œ œ œ
r
6
œ œ œ
? bb b œœ œœ œœ œœ œœ œœ œœ œœœ œœœ
b œ œ œ œ ‰
œ œ
bb nœ
& b b œ œ œ nœ œ œ œ bœ ‰ œ œ œ œ œ œ œ ‰ œœ
10 3
œ œ nœ œ œ
? b b n œ œœœ œœœ œœœ œœ œœ œœ

nœ œ œ œ œ
œœ œœœ œ œ
œ œ œ ‰
bb œ
bb nœ
& b b œ œ œ nœ œ œ œ bœ ‰ œ œ œ œ œ œ œ Œ
14 3
œ œ nœ œ œ
? b b n œ œœœ œœœ œœœ œœ œœ œœ

nœ œ œ œ œ
œœ œœœ œ œ
œ œ œ ‰
bb œ
045

‰ œj œ œ œ œ œ œ
Andante
bb 2 j
& b b 4 œj œ œ œ œ œ œ œ œ ‰ œ œ œ œ œ œ n œ œ œ
œœ œœ œœ œœ
œœ œœ œ œœœ œœœ œœ œœ œ œ œ
Piano
? bb 2 ‰ œ œ ‰ œœ‰ ‰ œ n œ œ ‰
bb 4 œ œ œ
œ
() %
bb b b œ œ œ œ œ œ œœœ œ ‰ j œ œ œœœœ œ œ œ ‰ œ
œ œ œ
6
& œ J
œœ œ œœœ œœœ œœ œœ œœ
? bb œ œ œ œ œ œœ œœ
œ œ ‰
œœ œœ œœ
bb œ œ œ
bb œ œ œ. œ
& b b œ œ œ œ œ œ œ œ ‰ œj œ œ œ œ œ œ œ œ œ  œ œ œ .
12
œ œ œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœœ œœœ
? b b œœœ œœœ œœœ œ œ œ œ œ œ œœ œœ œœ nœ œ
bb œ œ
œ œ
b œ j
& b b b œ œ ‰ œJ œ œ œ œ œ œ œ  œ n œ œ œ œ œ œ œ œ ‰ œJ
17
œœ
œ œ œ œ œ œ œ œœ œœ œœ œœ œ œ œ
? bb b œ œ œ ‰ œ œ œ œ œ œ œ œ œ œ œ
n œ œ œ œ
b
bb œ œ œ œ œ œ œ œ  %
&bb œ œ œ œ ‰ ‰
22
œ nœ œ œ nœ œ œ
œ œ œœ œ œ œœ œœ œœ œœ œ œ
? bb œ œ œ œ œ œ œ œ
n œ œ œ œ œ ‰
bb
046
# 2 œj œ . œœ œœ œœ œœ œœ .. œ  œ œ . œœ œœ œœ œœ œœ .. œœ œ  œ
Andante
& 4 œ œ. œ œœ œ œ . œ œ
p
œ œ œ œ œ œ œœœ œ œ œœœ œœœ œœœ œœœ
?# 2 ‰ œ œ œ œ œ œ œ
Piano
4 œ œ
# . œœ œœ œœ œœ œœ .. œœ œ œœ œœ œœ œœ œœ .. œœ œœ
& œœ .  œœ œœ ..  œœ
5
œ
œœœ œœœ œœœœ œœœœ œœœ œœœ œœœ œœœ
?# œ œ œ œ œ œ
# j j
& # œœ .. œœ œœ œœ œœ œœ œœ œœ œœ n œ œœ .. œœ œœ œœ œœ
9
œœ œ
œ œ œ œ œœœœœœœ œœœ œœœ œœœ œœœ
?# œ œ œ œ œ œ œ œ œ
#
& # œœ .. œœ œœ œœ œœ œœ œœ œœ œœ n œ œ
œ œ œ œ .. œœ œ
j
œ ‰
13
œœœ œ œœ œœœœ œœœœ œœœ œœœ œœœ

?# œ œ œ œ œ œ ‰
J
047
Andante
2 œ
& 4 œ œ œ œ œ œ œ œ œ œ œ . œ œ œ œ œ œj j r .. ‰ œr
œ œ œ
œ œ œ
? 2  ‰ œ œœ œœ œœ œ œœœ œœœ œœœ œœœ œœœ œœœ œ œœ œœ .. ‰ 
Piano
4 œ
j
& œ. œ œ œ #œ œ # œ  œr œ . j  r
6
œœ œœ œœ œ œœ œœ œ œœ œœ œœ œœ œœ
? œ œ œ # œœœ œœ œœ # œœœ œœ œœ œ œ œ
œ œ œ œ
j rD.C.
& œ. œ œ œ #œ œ # œ  œr œ . œ œ œ œ œ œ ‰
10
œœ œœ œœ œ œœ œœ œ œœ œœ œœ œœ
? œ œ œ # œœœ œœ œœ # œœœ œœ œœ œ œ ‰
œ œ œ œ
048
b 2 œ. œ œ œ œ œ œ œ œ. œ
Allegro
b œ œ œ # œ œJ ‰ # œ œ œ œ œj ‰ # œ œ œ œ œj ‰ œ œ
r r r r
& 4
f p
Piano
œœœ œœœ œœœ œœ œœ œœ œœ # œœ
? b b 42  œ œ œ œ J ‰ Œ œ œ œ œ‰ Œ  
œ J
b œ œœœœ #œ œ œ œ œ œ
œ œ œ œ œ œ œ œ #œ œ
&b œ œ œ œ œ œ œœœœ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ
r r
7
œœœ œœœ œœœ œœœ œœœ œœœ œœ # œ œœ œœ # œ œœ œ œœ œ œœ œ œœ # œœœ

3
3
? bb œ œ œ œ œ œ œ œ œ œ œ
œ
b b œ œœœœ œ œœ œ œ œœœœ œ œœ œ œ œ œ œœœœ œ œœ œ œ œ

14
& œ œ
#œ œ œ œ #œ œ #œ œ
f
? b œœ œœ œ œœœ œ # œœœ œœ œ œœ œ œœœ œ # œœœ œœ œœ œœ œ œœœ œ # œœœ
b œ œ œœ œ œ œ
20
b b œ œœœ œ œ œ œ œ œ. œ œ œ œ œ œ œ œ. œ œ œ œ #œ œ œ œ
r r
& œ œ #œ œ œ œ
#œ œ
p
? b b œœ œœ œœœ œ œœœ œ # œœœ œœ Œ œœ œ œœ
œ œ œ
œœ œœ œœ œœ # œœ
œ œ œ œ Œ
œ œ
œ œœœ œ œ œ
&b
b œ œ œ œ œ œ œ œ ## œœ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ
27
œ œ œ œ
3
f
? b œœ œœ œœœ œœœ œœ # œœœ œœ œ œ œœ œ # œœœ œœ œ œœ
b œ œ œ œ œ œ œ œ œ
œ œ œ œ œ œ
# œœ œœ # œœ œœœ œœœ œœ œœ œœ œœ
b œ œ œ œ #œ œ œ œ œ #œ œ œ œ œ
&b Œ œœ œœ œœ Œ
33
f
œ œ
œ œœ œœœ œœœ # œœ
? bb œ œœ œ # œ œ œœ œœ œœœ Œ œœ œœ œœ Œ
œœ œœ œœ
œ œ
049
œ œ œœ
Andante
b 2 œ # œ œ œ œ. œ œ. œ
b
& 4 œ ‰ .. œJ
f
œ œ œ œ œ œ œœ
? b 2 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ .. ‰
Piano
b 4
b j
&b œ œœœ œ œ œ ‰ œ œ œœœ œ œ œ ‰ .. .. œ
6
J
œ œ œ œ œœœ œœœ œ œ œ œ œ œ œœœ
? b œ œ œ œ #œ œ #œ œ œ œ . .
b J . .
b œ #œ œ œ œ œ œ œ œ ‰ .. ‰
&b œ œ œ œ œ œ #œ œ
11
œ œ œœœœœ
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ
b nœ œ J .. ‰
050
Score
Moderato
4
& 4 œ #œ œ œ œ œ œ œ #œ œ nœ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ #œ œ œ œ
3
3 3 3
#œ
œœ œœ œœ œœ
3
œœ œ œ
Piano
?4  œ #œ œ œ œ Œ œ
4 œ
%
œ œ œ œ œ #œ œ nœ œ
œ œ œ œ œ œ œ #œ œ œ #œ œ œ œ œ œ œ #œ œ œ œ
4 3
& œ œ œ œ #œ œ
3 3 3
œœ
œœ œ
œ
3
œœ œœ œœ œ # œ œœ œ œœ œœ
rit.
? Œ œ # œœœ
œ œ œ
Œ Œ œ œ œ œ œ
3
œœ œ œœ
‰Œ œ œœ ‰ j œ œ . œj œ œ œ œ œ œ œ Œ ‰ œj
8
& œœœœœ j œ
ƒ p
œœ œœ œœ œœ œ œ œ œœ œœ œœ
3
? œœ œœ œœ
œ œ Œ œ œ œ œ
œ œ œ œ
œ3
Œ Œ œ œœœ œ Œ Œ œ
œ œ. œ œ œ œ œ œ œ œ Œ j œ œ . œ œjUœ œ œ œ œ œ œ œ œ j
‰œ Œ ‰œ
12
& J J
3
œœ œœ œœ œ # œœœ œ œœœ
cresc. dim.
? œ œ œ œ
# œ œ œ œ œ œ œ
œ œ œ œ
jU Œ Œ œ œœ œ
. œ œ
œ œ œ œ œœœ œ ‰ j œ œ . œj œ œ œ œ œ œ œ Œ ‰ j
16
& J Œ œ œ
œ œ œ œœ œœ œœ œœ œœ œœ
3
cresc.
? œ # œœ œœ œœ œ œ œ œ œ œ œ œ # œ # œ œ œœ
œ œ œ œ œ #œ #œ œ
U
#œ œ œ œœœœ œ œ œ
œ
2 050
œ œ. œ œ œ œœœœ œ Œ ‰ #œ œ œ .
20
&
3
J J J œ
œ œ œ œ œ œœ œœ œœ
3
œ œ œ œ # œœœ œ œœœ
rit.
? œ #œ œ œ œ œ œ œ œ œ œ
Œ Œ œÓ œ œ œ œ œ œ
3
U œ œ # œ ###
& #œ œ œ œ œ œ œ œ œ œ œ Œ œ
24
U f f
œœ œ œœ œœ œ
3
#œ ###
accel.
? œ œ œ œ œ Œ Ó
œ œ œ œ
œœœ œœ œœ
### œ œ œ œ œ œ œ œœœ œœœœœ œœœ œ
œœ œ œœ œ œœœœ
27
&
3
œœ œ œ œ œ œ œœ œœ œœ
accel.
œ
œ œ œœ
3
? ### œ œ œ œ œ œ œœ
a tempo accel. a tempo
### œ œ œœ œ œ œ U œ œ nœ U
œ U
#œ œ œ œ œ œ œ œœ œœœ œ œ œ œ œ œ
31
& œ œ
œ
3
œœ œœ œœ œ œ
3
? ### # œœ œ œ œ
rit.
œ œ œ œ
n œ œ
œ œ œ
### œœ
œ œ Œ Œ œ œ Œ œ Œ
34 1, 2. 3.
& œ #œ œ
3
œœ œœ œœ
? ### œ œ Œ œ Œ
œ œ œ
Ao sinal duas vezes œ
051
bb
&bb c œ œ œ œ œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œœ œœ œ œœ œœ œ
Piano 
? bb c œ œ œ œ œ œ œ œ œ œ
bb œ œ
œ œ œ œ
bb
&bb œ œœœ ˙ œ œ œœ œ
3
œ œ œ œ œ œ J ˙ œ
œ œœ œ œ œœ œ œœ
œœ œœ œ
? bb œ œ œ œ ‰ œ ‰ ‰ œ ‰ j‰ œ œ œ
bb œ œ œ œ œ
œ œ œ œ
œ
bb U j
&bb œœ
. œ œ. œœ œœœœ œœœ ‰ œ œœ ˙ œ œ œœ œ ˙
6
œ J > œ œ œ œ œ œ œ J œœ œ
U̇
˙ œ
? bb œœ œœ œœ œœ œœ j œ œ
bb j‰ œ ‰ œ ‰ œ ‰ œ ‰ ‰ œ ‰ œ ‰ œ ‰ œ ‰ œ
œ œ œ
œ œ œ œ
œ
bb Uœ j j j œ . j œj œ œ œ > œ j j j
&bb œœ
. œ . œ œ œ œ œ œ œ œœ ‰ œ œœ ‰ œ œœ ‰ œ œ œœ œ œ œ œ œ ‰ œœ ‰ œœ ‰
œ
10
œ J > œ œ œ œ œ œ œ Œ œJ ‰ Œ œJ ‰ œ œ œ
U̇
˙ œœ gœœ gœœ
gg œœ gg œœ gg œœ gg œœ gg œœ g
? bb
bb j‰ œ œ ‰ gg œ ‰ gg œ ‰ gg œ ‰ œ ‰gg œ ‰ œ ‰gg œ ‰ œ ‰ ggœœ ‰ gggœœ ‰ gggœœ ‰ g
œ œ œ
œ œ œ
j j j j
bb œ . œ œ œ œ œ œ œ œ œ œ œœ ‰ œ œœ ‰ œ œœ ‰ œ . œ œ œ œ œ >œ œ œ œ œ g œœ ‰ g œœ ‰ g œœ ‰
&bb
14
J J > œœ œ œ œ œ œ JJ gg œœ gg œœ gg œœ
g g g
œ œ gg œœ gg œœ gg œœ œœ œœ g œœ gœœ gœœ
? bb œ œ g œ gœ gœ
bb œ ‰ œ ‰ j ‰ œJ ‰ œ ‰ gg œ ‰ gg œ ‰ gg œ ‰ œ ‰ œ ‰ œ ‰ œ ‰ œ ‰ ggœ ‰ ggœ ‰ ggœ ‰
œ œ œ œ
œ œ
b œ œ œ œ
& b bb œ œ œ.
œ œ œ œ œ œœ œœœ œœœ Œ ‰ œ œ œ . œ œ œ œœ œœ œ
18
œ. œœ œœ œ œ
g p
œœ œœ œœ
3
rit.
? bb œ œ
bb œ ‰ œ ‰ j ‰ œJ ‰ œ ‰ œ ‰ Œ Ó œ
œ œ œ œ
œ
2 051
bb œ œ œ œ œ œ œ œ
& b b œ . œ œ œ œ œœ œœ œ . œ œ œ œ œœ œœ œ . œ œ œ œœ œœ œ . œ œ œ œœ œœ
22
œ œ œœ œœ
? bb b œ œ œ œ
b œ
œ œ œ
U
bb œ œ œ œ
& b b œ . œ œ œ œ œœ œœ œ . œ œ œ œ œœ œœ œ . œ œ œ œœ œœ .. j ˙ œœ œœ
26
œ. œ J
œ œ œ
? bb œœ œœ
bb œ œ œ ‰ œ‰ ‰œ‰œ‰
œ
œ œ œ
œ œ
bb U j j j
& b b œ œœ˙ œœ
. œ œ . œ œ œ œ œ œ œ œœ ‰ œ œœ ‰ œ œœ ‰ ˙ œœ œœ
30
œ œ J > œ œ œ œ œ œ œ J
œ U̇
˙
? b b j‰ œ œ œ gg œœ gg œœ gg œœ œœ œœ
bb œ j‰ œ ‰ gg œ ‰ gg œ ‰ gg œ ‰ ‰ œ‰œ‰ œ‰
œ œ œ
œ œ œ œ
œ
U j j œj œ œ œ
bbbb œ. œ œ. œœ œœœœ œœ ‰ œœ
œ œ œ. œ œ g œœ
34
& œ œœ˙ œ J > œ œ œ œ œ œ œ Œ gg œ ‰ Œ ggg œJ ‰

œ
œ
gJ
œ U̇
˙ gg œœ gœ œ gœ gœ
? b b b œj ‰ œ œ œ
b j‰ œ œ ‰ gg œ ‰ ggg œœ ‰ gggg œœ ‰ œ ‰ ggg œœ ‰ œ ‰ ggg œœ ‰
œ œ œ
œ œ œ
j
b
& b b b œœœ .g œ˙œœ . gœœœ œœœ œ. œ œ œ œ. œ œœœœ œœ ‰ œœ
œ œ
œ. œ œJ œ œ œ
38
Œ gg ‰gg ‰gg ‰
J J gJ > œ œ œ œ œ œ œ J
œ œ gg œœ g œœ œœ œœ œœ
? bb ‰ ggœœœ ‰ ggœœœ ‰ ggœœœ ‰ ‰ œœ ‰ j ‰ œœ ‰ ‰ gg œ ‰ ggg œ ‰ gggg œ ‰ œ ‰ œ ‰ œ ‰ œ ‰
bb œ g g g œ œ
œ œ œ J œ œ œ
b >œœœœœ œ œ œ
& b bb gg œœœ ‰ gg œœœ ‰ œ . œ œ œ.
œ
œ œœœœ œœ œœ Œ Ó
42
gg gg œœ œœ œ œ
œ
 g
gœœ gœœ gœœ œœ œ œœ Œ Ó
smorzando
œœ ‰ ‰ œœœ
3
? bb
bb œ ‰ gggœœ ‰ gggœœ ‰ gggœœ ‰ œ ‰ j ‰ œ‰ œ‰
œ œ œ J œ œ.
œ

052
U
Muito Vagamente
 œœœœœ
rit.
& b 42    
R
œœ
j j 
 œ ‰ œ  œ ‰ œj  œ ‰ œ  œ‰ œ j
U
Piano
? b 42 œ œ œ œ œ œœ œ œ œ œ œ 
œ œ b œ œ œ. œ œ œ
œ bœ œ œ œ œœ œ œ œ œ œ
.
%
& b œ . œ œ œ œ œJ œ œ œ . œ œ œ œ œœ .. œœ œ œ . œ œ
œ œ œœ J
œ
6
J J
? œ œ œœœ œ œœ œ œœ œ œ œ œ œ œœ
b œ œ œœ
œ œ œœ œ œ œ œ œœ œ œ œœ œœ
œ œ
. œ
& b œ. œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ . nœ œ œ œ
12
œ œ
? b œ œ œ œœœ œœ
œ œœ
œ œ #œ œœ œ œ
œ œ œ œœ œœ
nœ #œ
œœ
œ œ œ œ #œ œ
œ œ œ
U
œ œ œ œ œœœœ œ œ U œ. œ
&b œ œœœ œ œ œœœ
17
œœ œ

œ œ œ œ œ
? Uœ œ œœ # œœ œ
œ œ bœ œ œ œ
rit.
b ˙ œ nœ œ œ œ œ ˙˙
œ
U
œœ
rit.
&b     œ œ
22
œ R
œœ
j j j
?  ‰
œ œœ  œ œ œ  œœ‰
‰ œ œœ U
b œœ œ œ œ œ œ œ 
œ œ œ b œ œ œœ œ œ œ
œ. œ œ. Al Segno
053
j j j
& c œ œ œ œ œ œj œ œ œ j ‰ ‰ j .. j œ œ
œ œœ
˙ œ œ
F F œ œ œ
? c ‰ œ œœ œœ œœ œœ œ . œœ œœ œœ b œœ
Piano
œ œ œ œœœ œœœ . œ œ œ #œ
œ
j
œ œœ œ œ œj ‰ œj œ œ œ œ œ œ œ œ
4
&œ œ œœ ˙ œ J J
œ ˙ œ
? œ œœ œœ œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
œ œ œ œ œœ œ œ œ œ
œ
œ œœœ œ œ œ
j j
& œœœœ ˙ œ œ œ œ œ œ œj œ œ ˙ ‰
8 3
œ j j œ œ
œ œœ
œ œ œ œ œ
? œœ œ œœ œœ œœ œ œœ œœ œœ œœ œ œœ œœ œ œ œœ œœ œ œœ # œ b œœ
œ œ œ œ œ œ œ œœœ œ
j j
œ œœ œœœ ‰ œj œ œ œ œ œ œ œ œ
12
&œ œ œœ ˙ œ J J
œ ˙ œ
? œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
œ œœœ œ œ œ œ œ
œœ œœ
œ j j ‰ . œœ œœ Œ
& œ œœ ˙ .
16 3 1. 2.
œ œ œœ
œœœ œ œœ ˙ œ j
œ œ
? œœ œ œœ œœ œœ œ œœ œœ œœ œœ œ œœ œœ œ .. œ œ œ Œ
œ
054
Moderato
#
& c Œ œœœœœ
Œ
œœœœœ
Œ
œ œœ œœ
œ Œ
œœœœ
f œ œ
? # c œ œ ggg œœœ ggg œœœ œ gg œœœ gg œœœ œœ gg œœœ gg œœœ œ œ œœ
Piano
g g œ gg g g œ œ g
g g g œ gg œ gg œ
œ œ œ œ
# œ
Œ œ gœ g œ Œ œ œ œ Œ œ œ œ Œ œ
5
& gg œœœ ggg œœœ ggg œœœ gg œœ gg œœ gg œœ gg œœ ggg œœ gg œœ gg œœ ggg œœ gg œœ gg œœ gg œœœ

gg g g gg g g g g g g gg gg p
?# œ œœœœœœ œœœœœœ œœœœœœœ œœœœœœœ œ œ Œ
œ œ œ œ œ œ
œ
# . œ œ. œ œ œ œ
œ J œ. œ œ œ ˙. œ ˙.
œ œ. œ . œ
10
& J J œ œ œ ˙.
?# gg œœœ gg œœœ gg œœœ gg œœœ gg œœœ œ

gg œœ gg œœœ œ
gg œœ gg œœœ gg œœœ gg œœœ gg œœœ gg œœœ gg œœœ gg œœœ gg œœœ œ
gg œœ
œ gg gg gg œ gg gg gg œ gg
œ
gg œ gg gg gg œ gg gg gg œ gg gg gg
# œ. œ œ œ œ
œ ˙. ‰ œ œ. œ œ. œ œ œ ˙ œ œ œ œ œJ œ œj ˙ . Œ
16
& J J J
?# gg œœœ œ
gg œœ gg œœœ ggg œœœ ggg œœœ gg# œœœ gggœœœ ggg œœœ ggg œœœ ggg œœœ ggg œœœ gg œœœ gg œœœ
gg œ gg œ gg œœœœ
gg ggg œœœ gg œœœ ggg œœœ
œ gg œ gg œ gg g g œgg g g œ œ œ g gg g
#
& ‰ œ œJ œJ œ œ # œ œ ˙ ‰ œJ œ œ œJ œ œj œ œ œœ Œ
22
œ œ œ œ œœœœœ
f
? # gg# œœœ ggœœœ gg œœœ gg œœœ ggg œœœ gg œœœ gg œœœ gg œœœ gg œœœ œ œ œ œ
gg gg g g gg œ gg œ gg œ œ œ gggg œœ gggg œœ œ œ gggg œœ gggg œœ
œ gg œ œ œ œ
#
& Œ œ Œ Œ
27
œ œ œ œ œ œ œ œ œ gg œœœ ggg œœœ

gg g
?# œ gg œœ ggg œœœ ggg œœœ œ
œ gg œ gg œœ œ œ œ œ Œ
œ œ œ
œ œ
055
œ œ œœ .. œ œ
b 2 œj œ . œ œ œ œ œ #œ œœ .. œ œ œ # œ œ . J .. œ. œ œ œ œœœœ
& 4
3
œ œ #œ œ œ #œ œœ œœ œ œ œ œ
3 3
? b 42 ‰ œ œœ œœ # œ œœœ œœœ œ œœ œœœ œœ œœ ‰ .. œ œœ œœ œ œœ œœ

Piano
œ œ œ œJ œ œ
œ œ œœœ œ # œœœ .. œ œ œ œ œ œ # œ œ œ # œœ œœ œœ œ œ œ œ œœ ... œ œ œ

b œ. J . . J œ œ
7
&
3
œ œ œ œ # œœœ œœœ œœœ œœœ œœœ œœœ

3 3
? œ œœ œœ œœ œ œœ # œ # œœœœ œœœœ # œœœœ œœœœ

3
b œ œ œ œ œœ œ œ œ œ œ
œ œ
œ œ œ œ œ #œ œ œ œ œ œ œ œ Uœ œ œ œ œ œ œ Uœ œ œ œ œ œ œ# œ
b œ œ
14
&
œ œ œœœœ œœœœ œœœ œœœ U œœœ # œœœ œœœœ
? œ œœ œœ œ # Œ  œ #œ œ
b œ œ œ œ œ
œ œ œœ .. œœ
j œ œ œ œœ .. œ œ œ # œ œ . œ . œ œ
&b
œ ‰ œ œ. œ œ œ œ # œœ œ J
20
J
3 3
œ œ œ œ œœ # œœ œœ œœœ œœ œœ œœ œœ
3
#œ œ
? b œœœ # œœœœ œœœ ‰ œ œœ œœ # œ œœœ œœœ œ œ œ œœ œ œ œ # œ œ œ œ œœœ
œJ œ œ œ œ
œ. œ œ œ œ œ. œ œ. œ œ œ œ œ. œ œ . œ œ œ œ œœ ..
# œœ .. j œ
&b J J #œ. œ œ. œ œ œ
27
3 3 3 3
œ œœœ œœ œœœ œœœ œœœ # œœœ œœœ œœœ œœœ # œœœ œœœ œœ œœ
? b œ œœœ œ œ #œ œ œ
œ œ #œ œ œ œ œœ œ œ œ œ œœ œ œ œ
œœ œ œœ .. œ j
b # œœ œ œ œ #œ œ. J .. œ >œ œœ ‰
34 1. 2.
& œ
>
3
? # œ # œœœœ œœœœ œ œ # œœœ

œ œœ œ . œœ # œœ œœ ‰
b œ œ œ œJ ‰ . œ J
056
œœ œœ
# # # 4 œ œ œœ œ œ œœ œœ œœ œœ œœ œ ˙ œ œœ œœ œœ œ œ œ œ œœ œ œ ˙
œ œ œ œ œœœ œ ˙
œU
œ ‰ ‰ j nnn
& 4 J œ
‰œœ J œ
œœœ œœœ œ œ œ œ œ œ œœ
Fine
# œ œ œ œ œ œœ œ œœ U nnn
Piano
? ## 4 ‰ Œ œ œ j‰ Œ
4 œ œ œ œ œ œ œ
œ œ
œ œ j œœ œ œ œ œ œ œ œ œ œ . œJ œJ œ œJ
& œ. J œ. #œ œ œ œ œ œ. œ
œ œ.
6
J œ.
Œ œ
˙ œ œœ œ œœ œœ œœ œœ œœ œ œ œ œœ œœ
? # ˙ œ œ œ œ œ œ œ #œ œ œ
œ ˙ œœ œ. œœœ œ ˙ œ ‰ j œ. œ œ. œ
˙ œ. œ
11
& JJ J œ J # œ J
œœ œœ œœ œœ œ œ œ œ˙ œœœ œœ œ œ œ
? œ œœ œœ œœ # Œw #œ œ
œ œ œ
#˙ œ œ œ œ
f
j j œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ
œ œ. œ œ œ ˙. œ‰œœœœ
16
& œ.
œ œ œ œ œ œœ œœœ œœ œ œ œ œœ œœ
? œ œœ œ œ œ œ œ œ œ œ œ œœ œœ œœ #œ œ œ
j # œ. œ. œ œ œ
& œJ ˙ # œ œ œ # # .. œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œw œ œ œ
œœœœ
21
œ œœ w œ w œœ œ œ
œ rit. œ œ œ œ œ œ œœ œœ
? œ # œ œ œ # # # .. œ œ œ œ œ œœœ œœœ œœœ œ œœ œœ œœ œ
Vagaroso
œ œ œ œ
U œ œ œ
## œ œ j œ œ œ
& # w œ œ œ œ œ œ œœœœ œ œ œœ œœ œœ .. œ œ œ œ œ œ œ œ œ
26 1. 3 2. 3
œ œ
U J rit. J
œœ œœ œœ œœ œ
? ### œ œ œ œ œ J ‰ Œ .. œ œ j
D.C. al Fine
œ œ ‰ Œ
057
# # # c œ. œ œ %œ. œœ- œœ- œ. œ # œ n œ. œœ- œœ- œ. œ œ œ. œœ- .

œœ- œ œ œ œJ œœœ- œœœ- œ. œ œ
Allegretto
& Jœ œ Jœ œ J œ œ
F
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? ### c ‰ Œ
Piano
œ
.
# # # œ. œœ- œœœ- œ. œ # œ n œ. œœ- œœ- œ. œ œ œJ. œœ- œJ œJ. œœ- œJ ˙˙˙ œ U n n Meno j
œœ Œ n œ . œ œ œ œ œ
Jœ
5
& Jœ œ œ œ
p
œœ œœ œœ œœ œœ œœ œœ œœ œœ .œ . U n . œ. œ. œ. gg œœœ
Fine
? ### œ œ Œ n n œ. œ J ‰gg J ‰
œ
œ œ œ
.
œ œ œœœœ ˙ œ Œ œ. œ œ œ œ œ œ œ œ ˙
& œœ˙ Œ Œ
10
. .
œ
. œ
œ œ . . œ. g œœ
œ . . œ. œœœ
œ . . . œ
œ . . œ. œ. œ. œ.
œ
? . œ. œ g œ . œ œ
J ‰gg J ‰ œ. œ J ‰ ggg J ‰ . œ. œ œ
J ‰ J ‰ œ œ. # œ J ‰ œœ ‰ . œ. œ Jœœ
œ œ œ œ J œ œ
.
j j œ. œœœœœ
& œ. œ œ œ œ œ #˙ œ Œ œ. œ œœœœ œœ˙ Œ
15
J
œœ # œœ œ. œ. œ. . . œ. g œœœ . œ. œœ œ œ œœ
œ
œ
? œ œ œ œ œ œ. # œ œ. œ. . œ. œ. œ J ‰ ggg J ‰ . œ.
.
œ œ J ‰gg J ‰ œ œ J ‰ œJ ‰
g œ œ
œ œ œ œ. œ
œ
j
˙ œ Œ œ. œ œœœœ œ Œ œ. U ###
˙ ˙ œ ˙ œœœœœ
20
& J
. .
œ . œœ f
œ œ œ œ œ œœ œ . . . œ œ
œ g œ œ œ J ‰ œJ ‰ œ œ œ œ œ œ œ œœ œœ œ # œ œ œ œ J ‰ œœ ‰
dim.
? . œ. J ‰ gg J ‰ œ ###
œ œ œ #œ J
œ
### j
2 poco più
057
œ
œ œ ‰ ‰ œj œ . Œ œ. œœœ œœ
24
& #œ œ œ. œ #œ œ ˙ J
.
œ .
œ .
œ .
œ .
œ .
œ .
œ .
œ
# œ . œ
? ## œ œ œ œ œ œ œ œ . œ . œ . œ œ
œ œ œ œ œ œ œ œ.
. . œ . œ
28
### Ó œœœ œœ œ ˙
œ. Œ œ . # œj œ n œ œ œ
& ˙ J
# œ œ. œ. œ œ. œ. œ œ. œ. œ œ. œ. œ œ. œ. œ œ. œ. œ œ. œ. œ œ. œ.
? ## œ œ œ œ œ œ œ œ œ
##
& # ˙ ‰ œj œ .
Œ Œ
32
#œ œ œ. œ #œ œ ˙
# œ œ. . œ œ. . œ œ. . œ œ. . œ œ. . œ œ. .
? ##
35
### œ . œ #œ nœ œ œ Œ œ. œ œ œ œ œ
& J ˙ œ J
œ. . œ. . . . œ. œ.
? ### œ # œ œ œ œ œ œ œ œ œ œ. œ œ œ œ œ. œ œ œ œ. œ n œ œ œ.
#œ œ œ
œ
j
### œ
‰ œ œ. œ œj œ j
œ ˙
j .
œ œ œ œ
38
& ˙ J ˙ ˙
f
œ œ. œ œ. œ. œ. œ. œ. .
œ
? ### œ œ œ. œ. œ œ j œ œ J œ. j ‰ Œ
D.S. al Fine
œ œ œ
œ œ œ allarg. œ œ J œ
058
> j %
œ œ œj œj œ j j œ œ nb b b b œ œ œ n œ b œ œ œ n œ b œ œ n œ œ œ œ b œ n œ
Fox Medium
b œ
& b C b ˙˙˙ n˙ ˙ œœ ‰ Œ
˙ œ b ˙˙ œ
F Jp œœ œœ n œœœ
œ œœ œ œœ œ œ œœœ
Piano
? b C #˙ œ œ œj ‰ Œ nb b bb œ œ
œ œ
bbb b œ œ œ n œ b œ œ œ n œ b œ œ œ œ ˙ œ œ œJ œJ œ J œ œ ˙ . n œ œ œ œJ œ œJ
b J J J
5
&
n œœœ œœœ œ œ œ n œœœ œœœ nœ œ
3
? bb b n œ œ œ œ œ œ œ œ œ œ œœ œ œœ
b œ œ œ
œ
bb b b œ œ œ œ œ nœ bœ œ nœ bœ œ nœ bœ nœ œ œ bœ nœ bœ œ nœ bœ œ nœ
J œ œ œ œ œ œ œ
10
&
œœ œœ n œœœ œœœ n œœ œœœ
3
œ
? bb b œ
b
œ J ‰ Œ œ œ œ œ nœ œ œ
bb b b b œ œ œ œ œ œ œ œ b œ œ œ n œ b œ œ œ n œ b œ œ œ
œ œ œ œ ~~ œ œj œ n œj œ Œ
œœ œ n œ n ˙˙ nnnnb
14
& œ œ œ œ ˙
œœ œœ œœœ œœœ œ œ nœ
? bb b œ œ œ n œ œ
b œ œ n œ œœ Œ ˙ nnnnb
œ
j j j
œ œ n œj ˙˙ j œ œ œ œ œ œ œ œ œ œ œ œ œœœ œ œ b œ
& b œœ b ˙˙ œ n œ n ˙˙˙
œ œ œ œ œ
œœ ‰
19
˙˙ œœ Œ ˙ ˙ ‰ Œ
˙ J
Œ œ Œ œ Œ œ
? b œ œœ # ˙ œ œœ ˙ œ œœ #˙ œ j
œœ œ œ œ œ ‰ Œ
˙ œ b ˙ œ ˙ œœ Ó
œ
Œ
>
œœ j j j
b œ œ ‰ œ œ bœ ˙ œœ œ œ˙ œ n œj ˙˙ œ œ œ œ ˙. œ
˙˙ œœ Œ ˙ n œ ˙˙ ..
25
& œ b˙ n˙ ˙˙
˙ ˙
œ œœ
? œ œ œ Œ œ œ ˙ œ œœ # ˙ œ œœ
˙
œ œœ # ˙ œ
œœ œ œ b œ
b
> >
%
Œ >˙ j j j fi fi
&b œ
œ œ œ œ œ œ œ b ˙˙ œœ œœœ j ˙ jœ bœ œ Œ œ œ Œ œœœœ Œ Ó
31
bœ Ó n˙ ˙ œ b ˙˙ œ
œœ ‰ Œ œ œ
œœ
J S S
? b b wwww œ Œ #˙ œ œœ œœ œ œ j‰ Œ œ Œ œ Œ œ Œ Ó
œœ œ œ œ
059
.. œ œ œ >œ œ œ >œ
Moderato
2
&b 4 Œ ‰ œœ œ œœ œ œ œ œ œ œ œ œ œ œ . œ œ œ œ œ
p > >
? 2 œœœœ œœœ .. # >œ œ œ œ >œ œ œ œ >œ œ œ>œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
b 4 œ œ œ #œ œ œ
>œ œ
b
& œœ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ # œ œ ‰ œ œ ‰ œ
6
> J œœœ œ œ J
? b #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ
œ
œ œ œ œ ‰ j >œ j n##
b
& œ #œœ œ œ œ œ œ ‰ œj œ œ œ œ œ œ ‰ œ
12
>
œ œ
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ >œ œ >œ œ # œ œ œ œ œ œ œ œ œ œ œ œ # œ œ
dim.
œ œ œ œ n##
# j j
& # œ œ œ œ œ œ. œ œ œ œ œ œ œ. œ œ œ œ œ œ œ. œ
18
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ œ
## . nn >œ >œ
& œ œœœ œ œ ‰ œœ. b œœ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ
24
F> >
? # # œ œ œ œ œ œ œ œ œ œ œ œ n œ œ .. n n b # >œ œ œ œ > œ œ œ >œ œ œ>œ œ œ œ œ œ œ œ œ œ
œ #œ
œ œ œ œ
>œ
& b œ. œ œ œ œœ œ œœœ œœ œ œ œ œ œ œ œ Œ
29
œœœœ œœ œ # œ œ
>
? b œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ Œ
œ œ œ œ
060
˙˙ .. œœœ œœ ˙˙ œ œ #œ œ œ
& b C ˙. œ ˙ ˙˙ .. ˙˙ .. Œ
f
œ œœ œ œœ # œœ œ
3
? C œ œœ œ œœ œ # œ œœ œœœ œœœ
Piano
œ œ œ # œ
b œ œ œ œ
œ œ
&b Œ ‰ ‰œœœ
œ ˙ ˙
5
œ œœ
œ œ œ œ ˙.
3
˙. #œ œ œ
œ œ # œœ œœ œ œœ œ œœœ œ œ
?
b œ œ #œ œ œ # œ œœ œ œ œ Œ œ œ œ œ
œ œ œ
&b ˙ ‰œœ œ œ œ ˙. ˙ ‰ œœ ˙ ‰œœœ ‰œœœ

10
œ œ ˙
œœ
3
? œ œœ œ œœ œ œ œ œ œ œœ œ œœ
b œ œ œ œ œ œ
œ
œ œ
& b œ œ œ ˙. ˙ ‰ .. ˙ ‰ ˙ ‰œ ‰œœ
15
œ œ œ œœ ˙ œ
œ œœ
œœ œ œ œœ œ œœ
3
?b œ œ œ œ œ .. œ œœ œœ œ œ
œ
œ œ œ œ œ
U
U œ j U 1.
& b ˙Œ . œ œœ œ œ œj œ œ œ œ Œ ‰ ..
20
#œ œ œ w œ œ œ
gg œœ ggg œ œ
Ug
? g ˙˙ ... œ œ œ œ œ œ
b ggg ˙˙ . Œ œ œ œ œ œ œ œ œœ œ œ œ œ œ œ œ Œ ..
g œ œ œ œ œ
A060
#
2
&b w Œ Œ # œ n # .. œœ ˙˙ œœ œœ n œœ ˙˙ œœ œœ œ œ œ ˙ .
25 2.
œ ˙˙ ..
œ œ œœ œ œ
3
? œ # œœ œ œ œ # œœ œ # # .. œ œœ œ n œ # œœ œ œœ œ n œœ œ œœ
b œ œ n œ
œ #œ œ nœ œ
##
˙˙ .. œ ˙ œœ œœ œœœ œœœ œœœ œ œ œ œ œ œ ˙ . Œ œ ..
30 1.
&
3
˙. œ œ ˙ ˙
? ## œœ œœ œ œœ œœ œœ œ œœ œœ œ œœ œœ œ œœ .
œ œ œ œ œ œ .
œ œ
## Œ .. n ˙ .
35 2.
& œ œ œ ˙. œ œ n˙ œœ œœœ .
3 3
˙. ˙
? ## œ œœ œ œœ œ œœ œœ . œ n œœ œœ n œ œœ œ œœ œ œœ œ
œ œ . œ œ
œ œ œ nœ œ
## Œ ..
œ ˙ œ œ #œ œ ˙. ˙.
40 . 1.
&
˙. ˙. œ
œ œœ œ œœ
? ## œ œ œ œœ œ œœ œ œ # œ
œ œ œ œ
#œ œ œ ..
## fi Coro ao A e 3ª vez
U
œ ˙ œ ˙
45 2. 3.
& nœ #œ nœ
bœ œ œ œ ˙. bœ œ œ œ ˙.
3 3
˙.
œœ
? ## œœ
œ
œœ œ n œœ œœ
œ # œœ œ œœ
œ
œœ œ n œœ œœ
œ œ œ œ œ
œ œ
## fi
Œ ‰ Œ œ Œ
n œœœ
50
& œ œ nœ œ
œ
? ## œ n œœ œ Œ œ Œ œ Œ
œ œ œ œ
Do % ao fi
061
U j
j
&C ‰ œ œ œ b œ œ œ œ œ œ œ œ b œœ œœ œœ œœ œ œ œ œœ œœ œœ œœ œœ ˙˙ œ œ
œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœ œ
U œœ
œœ b n œœœ œœœ
a tempo
œœœ œœœ
Piano rit.
?C  #œ œ œ ‰ Œ œ œ œ œ j‰ Œ
œ J œ
#œ œ œ œ œ œ
œ
% ‰ œ œ ˙ U ‰ ˙
œ
œœœ œœœœœœœœ ˙
œ œ œ œœœœœœœœ
œœœ
5
& œœœœ œ œ
œœ œ ˙ œ
œ U œ œœ œœ
? œ œœ œ b œ œ œœœ œœœ œ œœœ œ
# œœ œ œœ œ œ œ œ bœ
œ bœ œ œ œ œ œ œ bœ
œ œ œ œ œ œ
‰ œ œ ˙ U ‰ œ œ
œœœœœœœœ ˙
˙
œ œ
œœœœ œœœœ œœœ œœœœœœœœ
10
& ˙ œœœœœ œœœ ˙ œ

U œœ œœ œœ
? œ œœœ œœœ œ
œ œœ œœœ œ œ œJ ‰ Œ
œœ
œ œ œ bœ œ œ œ œ
œœ œ
œ œœ œœœ
œ œ œ œ œ bœ œ œ œ
œ œ œ œ œ œ
‰ œ œ ˙
œœœœ œœœ œ  
œ œ n œ œj n œ œ œ jœ œ œ jœ œ
œ
& ˙ œ œ #œ œ j œ
16
#œ œ œ œ ˙ œœœ
#œ œœœ n œœ œœ œ U œ
? # œ œœ #œ œ œ œ œ œœ #b œœœœ ‰ Œ # œœœ ‰ Œ n œœœ ‰ Œ œ
j‰ Œ
œ #œ œ J J J œ
#œ œ œ œ œ
œ œ œ ˙
& .. œœ œœ œœ œœ œœ œœ œœ œœ ˙˙ œœ œ œ œœ œœ œœ œœ œœ œœ œœ œœ ˙˙
œ œ œ ˙ bœ
œ œ œ œ # œœ œœ œœ œ œœ œœ œœ œœ
21
œ œ
œ œ
? .. œ œœ # œ œœœ œ
œœ œœ
œ œœ œœœ œ
œœ œœ
#œ
œœœ œœœ
œ œ œ œ
U j œ
j . œœœœ ˙ œ œ œ œœœ œœ Œ
.
& œœ œœ œ œœ b œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ ˙˙ œœ œ œ œ . œœ œœ œœ œœ œ œ œ œ ˙
26 1.
œœ œ œ
U
% FIM
œœ œœ
? œ œœ b n œœœœ ‰ Œ œœ œœ
a tempo
œœ œœ . œ
rit.
œ œ œ œ œ . œ œ œ œ œ j‰ Œ œ œ œ Œ
J œ œ
œ œ œ
062
Lento
& b c ‰ œœœ œœ œœœ œœœ .. œœ œœ œœ œ œœ œœ œ œ œ œ œ œœ œœ œ˙ # œ œ œ œ ˙˙ œœ œœ œ œœ œœ

œ œ œ œœ œœ œ œ œ œ œ œ œ œ œ ˙ œ œ œœœ œ
œ
 ß#œ
œœ œ œ # œ œœ œ œœœœ
? c œ œœœ œ œ œœ œ œ œ
Piano
b .. œ œ œ œ œ #œ œ œ œœ œœ Œ
œ œ œ œ
œ œ # œ œ œ
Lento
œ œœ j
& b œœ œœœœ œœœœ œœœ œœ œ œ˙ œ œœ œœ œ œœ œ œ œ œœœœ ‰ œ œ. œ œ œ œ
5
œ œœœ ˙˙ œœ JJ J
J p
œ œ œ œ # œœœœ œœœœ ˙ j œ œ œœ œ œ œœœ
? œ œœœ œ œœ œ œœ œ œœ #œ œ œœ œ œ œ Œ œ
b œ œ œ
œ œ #œ œ œ
œ
œœ œ œœ œœ ˙ œ ‰œ œ œ œœ œœ œ ˙
& b œ œ #˙. Œ
10
œ œœ # œœœ œ œœœ œ œœ œ œœ œœœ œœœ œœ œœ

?
b œ # œ œœ œœ
œ #œ œ œ œ œ œ œ œ œ
œ œ œ
#œ œ œ œ œ œ
œ œ œ œ œ
œ œ ˙. œ œ œ. œ œ #œ œ œ ˙ œœ
&b œ œ œ #œ J J
15
J
p
œœ # œœœ ‰œ œ ‰ œœ # œœœ œœ œœ
? œ œ
b œ œ œ œ œ œ œ ‰ œ ‰ œ ‰ œŒ œ œ
œ
œ œ
œ œ œ
œ œ
œ œ œ œ
œ œ œœ œ œ ˙ œ j j j #Uœ œ
b œ œ œ
19
& J J œœ œ œ œ ˙ œ œ œ œ J
œ œ œœ œœ œœ œ # Uœœ
? œ œœ œœ œ œœ œœ œ œœ
œ #œ œ œ œ œ œ
œœ
œ œ œ œœ œ
œœ œ œ
b œ œ œ œ #œ #œ
œ œ œ #œ œ œ œ œ #œ
^^ œ ^ ^^
‰ ^
œ œ œ œ œ œ œ œ œ œ œœ œ œœ œ œœ œ œ ^ œœ œ # œ œ œ œ œ˙ œ œ œ n # #
&b œ œ œ œœœ œ œ œ
24
˙
Œ œ œ œ Œ œ œ œ
 œ œ œ œœ œœ œœ # œœœ œœœ œœœ # œœœ œœœ œœœ
dolce
? œ œœ œœ œœ œ œ œ œ #œ œ œ œ œ œ œ #
b œ n#
œ œ #œ œ
## œ œ . œ œ . œ œ . œ w
2 062
œœ œ . œ œ . œ œ . œ œ œ .
œ œ œ œ. œ œ. œ w
28
&
Ó Œ
œ œ œ Œ œœœœœ œœ œœ œœ œœ œœ œœ Œ œœœ œ œœœœœœœ œ
? # # œ œœ œœ œœ œœ œœ œœ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ
œ œ œ
œ œ
##
œ œ . œ œ . œ œ. œ œ. œ œ. œ œ. œ œ. œ w œ œ œ œ œ œ œ œ. œ œ. œ œ. œ w
œœ œ œ. œ œ. œ
33
&
Œ Œ Ó Œ ‰. R
œ œ œ œ œ œ
? # # œ œœœ œœœ œœœ œ œœ œœ œœ œ œœœœ œœœœ œœœœ œ
œœœœ œœœœ œœœœ œ œœ œœ œœ
œ œ œ œ œ
U
œ œ œ œ œ 1.œ ˙ Œ
## œ nn
œœœœœœœœ œ ˙ ‰œ œ b œœ œœ .. œ ˙ Œ b
38 2.
& J n œ œ
Œ Œ ‰ œœ œ œ
? # # œ œœ œœ œœ œœ œœ œœ
œ n œœ œœ
œœ . n œœ œ œœ
œ . œ œ ‰ œ ‰ nnb
œ œ
œ œ œ œ
œ œ œ #œ œ.
Piu Lento
œ œœœœœœœœ
& b Œ œ œ œ œ œ œ #˙. ˙. œœ
43
J
^ œ œ ^ ^ # œœ œ œœ ^ ^ # œœ œ œœ ^ ^ œœœ ^ ^ œœœ ^
? œ
b œ œ œ‰œ‰ œ ‰œ‰ œ ‰œ‰ œ œ œ ‰ œ ‰ #œ
œ œ‰ ‰
œ
œ œ # œ
œ œ
jU j 3
œ œœ œœ œ œ œ œ œ ˙ Œ
œ œœ œœ œœœ œœ œœ œ œœ œ œ
&b J œ œœ œœ œœ œœ œœ
48
J Ó ‰ œ œ
œœœ œ œ œ œœ œœ œ œœ œ œœ
œ

? b ^ œœ œ œœ ‰ ^ ‰ œ œœ œœ œœ œœ
œ œ œ ‰ œ œ ‰ œ #œ œ
œ œ œœœ
œ œ œ œ œ œ #œ
&b Œ Œ ˙˙ œœ Œ
gg œœœ gg # œœœœ
52
gg œ ˙˙ œœ
g gg œ
g p
? œœ œ
b œ œ # œœ œ œ Œ
œ œ œ œ œ
œ
063
% œ œ œœ
Moderato {q = 75} r
œ œ œr œ œ œ œ œ œ œr œ œ œ. . œ. . . .
œ
œ  œ
œœ 
œ
œ œ œ œ. œ œ.
& 42 œ œ 
ƒ com animação œœ œœ œœ œœ œœ
Piano
? 42 œ œ
œ œ œ
œ^ U Não depressa, bem ritmado

œ. ggœœœ^ œœ ‰ b b b  œ œ œ nœ œ œ
œ œ œ œ œ œ #œ œ œ œ ‰ œj œ œ œ œ n œ
5
& Jgg
F
œœ ^ œœ œœ œœœœ œœœœ j œœ n œ œœ œœ j œ œœœ œ n œœœ
? œ Û b # œ œœ
œ œ‰ bb œ œ œ œ œ nœ œ œœœœ
œ œ œ œ
b  œ œ œ œ œ œ œ œ ‰ œj œ œ œ œ œ œ œ  œ œ
rit. a tempo
&bb œ Œ œœœœœ œ œ
11
? b b œ œœ œœ œj ‰ Œ œœœ œœœ j œœœ œœœ œ œœœ œœœ œ œœœ # œ œœ œ n œœœ œ œœœ

b œ œ œ œ J ‰
œ œ œ œ œ #œ œ œ
b bb œ  œ œ œ œ œ œ n n n œ .œœ # œ‰.œœ œœ œ œ ‰
œ œ œ œ œ‰ .œœœ # œ .œœœ
18
& œ œ œ œ œ œ nœ ‰œ œ
œ J J J J
œ œ œ f
œ
? b b œ n œœ œ œ œ n œœ Œ Œ  œ n œ œ #œ œ œ n œœ œœœ œ #œ
b œ œ œ œ n n n œJ ‰ œ ‰ œ Rœ œ œ œ œ
J‰ J‰
J
œ. œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ 
& œœ œ œœ œ œ œ nœ œ œ‰ . œœœ # œ‰. œœœ
24
#œ œœ
jF œ œ œ J J
œ œ f
œ
? œœ œ n œœ œ ‰ œ
œ œ #œœ œ œ
œœ n œœœœ œ œ
œœ œ œœ œœ œ œ œ #œ
œ œ œ œ œ Rœ œ œ
œ J‰ J‰
œ. œ œ œ œ œ œ œ œ œ œ
2 063
&œœ œœœ . ‰ . œ œ  œ. œ œ œ œ #œ #œ œ œ œ œ œ
30
œ œ
‰ œœ œ # œ œœ œ œ œ œ œ
œ œ J J j F œ
? œœ œ n œœ œœœ œ œ œœ # œ œ œ n œ œœ ‰ œœ # œœœ # œœ œœ
Rœ J‰ J‰ œ œ œ
œ œœ œ œ œ œ œ œ œ œ œ œœ
œ
& œœ œœ œœ œ œ # œ n œ œ œ  œ œ œ œ œ œ œ œ œ œœ  œ œ
35
œ œ œ œ
f ^ ^ƒ œœ œœ œœ
? œ
œ
œ œ œ œ
œœ œ ‰ œ œ œ
œ œ ‰
œ œ œ œ J
J œ œ œ
œ
œ œ
œ œ œœœ œ œœ œ
œ œ œ œ
& œ œ œ œ œ œ œ œ ‰ j Œœ œ œ‰ œœœœ œ‰ .œ œ ‰j œŒ œ œ‰ œœœ œ‰ œœ œœ  œ ‰ œ
39
œ œœœ œ ƒ
œœ J J
œœ f
? œ œ œœœ œ œœ œœ j œ œ j œ œ œ œj œ . j œj œ œ j œ œ œ œœ œœ ‰
marcato
œ œ œ œ œ J
œœ œœœœ œœ œ œ œ œ œœ œœœœ œœ œœ
& œ œ œ œ œ œ œ œœ  œ œ œ œ œ œ œ œ œ œ ‰ œj œ . œ œ œ œ
45
œœ œœ œœ œœ œœ œœœ
? œ œ œ œœœ ‰œ œ œœœ œ œœ # œœœ
œ J œ œ œ J #œ œ ‰
œ œ
Ÿ ^ ^%
œ œ œ
œœœ œœ ‰ œ œ œ  œ œ œ œ œ œ œ  œ œ œ œ
r r
& #œ œ œ œ œ œ œ œ œ œ œ œ œ #œ nœ œ j
50
œ
Z^ Z^ ƒ
? œ # œœœ œœœ ‰ œœ
œ œ
œ œ œ
œ œœ œ ‰ œ œ œ œ œ œ œ
œ œ J œœ œ
œ œ œœ  œ œ œ œ. . œ. . œ. . . . . œœ^ œœ^ U
r
D.C.
œ
œ œ œ œ œ œ œ ggg œ œ ‰
55
& J g
œœ œœ œœ œœ œœ œœ ^ ^ U
?œ œ œ œ œ œ œ œ ‰
œ œ œ œ
064
# 2 œ œ j œ j j j jn œ œ œ nn
& # 4 œœœ œ œœ
˙ œ œ œ œ œ ‰ œ œ œœ œœ ‰ œœœ ‰ œœœ
œ ‰ œ œœœ b
f j j
n ˙˙ œœ ‰ œœ ‰ # œœ n œjœ ‰œ Œ œ œ j j j
# ˙˙ œ
Piano
? ## 2 ‰ Œ Œ œ œ œ œ œ œœ ‰ œœœ œ ‰ œ ‰ œ ‰ œj ‰ nn
4 j‰Œ b
J œ œ œ œ
œ œ
%
œ œ œ #œ œ #œ œ j
&b œ œ ˙˙ œ œ œ œ œ œ œœ œ # œ œ œ˙ œ œ
9
˙ œ œ
‰ # œJ ‰ œJ ‰̇ œœœ œœœ œœœ œœœ œ
‰̇ œœ ‰ # œœ œœ
J J J
œœ œ
? œœ ‰ Œ œ # œœœ œœœ œœœ œœœ œj ‰ # œj ‰ j ‰ Œ œœ # œœ œ ‰ œ ‰ œj ‰ Œ œœœ ‰ Œ
b J œ J‰ J‰ œ Œ
œ œ œ J
œ #œ œ œ
j‰ jœ œ œj . j‰ œ
&b ˙ œ œœ œ œ ˙ œ œ œ œœœ œ œ œœ œ
œ œ œ œœ . œœ œœ œ
3
˙
18
Œ̇ œ œ Œ ˙ œ œ œ ‰œ œ œ œ #œ ˙ œ J J
J
œœ œ œœ œœ œ œ œ œ œœ œ œœ œœ œœ œœ œœ œœ œœ
?b œ œ œ œ œ œ œ œ œœ œ œ œ œœ œ œœ œ œ # œ œ j œœ ‰ Œ œœœœ œ œ
œ #œ œ‰Œ J œ œ œ
j j j
œœ œ œj # œ œ œ ‰ ‰ œj n # # .. œœ ‰ œj ˙ œ œ œ˙ œ œ œ
&b œœ ˙˙ œœ œ
28
J œ œJ œ œ #œ œ œ ˙
œœ œœ # œœ œœ œ j j
œ œ # œœœ ‰ Œ œœœ œ ‰ œœ ‰ œœ œœ œœ œœ œœ œœ œœ
? œœ œ #
n # .. # œ Œ œ œœ œœ œœ œœ œ œ œ
b œ #œ œ J #œ œ œ ‰œ ‰ œ
œ œ
# ˙ j j
œœœœ œ
& # ˙ œ œ ˙˙ œœ œ œ œ ˙˙ œœ ‰ j
36
œ̇ œœ ‰ œœ œœ ‰ # œœœ ‰ œœ œ
J J
œœ œœ œœ œœ >œ œ J
? ## œ œ œ œ œ
œ œ ˙
˙ n˙ œ œœ œœ œœ œœ œ œœ œ œœ œœ œœ œœ œœ œ œœ œœ
n˙ œ œ œ
# j ˙ œœœœ œ œ œ œ
& # ˙ œ ‰œ œ œ ˙ œœœœ œ œœ
44
˙ ˙ œ ˙
j j j
‰ œœ ‰ œœ
? ## œ œ œ œ ‰ œjœ ‰ œœœ œœœ œœœ œœœ œœœ œœ
œ‰Œ œœ œœ œœ œœ œœ œ
œ œ œ œ œ œœ ‰ Œ
n˙
# ˙˙˙
œœ
œ
œ nœ œ œ œ œ œ œ ‰ Œ œ œ œ J Œ
œ J J
## œ . œjœ ˙˙
j
œœ ‰ ‰ œj .. œ œ ‰ œj ‰ œj ‰ j jfi n œ œ œ
œ œ œ œ œœœ œœœ
55 1. 2.
& ˙ œ ˙ œ œ œ œœ œœ
‰ œœj ‰ # œœj œjœ ‰ Œ œœœ
? ## œ œ œ
œ œ œ œ œ ‰ œ œ œ ‰ Œ .. œ ‰ œœ ‰ œ ‰ œ ‰ j
œ ‰Œ
J œ œ
J œ œ D.S. al Fine
065
%
> >œœ ˙ j
Allegro animato
# # # # 4 œJ >œ >œ œœ ‰ ˙˙ ... œœ ‰ œœ œœ œ n œ œ j ‰ >

œ œœ œœ œ œœ œœ œœ œ # œœ n œ
& 4 Ja tempo
F ƒ >
rit.
? #### 4 ‰ Œ œ. œ. œ. œ. œ ggœœœ ggœœœ gg œœœ j . œ. gg œœœ

Piano
‰ œ . j œ J ‰g g g .
‰ œ #œ œ g
4 œ .
ƒ
. . j . . j. . j
.œ œœ. œœ œœ œ œœ œœ œ œœ œœ. œœ . œ. œ. œ. œœ œ œœ œœ. œœ . œ. œ. œ. œœ œ œ. œœ.
j
# # # # œj œ œ .
.œ œ. œ œ œ œ
œ‰œ œ
4
&
> fF
. .œ
œ gg œœœ œœ œœ gg œœœ œ gg œœœ œ
#
? ## # ‰ œ . œ gg g
g
œ gg œ g
g œ œ ggg œœ œ ggg œœ
J œ gg g œ g œ
œ
f F
œ. . >œ > > >œ j
# # # œ œ.
# œ .œ U œ œ œœœ ‰ ˙˙˙ ... œœ ‰ œœ œœ œ n œ œ j ‰ >
rit.
œ œœ œœ œ œœ œœ œœ œ # œœ n œ
8
& œ Ja tempo
p F ƒ . œ. œ. œ. >
? #### œ œ. œR U ‰ Œ j œ œ œ ggœœœ ggœœœ ggg œœœ j . . œ. gg œœœ
œ
‰ œ . J ‰g g ‰ œ #œ g
œ. œ. . œ .
ƒ
# # # # œj œ . . œ. œ. œ. œ. . œ.
œ >œœ >œ œ œ œj j
œœ œ. œ. œ # œ œ. œ œ Œ œ #œ œ œ œ œ œ œœ œœ œœ
œ‰œ
12
& J J J
>
œœ f > >
. œ gg œœ # œœœ œ
œ g ggg œœœ > gg œœœ
? # # # # ‰ œ. œ. œ gg œ g
g
#œ g g
œ gg
œ œ ggg œ Ó œ g
J œ
> f >
. . . .j
#### ˙ œ U‰ j n n n n # ‰ œ œœ œœ ‰ Œ
Cantabile suavissimo
j j
œ . œj œ # œ œ ˙ œœ œ œ ‰ œJ œ . œj œ # œ œ
16
& ˙ œ œ J J
ƒ p
ggg œœœ œ ggg # œœœ
Fine
œ
? # # # # >œ œ # œ œ œ œ œ UŒ n n n n # gg œœœ œœ
g
g # œ gg œœœ gg œœœ Œ g
. œ g g
g œ g g œ
œ >
ƒ. . . œj
œ œœ œ ‰ Œ
# ‰ œœ œ œ œ . j œj œ j # ˙ œ ‰ œ
j
œœ .. j œj œ j gg # ˙>˙˙ U œ
œœœœ œœ
& ˙ œ œ œ œœ œ œ œœ gg ˙
21
‰ J œ. œ œ œ œ ˙ œ œ
œ œ œ ƒ U
œ
œ œ œ #œ œ gg# œœœgg œœœgg œœœ #g œœœ g œœœ # œœœ œœ
cresc. cresc.
? # ggg œ ggg œ ggg œ # œggg œœ œ ggg œœ # œgg gg

œ œg g g œ œ œ
> > œ œ
œ #œ œ œ
# g œœ ... # œ œ œ œ œ œœ ˙ œ j œ œ œ
#œ œ œ
2 065
œ
& ggg œ . J J J ˙ œ ‰ œœ # œœ .. œ œ œ œ œ œj ˙ ‰ œ œ œ
26
J œ J J J
g a tempo J >
Ï œ œ œ F œdim.œ œ p
? # œ œœ gggg œ ggg œ
œ œ gg œœ gg œœ gg œœ g
g
# œœœ g œœœ g œœœ
g g ggg œœœ gg œœœ gg œœœ g
g
œœ g œœ g œœ
œ gg œ gg œ
g g g g g g œ g g
g g œg
œ œ œ g œ œ
Ï˙ F p >œ œ
U œ œœ œ œœ
# a˙tempo œœ ‰ œ œ # œ œœ œ œœ ˙ ˘ ggg œœœ œJ ‰ n# # # # œ œœ
œ g œ
Piu mosso
œ >œœ g œœ J J
31
& J J J
F
œ œ #œ gg œœœ gg œœœ
? # œ gggg œœ gggg œœ œj œ # œ œœ gg œœœ œ œ Œ n# # œ œ
œ œ œ œ gg œ gg
œ œ œ g ## œ
œ œ œ
fl
# # # # ˙˙ œœ œœ œœ œœ> œœ ‹ œ œ ˙ œ
‰J J ‹ œ œ ˙˙ œœ ‰ œ œœ œœ- ‹ œœ- œœ- œj
35
& œ J -
J J
? #### œ œ œ œ gggg œœ ggg œœ œ œ œ œ gggg œœ ggg œœ œ œ œ œ œ œ ggg œœ ggg œœ œ œ œ œ œ ggg œœ ggg œœ
œ œ œ œg œ œ œ œg œœ œ œœ œ
œ œ œ œ
#œ œ œ œœœ ˙ œ œ
#### ˙˙ œœ ‰ # œ œ œ #œ œ œ
˙ œ ‰ J
39
& J œ #œ œ
J J
? #### œœ œ g
g œœœ gg # œœœ œœ œ g
g œœœ gg œœœ œ œœœœ œœ œ œ
œ œ œ œ g œg œ œ œ œ-g œg œœ œ œœ œœ
œ œ
œ œ œœ œ œœ ˙˙ œœ nœ œ œ œ œ
#### œ œ n œ # œ œ œ œ œ
J J ‰ J
42
& J J
g œœ œœ g œœ # œœœ œ œ œ œ œ gg # œœœ g œœœ

? #### g
œ œ œ œ ggg œ gg œ œ œ œ œ œ ggg œ n œ #œ œ œ œg gg
œ œ œ œ g œ œ œ œ
œ
˙˙ œ nœ
## œœ œ j
& ## ‰ œJ n œ œ œœ œ œ ˙ œ ‰ œj
45
œ nœ ˙ œ
J
œ
œ œ œ œ œ œ gg œœœ gg n œœœ ng œœœ gg œœœ
g g œœ
œ g œœ œœ
œ œ ggg œ
rit.
? #### œ œ gg n œ g g œ g
a tempo
gœ n œ g œ g œ œœ œœ œ œ g
g g
œ
%
>
œ >
œ >œ
## œ >˙ œ U
& ## œœ œ œ
œ œ œ
48
œ ˙ œ
J J R
œ œ œ
œ œ œ œœ gg œœ œœ œ œ œ œ œ U
? #### œ œ œ œ ggg
rit.
œ œ g œ r  ‰ Œ
. œ. œ
066
b œ œ œ œ œ œœœ œ œ œ œ œœœ œ œ œ œ œ œœœ œ œ œ œ œ ˙

& b c œ œœœ œ œ œ œ œ ‰ ‰ œ #œ
J
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
Piano
? b c œ œ œ œ œ œ œ œ œ
b œ œ œ œ œ
b j ˙ œ ‰ ‰
& b .. œ œ œ #œ œ œ bœ nœ bœ ˙ œ ‰ ‰ œ #œ œ œ œ #œ œ œ œ œ #œ
6
J œ #œ
œœ œœ œœ œœœ œœœ œœœ œ œ œ œœ œœ œœ
? b .. œ œ œ œœ œœ œœ œ œ œ
b œ œ œ
œ
b
& b œ œ œ #œ œ œ bœ nœ bœ ˙ j œ #œ ˙ œ ‰ ‰ œ œ œ œ œ œœœœ
œ ‰ ‰ œ #œ œ œ œ #œ œ œ œ
10
J J
œ
? b b œ œœ
œœ œœ
œ œ
œœ œœ œœ œœ œœ œœ œœ œœ œœ n œœœ œœœ œœœ
œ
b œ œ œ œ œ œœœœ
& b œ œ œ œ #œ œ #œ œ œ œ #œ nœ œ œ œ bœ ˙ œ‰‰
15
œ œ œ œ œ #œ œ #œ
œœ œœ
? b œ œ œ
œœ
œ
œœ
œœ
œœ
œœ
œœ
œœ œœ œœ œœ n œœœ œœ œœ
œ œ
œœ œœ œœ
b œ
œ œ œ
œ œ œ œ œ
œ
-
b œ
& b œ œ œ #œ œ œ œ œ nœ ˙ œ‰Œ Œ œ œ œ œ œ #œ œ œœ œœœœœœœœ
20
J œ œ #œ œ #œ
œœ œœ œœ œœ œœ œ œ
œœ œœ œœ # n œœœ œœœ # œœœ œ œœ
accel. a tempo
? b œ œ # œœ ggg œ Œ Ó œ
b œ œ gg œ #œ œ
g
b œ œj œ ˙ . œœœœœ œ
b Œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
25
&
œœœ œœœ œœ œ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ
? b œ œ œœ # œœœ œœœ
œ
bb œ œ œ n œ œ . œ œ œ œ œ
œ œ Œ Œ ‰ .
. œ œ œœœ œœ œœœ Œ
31 1. 2.
& œ #œ œ œ #œ œ #œ œ #œ œ œ #œ œ
œ œ œ œ œ œœ œœ œœ œœ œœ œœ œœ
? b # œœ œœ œœ # œœ n œœ œ œ œ .
. œ # œ œ œ œ œ Œ
b œ
œ œ nœ œ œ
067
Muito devagar
# j j j
& c Ó Œ
œ œ
œ œ œ œ œ œ œœ œ œ œ œ œ œ œ œœœ
3
œ œ. ˙ œœ
p
? # c œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œ œ œœ œœ œœ
Piano
œ œ œ w œ œ œ ˙ œ ˙ œ ˙ œ
w w
# j j j U
œœ œœ œ œ
œ œ œ œ œ œœ # # œœ
6
& œ œ œ ˙ œ œ œ œœ œ œ œ œ œ œ œ œ œ
3
œœ œ œ. œ œ œ œ. J U
f ƒ ˙˙ ..
?# œœ œœ œœ œ œ œ œ œ œœ œœ œœ ˙.
œ œ œ w œœ œœ œœ ˙ œ ˙ œ ˙
w w
w
# œœ œœœœ œœ œœ j œ œ jœ j œ
Fine
œœ œœ œœ ˙
11
& œ œ J œ œ œ œ œ. J
œ #œ
j pj
œœ œœ œœ œ œ œ
cresc.
œœ œœ œœ œœ œ œ œ
?#
œ œ
œ œœ œœ w œœ œœ œœ œ œ œ œ # ˙ . œœ œœ
œ w œ œ œ
˙ œ œ œ
J J
U̇ œ œ nœ œ
# œ œ œ œ œœœ . œœœ
& œ œœœœ œœ œ œ
17
œ œ œ œœ œ œ
U
f ˙˙ .. f
œ œ œ œœ œœ œœ œ œ œ
? # ˙. œ œ œ ˙. œ œ
˙.
œ
w # œœ œœ œœ
#w
# j
œ œ œ. œ œ œ œ œ œ œ œ œ ˙.
22
& œ œ # œ
p œ
œœ œœ œœ
# œœœ œœœ œœœ œœ œœ œœ
D.C. al Fine
?# œ œ œ
w w œ œ œ
w
068
#### 2 j œ œœ ‰ g œœœœ ‰ œj œ œ ‰ g œœ ‰ g œœœœ ‰ œj œ 

& 4 ‰ œ œ
œ œ œœ ‰ g
ggg œ ggg J œ œ œœ ggg œ ggg J œ œ
p J J œ
? #### 2 œœœœœ œœœ œœœ œœœœœ œœœ œœœ œœœœœ
Piano
4 œœ œ œ œ œ œ
œœ œ œ œœ
U
#### j j
œ œ
œ œ œ œ œœ œ œœ œ œ œ œ œœ œœ œ
œ œ œ œ œ œœ
œ œ œœ œ œ œ œ œœ œ
6
& J J
œœœ œ
? #### œœ œœ œ œœœœœ
œ œœœœœ œœœœœœ œœœœœœ
œœ œœ œœ œ
œ œ
U
œ œ
#### œ j j œ œœ œœ œœ
œ œœ œœ œ œ œœ œœ .. œ œ œ œ œœ
œ œ œœ œœ œ # œœ œ
12
&
œ œ J J
f
œ œœ œ œœœœ
? #### œ œ œ œ œ œ œœ œœ œœ œœ œ
œ œœ œœ œ
œœ œ
œ œ œœ œœ œœ œ œ #œ œœ œ
œ œ ‰ œ
J
#### œ œ œ œ œ œ œ œ œœ œ œ œœ  œ n n n n œ œ œ œ œ œ œ œœ œœ
œ œ œ
18
& œ œ œ œœ œ œ œ œ œ J J œ
p
œ
? #### œ œ œ œ n œ œ œ œ œ œ œ œ
œœ œœ œœ œœ œ œ œ œ œ œ
œ œ n n n œœ œ œœ œ œœ œœœ œœœ œœœ
rit. a tempo
œ œ œ n œ œ œ œ œ œ
œ #œ
U
œ ^ j
j œ œ œ œ œ œ #œ œ œ œ œ #### j
rit.
&œœ œœ J J œ # # œœœ œœ  œ œ
œ œ œœ œ œ œ
œœ
25
J œ œ
œ œœ f œ j
? œ œœ œœ œœ œœ œœ œœ œœ # œ œœ ‰ œ œœœ
a tempo
œ #### œ œ œ
œ œ œ œ J œ œ œ œ # œœ # œ
‰ œœ œœ œœ œ
œ ‰ œ œ
U J
# # # # œ œœ œ œ œ œœ œœ œœ œ œ œœ œ œœ
œ #œ œ œ œœ œœ œœ œœœ œ
œ œ œ œ
œ œ
32
& J J œ œ œ ggg œœ.

f œœ
œ œ œ œ œ œ œœ œœœ œœœ œœœ œ œœœ
? #### œ œ œ œœ œœ œ nœ œ œ
rit.
œ œ œ
œ œ #œ œ œ #œ œ œ œ œ œ
j U
#### œœ œ j œœ j j
‰ gg œœ ‰ ggg œœœ ‰
œœ œœ œœ ‰ œ ‰ œœ
38
& ggg gg J œ œ œ gg n œœ œ œœ
gg œ

œ œ œ œ œ U
morr. poco a poco
? #### œ œ œ œ œ œ œ œ œ œ œ
n œ œ œ œ œ œ œ œ
œ œ œ
069
>œœ n œ %
œœ œœ œœœ >œ j >œ œ
Tempo de Marcha
j
2 ‰ œœ n œ
œœ œ
œ œ œ
œ œ œ œœ >œ œ œœ œ
&b 4 J J œ œœ œœ œœ œ œœ œœœ œœœ œ J
J J œ œ
ƒ
œ œœœ ‰ œj >œ œœœ œœœ œ œœ ‰ j >œ œœ œœ œ œœœ ‰ œj >œ œœœ œœœ
Piano
? b 42  œ œ œ
> >
j j
˙ œ. œœ œœœ n œ œœœœ œœœ œœœ >œ œ œ œ œ >œ œj œ j >œ œ
& b œ. œœ œ n œ œ œ œ œ œœ œ œ œ œœ œœœ œœœ œ œ
8
‰̇ J J J œ œ œ œ œ œ J
 J
œ j œ œ œ œ œ œ j œ œ
cresc.
?b œ œ œ œ œj ‰ Œ œ œœ ‰ œ >œ œœ œ œœ œ œ ‰ œj œ œ œ œ œœ ‰ œ œ œœ œ œœ
œœœ œ œ œ œ œ œ œ œ œ œ
>> > >
>œ >œœœ
œ œœœ œ bb j j‰ ‰ . j
&b ‰ ‰ J ‰ ‰ j nb b œ . œ œ j œ œ
16
œ J œ œ œ œ
J œ ‰̇ n œœ .. œœ œ
ƒ FIM F J
œ œ œ œ œ œ
? œ ‰ œj ‰ j ‰  b b œ œ œ œ œ œ œ œ ‰ nœ nœ bœ œ #œ nœ #œ nœ œ œ œ
b J œ nb b
œ
œ
j
b j ‰ ‰ œj œ . n œ n n œ˙ n œ œ œ b n œ˙ . n œj œ˙n œ n œ œ
& b bb œ œ œ œ œœ n˙
24
‰̇ œœ .. œ
J j j
œ œœ œ œ b œ j œœ œœ j n œœ œœ nœ œ nœ œ
œ
? b b nœ œ ‰ œ #œ œ #œ n n œ˙ b œ˙ œ œ œ œ nœ œ œ œ
bb
b b nœ. j j‰ ‰ j
&bb ˙ œ œ œ nœ œ nœ. œ j œ. œ œ œ œ œ nœ.
31
‰̇ œ . n œœ œ ‰̇ œ .
J
n œœ œœ œ n œœ œœ
œ
? bb J J J J ‰ n œ œ n œ œ b œ œ œ œ œœ œ œœ œ œœ œœ
œ ‰ nœ nœ bœ
bb
bb b j ‰ ‰ j j ‰ ‰ j œ. j
2 069
& b œœ j œ. œ œ œ œ œ œ. œœ œ œ œœ n œœ œ œ œ
38
œ œ ‰̇ œ . œ J
J J j
j œœ œœ
œ œ œ œ
? b b œ #œ nœ #œ nœ œ œ œ nœ œ œ œ ‰ œ œ bœ œ # œ œ œ n n œ˙ n œ œœ œ œœ
bb
b j ‰ j j
. nn
& b bb œ . j
œ œ œ œ œ œ.
œ nœ œ œ œ ˙
n œ œ œœ œœ
œ œ nœ nœ n n b
45
J
œœ œœ œ œ œœ œœ œ œœ œœ œœ œœ p
? bb œ œ œ œ j n
œ œ œ ‰ Œ
bb œ œ œ J œ J nnnnb
j- - -
& b .. œ œœ œ œ œ ˙ œœ œœ œœ
œœ œ œ œ
œœ œ‰ œŒ œ ‰̇ œœ ‰ œœ œ
51
‰̇ œœ ‰ œœ œœ œ‰ Œ ‰̇ œœ ‰ œœ œœ ‰ Œ ‰ œ ‰ œ
œ J J J J J J J J J‰‰‰
J J J
? .. œ ‰ œ ‰ œ‰œ‰ œ‰œ‰ œ‰œ‰ œ‰œ‰ œ‰œ‰ œ‰œ‰ œ œœ
b œ j
œ œ œ œ œ œ œ œ ‰Œ
j j
œ . œ œ . œ œ # œj œ œ œ
& b œœ œœ n œœ œœ œœ œœ # œj œ œ œ œ œ
59
‰ ‰ ‰ ‰ ‰̇ œ œ ‰ ‰̇ œœœ œœœ œœ ‰ Œ œœ œ œœ
J J œœœ
>>>
? ˙ #˙ n œ ‰ n >œ b >œ
‰ œ œ bœ
b ˙ #˙ n œ n œJ bœ ‰ Œ œ ‰œ‰ œ‰ ‰
œ
J
œ ˙ œ œ œœ œ œ ˙ œ œ œœœ ˙ œ. œœ
œ œ œ
˙ œ œ œœ œ œ œ œœ œ œ œ œ ˙ œ. œœ œ œ œ
& b œ œ œœ œ œ œ œœ œ œ J œ œ œ œ œJ œœ œ œ
66
œ >œ > > ‰̇ > J ‰̇ > ‰̇ ‰

œ
> Fœ œ œ œ œ œ œ œ œ œ œœ œœ
? b j ‰ œœ œ œ œ œ œ œ œ œ œ œ œ œ œJ ‰ œ œ œ œ œ œ ‰ œ œœœœ œ‰Œ
œ œ J J
bœ œ œbœ j j œœœ n %
œ
bœbœbœnœ œ œ jœ n œ
&b œ .œ œ œ . œ œœ œ œ œ. ..
75 1. 2.
œœ œ œ œ ˙‰ œ œœ J
#œ ‰œ .œ ‰œ ‰œ .œœ ‰œ œ ..œ.. œ
f œ f ƒ
œ œœœ j
? j ggg œœ œ j œœ œœ œœœ œ. œ j ˙ ‰ ‰ œ ‰ Œ .. œ ‰Œ
b bœ ‰ g J ‰ Œ œ J J‰ Œ œ . œ
bœ œ. œ . œ œ œ œ
œ œ J J œ œ
J J
070
Lento
# .. Œ œ œ œ œ œ œ
& # c Œ œ œ œ #œ œ œ ˙ ˙ œœ Œ Ó 
˙˙ œ
œ. œœ. œœ. . . . œ œœ
Œ œ œ œ œ œ œ œ œœ œ œœœ b œœœ œœœ .. œ œœ
Piano
? ## c  œ œ œ
œ œ
œ
#
& # œ œ œ ˙œœ#. œœ œœ Œ œ œ œ w œœ œ œœ Œ œ œ œ œ œ œ œ ˙œ . œ œ Œ œœœœœ
6
œ œ œ œ n # œœ # œ œ œ œ œ
œœ œœ œœ œ œœ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œœœ
œ œ œ œ œ
## œ
œ œ œ ˙. Œ œ #œ #œ œ œ œ œ œ œ œ ˙. Œ œ # œ # œ œ œ œœ œ œ ˙.
12 3
&
œœ. œ œ œ œœ œœ œœ # œœœ œœœ œœœ œœ œœ œœ

œœ. œœ.
3 a tempo
rit.
? ## œ # œœ œœ œœ œ œ œ œ œ œ œ
œ œ œ œ œ #œ
œ #œ œ
# œ œ œ œœœœœ œ œ ˙ œ œ œ. œ œ œ œ.
& # Œ œœ œ œ œ œ ‰ œ
17
œ œœ œ œ œ œ #œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ
? # # œ œœ œœ œœ œ œœ œœ œœ œ œœ œœœ œ œœ œœ Œ œ œœ œœ Œ œ œœ œœ Œ
œ œ œ œ
œ œ
# œ œ œ œœœœ œ œ œ œ œ œ œ
& # œ œ. œ œ œ œ œ œ œ. œ œ œœ œœœ Ó
23 1.
œ œ
œ œ œ œ œœ.
? # # # œ œœœ œœœ Œ œ œœœ œœœ Œ œ œœ œœ Œ œ œœ œœ Œ œ œœœ œœœ œ œœ.
œ œ
œœ.
œ
#œ œ œ œ œ œ œ
## .. œ œ ‰ œ œ œ
rit.

œœ œ œ œ œ Œ n b œœ n œœ œœ Œ œœ n œœ œœ Œ n b œœ n œœ œœ
29 2.
& œœ œœ œœ œ œ œ œ œ œ œ œ œ
? # # œ œœœ b œœœ œœœ .. œ œœ œ

œœ œœ Œ Ó œ Œ Ó œ Œ Ó œ Œ Ó
œ œ
œ œ œ
œœ
# œœ
& # Œ œœ n œœ œœ Œ œœ b œœ œœ Ó Œ Œ Ó
35
œ œ œ œ œ œ w œ
œ
animando
? ## œ Œ Ó œ Œ Ó œ Œ œ Œ w œ Œ Ó
œ œ œ
071
j j ‰ j
#
& # 42 œ
j .. œ œ œ œ œœ œ œ œ œ œ œ œ œœj œ˙ œ œ œ j
œœ œ œ
˙˙ œ œœ ‰̇ œœ ‰ œœ ‰ œœ ‰ œœ Œ œ
J J J J
œœ œœ œœ œœ œ œ œ œ œ
? # # 42 ‰ .. œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œœ œ œœ œ œœ œ œœ
Piano
œ œ œ
## ‰ j œj j ‰ j j
œœ ‰ œœ œ‰ œœ œœœ ˙˙ œ œ œ œœœ œ œ œ œ œ œ œ œ œœ œ˙ œ œ œ
7
& ‰̇ ‰J œ ‰̇ œœ ‰ œœ ‰ œœ ‰ œœ
J J J J J J J
œœ œœ œœ œœ œ œ
? ## œ œ œ œ œ # œ œœ
œœœ œ œ œ œ œ œ œ œ # œ œ œ œ œ œœ œ œœ
œ œ
# # œj ‰ j j j j
œ œ œ œ œ œœ œ œ œ #œ.
3
œ œœ œ œ ˙ œœ .
14
& œ œ Œ œ œ ‰̇ œœ ‰ œœ ‰ Œ ˙ ˙ . œœ
J J J J
œ œ # œœœ œœœ # œœœ œ œœœ
? # # œ œœ œœ œ œ œ œ œ ‰ # œœœ
œœ œœ
œœœœ
œœ
œœ
œœ
œ œ
œ J > œ #œ œ œ
## j ‰ j j j
& œ œ œ œ œœ œ œœ . œ œ˙ œ œ œ . œ œ œ œ˙ œ œœ
21 3
œ Œ œœ Œ œ œ ˙
? # # œœœ œœœ œ œœœ œœœ œ # œœœ œœœ

œ
# œœœ œœ œœ
œ
œ œ œ
œœ œœ
œ œ œ œ
œ œ œ #œ #œ œ
# j j j ‰ j
& # œ˙ œ œ # œ . œ ..
3
œœ . œ
27 1.
Œ œ œ œ œ œ œ
œ œ Œ œ œ ‰ Œ # œ œ̇ œ œ Œ œ œ
J
? ## œ # œœœ œœœ # œœœ œœ œœ
œ œ # œœœ œœœ œœ œœ œœ œ
œ œ œœ ‰ œœœ ..
#œ œ # œ œ œ œ œ œ œ J
œ
2
# #### . j071> j >
œœ œ œ
œœ œœ œ
& # œœ # œœœ # œœœ n œœ œ . œœ œ œœ
33 2.
œ œœ œ œ #œ ‰̇ œ
œ J J J
f œ
? ## #œ #œ # # # # .. œ
œ œ œ
œœ œ
œ œ œ # œœ œ œœ œ œœ
œ œ œ œ
œ #œ #œ
#### œ ‰ j œj j ‰ j
œœ œœ œœ œ œœ œœ œ œœ œ œœ œ œœ œ œœ ..
38 1.
& œ J œ œœ
J J J J œ J ‰̇ œ J
œ œ
J
œ
J
œ œ
? #### œ œ œ œœœ œœ
œ œ
œœ
œ œ
œœ
œ .
.
œ
#### j‰ j j ‰
œœ œœ œœ œœ œœ œœ œœ œœ œ
˙œ˙ œ # œ .
3
œ œ #œ œ œ œ œ œœœ œœ œœ œœ
43 2.
&
‰̇ œ œJ ‰ œ œ œ œ œ œœ œ œ œ œ
J J
? #### œ œ œ œ œ œ œ œ œœ œœ œœ œœ #œ œ #œ œ
œ œ œ œ œ œœœ œ œœ œ œœ œ œœ œ œœ
#### j j j j
œœ œœ œœ œœœ
œ œœ œœ œœ œœ œœ œœ
œœ œœ œœ œœ œœ œœ œœ œ œ˙ œ œ œ .
49 3
& œœ œœ œœ œ œ œœ
J J J J
œ œ œ œœ œ œœ œ œœ
? # # # # # œ œœœ œœœ # œ œœœ œœœ œ œœ > œœ >œ œœ œ œ œ œœ œ œ œœ œ
>
j
#### œ œ œ # œ . ‰ j‰ ‰œ
j j ˙ œœ œœœ
3
œœœ œœ œœ œœ œ œ œ
55
& ˙˙ œ œ œœ œ œ œœ œ‹ œ œ œ œœœ œœœ

œ œœ
J J J J J
? # # # # œ # œœœ œœœ # œœœ œœœ œœ œœœ œœœ œœœ œœœ œ œ œ j‰ j‰
#œ œ
FIM
œ #œ œ œ #œ œ œœœ œ œ
œ œ
072
œœ j #œ
bb c ‰ j œ œœ gg œœœ ‰ Œ j œ œ
œ g
g œ
‰ g œ ggœ gggœ gg ‰ Œ ‰ j œj œœ ‰ œœ
& œ g œ g
g
gg œœ ggœ g Jœ g
g œ
gg œ gg œ J œ # œ
œ
gg g g œ œ
œ œ œ œ œ
j
legato
? b c ˙ œ œœ œ ˙ œ œ œ
Piano
b œ œ œ œ œ

b œœœ œ œœ gg œœœ ‰ ‰ œj . j
b j œ œ ˙œ œ œ œœ œ œ œj
gg œœœ gggœœ gggœ gg J œ œ. œ œ˙. # œ . œ œ œ œj
4
& œ
gg g
? b œ œ œ j‰ Œ ggg œœœ gg œœœ œ gg œœœ gg œœœ œœ œœ
b œœ g œ gg gg œ œ #œ œ œ
œ g œ œ œ œ œ
œ
‰ œj . j
œ œ œ œj
bb œ.
j
œ œ˙. œ œ ˙œ œ œ œ œœ œ # œœ .
8
& œœœœ ˙ œœœ œ

3
œ J
œ gg œœœ gg œœœ œ gg œœœ gg œœœ œœ œœ
? b
b œ œ œœ œ œ œ œ gg g œ gg gg œ œ # œ œ œ
œ œ œ œ œ œ
b œœ œœ œœ œ œ œ œ œ œ œ ˙
& b œœ .. œ ˙ # n œœœ œ œ œœ œœ œœ œœ œ œ # œœ ˙˙ œœ œœ ˙˙ # n œœœ
12
? bb gg œœœ gg œœœ Œ gg œœœ gg œœœ gg œœœ gg œœœ Œ œœ œ˙ œ

gg œœœ gg œœœ Œ
g œ gg g Œ g œ œ ˙
# g
œ gg œ gg œ œ gg
œ œ œ
U
bb œœ œœ œœ œ œœœ œœœ œœœ œœ œœ œ b˙ œœœ #
n n ˙˙ œœ œ
& œ œ œ œ œ œ œ œœ œœ # œœ ˙˙
17
œ œœ œœ œœœ œœ # œœ œœ # œœ œ œ œ œ œ œœ œ œ œ œ
? b œ ggg œœ ggg œ Œ #n œ œ œ œ œ œ œ œ
rit. a tempo
b g œ œ n
œ
œ œ
#
2 072
21
& œ
3
˙ ˙
œ œ œ œ œ œ œ œœ œœ œœ œœœ œ œœ œ œ œœ œœ
? # œ œ œ œ œ œ œ # œœ œ œ # œ œ œ
3
# œ œœ œœ œœ # œœ œœ œœ ˙˙
œ˙ œ œ˙ œ œ œ ˙˙ œ
25 3
& œ ˙
3
#˙ nœ Œ Ó œ œ œ œ œ œ
? # œœ œ œœ b œœ œœ œœ œœœ œ œœœ œœœ œœœ œœ œœœ œœœ œ œœ œœ œœ œœ

œœ œ œ œ œ
#
29
& œ
3
˙ ˙
? # œ œ œ œ œ œ œ # œœ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œœ
3
œ œ œ œ
U
# ˙ œ
‰ œJ œœ œœ
œ œ œ œgœ ggg œœœ b œ œ nb
& œ œ œ œ œ œ œ ˙˙˙ œ b
33 3
œ œ œ œgg œœ ggœœ ggœ g œ

3
œ
gg œ gg
U #œ œ œ œ œ œ œœ œœ œ 
?# œœ œœ n œœœ œœœ ˙˙
j j j
œ œ œ œ œ œ
˙ Œ œ œ œ œ #œ œ œ œj ‰ Œ nb
b
j #œ œ
bb ‰ j œ œœ gg œœœ ‰ Œ ‰ j œ gœœ ggg œœœ ‰ Œ ‰ j j œœ ‰ œœ œœ ‰ ‰ ‰ Œ
g g œ gœ g œ œ
37
& œ g œ gœ
ggg œœ ggœ g J g g œ g
ggg œ ggœ g g J œ œ
œ œ # œ œ j
g g œ
œ œ œ œ œ
j
œ œœ œ
pizz.
? b ˙ œœœœ ˙ œ œ œ œ œ ‰ œ ‰ j‰ Œ
b œ

073
Uœœ œœ œ
2 œ j œ. œ œ œ œ œ œ œ œ . œ œ
& 4 œj .. œ . œ œ œ œJ œ œ J J
œœ œœ œœ œœ œ œ œœ œœ œœ #Uœœ œœ œœ
rit.
œ œ ‰ # œœ œœ
Piano
? 42 ‰ .. œ œ œ œ‰ œ œ œ œ‰
œ œ #œ J œ œ œ J œ œ œ J
#
‰ œj # # œ . œ œ œ œ œ œ œ œ . œ œ œ œ œ œ œ œ . œ œ œ œ
a tempo
& œ. œ œ. #œ œ
7
? œœ
œ # œœœ œœ œœ # # œœ œœ œœ œœ œœ œœ œœ œœ œ œ
œ ‰ # œ œ œ œ œ œ œ ‰ œ œ œ œ œ ‰ œ œœ œœ
œ œ œœ J J œ œ J œ
### œ œ œ œ. œ œ œ œ œ œ œ ‰ œ œ. œ œ œ œ œ œ œ œ œ œ œœ
14
& œ J
œœ œ œ œœ œ œ œ œ œœ œœ
? # # # œ œœ œœ ‰ # œ œœ œ œœ œ œ œJ ‰ œ œœ œ œœ œ œ œJ ‰ œ œ # œ œ
J
### œ œ . œ œ œ œ œ œ œ œ j . œœ œœ j n n n
‰
œœœ J œ œ œ œ œ #œ œ œ œ œœ œ œ œ œ . œ œ œ œ
20 1. 2.
&
œ œ œ
? # # # œ œœ œœJ ‰ œœ œ œœ œ œœ œœJ ‰ œ
œœœ # œ œœ œ œœœ œœ ‰ .. œ œ ‰ n n n
œ J œ
œœ œœ U
œ
œ œ œ œ. œ œ œ œ œ œ œ. œ.
a tempo
& œ. œ œ œ œ œ œ j œ . œ œ œ œ œ #œ œ œ œ
26
J œ J J
? œœœ œœ œœ œœ œœ œœ œœ œœ œœ #Uœœ œœœ œœœ œœ # œœ œ

rit.
œ œœ‰ # œ œ œœ‰ œ œ ‰ œ œ œ œ
œ œ #œ J œ œ œ J œ œ œ J œ œ
074
œ œœ œ œœ œ ‰ œ œœœ œœ .. œ . j
& c ‰ œ œ œ œœ œ œœ œ ‰ œ # œ œ œ œ œ œ ‰ œ œœ œœœ œ œ œ œ
# œ œ œ. œ œ.#œ
f p
œœ œœ #œ œ œ # œœœ œ œ œ œœ œ
œœœ œ œœ œœ œœ .. œ
Piano
?c
œ œ œ œ œœ œœ œœ œ œ œ œ
œ
j j j
& œ œ œ œ œ ‰ œ # œ n œ œj œ œœ œ ˙ Œ œ œ œ œj œ œ œœ œœ œ. # œ. œœ œœ œ. œ. œœ œœ n œ œ œ œ
6
#œ œ œ œ #œ
? #œ œ
œ œ œœ œœ
#œ
œ œœ œœœ œ œœ œ œœ œœ #œ œ œ
œ œœ œœ œœ
#œ
œ œœ œœ
j U
### j œ j
jj j j
‰ œ œ œ œœ œj œ œ œœ ‰ ‰ œj œ œ j œ œ
œ
12
&
œ œœœ œœœ œœœ # œ œ . œ œ œœ œœ œœ œœ ˙˙ œ œ œ œ
œ œ œ
F
? ### œœ œœ œœ œœ œœœ œœœ œœ œœ œœ œœœ U œ œ œ
w œ œ œ œ œ œ œ œ œ œ œ J ‰ Œ œ œœ œœ œœ
œ
## j j
& # œœ œœ œ. œ. œœ œœ œ. œ. œœ œœ œ œ œ œ Œ ‰ œj œ j œ œ œ œ. œ. œ œ. œ. œj œ œ
18
œœ ˙ œ œ œ œœ œ œœ œ œœ œ œ œ
œ
? ### œ œ œ œ œœ œœ œœ œœ œœ œœ
œ œœ œœ œœ œ œœ œœœ œ œ œ œ
œœ œœ œœ
œœ œœ
œ
### ^ ^ ^ ^ > .œ
œ œ œ œ œ
24
& œ œ œ
œ œ. œ
œ. . œ. . . œ. œ. . . œ. œ. œ. œ. œ œ œ œ œ. œ. œ . œ. œ. œ œ œ
. . œ. ˙ f œ œ >œ > œ
>
? ### œ œ œ œœ œœ œœ
œ œ œ œ œ œ œœ œœ œœ œœ œœ œ œ œ
œ œ œ
œ œ œœ œ
>>>
### j j
œ œ ‰œœœ œ œœ œ œ œ ‰ œœœ œ œ
30
& œ œœ œœ œœ ˙ J J œ . œj
˙ œ
œ œ >œ >
>
? # # # œ œœ œœ œœ œ œ>>> œ œ œ œ œ œœ
œ œœ œœœ œ œœ œœ œ œ œ œ œ œ œ œ œ œ œ
œ œœ œœœ
œ œ
### œœ^
Œ Ó .. Œ œœ ‰ Œ
‹
œœœœ œ œœœœ
36 1, 2. 3.
& œœ œœœœœ œ œœœœ

œ œœ œœ œœ œ œ J
? ### œœ œœ œœ œœ œœ œ œ œ œ œœ œœ œœ . œœ^
œ œ œ œ œ œ œ œ œœ œœ œœ œ œœ œœœ œ n œ œ œ . œ œ ‹ œœ ‰ Œ
œ J
075
### %
Allegretto
c œ œœ œ œœœ œœ œ œœœ œ œœœ œœ œ œ œœœœ œœœœœœœœ

& œ œ œ œ œ œ
f
œœ œœ œœ œ œ œ œœ œœ œœ œœ œœ
? ### c ‰ Œ œ œ œ œ œœ œœ Œ œ œœ œ
Piano
œ œ Œ œ œ Œ
œ œ œ œ
## œ œ œ œœ œœœœœ œ œU
Lento
& # œ œ œ œ œœ œœœ  œ œœ œ œ œœœœ œ œ œœœœ

6
R
p
œ œ œ œ
œœ œœ œœ œœ œœœ œœœ œœœ
FIM
? # # # œ œœ œœ Œ œ œ œ rU ‰
œ œ œ œ Œ œ œ œ œ œ
œ
### œ œœœ œ œ œ œœ œ œ
œ œ œ œ œ œœœ œ œ œ œ œ œ œ œ œ œ œœœœ œ œ œ
12
& œ œ œ œ
œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? # # # œ œœ œ œ œ Œ œ œ œ œ
œ œ œ
œ
œ œ œ œ Œ
œ œ œ œ œ
# # œj œ œ œU œ œ œ œ œ œ œ œ œ œ œ œ œ œ U
œ œ œ œ œ œ œ œ œœ œ œ
& # œ œ œœœœ
j
18
œ
œ
p p
œ
œœ œ œ œ
œœ œ œœ œœ Œ œ œœ œœ œœ œœ œœ œ œœ œœ
? ### œ œ œ œ Œ œ œ œœ œ œ œ Œ
œ œ œ œ
œ œ
## œ œœœ r . . . j œ. œ. œ. œ
& # œ œ œ œ œ œ œ œ œ œœœœ œ œœœœ
24
œ œ œ œ
œœ
? # # # œ œœ œœ œœ œœ œœ œœœœœ œœœ œœ œœ
espress.
œ œ œ
œ Œ œ #œ œ œ
œ œ #œ œ œ
œ
##
& # œ œ œ œ
r . œ. œ. œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ
œ œœ
29

œ œœœ œœ œœ œ f œœ
? ### œ œ œ œ œœ œœ
œ œœ
œ
œ œ
œ œ #œ œ œ œ #œ œ œ
### œ j j œ œ %œ
œ œ
œ œ œ œ œ œ œ. œ. œ. œ œ œ œ œ œ œ œ œ
34
&
p f
œœ œœ œœ œœ œœ œœ œœ œœ œœ
? ### œ œ œ œ œ œ œ œ œ Œ
œ œ œ œ œ œ
076
j j
Devagar
œ ˙. œ . œ œ œ œœ œ œ œ œ. œ œ œ œœ œ œ œ œ
œ
& 43 Œ Œ .. œ œ ˙. œ
> Œ
p
^ ^ ^ ^ ^
Piano
3
&4 œœ œœ .
. œ œ
˙. ˙. œ œ ˙ . œœ œœ œœ œœ œ œ
˙. œ œ
˙.
> j j
œ. œ œ œ œ œ œ. œ œ œ œœ œ œ
6
˙. œ œœ œ œ ˙. œ œœ œœ
&
Œ
^ ^ ^ ^
& œœ œœ ˙. œœ œœ œœ œœ œœ œœ
˙. ˙. ˙.
œœ .. œœ œœ œ œ œ œ œœ œœ œœ .. œœ œœ # œ n œœ œœ œœ
J œ œ œ œ J # œ œœ œœ
10
&
^ ^ ^ ^
& ˙ œœ œœ ˙. œœ œœ œœ œœ ˙. œœ œœ
. ˙.
œ œ œ 
œ œ
œœ .. œœ œ œ œ œ œ œ œ œ œ œ œ œœ
‰ J œ œ œ œ œ œ ‰ œ .. œ œœ ‰ Œ
J
14 1.
& J J
f
^ œj ‰ Œ
rit.
^ œ œ ^ œ œ ^ ^
a tempo
.
& ˙. œ œ ˙. œ œ œœœ œœœ ˙
œœ œœ .
˙
œœ
˙. .
077
œœ œœœ j j
Andante
b œ œ. œ œ j œ œ. œ œœœ
& b c œj œj œ œ œ œ œ œœ œ œ œ #œ œ œ nœ œ œ
j œ . # œ œ œ # œ # œœ œ œ n œ œ œ
œ
J
F F
œ œ œ œœ œ œ œ œ œ œ
œœ œ # œœ œ œœ œ œ j ‰ Œ
Piano
? b c Œ ‰ œ œ #œ
b œ œ œ œ œ œ
œ œ
%
b œ œ. œ œ j œ œ. œ j œ œ œ œ. œ œ œ j œ nœ œ œ œ
&b œ œ
5
œœ œœ œ œ œœ
œ œ œ œj œ œ œ #œ œ œœ œ œ œ œ œ
œ g œœ œ
œœ gg œœ gg œœœ gg œœ œ œ œ œ ggg œœœ nœ œ œ œ œ œ
? b gg œ gg gg œ Œ œ # œ œ œ gg Œ
b œ œ
œ œ
œ œ
œ g œœ ˙˙ œœ fi
b œ g œ g œ ˙˙ œœ ‰ ‰ j œ œ œ bœ. œ. œ. jœ n# j
& b œ gggg œœ gg œ gg œ œ #œ ˙ œ n œ œ n œ . œj œ . œ
10
J œJ
œ
œ œ œ ˙˙ œœ œ
œœ
œœ œœ œœœ
? b
b & ˙ œ‰Œ ? œ œ œ œ œ œJ ‰ Œ nn
# œ Œ
J œ œ #œ œ œ
œ œ
# œ œœ œ ˙ œ œ œ #œ œ œ œ œ œ œ œœ œœœœœ
œ œ œj œ œ œ . œj œ . œJ œœœ œœœœ œœ œœœ
15
j
& œ
œ œœ œœ œœ œ œœ œ œ
?# œœœ œ œ Œ œœ œ Œ œœ œ œ Œ œ œ Œ œœ œ
œ
Œ œœ œœ œœ Œ
œœœ œœ œ œ œ œœ œœ
œ
œ
# œœ j
œ œ jœ œ œ . jœ. œ œ œ œ œ œ jœ œ œ œ. j œ. œ ˙ œ œ œ #œ œ œ œ œ œ
21
j
& œ œ œ œ œ J
œ œ œ nœ
œ œœ œœœ œ œœ
? # œœ œ œ œ Œ œ Œ
œ
œœ œ œ œ Œ œœ œ Œ œœ œ œ Œ œœ œœ
œ œ Œ
œœ œ œœ œœ œ œ œ
œ
œ œ
U
# œ. œ. n b fi œ œ. œ œ j
& œ œ œ œ œ . œ œ . œ œ . œ œ œ . œ œ . œ œj ˙ b ˙
27
œ œ n œj b œ œ œ œ œj œ œ œ œœ œœ œ œ
œ œœ œ œ œ œœ
rit.
œ œ
Œ œœ œ œœ œ œ
D.S. al O
?# œœ œœ Œ nb
b œ œœJ‰ Œ œœ œ
dim.
œœ œ bœ œ
œ œ œ
b œœ
&b œ œ œ. œ #œ œ. #œ œ œ œ œ œ #œ œ
# œœœ œœ Œ
32
œ œ nœ
j
œ œ œ

œ œ œ œ œ œœ
? b œ #œ œœ œ # œœ œ œœ œ Œ
b œ
œ œ œ œ
œ
078
# 2
& 4 œ œ œJ œ œ œ œ œ œ œ œJ œ œ # œ œ œ œ œ œJ œ œ œ ˙ ˙
F
œ œ œœ œ œœ œœ œœ n œœ œœœ œ œœ œœœ œœœ œœœ
? # 42 Œ b œ œœœ œ œœœ œ œœ œ œœ
Piano
œœ œœœœ œ œœ œ œ œ
#œ œ . œ œ œ œ œ
. œ œ œ œ œ .
8 3
& œ œ œ œ ˙ œ œ œ œœœœ œ ˙
3
p
œ œ œœœœ œ œœœœ œ
? # Œ .. œ œ œ œ œ œ œ œ œ œ œ œ # œ œ Œ œœ œœ
œ œ œœ œœ œœ œœ œœ œ œ # œ
œ œ œ œ œ #œ
œ œ #œ œ œ
# œ œ œ œ œœœ
& œ œ œ œ œ œœœœ œ ˙ œ œœœ ˙
16
˙
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ
3 3
?# œ Œ œ œ œ œ œ œ œ #œ œ Œ œœœ œ
œ œœœœ œ œ œ #œ nœ
œ œ #œ œ
# œ œ œ œ Uœ œ œ œ ˙
œ œ œ œ œ œ œ œ œ œ #œ
24 3
& œ ˙ œ
œ œ œœœ œ œ œ œœ œœ U œœ œ œœœ œœœ œœœ œœœ
3
?# œ Œ œ œ n œ œœ œ # œœ œœ œœ œœ # œœ œ œr n œ œ # œœ œ #œ œ œ
œ œ œ
#
˙ œ œ œ œJ œ œ œ œ. œ œ œJ œ œ # œ œ . œ œ œJ œ
31 1.
& œœ ˙
bœ œœœ œœ œœ œ œœ œœ n œœ œœœ œ œœ œœœ
? # œ œœ œ œœœ œœœ œ œœœ b œ œœœ œ œœ # œœ
œ œ œœ
œ
œ œ œ œœ œ œœ œ
# œ .. ˙ œœ
œ œ Œ
39 2.
& ˙ œ
œœœ œœœ œ f
?# œ œ Œ .. œœ œ œ Œ
œ œ œ
079
œ œ œ œ œ œœ œœ œœ œ œ œ œ

œ  œ   œ œ #
J ‰  # œ  œœ œ œ
&c Œ œœ 
œ œœœ œœ
 J J J
œ œ
œœ # œœ œ œ j # œœœ œ
? c j ‰ œœœ ‰ œœJ ‰ œœ
Piano
j
‰ œ ‰ J ‰ J ‰ J ‰ œ ‰ J ‰ ‰ ‰
œ J œ
œ œ
œ  œ œ  œ œ  œ #œ  œ œ œ œ œ œ œ œœ œœ œœ œ
J ‰ œ  œ  œœ  
4
&
œ
f J
j œ œ œœ œœ j œ œ
œ j œœ œœ # œœ
? œ ‰ # œJ ‰ œJ ‰ œJ ‰ œ ‰ œœ ‰ œJ ‰ ‰ œ ‰ J ‰ œJ ‰ œJ ‰
œ
œ œ J œ
°
œ  # œ œœ  œ œœ  œ œ  œ œ  n œ œ  j ‰ Œ ###
œ œ œ ‰ œ œ œ ‰ œj ‰ œj
7
& J œ œ
œ # œ
œ
œœ # œ œ œ œ œ œœ œœ
J J J f œ
œœ œœœ œ œ j j
? œ ‰ ‰ ‰ ‰ ‰ Œ # œ   #
œ ## œ œœ œ œ œ œ
J œ
*
### Œ
œ œ œœ ˙œœ . œœ œœ œœ Œ œ # œ œ œ œ œ w œ œ œ œ œ
10
& œ œ œ
œœ œ œ œ œ
J J . œ œ œ œ
. J J
‰ ‰ ‰ œ œ œ œ œ
? # # # œœ œ œœ œœ œœœ œœœ œ œ œ œ œœ œœ ‰ ‰ ‰
œ œ
œ˙ œ œ ˙ œ œ œ˙ œ œ ˙
### Œ œ w œœ œœ œœ œœ œœ
œœ œœœœ œ œ ‹œ œ œ œ œ #œ œ œœœ œ
14
& ‰ œ œ ‰œ‰œ
. . . ƒ .
? # # # œœ œ œœ œœ œ œ
œ œ œœ œ œ˙ œœ œ œœ œœ œ œœœ œœœ œ˙ œœ œ œœ œœ œœœ œœœ
œ œ œ
œ˙ œ œ ˙
œ w œœ œœ œœ œ # œ œ œ œ œ wœ œ œ œ œ
## w œ œ œ œ œœ ‰ œœ Œ œ # œ œ œ œ œ
& # Œ œ œ w œ w œ œ œ œ œœ ‰ œœ
18
œ œ œ œ œ ‰ ‰ ‰
. . .‰
. œ œ
? # # # œœ œ œœ œœ œœœ œœœ ˙ œ œ œ˙ œœ œ œœ œœ œœœ œœœ ˙ œ œ
œ˙ œ œ ˙ œ œ ˙ œ œ ˙
### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ
2 079
Œ œ œ œ œ œ ## œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ
22
&
. . . .
# œ œœ œœ œœ. œœ œœ œ œ œ œœ # œ œ œ œœ œ œ œ œœ œ œ œ œœ a ƒtempoœœ œœ œœ
? ### #œ œ
rit.
# œ œ œ˙ œ˙ œ˙ œ˙ œ œ œ œ
# # # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ n n n œœ œœ œœ œœ œœ œœ œ
œœ œ œ œ œ œ œ œ œ œ œœ œœ œ œ œ œ œ œ œ œ œ œœ œœ œ œ œ œ œ œ œ œ œ œœœ œœ œ œ œ œœ
26
& œ œ
œ œ œ œ œ œ 
? ### œ œ
œœ œœ œœ œ œœ œœ œœ œœ œœ œœ n n œœ œœ œœ
œ œ n œ œ œ œ
œ
œ œ
& œœœ œœ œ œœ œœ œ œœ œœ œ œœ œœ œœœ # œœ œ œœ œ œ œ œœ œ œœ œœœ œœœ œœ œ œœ œœ œ œœ œœ œ œœ # œœœ œœ Œ Œ ‰ œj
30
œ œ
œœ œœ œœ œ œ œ œ œ œ œœ œœ œœ
? œ œ œ # œœ œœ œœ # œœ œœ œœ
Lento
j j œ œœ œœ œ œ œ œ œ ˙ . œœ œ œœ œœ œ œœ œœ œ œœ œœœ œœ œ œ œ œ
& œ œ œ œ œ ˙‰ #. œ œ œ œ œ œ œ œ œœ œœœœ œ
34
J J ‰ œ œ J J
œœ œœ œœ œœ œœ
? œœœ œœ
œ # œ œ œ # œ œ
œ œ #œ #œ œ œ œ œ œ œ œ œ œ
˙. œ‰œœœ j j
&œ J œ œ œ œ œ œ ˙ .# œœ œ œœ œœ œ œœ œœ œ œœœ œœœœ
39
#œ œ œ œ œ œ ˙ œ ‰
œ œ œ # œœœ œœœ œœœ œœ œœ # œœœ œœœ œœœ
? œ œœ œœ œœ œœ œœ œœ œ œ
œ œ œ œ #œ
œ œ œ œ œ œ
& œ œ œJ œ œJ ˙ . œ œ œ œ œ œ œ œ œ œœœ œœ œ œ œ œ œ ˙ . œ‰œœœ
J
44
œ œ J J #œ œ œ
‰
œ œ œœ œœ œœ œœ œœ œœ œ œ œ œ œ œ
? # œœ
#œ
œœ
œ œ œ œ œ œ œ œ œ œœ œœ œœ # œœ œœ œœ
œ œ
j
079 3
œ. œ œ. œ œ. œ œ. œ
& ‰ œ. œœ. œ
œ œ. œ œ. œ œ. œ #œ. œ œ. œ œ. œ œ. œ œ. œ #œ. œ
49
œ œ
œ
œœ
œ œ
œ œ œ œ
œ œ œ œœ # œœœœ œœ œœ œœ
? j ‰ œœ ‰ œJ ‰ J ‰ j ‰ œJ ‰ # œJ ‰ œJ ‰ j ‰ # œœJ ‰ œJ ‰ J ‰ j ‰ # œJ ‰ œJ ‰ œJ ‰
œ J œ œ œ
œ
œœ œ j j jœ œ œ œ œ ˙ œœ
& ˙. œ
53
œ œ œ œœœ˙ œ œ #œ œ # œ J
œœœ œ œ œ œœ œœœ œœ
? j ‰ œœœ ‰J
œœ
‰ œ ‰ œj ‰
œœ œ œœ
œ ‰ œœ œ œ #œ œ œ # œœœ œœ
œ œ œ
œ J J J #œ œ œ
œ œ œ œ
&œ œ j œ œj œ œ œ # œ œ œ œ œ # œ œj œ œ œ  œ  œ 
58
œ œ #œ œ œ J œœ
 J
œœ œ œ œ œ œ
œ
? œœ œœœ œ # œœ œœ # œœ œœ j ‰ œœ ‰ œJ ‰
œ œ œ #œ œ œ œ J ‰
œ
œ œœ œœ œœ œ œ # œ œ œ œ œ  œ œ  œ œ œ #œ œ
 J œ  œ  
‰ # œœœ  œœœœ 
62
&
œ J
J
œœœ œœ œœœ œ œ j # œœœ œœœ œœœ
? j ‰ J ‰ # œJ ‰ J ‰ œj ‰ # œœJ ‰ ‰ ‰ œ ‰ J ‰ J ‰ J ‰
œ œ œ
œ
œ œ œ œ œœ œœ œœ œ

œ  œ  œ  # œ œœœ  œ œœœ  œ œœ  œ
& J ‰ œ œ œœ 
65
œ J œ œœ
f J J J
j œ j œ #œ œ
? œ ‰ œœœ ‰ œœJ ‰ ‰ œ ‰ œœœ ‰ œœ ‰ œœ ‰ œœ ‰ ‰ ‰ ‰
œ J œ J J J J
°
>œ
œ  nœ œ  #œ j ‰ Œ Ó ggg œœœ ‰ Œ
g
68
& # œœœ œ gJ
œ
J ƒ
œ œ œ œ ‰ œj ‰ j ‰ Œ
? ‰ Œ # œœœ   œ œ œ œ
J œ
>
080
Vagaroso %
œ œ
& b c œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ
? b c ‰ Œ œ œœ œœ œœ œ # œœœ œœœ œœœ # œœœ œœœ
Piano
œ #œ œ œ œ œ
œœ œœœ œ ˙ œ ‰ j n##
&b œ œ œ #œ œ œ œ œ œ œ œ œ œ œ ˙
5
œ œ œ #œ
œ œœ œœ œ œœœ ˙˙˙ œœ
? b œ œœ œ œ œ œ œœ
n##
Fine
œ œ œ
# œ œ ‰ 
œ œ œ
# œ œj œ ˙ œ œ œ œ œJ œ œj ˙ j œ œj
& # œ œ œJ œ ‰ œ œ œ œJ
10
œ œ œœ œ œœœ œœœ œœœ œ œ œœ œœ œœ œœ œœ œœ

? # # œ œœ œœ œ œ œœ œ œœ œ œ œ œ œ œ œ œ
œ œ
## œ Uœ œ œ. œ. U
œ ˙ œ œ œ œ œJ œ œj ˙ j
œ ‰œ œ œ J J œœœ
œ œœ
15
&
œ œ œ œœ œœ œœ U U
? # # œ œœ œœ œœ œ œ
œ œœ œ œœ œ œ œ œ œ œ œ œ œ œ œ œœœ
œ
œ œ
œ œ
œ œ
œ œ
œ
œ œ œ œ
%
# œ Uœ U
j œ œ J œJ œ. œ. œ œ œ œ œ œ œ œ œ œ
œ
& # œœœœœœœœ ˙ ‰ œ œ ˙ œ œ bœ œ
20
œ œ œ
œœ œœ œœ œœ œœ Uœœ œœ œœ Uœœ
? # # œ œœœ œœœ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ
œ œ j‰ Œ
œ œ œ œ œ œ
081
Tempo de fado vagaroso

. . œœ œœœ œœœ œœ ˙˙ j
& b c œ œ œ œ œ œœœ .. œ œ œ œ œ œœœ .. œ œ œ œ œ # œœœ ... œœœ œ œ ˙ ‰œ
p
œœ œœ œœ # œœ œœ œœ f œœ œœ œœ
œ œ œ
Piano
?b c ‰ Œ œ œ œ œ œ œ œœ œœ œœ
j
& b .. œ . j jœ j œ œ . ‰ œ œ. j jœ j
5
œ œ œ œ #œ. œ œ œ œ œ œ œ ˙ œœ œ
œœ œœ œ œ œ œ œ œ œœ œœ œœ œœ œœ
? b .. œ œ œ œœ œœ œœ œ œœ œœ œœ œ œ œ œ œ
œ œ œ œ œ
j j
& b œ #œ. œ œ. ‰ œJ œ . # œj œ œ œ œ œ . œ
10
œ œ œ œœœ œ ˙ œœ
F
?b œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œ # œ œœœœ œ œœœœ œœ
œ œ
œœ
œ
œ œ œ # œ œ
œ œ
& b œ œ. ‰ œJ œ. j œj œ œj œ œ . œ
œœœ œ ˙ œ œ œ œ.
15
#œ œœœ
# œœœ œœ œœ œœ œœ œœ
?b œœœœ œœœœ œœ œœ œœ œ # œ œœ œ œœ œœ œœ œœ œœœœ
œ œ œ œ #œ œ œ œ œ œ œ œ œ #
U
# # œ œ œ . œj œ œ j j
Estribilho Mais Depressa Vagaroso
œ œ œ j
b n œœ œœ œœ .. œœ œœ œœ œœœ œœœ œœ .. œœ œœœ œœœ
œ . œ œœœ œœœ œœœ ... œœœ œœœ œœœ œ œ œœ œœœ œ b œ œ
20
& œ œ . # œœ œœ œœ œœœ
f p
U
? b # œœœ œœœ œœœ n # # œœœ œœœ œœœ œœœ œœœ œœœ œœœœ œ œœœœ œ œœœ œœœ ‰ Œ
œ œ J
œ œ
œœ
## œ œœœ œ œ ˙ ‰ œ .. œœœ œœœ œœœ œœ Œ
j
œ œggg œœœ .. œ.
œ œ œ œ n œgggg œœ .. œ œ œ n œ œ œœ ... œœ œœ œœ œœ n œœ ˙˙
25 1. 2.
& gg . g
œœ œœ œœ f œ œœ œœ
? # # b œœœ œœ œœ
œ œ œ n œœœ œœœ œœœ œ œ œ n œœœ œœ œœ .. œ œ Œ
œ œ œ œ œ œ
082
# j j œ œ œj
& c œ œ œ œ œj œ œ œ œ œ œ
j œ œ œ œ œ œ ˙.
œ œ œ œ
œ œ œJ œ œ œ œ
J
œ œœ œœœ
œ
œ F
œœ œœ œœ œœ œœ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ
Piano
?# c
œ
œ œ œ
œ
œ
œ
œ
# j j j j
& œœœœ˙ œ œ œ œj œ œ j
œ ‰˙ œ œ œj œ œ œ œ œj œ œ œ œ œ œ œ
6
œœ
œ œ œ œ œ
œœ
œœ
œœ œœ œœ œœ œ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œ

?# œ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œœ
# j j j j
˙. œ œ œ œ œJ œœ œœ œ œ œJ œ œ œ œ œ œ ˙ œ œ œ œj œ œ j
œ ‰˙ œ
12
& œœ
œ œ œ œ œ œ
œœ œœ
?# œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ
œ œ œ œ œ œ
# j j
œ œ œ œ ‰ ‰ œJ œ œ œj œ œ œ ˙ # œ œ œ j
œ œj œ œ œ‰
18
& ˙œ œ œ œ œœ
œ œ ‰ œ œ œ œ œœ # w œ
?# œœ œœ œ œœ œœ œœ œ œœ œœ œ œ œ œœ œœ œœ
œ œ œ œ œ # œœ # œœ œœ œœ œ œ œ œ
œ œ œ œ œ
œ œ œ œ
œ œ
# œggœœœ ‰ ‰ j
œ œ œj œ œ
j œ œ œj œ œj
23
& œœ œ œ œ #˙ œœ œ œ œ œ œ œ#œ œ œ ggg œ œ œ

J
œœ œœ œœ œœ œœ œœ œœ œœ œœœ j œœ œ
?#
œ œ œ œ œ # œ œ œ œ # œ œ œ œ ggg œ ‰ Œ œ œ œ œœ
œ œ œ œ gg œ
œ
# j œœœ j j j
˙. œ œ œ œ œ œ œ œ œ œ˙ œ œ œ œj œ œ j‰ œ
28
& œ œ œJ œ œ J œœ
œœœ ˙ œœ œ œ œ œ
œœ
?# œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ
œ œ œ œ œ œ
j œœ œœ
# œ œ œ œ œ œœ œ œ œ œ œ œ ‰ ˙ œ œ œ œœ œœ œœ j œ œœ ‰ œœ
œ œ œ œ œ œ ‰ J ‰ Œ
34
& œ œ ˙ œ œ œ œ J
J J J J
f j
œ
œœ œœ œœœ
œ œœ œœ œœœ œœœ œœœ œ œ j
?# œ œ œ œ œ œ œ œ œœœ œ œœœ œ ‰ œ ‰ œj ‰ Œ
œ œ
083
Allegretto Moderato
b œ .
& b b b 42 r œ œ œ œ œ œ œ œ œœœœ ! œ œ œ œ œ œ œ !
œ . œ œ
œœœœ œ j œ œœ œœ œ œœ œœ
? b b b 42 œr œ œ œ œ
Piano Com sentimento
œ ‰ œ œ œ n œ
b œ œ
œ
b œ j
& b b b n œ œ œ œ œ œ œ œJ ! r œ œ œ œ œ œ ! r nœ œ œ œ œ
5
œ œ
œ œ œ œœ œœ œ œœ œœ œ œœ œœ œ œ œœ œœ
? bb b œ n œ œ œ œ œ œ œ n œ n œ
b œœ œœ œœ œ
bb b b œ œ œ ! œ œ œ œ œ œ œ œ œ ‰ œr œ . n œ n œ œ œ œœ œ œ œœ œœ ! œR
10
& J R R œ œ. nœ nœ œ œ œ œ J
? b b œ œ œ œœ œœ œ œœ œœ œœ œœ œ f œ
cresc.
œœ œ œ œ œ n œ œ œœ œœ ‰
bb œ œ œ œ œ œ œ œ œ
œ œ J
U
b b œœ .. œœ œ œ œ œ œ ! œr œ . œ œ œ œ œ œ œ œ œJ ‰
r
b
& b œ
15
J nœ. œ œ
œ œœ œ œœ œœ
dim.
? b b b œ œ œ œ œ œœ œ n œ œ œ j‰
b œœœ œ œ œ œ
084
%
## ‰ ‰ j 3 ‰ ‰ ‰ j j nn
rit.
n œ . j
a tempo
œ œ œ œ œ œ œ j œ
& c œœ œœ œ œ œ œ œ œ # œ œœ # œ- . # œœ n œœ œœ œœ œœ œœ œœ œ œ œ œ œ œ œj œ œ œ œ gg œœ gg œ œ œ b
œ œ g
g œ
œ œ œ œ
3
œ ˙ œ œ œ gg œ g œ
f J J J g
œ œ -
œ œ œ œ œ
#œ œ œ œ œ ‰ ‰ œ œ œ ‰ ‰ œ œœ
Piano
? ## c œ œ œ ‰ œj j œ œ œ n n b
œ œ J J œ J œ
J œ
&b j œ œ n œ # n œ˙ # œ œ œ œ˙ œ œ œ
5
j
œ˙ œ œj œ œ œj œ œ œj œ j œ œ œ œj œ œ œj œ œ œj œ j œ
# œ œ˙ # œ œ œ œ œœ
3
bœ
˙ œ nœ #œ #˙ ˙ #œ œ œ ˙ œ
J
? œœ œ œœ œœ œ œ œJ ‰ Œ
œ
œ #œ œ œ
œ
b œ œ # œ œ œJ œ œ #œ œ
œ œ œ œ œ
œ
& b # # œw œ œ œ ˙ œ œ #œ œ œ œ ˙ œ œ œ # œ œ œ œj œ œ œ
10
j
œ˙ œ œj œ œ œj œ œ œj œ # œ œ # œ œ œ jœ œ œ
3
˙ œ œ œ nœ #œ
œ
# w ˙ ˙ œ œœ œ n œ # œ
J J
? œ œœ œœ œ
œ œœ nœ œ œœ œœ œ
b œ nœ #œ #œ œ nœ œ œ œ j œ œ œ œ
#œ œ J œ œ #œ
œ œ
& b œ#j œ˙ œ œj œ œ œj # œ œ œj œ œœ ˙˙ œœ œœ œœ # œœ
j
œ œ œj œ œ œj œ œ œ œ œ . œœ œj œ ! œ !
15
œ œ œ# œœ œ œ # œ
3
œ
˙ œ œ œ
3
w œ ˙
œ. J Jœ
? œœ œœ œœ œœ œœ œœ œœ œ œ œ # œ œœ œœ
b œ œ œ œ œJ ‰ Œ œ œ
œ œ œ
# j
&b œ ‰ ‰ # œ n # œ œœ œœ œœ œœ œœ œœ œœ œ œœ œœ œœ œœ œ œ j
œœ ‰ ‰ œ œ œ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ ‰ ‰ œœ œœ
20
œ œœœ œœœ œ œœ
j p
? œœœ œœ œœ ‰ Œ # œœ œœ œ œœ œœ œ œ œ œœ œœ
b œ n# œ œ œ œ œ
œ
J
## œ œ œ œ œ œ œ œ n œ . œ œ
œ œ j‰ ‰ j 3 j
‰ ‰
j 3
‰‰
. œ œ œ œœœœ œ œ œœœ œœœ œ
25
& œ œ œ œ œ œ # œ œ # œ- # œ n œ œ œ œœ œœ œœ œ œœœ œ œ j œ œ œœ œœ œœ œœ œœ œœ œœ œœ œ
œ ˙ œ ˙ œœ
J J J
- J
œ œ œ œ œœ œœ ‰
? ## œ œ œ
#œ œ œ œ œ ‰ ‰ œ
œ J œ J ‰ œj œ œœ œ œ œj ‰ œ œœ œ œœ
J œ
fi
## j œ œ œ œ œ œ œ nœ. j‰ ‰ j
œœ œœ œœ œœ œœ œœj ‰ ‰ œ œ œ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ ‰ ‰ œœ œœ œ œ œ œ œ œ # œ œœ # œ- . # œœ n œœ œœ œœ œœ œœ œœ œ
30
&
œ
J
-
? ## œ œœ œœ œ œ œ œ œœ œœ œ œœ œœ œ œ œ œ œœ # œ œ œ œ œœ œœ ‰ ‰ œ
œ œ œ J J
j
‰ ‰ j 3 ‰ j j fi 3 ‰ ‰ j œœœ gg œœœ ‰ Œ
#
3
& # œ˙ œ œ œ œ œœj œœ œ œ œ œ g œœ ggg œœœœ œ œ œ œ œ œ œ œ œj œœ ˙ œ œ ggg œ

35
3
J œ g œ g
gg œ g œ œ œ ˙ œ J
J g J
œœ œœ ‰ ‰ œ œœ œœ œœ ‰ ‰ œ œœ j
? ## ‰ j œ œ g œ ‰ Œ
œ J œ œ j œ œ J œ œ ggg œ
J œ J gg œ
% fi
085
œ œ œ œ œ.
œœ œ œ
œ œ‰Œ ˙˙ % b
2
&b 4
J
J J n ˙˙˙ nb b œ . œ œ œ. œœ
F 3
p
œœ œœ œœ œœœ œœœ œœ œœ œœ œœ œœ
Piano
? 2 ‰Œ œœœ b œœœ œ œ
b 4 œ #œ œ œ œ œ #œ œ œ œ n œ nb b œ œ
œ
b œœœ nœ œ œ. œnœ j œ œœ œ
&bb nœ. ˙ œ. œ œ œ n œ œJ œ
7
˙ J J J
3
œ œœ œœ œœ
? b œ # n œœ œœ œœ œ # n œœ œ n n œœœ œœœ œœœ œ n œœœ œœœ
œœœœœ œ‰Œ
j œœœ
œ œœœ œ œ œ
bb œ #œ œ nœ œ œ œ bœ œ
b œ œ œ œ œ œ œ. œ œ œ œ œ nœ.
& b b œ œ œ œJ œJ œ J œ œ œ œ œ œ œ œ œ . œ.
15
‰̇ œ . œ .
‰̇ n œ .
F p 3
3 3
? b œ œ œœ œ œ œœœ œ œ œ nœ œ œ
œ
b b œ œœœ œ œ œ
œœœ œ œ œ œ œ œ œ œ œ n œ n œ œ œ œ œ œ œ œ œ # œœœ œ
œ œ œ
œ œ œ œ
b j j
&bb ˙ œ. nœ œ œ. œnœ j œ œJ œ œ œ . œœ
24
˙ ˙ œ œ
œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? b œ # n œœ œ n n œœœ œœœ œœœ œ n œœœ œœœ
œœœœ œ n œ œ œœœ
bb #œ œ nœ œ œ œœ œ œ
œ œ œ
b œ œ œ œ œœœœ j
&bb n œ œ œ œ œ n œ ˙˙ œœ n œ n n n b .. œ œ œ œ . œ
32
n˙ œ œœ œ œ œ œœœ œ œ
3 3
œ œ œœœ œœ œœ œœ œœ œœ œœ œœ œœ
? b œ œ œ œ œ # n œœ œœ œœ n n œœœ œœœ œ œœœœ œ œ œ œ œ
bb œ nœ bœ œ j‰ Œ n n n b .. œ œ œ œ œ
œ nœ œ
œ œ œ œ. j
&b œ œœ œ œ œ œ œœœ œ œœœœ ˙ œ œ œ ˙ œ œœ
40
œ J œ
œ 3
? œœ œœ œœ œœ œœ œœ œœ œœ
œ œ œ œ j ‰ j ‰ # œœ œœ œœ # œœ œœ œœœ œœœ œœœ
b j‰ Œ œ œ œ œ œ j‰ Œ
œ œ œ œ œ # œ œ œ
œ œ œ bœ œ œ œ
fi
œ j j % fi j j œ
& b œ œ œ. œ œ j j œ .
. œ ‰ œœœ
œ œœ ‰ Œ
49 1. 2.
J œ ‰̇ œ ‰ œ œœ œ œ œ n n ˙˙˙ œ
œ œ œ ‰ Œ n˙ J
Jf S
œœ œœ œœ œœ œœ œ S j
3
? œ #œ œ œ œ œ œ j ‰ Œ .. œ ‰ œj ‰ œ ‰ Œ
b œ œ œ œ #œ œ œ nœ nœ J œ
086
j
œ œ œ œ œ œ œ œ œ œ œ œ œ œ n œ %
Moderato
œ œ œ œ # œ j #
rit.
2 œ œ œ #œ œ #œ œ œ œ #œ œ #œ œ œ œ
&b 4 ‰ # œ œ nœ ‰ ‰ œ n #
œœ J
f 3 3
œœ œœ # œœœ œœœ œœœ œœœ

3 3
œœ œœ œœ œœ
Piano
? 2 Ó œœ œ œ œ œ œ œ‰Œ œ œ œ œ œ ‰ Œ n##
b 4 œ œ œ œ œ J œ œ #œ œ
## jU j jU
a tempo
œ œ œ œ œ œ j j j‰ ‰ œ œ œ œ œ œ
œ œ j
10
& œ j œ
J œ œ œ œ œ ˙ œ J
U U
? # # œ œœ œœ œ œœ n b œœœ œœœ œœœ œœœ œœœ œœœ œœœ œ œœ œœœ œ œ œ œ œ œ œœ œœ œ œœ n b œœœ
œ œ
## j j
œ ‰ bœ œ œ œ nœ œ
œ œ j œ œ n œj œ œ œ œ œ
20
& œ j œ
3
œ œ œ œ œ ˙
? # # œ œœœ œœœ œ œœœ œœœ œ œœœ œœœ œ œœ œœœ œ œœ œœ œœ œœ œ b n œœœ œœœ b n œœœ œœœ n œ n œœ œœ œœ œœ
œ œ nœ œ nœ œ
œ œ œ
##
rit.
‰
œ œ # œj œ œ œ œ n œ œ œ n œ œ œ œ j ‰ œ œ œ œ œ bœ nœ œ ‰nœ nœ œ ‰nœ
29
&
3
œ œ nœ nœ nœ
nœ œ n œ œ nœ nœ
? # # œ œ # œ œœœ œœœ # œ œœœ œœœ œ n œœ œœ œœ œœ œ œ n œ b œœ œ œœ n œ b œœ œ œœ n œ œ n œ œœ n œ œ n œ œœ
œ œ
## nn > j j j
a tempo
b œ œ ‰ œ b œ œ ‰ œ n œ œ œœ # œ ˙˙ b ‰ œ œ .. œ œ œ œ œ œ œ œ œ œ
38
& œ J J J J J J œ œ
nœ nœ ˙
nœ œ b n œœ œ œœ n œ ‰ Œ
? # # œ b œ œ œœ œ n n œ œœ œœ .. œ œœ j >œ œœ œœ œ œœ j # œ œœœ œœœ
J œœ œ b œ J >œ œ J >œ œ
j j j j j j j j j j
&b œ œ œ œ œ œ œ œ œ œj œ œ œ œ œ œj œ œ œ œ œ œj œ j j j
47
J J œ œ œ œ
? # œj œœœ œj >œ œœ œœ j œœ j > œ œ >œ œ œ

œ œœ œj œ œ œ œ œj œ j > # n œœ œœ # # œœœœ
b œ #œ œ œ œ œ œ œ J > œ nœ œ #œ œ j j
> > > œ œ
>
fi 1. j
œ n œ fi j
j j j #œ j ## j
& b œj œ œ œ œ œ œj œ œ .. œ œ œ # œ œ n œ ‰ ‰ œ n œ‰ œœ ‰ Œ
56 2.
J œœ J œ
œ œœ œœ
> >
> n # œœœ œœ n # œœœ œœ œœ œœ j
? œ œ œ œ œ œ .. œ ‰ Œ # j
b nœ # œ #œ œ n# œ œ ‰ œœ ‰ Œ
% fi > œ
>
087
Fado Marcha
# 2 j œ # œj œ œ œ œ
j œ œ œœ
& 4 œ #œ nœ œ #œ œ. œ œ œœœœ œœ
3
œœ J
3
œ #œ œ J
f
œœœ œ œ
n # œœœœœœœœœ # œœœ œœœ nœ œ œ œ œ
3
# œœ œœ œœ # œœ œœ
Piano
?# 2 Œ ‰ #œ œ œœœ œœ œ
4 #œ #œ œ œ œ
œ œ œ œ
# j œjœ œœ %n j
‰ ‰ œ ‰ Œ b b œ. j
œ œ . œ .
7
& œ œœ œ œœœœ œ œœœ ˙ œ #œ

3 3
J œ œ œ œ. œœœ
ß ß p œœœ
? # j ‰ œ ‰ j ‰ Œ n b œ œ œ œ œ œœ œœ œ œœ œœ œœ
j
œ
œœ œœ # œœ œœ œœ
œ œ œ # œœ œ œœ œœ œœ œœ œ
œ œ œ b œ œ œ œ œ
œ
b j 3 j
&b j œ # œj œ œ œ œ œ. œ
16
j j jœ œ œ œ œ œ . j j
3
˙ ˙ œ . ˙
? b œ œ œ j ‰ Œ ˙˙
œœœ œ œ ˙ œœ œœ œœ œ œ œœ œœ œœ
b œ œ ˙ œ
# œœ œœ œœ # œœ œœ ‰ Œ
œ #œ œ j
# ˙˙
œ œ œ œ œ œ œ œ #œ œ œ
œ
b j j
& b œ œ œ œ œ. œ. j œj œ œ œJ œ œ œ œ
26
œœœ ˙ œ #œ
3
œœœ ˙ ˙ œ œ J
œœœ œœ œœ œ œ œ
? b œœ œœ œ œ œ œ œ œ œ #œ œ œ œ
œ œ œ œ œ œ
œ # œ œ œ œ œ œ œ œ œ n œ œ œœ œœ œœ œœ œ œ œ
b œ œ œ œ œ œ œ
œ
b j j- # 3
& b œ œ œ œ œ #œ
œ œ œ œ œ # œ ‰̇ œ ‰ œ œœ œ j n n .. œ œ œ j œ œj œ œ œ œ . j
36
œ œ œj
3
nœ œ nœ #œ œ. œ
J
œ œ œ œ œœ œœ œœ œ
? b œœ œœ # œ œ œ # œ œ n œ œ j‰ Œ # œœ œœ œœ œœ
b œ œ œ #œ nœ n n .. œ œ œ œ œ œ
œ œ œ œ œ œ
# œ œ #œ œ œ œ œ
j nœ œ œ . œ œ
45
& œ œ œ œ. œ œ œ œ œ œ œ œ. j œœ
3
œ œ #œ œ œ #œ œ J
3 3
œœ œ # œ œ œ œ # œ œ n # œœœœœœœœœ # œœœ œœœ
3
?# œœ œœ œ œ œ œ œ
œ œ œ #œ nœ #œ œ œ #œ
œ œ œ #œ œ œ
# j >œ %
. œ j ‰ œœ ‰ œœœ ‰ Œ
œœœœ j j . œ œ œ
53 1. 3 2. 3
& œœ œ œ œ œ. j œ œœ œœ J œ œ œ œœ
3
œœ œ œ œ œ #œ > J
f ß ß
nœ œ œ œ œ œœ œœ œœ œœœ œ œ œ j
?# œ œœœ œ œœ œ j ‰ Œ . œ œœœ œ œ œ j‰ Œ
œ œ œ . œ J‰œ ‰ œ
œ œ œ
œ
088
% j j
accel. rall.
& b 44 Ó Œ œ
..
œ ˙ œ œ œ œ œ . œ ˙ . œ
œ
œ œ œ.
œ cresc. œ ˙
œ #w ˙ ˙. #˙
>
w
!
œ j>œ œ j>
œ œ œ f
œ œœ œœ œœ œ œ
œ œ œ
Piano
? 4 .. œ œ œ œ œ œœ
œ #œ œœ œ
# œ # œ
b 4
Í œ œ œ
‰ j
rall.
œ œ
a tempo
˙ œœœ ˙ œ œœœ # n œœ n b ˙˙ œ œ # œ
rall.
&b w œœ œ . j œ. j
6
œ w œœ ˙ œ#˙ œ ˙ œœ Fim
p ˙ p - >
œ jœ
œ fœ S œ
# œ Œ Œ # œœ Œ˙ n œœ
dim.
? œ œ #œ
œ œ jœ
œ œœ œ œ œ !
œœ
b œ œ œ œ #œ
œ œ œ œœ ˙
œ œ œœ >
a tempo
œ œ œ #œ œ œ œ nœ œ œ œ œ j j œ
‰ œœ œ . œ œ œ œ œ œ œ œ œ œ œ # œ
accel. rall. a tempo
&b œœ œ œ œ œ ˙.
11
pŒ œ Œ n œœ Œ œ Œ # œ p
œ œ
?b ˙ œ #˙ œ ˙ œ ˙ œœ œ œ œ œ œ œ œœ œ œœœ œœœ œœœ œ œœœ œœœ œœœ
œ œ œ
> >
œ
U a tempo # # ˙˙ ˙ ˙˙ ˙ ˙˙ ˙ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ
rit.
&b # œ œ œ œœœœ Œ n
16
˙ œ
p œœ S œœ
œ œ #œ œ œ œ œ œ œ
? b #œ œœ œœ g œ
ggg œ œ # œ œ œ n # # œ œ œ œ œ œ œ œ œ œ œ œ œ œ
gg œ œ œ œ
>
# # ˙˙ ˙ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ ˙˙˙ ˙˙˙ b ˙˙
˙ b ˙ ˙
˙˙
˙ n ˙˙˙ ˙˙˙
˙ ˙ ˙ ˙
20
&
S œœ ƒ œœ p œœ œœœ œœ œœ n œœœ œœ
œ œ œ œ
œ œ œ œ Œ b œ
œ œ œ
œ Œ
? ## œ œ œ œœ nœ œ
œ œ w w
U n ˙˙ U˙˙
> œ œ
œ .
. œ œ b œ n œ b
1.œ . œ œ
œ ˙˙ œœ
# # # ˙˙˙ ˙˙˙ n ˙˙˙ .. œ . œœœ œ . # œ œ œ n œ œ . œ œ œ œ œ n ˙˙ ˙˙˙
œ œ œ n ˙
˙ œœ ‰ .
. œ ˙. œ œ œ b œ b œ . œ œœ œ œ .
24
& #˙ ˙ J
jp cresc. J J J
U J
f dim. p j !j œ
SŒ œ œœ œ ƒŒ œ œœ œœœ ‰ Œ n œœ Œ œœ Œ b œœœ Œ œœœ Œ n œœ UŒ œœ ‰ n œœ œœ œœ ‰!Œ
? # # n ˙ # œœ œ œœ œœ œ œ # ˙ œ ˙ œ ˙ ˙ œ ˙ œ œ œ œ œ œ
˙ ˙ ‰ ˙ ˙ ˙ ˙ œ ‰ œ ‰ œ ‰ Œ ..
œ œ U
œœ ˙ œ D.S. œ ao Fim
J œ n ˙
˙ œ
œ œ ˙ œ
b œ
2..
# # b œœ .. œœ œ œœ œ œ n ˙˙ œ œ œœœ œ n ˙˙˙ œœœ J
œ œœ œ œ J ‰ !
30
&
J j j œ
œ f ƒœ p !
œ œœœœ Œ˙ n œœœ UŒ ‰ n œœœ œœœ œœ ‰ !Œ
? # # Œ˙ b œ
Œ œœ œ
˙ œ
˙ ˙
˙ ˙
˙ œ ‰ œœ ‰ œ ‰ Œ
œ
J
!
089
& 42 œj .. œ ! œ œ ‰ # œj œ # œ œ œ ! œ œ ‰ œj œ !
œ œœ p œ œœœ
? 42 ‰ .. œ œ œœœ œ œ œ # œ œJ ‰ # œœœ œœœ œœœ œ œJ œœ
Piano
œ œ œ œ ‰ œ œ œ œ
œ
‰ # œj œ # œ œ ! œ ‰ œj !
6
&œ œ œ œ œ œ œ œ œœœ œ !œ œ œ œœ!œ

p
? œ œ œ # œ œJ ‰ # œœœ œœœ œœœ œœ j‰
œ œ œ œ œ œ œ œ œ œœ œ œœ
œ œ œ
œ œ œœ
‰ œj œ œ ! œ ‰ œJ œ œ œœ!œ
& j‰ œ œ œ œ !
12
œ œ œ œœœ œ !œ œ
œ
p œœ œœ œœœ œœœ
Sentimental
? œ œœ
œœ ‰ œœ œœ œœ œ œ œ œœ œœ œ œ ‰ œ œ
J œ œ œ œ
œ J
œœ œ œ œ œœ! œ œ œ œ œœ !œ œœ œ
‰ œ ‰ ‰ œ
18
& J #œ J J
œ œ œ œ pœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ pœ œ
? œ œ œ œ œ œ œ œ œ œ
œ œ œœ! œ
‰ œJ œ œ œœ !œ œ œ œ œ œ œ œœ!œ œ ‰ œJ
23
& #œ
S S
œœœ œœ œ œœœ œ œœœ œœ œœ p œœœ œœœ œœœ œœ j
?œ œ œ œ œ œ œ œœœ ‰ œ œ œ œ
J ‰ œ œœ J œ œ œœ ‰
œ œ ! œ œ œ œ œ ! œ œ1. œ œœ
& œ œ œœ!œ œ ‰ œj .. œ œ
29
p f
œ œ œœ œœ œœ
? œœ œœ œœ œ œ œœœ ‰ œ œœœ œœœ œœœ œ œ ‰ .
œ . œ œ
Fim
œœ J œ J
œ
090
% n
# # # 4 j œ. œ œ œ. œ œ œ œ œj œ œ œ œ. œ œ œ. œ. œ. œ. œ œ œ œ œgg œœœœ U‰
Allegretto Moderato
œ. œ n
j
œ ggg œ n
œ
& 4 œ. œ œ
. . œ. . . J
p f f p
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œ U
? # # # 44 Œ ‰ œ œ ggg œœ ‰ Œ n n n
Piano
gg J
œ ˙ ˙ œ‰ œ
& œ œ œJ œJ œ œ œ œ œœœœ œ œ J œœ œ œ œJ œJ œ œ œ œ œ œœœ
5
m.d. œ
? œœœ # œœ
œ œœ œœ œœ Œ˙ œœ Œ œœœ Œ œœ # œœœ œœ # œœ œœ Œ˙ œœ
œ œ œ œ œ œ ˙ ˙ œ œ œ œ œ œ œ œ
œ œ ˙ œ œ œ œ >œ # ˙ œ
& ˙. œœœ ‰ œŒ œ
11
˙˙ .. œœ œ œ˙
˙ œ
j p #œ J
? Œ˙ œœ œœœ ˙˙
˙œ
œœ ‰ Œ
œ œ ‰ Œ œ œ œ œ œ œ œ œ
œ
J
>
& œ˙ . œ wœw œ œ œ #œ œ
15
#˙ œœ œ œ˙ ˙ Œ
˙. #œ ˙ ..
? œ œ œ œ œ œ
œ œ œ
œ
œ
œ
œ
˙˙ .. œœ œ # œ œ œ œ œ œ œ ˙ . ## œ œ. j œ œ. j
œœ # œ. œ œ.
18
& ˙. œ # ˙˙ # œ n ˙˙ ˙˙ ..
œœ ‰ J J œ
J œ œ
œœ œœ
com sentimento
? œ œœœ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ ### œ œ œ œ œ œ œ œœ
œ œ #œ œ œ œ œ
### - œ œ.
œ j jœ œ ˙ œ ˙
rall. a tempo
œœ
œ œ # œ œ œ . œJ œ œ œ œ œ . œJ œJ
23
œœ
& J nœ œ J
œœ œœ œœ œœ œœ œœ œœ œ œ
? # # # œ œœ œ œ # œ œ œœ œœ
œ œ œœ œ
œ œœ œ œ œ
œ
%
### œ œ œ. œ œ œ œ œj œ œ œj œ œ. œ œ œ œ . . œ œ œ
œ
œ œ. œ œ. œ œ œ ggg œœ ‰ U
˙ œ œ œœ œ œ . œ. œ. œ. œ. .
28
& ‰œ‰œœ‰Œ gg J œ
J f p p
œœ œœ œœ œ œ œ œ œ œ
œ œœ œœ œœ œ œ œ gg œœ U‰ Œ
? ### œ ‰ œ ‰ j‰ Œ œ œ œ œœ œœ œœ œ
Fim
œ œ œ œ ggg œ
J
091
Score
# 2 œœ œ œœœ œ œ œ. œ %
Fado Marcha, Não Depressa
j nb
& 4 œœ œ œœœ œ œ œ. ‰ Œ ‰bœ œ b œ ‰ œœ j œœœœœœ j œ œ œ œ #œ œ
3 3 3
œ
3
J J œ œ œ. œ œ.
f jF j
Œ ‰œ ˙ Œ œ œ œ #œ
œœœ œœœ œ ‰Œ
Piano
?# 2 ‰ œœœ nb ! n œ
4 œœœœ œœœœ œ b œ œ œ œ œ œ
œœœ >> > œ >> >> > œ
> >> >
j gg# ˙˙˙
bb
œ œ. ˙ œ œ œ œ œ ‰#œ œ b˙ œ œ # œ œ œ œ # ˙˙
j gg ˙
10
& œ g˙
3 3
œ œ
3 3
J J ˙ œœ œœ œ.
j ˙
n˙ œ‰Œ Œ œœ Œ œœ ˙ œ œœ
? b # œœ n œœ # œœ̇ œ n œ # œ # œ ! œ œ œ œ œ œ œ̇
b J‰Œ
œ #œ œ œ œœ œœœœ
> œ œ œœœ
#œ
b j jœ œ ‰
& b œj œ œ œ œ ‰ n œ œ j œœœœœœ j œ bœ œ œ œ œ ˙ œ J œ˙ œ # œ œ œJ œ .
20 3 3
˙
3
œ œ. œ œ.
3
˙
œ
? b Œ Œ œœ ˙ Œ œ œ n˙ ˙ œ̇œ
b œ !
œ œ œ œ œ nœ
œ œ œ nœ œ ‰ Œœ̇ œœ !
œ œ
J ‰ Œ œ nœ œ œ >
œ œ œ nœ J œ
>> > >
" nœ œ œ j j
œ œ œ œ j ‰ œ œœ ‰œ œœ
b œ #
n .
& b œ œ œ œ œ œ jœ . œ œ œ n . œ̇ œ œ œ œ œ œ ˙˙
31 3 3
#œ #œ œ œ œ ˙˙
n˙
œœ
œ nœ œ œ ˙ ˙˙
n˙
? b œ #œ n œ œ n n # ..
b œ œ nœ œ # œ œ œ œ œ
œ œ œ œ > nœ œ œ œ
œ > > œ œ œ
j j b˙
rall.
˙ ‰ œ œ œ œ #œ nœ œ œ œ œ
j
œœ
#
œ
U #U #
a tempo
& œ œ œ œ œ œ œ ˙˙˙ œœ œ œ œ ˙˙ œ œ n œ b œ œ # œ j n œj
39
˙ œœœœœ œ n ˙˙˙ ˙ œ œ œ œœœœ

œ---
J
œ #œ 3 œœ j gg œœ gg # œ
3
?#
3
œ œ œ b˙
3
œ œ œ œ œ œ œœ œœ J‰Œ ‰œ œ œ Œ
J œ
gg œ gg # œœ
gg g
œ œ œ œ œ
j j
œ œ œ fi
"œ œ > > > 2. % fi
# "œœœ j jœ œ œ . j j œœ
j >œ
œœ ‰ Œ
œ œ œœ œ œ œ . œ œ. œ ‰ Œ œ œ
1.
œ œ œ ‰ œœ . œ œ œ
48
& œœ œœ œ
œ
J J œ œ > J
œ œ j
? # n n œœ œ œ
gg œ
gg œ œ œ œ .. ggg œ œ œ œ œ œ œ œ ‰ œj ‰ j‰ Œ
3
œ œ ggg œ œ œ œ g
g œ œ œ œ œ œ
œ g > >œ g J >œ >œ œ > >œ >œ > > œ
> >
092
œ œœœœ œ
œœœ œ œ œœœœ œ œ œ œ œ . œ j
Allº Moderato
2 j œ œœ œœ
b
& 4 ‰ œ œ œ . œ œ œœ ‰ ‰ œ ‰ œ.
œ œ œ œ œœ J
œ œ œ œ œœ f p
œœœ œ œœ j
3 3
? 2 œ œœ œ œœ œ œœ
ligeiro
j œ ‰ ‰ œj œ œ œ œ œ œ œ œ
Piano
b 4 œ œ œ # œ œ œ .. œJ ‰ ‰ œ œ œ œ
œ œœœœœ
œœ œ œ
& b œœœ œ œ œ œœœœ œ œ œ œ œ œ œ œ nœ œ œ œ nœ œ #œ œ œ nœ œ œ
9
œ. J œ
3
œ œœ œœ œ œ
3
? œœ œœ œ œ œ œ œ œ œ œ œœ œ œ
œ œ # œ œ œœ
œœ
b œ œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œ
œ œ œ œ œ
œ œ œ œ œ nœ
&b œ œ. œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œ œ
œ .
18
œ œ œ œ œ œ
3
œ
? b n œ # œœœ œœœ œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œœ œœ œœ
œœ Œ œ œ œ œ œ œ œœ œ œœ œ œ Œ
œ œ œ
œ
&b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
26
œ œ œ œ œ œœ œ œ œ œœœœ œ œ œ
3
œ œ
3
? œ œ œ œ œ œ œ œ œœœ œ œ œœ œœ œœ œœ œœ œœ œ œ Œ
b œ œ œ œ
œ œ œ œ #œ œ
œ
b œ œ œ œ œ
34
& œœ œœœ œ œœ œœœ œ œœ œœ œ œ œ œ œ œ

3
œ
œœ œ œ œ œ œ œ œ œœ œ œ œ œ
? œ œœ œœ œ œ œ
œ œ œ œ
œœ œœ œœ œœ œœœ Œ œ œœ œœ œ œœ œ œœ
b œ œ œ
œ œ œœœ œœ œ œ œ œ œ œ œœœ œœ œ œ œ œ œ
œ œ œ œœ œ œœ œœ œ œœ
& b œœ œœœ œ œ œœ œ œœœœ œ
42
3
œ œ œ
? œ œœ œœ œœ œ œœ œœ œœ Œ œ œœ œœ œ œœ œ œœ œ œœ œœ œ œœ œ œœ
œœ
œ œœœ œ œ œ œ œ œ œœœœ œ œ
œ œœœœ œ œ œ œ œ. œ œ œ œ
b œ œœ œ œ œ œœ œ œ œœœœœ œ œœ œ œ œ
œ
J ‰Œ
œ œ.
50
&
ƒ
œœ œœ œœ œ œœ œœ
3 3
œ œ œ œ œ œ œ j
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ
œ œ œ œ # œ œ œ .. œJ ‰ Œ œ ‰Œ
b œ
093
b œ œj œ œ œ œj œ œ œj œ œ œ œj œ ˙
4 œ œ œ œ nœ œ
j j
nœ
œ œœ œ œœ
j
b
& 4
œ
J ‰ ‰ œœ
3
œ
f
œ œ œ
œ ‰ œœœ ‰ œj ‰ œœœ ‰ œ ‰ œœœ ‰ œj ‰ œœœ ‰ œ ‰ œœœ ‰ œj ‰ œœœ ‰ œ ‰ gggg œœœ ‰ gggg œœœ ‰ gggg œœœ œ œ œ
Piano
? 4 ‰ Œ
b 4 œ J œ J œ J œ g gJ g
Œ ‰ j œj ‰ Œ Œ
3
œ ˙.
& b w œ œ œ œ n œ œ .. œ‰ œœ œ œ
J œ œ œ œ nœ œ ˙ œ œ œ w œ œ œ œ #œ œ
œœœ ˙ ‰œ
5
J
S p p . . p
œœ œœ œœ œ œœ œœ œ œ œ œ œ #œ œ œ
? œ ‰ œ ‰ œ ‰ ‰ œ .. ‰ œ ‰ œ œ œ ‰ œœœ ‰ œj ‰ œœœ ‰ œ ‰ œœ ‰ œ œ œj ‰ œœ ‰ œ ‰ œJ ‰ œ ‰ œœ ‰ œœ ‰ œœ ‰
b œ œ œœ J J œ J œ. œ œ J J J J J J
œ J f .
J
j f. . Œ . .
& b œ. œ œ. œ œ ˙ œ œ œ œ œ œ nœ œ œ ˙. œ‰
11
œ œ œœ w œ œ œ œ nœ œ ˙œ œ œ œ
S p
#œ œ œ œ. œ œ œ œ. œ
? œ ‰ œœ ‰ j ‰ œœ ‰ œ ‰ œœ ‰ œ œ œ œ œ œ ‰ œœ ‰ œœ ‰ ‰ œœ œ ‰ œœ ‰ œœ œ œ. œ. œ ‰ œœœ ‰ œj ‰ œœœ ‰ œ ‰ œœ ‰ œ œ
b J J œ J J J œ œ J J œ œœ J œ J œ œ
. . œ œ J
J J
Œ j ‰ j
& b w œ œ œ œ #œ œ œ œ œ ˙ ‰ œ œ. ˙ . ˙ œ œ . œ œJ œ œ œ œ œ ˙ ‰ œj
17
J œ œ. œ œ J J
œ œ #œ œ œ #œ œ œ œ j œ œ œ œ œ
? œj ‰ œœ ‰ œ ‰ œJ ‰ œ ‰ œœ ‰ œœ ‰ œœ ‰ œ ‰ œœ ‰ j ‰ œœ ‰ œ ‰ œœ ‰ œœ ‰ œ œ ‰ œœ ‰ œœ ‰ œœ ‰ œ ‰ œJ ‰ œ ‰ œœ ‰
b J J J J J J J J œ J J J J J J J J
œ œ. œ
j j j # # ‰ . œ œ œ . œ œ œ œœ . œ œ œ œ œ œ œj ‰ Œ
b œ . œ œ n œ œ œ œ œ œ ˙ œ œ #œ œ
23
& œ œ #œ œ ˙ œ #œ
> >
#œ œ œ œ œ œ œ œ œ
? œ ‰ œœ ‰ j ‰ œœ ‰ œj ‰ # œœ ‰ œœ ‰ Œ # # œj ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œœ ‰ œœ ‰
b J J œ J J J n J J J J J J J
œ œ
j j
‰ . œ œ
œ œ . œ œ œœ œ œ œ œ œ œ œ ‰ Œ ‰ . œ œ œ . œ œ œ œ œ #œ œ
3
## œ
n œ œ # œ œ œ œ n œ œ œJ œR # œ œ œ œ œ œ
3
& œ œ œ œ
œ œ œ œ œ œ œ #œ œ
27
‰
œ œ œ œ
œ œ
? # # œ ‰ œœ ‰ j ‰ œœ ‰ œ ‰ œœ ‰ œœ ‰ œœ ‰ # œ ‰ œœœ ‰ j ‰ # œœœœ ‰ œœœ U‰ Œ Ó
J J œ J J J J J J J œ J J
. . . . j
œ ˙ œ ‰ Œ
# # ‰œ œœ œœ œœ œœ ‰ œœ œœ œœ œ ˙ œ œ œj œ
œ #œ œ nœ œ #œ œ œ œ nœ œ J ‰ #œ œ œ œ œ
31
& œ œ œ œ œ œ ˙ œ
œ
œ œ œ œ f œœœ #œ œ
? # # j ‰ œœ ‰ j ‰ œœJ ‰ œ ‰ œœ ‰ œœ ‰ Œ # œ ‰ ‰ j ‰ œœœ ‰ œœœ ‰ Œ Ó
œ J œ J J J J J œ J J
## œ nnb Œ j j
œ œ œ œ n œ œ .. n # # j ‰ œœœ ‰œœ ‰ Œ
35 1. 2.
& ˙˙ œ nœ œœw œ œ
œ œ > >
f
œ œ œ. œ œ œ
? # # j ‰ œœ ‰ j ‰ œœ ‰ ˙˙ œ nœ n n œ œ œ œ ‰ œœ ‰ œœ ‰ ‰ œœ .. # # œœœ ‰ j j
œ J œ J œ nœ
b J J n J œ ‰ œœ ‰ Œ
>
094
œ œœ œ œœœœœ œ œœœ
### œœœ œœœ !œœœ œ œ !œœœœ œ œœœ œ
c œ œ
& œ œ. œ œ. œ œ. œ
F
? ### c ‰ ! œ œ œ œ œ œ œ
Piano
œ œ
###
œ œ. œ œ. œ œ. œ œ. œ œ. œ œ. œ œ. j‰
œ œ. œ œ. œ œ. œ
5
& œœ. œœ. œ œœœ œ ˙.

3
œ w œ œ.
œ œ œ œœ œœ œœ œ œ œ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? # # # œ œœ œœ œœ œ œ œ œ œœ œœ œœ œ
œ œ œ œ œ œ
œ
œ œ œ
œ œ
## nn
œ œ œœœœ œ œœœ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œœœ
mosso
& # œ. œ œ. œ œ. œ œ. œ n ˙ œœœœ
11
œ #œ œ #œ œ œ œ
œ
œ œ œ œœ œœ œœ œ œœ œœ œœ œœ œœ œ œ œ
? # # # œ œœ œœ œœ nnn œ œ œ œ œœ œ œ œ #œ œ œ # œœ œœ œœ
œ œ
%
&œœœ ! œ ˙ œ #œ œ œ ˙ œ ‰ œ. œ #œ œ. œ ‰ œœ
16
œ œ #œ J œ œ œ œœœ œ ˙ ˙
3
œœ J œ
œœ œœ œœ œ œ œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
menos
? œ œ œ œ œœ œœ œœ œ #œ œ œ œ #œ œ œ œ œ œ œ œ œ œ
œ œ
rU
rit. a tempo apress.
& œ ‰ œœœ ‰ œ ‰ œœœœ ! œ ˙ œ #œ œ œ œ œ œ œ œ œ œ #œ œ œ œ

œ #œ œ œ #œ œ
22
œ ˙ œ œ #œ œ œœœ
œ œ œœœ œœœ œœœ œ # œœ œœ œœ œœœ œœœ œœœ œ œ œ œ œ œ
? œ œœ œ # œœ œ # œœ œœ œœ # œœ œœ œœ
œ œ œ
œ U! ###
a tempo
‰ œj œ j j ‰ œj œ j j
œ œ œ œ œ œ œ œ #œ œ
28
&œ œ ˙ œ œœ œ
˙
œœ œœ œœ œœ œœ œœ œœ œœ œœœ œœ œœ œœ œœ œœœ œœœ
più vivo
? œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ #œ ###
œ
## œ œ #œ œ œ œ œ œ
& # œœ‰ œ œ œ œ œ œ œ œ œ œ œ œ
33
œ œ œ œ œ œ œ nœ
menos
? ### œ œ œ œ œ œ œ
œ
œ
## œ œ #œ œ œ #˙ œ. œ œ œ U %
rall. a tempo
& # .. n œ œ œ n œ œ œ œ œ œ œ œ # œ œ .
J œ œ œ œ œ œ œ nœ . œ œ œ ! œ œ œ
37 1. 2.
n œœ œœ œœ # œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
vivo Fim
? # # # .. n œ œ œ œ œ œ œ œ œ œ œ œ œ œ .. œ œ œ
œ œ œ œ
095
Moderato
& 42 ! œ œ œ œ œ œ œ .. œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ ‰ œj œ œ œœœ
œ
œ œ œ œ œœœ œœ œ
3
.. œ œ # œ œ #œ œ
Piano
? 42
& œJ œ œJ œ œ œ œ œ œ œ œ œj œ œ œœœ j œœ œ œ
œ. œ œ œ œ œ œ œ œ œ œ œ J œ J
œ
7 3 3 3
œœœ œœ œ œœ œ œœ œœ œœ œ œ œ œ
œ œ
œ œœœ œœ œœ œœ
? œœ œœ œœ œœ œœ œœ œ œœ œœ œœ œ œ œ œ œ œ œ œ œ œ
3
œœ œ œ # œ œ œ œ œ œ œ # œ
œœœœ œ œœœ
œ œ œ œ
& œ œ œœœ œœœ j j # œ œ œ œ œ œ œ œ œ œ .. # # # œ ‰

œ œ œ œ œ œ œ. j
16 3 3 1. 2.
œ œ œ œ œ
œœ œ # œœœ œœœ œœœ œœ œ œ œ .. # # # œ œ œ œ œ œ
? œœœ œœœ œœœ œ œœœ œœœ œœœ œ œœ œœ œœ œ # œœ œœ œœ
3
œ œ œ œ
œ
### œ j j
j j j œ œ œ œ œ
œœ œœ œœ œ œ œ œœ
23
& œ œ œ œ œ œ œœ
œ œ œ œ œ œ œ œœ œœ œœ œœ œœ .. œ
3 3
œœœ œœœœœœœ
3
? ### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
### œ œ œ œ j j j
3
œ œ # œœ œ œ œœ .. j
29
& œ œ œ œ œ œ œ œœ
œ œ œ œ œ œ œ œœ œœ œœ œœ œœ .. œœ
3 3
? ### œœœ œ œœœœ œœœ œœœœ œœœ œœœœ œœœ œœœœ œœœ œœœœ œœœ
œ œœœ œ œ œ œ œ
### œ j j
œ œ œ œ œ œ œ #œ œ œ nœ œ œ œ
œ n œ œ œ œ œ œ œ œ n œ œ œœœ Œ
3
œœ œœ œ œ œ œœ
35
& œ œ œ œ œ œ œœ
œœœ œœœœœœœ œœœ œ œ œ œ œ

3
? ### œ œ œ œ œ œ
œ
œ œ œ nœ œ œ Œ
œ œ œœ
096
## œœœ œ œ% œ œ œ œ œ œ œ œ .
œœ ! œœ œœœ !œœ
& c ! ! ! ! œ œ œ œ !œ œ !
p
œœ œœ j œ œ
j œœœ ‰ œœœ ‰
Piano
? ## c ! Œ œ ‰ œ ‰ j œ ‰ œ ‰
œ J œ ‰ J œ J J Œ
# # œ ! œ œ œ ! œ œ œ ! œ œ œ # œ œ n ˙˙˙ œœœ œ œ œ ! œ œ œ ! œ œ œ œ œ œ œ œ œ œ
! ! ! œ
3
&
3 3
? ## j ‰ œ j œ j œ œ j œ‰ j œ‰
œ J‰ œ ‰ J‰ œ ‰ J ‰J‰ Œ œ ‰ J œ ‰ J
œ œ
œ œ
# # œJ ‰ œœœœ œœœœ œ œ œœ œ œ œ œœ ‰ œj ‰ ˙˙
j
œœ ‰ Œ nnb ˙ ˙
œ œœ ˙˙ œœœœ ˙˙ ˙˙
6
& ˙ œ
J
œœ œœ œœ j ‰Œ p
œ ‰ œ ‰ œœ ‰ n n œ ‰ œœ ‰ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰
Fim
? ## j‰ ‰ ‰ Œ œ œ ‰ œ œ j
œ J J œ œ œJ ˙ œ
b œ
J
j‰ ˙˙ ˙
&b œ œ œ ˙˙
# ˙ œ ˙˙˙ œœœ œœœ œ œ œ . œ ˙˙ ˙˙ #˙
œ œ œ ˙˙
œ
11 3 3 3 3
˙
J
?b œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ œœ ‰
j‰ n ˙ ˙ #˙ œ # ww
&b ˙˙ œœ œœ œœœœ # ˙˙ ˙˙ œ œ œ ˙˙
16
˙
3 3
˙ œ œ w
J
n œœ œœ œœ œ
?b œ ‰ œœ ‰ œ ‰ œœ ‰ œ ‰ ‰ œ ‰ ‰ œ ‰ ‰ # œ ‰ œ ‰ œ ‰ œœ ‰ ‰ œœ ‰ œ ‰ # œœ ‰ œœ ‰ œœ ‰
œ
> > j nœ
# ˙˙ ˙ œ œ˙ œ . # œj ˙˙ œœ ! œ œ ! œ œ
œ œ œ œ # ˙˙˙
3
& b n ˙˙ ˙ ˙˙
21
˙ #˙ œ ‰
J
œ œ
? b ‰nœ ‰ œ ‰ œ ‰ #
œœ
œ
œœ
œ‰ # œœœœ œ
œœœ œœ œœ
œ‰ œ‰
# œ ‰
D.S. ao fim
œ œ‰ ‰œ‰ œ‰ ‰œ‰ ‰
œ J Œ
097
# # c j œ œ . ˙˙ œœ œ œ œ ˙˙ œœ n œœ œœ œœ œœ œ # œ œ œ œ œœ œ œ
Allegretto
& œ .˙ œ ˙ œ œ œœ œ œ œœ œœ
œ œ œ œ œœ œœ œ œœ
œœ œ œ œ œ œœ œ œ œ # œ œ œ œ œœœ œ œœœ œ œ œ œ
Piano
? ## c Œ ‰ .. œ œ œ #œ œ œ œ œ œœœœ
3 3
œ
# œ œ œ œ œ œj # œœ œœj œœ >œ j j
3
& # œœ œ œ œ œœœ œœ
# œœœ œœ œ Œ ‰ ‰ ‰ ‰
3
œ œ
4
œ œ # œ œ >œ >œ œ œœ œ œ
3 3
> > >> > œ œ œ œ

œœ œœ j > >œ
? ## œ œ
œ œ œ œ œœ ## œœ œœ œœ œ œ œ œ œ œ œœ œ ‰ œj œ ‰ œ œ ‰ œj
œ œ œ J
J > >œ œ œ
> >œ œ >œ
>
# œ
& # Œ œ œ œ œ #œ œ œ Ó Œ œ œ œ œ #œ œ
8
œ œ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ œ
œ œ œ œ œ œ œœœœœœ œœœœœœœ
3
# œ j‰ !
3 3
& # œ œ! œœ ‰ œœ œœœœœ œœ ! œœœ ! œœœ

11
œ œ n œ # œ œ œ
œ œ œ œ œ œ œ œ œ œœ œœ
? ## œ œ œ œJ œ œ ˙.
J ‰ œj œj œ j
œ œj œ œ
J
# Œ œ œ œ œ #œ œ œ
& # ‰ œœ œœœœœ œœ ! œ œ. ! œ œ. œ. ! œ œ œ.
14
œ œ œ œ œ œ œ œ. œ œ œ. œ œ œ œ. œ œ.
? ## ˙ . ‰ œj œ œ œ œj œ j
œ ˙. ‰ œj œj œ œj œj œ œ
j
J J
#
& # ‰ œ œœœœœ œœ ‰ ‰ ‰ œ œ œ œ n # œœ œœ œ
18
œ
œ œ œ œ b œ œ nœ œ œ œ œœœœ œ
œ œ œ
œ nœ œ
j œ œ œ
? ##
˙. ‰ œj b œ . œ nœ. œ ˙ œ
J
# U
& # œ œœœœœ œ œœœ Œ œ œ œ #œ œ nœ ˙.
21
œ œœ œ œ œ ! œ
œ œ œ œ œ œ œ. J
? ## œ œ œ œ b œ n œ n œ U! œ œ œ œ œœ œ œ œ œœ œ œ œ œ œ œ œœ
J J œ Œ œ œ œ #œ œ
œ
#
2 097
j j œ œ œ œ.
& # œ. œ œ #œ œ œ œ œ œ œœ ‰ œ
25
˙. œ. J J
œ œ œœ œ œœœ œ œ œ œ œ œ œ œ œ œ œ œœœ œ œœœ
? ## œ # œ œ œ œ œœ œ
œ œ œ œ œ œ
# œ œ. œ ˙.
& # œ œ œ œ nœ œ #œ œ œ.
29
J ˙ J
œ œ
? ## œ œ
œ œœœ # œ œ œ. œ œ œ # œ œ œ # œ . œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ #œ œ œ œ œ œ
# j
& # œ. œ œ œ #œ œ œ j
˙.
œ œ œ œ œ œ ˙ œœ.
33
œ œ œ œ œ œ œ œ œ œœ œ œ œ œœ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
# œ œJ œ œ ˙.
& # #œ Œ ‰
œœ ‰ œœœ ‰ œ œœ œ œœœœœ œ œœ œ
37
J
3 3
œ œ œ œ œ œ
œ œ œ
? ## œ # œ œ œ œ œ œ œ œ œ œ œ ‰ j œ œ œ œj œ œ
j
˙. ‰ œj
œ œ J J
#
& # ‰ ‰ ‰ œ œœœœœ œœ ‰ ‰
41
œœœ œœ n œ œ nœ œ œ œ œ œ œ
3 3
œ œ œ œ œ œ
œ œ œ œœ bœ œ œ
? # # œj œ j
œ œj œ
j
œ ˙. ‰ œj b œ . œ nœ. œ
j
J
# U œ ˙˙ œœ œ œ œ
& # ‰ œ œ œ œ n # œœ œœ œ œ ‰ œ œ .
. ˙ œ
44
œ œœœœ
œ
œ œ > > >œ œ œ œ œ œ œ !œ œ œ
f œ œ œ œ œ œ œ œœ œœœ
? ## œ œœ U œ
˙ œ œœ œ œ bœ nœ nœ ! œ œ œ Ó .. œ œ
J J 3
3
# ˙ œœ n œœ œœ œœ œœ œ # œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ j j œ >œ
& # ˙˙ œ œœœ œ œ œœ œœ œœ œœ # œœœ œœœ œœœ # œœœ œœœ œœ œœœ œ Œ
48
œ œ #œ >
œ œ œ œœ œ œ œœ œœ
œ œ œj œ œ >œ
? ## œ œ œ œ # œ œ œ œ œœ œ œœ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ Œ
œ #œ œ œ œœœœ œ œ œ œ #œ œ œ œ œ œœœ œ œ œ
3 J œœ >
3
098
^ %
œœ. œœ. œ œœ œœ œœ. œœœ. œ^ œ œœ œœœ.
C
& œ #œ œ œ œ #œ # œœœ. œ œ œ œ œœ œœ Œ Ó œ
œ œ œ. œ œ. œ œ
F p
# œœ œœ œœ œœ # œœ œœ œœ œœ œœ œœ
Piano
?C œ œ œ œ œ
Œ œ œ œ Œ œ œ œ Ó Œ œ
6
&œ œ œ œ. œ œ. œ œ œ
œ œ. œ œ. œ œ œ
p
? #œ œ # œœ œœ œœ œœ œœ œœ # œœ œœ # œœ œœ
œ œ œ œ #œ œ œ œ œ œ œ œ #œ œ
^ ^ rall. U
a tempo
œ. œ œ. œ œ œ œ. œ œ. œ œ œ œ. œ œ œ œ
12
& ˙ ˙ œ
w œ. # ˙˙ œ. œ œ. œ œ
? ww œœ œœœ œœœ œœœ œœœ œœœ ˙˙ œœ œ
œ œ œ œ œ ˙ ˙˙ œ œ œ
œ œ œ #œ œ ˙ ˙
œ
# # j jj
Œ œ ‰ œj # .. œj œ œJ œJ œ œ œJ œ œ œ œ œj
18
&œ œ. œ œ. œ œ œ
œ ˙.
? œ # œœ œœ # œœ œœ Œ # # # .. œœ œœ œœ œœ
˙˙œ .. œ œ œ œ œ œ
œ #œ œ œ œ Œ œ œ
## j ‰ j j œ œ œ œj œ œ œj œj œ œj j
& # œ œ œj œj œ œj ˙ . œ œ œj œj œ œj ˙ . ‰ œj
23
œ œ œ JJ J
œœ œœ œœ œœ œ œ œ œ œœ œ
? ### œ œ œ œ œœ œœ œœ œœ œœ
œ œ #œ œ œ œ œ nœ œ œ œ œ œ œ
œ
29
### j œ œ œ œj œ œ œj œj œ œj œj œ œj œj j ˙ ‰ œj j œ œ œ œ œ œ œj œj œ œj
& œ œ JJ J œ œ . œ œ JJ J J
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? ### œ œ œ œ œ œ œ œ œ œ œ œ n œ œ œ œ
œ œ œ œ œ œ # œ œ
### œ œ j j ^ ^ %
œ œ œ œ ˙. ‰ œj .. œ œœ Œ
35 1. 2.
& J J œœœ œ
# œœœ n œœ ˙˙ .. Œ ^
? ### œ œ œ^
Fim
œ œ œ .. Œ
œ œ œ œ œ
099
Allegretto Agitato
b j
& b b 44 ‰ œœ ‰ œœ ‰ œœ ‰ œœ ‰ œœ ‰ œœ ‰ œœ ‰ œœ .. œ . œ œ œ œ œ
œ œ nœ œ œ œ nœ œ
œœ œœ n œœ œœ
Piano
? b b 44 œ ‰ œ ‰ œ ‰ œ ‰ ‰ œ ‰ œ ‰ œ ‰ .
. œ œ œ œ œ
b J J œ œ
b j j œ œ nœ
& b b œ œ œ œ˙ œ œ œ ẇ œ œ œœœœœ œœ œ
4
˙ œœ œ
? b œ œœ œ œœ œ œœ œœ œ œœ œ œœ œ n œœ œ œœ œ œœœ œœœ n œœœ œœœ œ œœ œ œœ œ n œœœ œœœ

bb œ œ nœ œ œ
b j
& b b w œ œ œœ œœ w œ œ œ œ œ . œ œ œ œ œ œ œ œ œ œ˙ œ
8
‰ œ ‰ œ ‰ ‰ ‰ œ ‰ œ ‰nœ ‰ œ
œ œ œ œ
? b b œ ‰ œ ‰ œ ‰ œ ‰ ‰ œ ‰ œ ‰ œ ‰ œ œ œ œ œ n œœœ œœœ œ œ œ œ œ œœ œ œœ
b J J œ œ
b œ œ nœ
& b b ẇ œ œ œ œ œ œ œ œ œ œ w‰ œœ ‰ œœ ‰ œœ ‰ œœ
12
˙ œœ œ
? b œ œœ œ œœ œ n œœ œ œœ œ œœœ œœ n œœœ œœœ œœ œœ n œœœ œœœ

œ nœ œ œ œ œ ‰ œ ‰ œ ‰ œJ ‰
bb œ œ œ J
b n j j
& b b w‰ œœ ‰ œœ ‰ œ ‰ œ n n œ ‰ œ œ ! œ œ ! œ œ ‰ œ œ ! œ œ ! œ w‰ œœ ‰ œœ ‰ œ ‰ œ
16
nœ œ œ œ
? b ‰ œ ‰ œ ‰ œ ‰ n œ œœœ œ œœœ œœ œœœ
œ
œ œ
œœ œœ
œœœœ
œœ œœœ
œœ œ œ ‰ œ ‰ œ ‰ œJ ‰
bb œ nn J
2 099
‰ œj œ œ œ ‰ œj œ œ œ
20
&œ œ œ œ œ œ œ œ œ œ œ œ
œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
?œœ œ œœ œ œ œ œœ œ œ œ œ œœ œ œ œ œ œœ œ
‰ œj œ ! œ œ ! œ œ ‰ œj œ ! œ œ ! œ w œ œ œ œ œ ‰ œj œ œ œ œ
24
&
œ ‰œ‰œ‰œ‰œ
? œ œœ œ œœ œ œœœ œ œœœ œ œœ œ œœ œ œœ œ œœœ œ ‰ œ ‰ œ ‰ œ ‰ œ œ œœ œ œœœ œœœ
J J œ œ
‰ j bbb ‰ ‰ ‰
28 1.
&œ œ œ œ œ œ œ œœ œ œ œ œ œ œ œœœ œœœ n œœœ œœœ
? œ œœ œ œœ œ œœœ œœœ œ œœœ œœœ œ œœœ œœœ œ œœœ œœœ œ œœœ œœœ b œ ‰ œ ‰ œ ‰ œ ‰
œ œ œ œ œ bb J J
œ
b
& b b ‰ œœœ ‰ œœœ ‰ œœ ‰ œœ .. œ ‰ œj n œ œ œ œ œ ‰ œj n œ œ œ
32 2.
nœ œ
? b ‰ œ ‰ œ ‰ œ ‰ .. œ œœ œ œœ œ n œœœ œœœ œ œœ œ œœ œ n œœœ œœœ
bb œ œ œ
b nœ
& b b œ n œœœ ‰ œœœ ‰ œœœ ‰ œœœ n œœœ Œ œœ Œ Œ Ó
35
w œ
j
? b œ ‰ n œ ‰ œ ‰ n œj ‰ œ
œ Œ œ Œ n ww ˙ ˙ n œœœ
Œ Ó
bb nœ œ nœ
œ ˙ ˙ œ Œ Ó
100
^ ^ ^
Tempo de Fado Marcha
2 œ œ œ œ œ œ œ. œ œ œ œ œ œ
œœ œ œ œœ œ œœœœ œ œ œœœœ
3
&4
3 3 3
˙ œ œœ
f
œœ œœ œœ œœ œœ œ n œœ œœ œœ
b
3
œ œ œ œ œ œ
3
œ œœ œœ œœ œ œ œœ œœ œœ œ
Piano
? 42 Œ ‰ œ œ #œ œ bœ
œ œ
%
& b œ ‰ œj ‰ j ‰ j bbb
7
œ œœ œœ b œœ œœ œœœ œœœœ ˙ œ œ œ œ œ œ #œ ˙ œœœœ œ œœ

œ œ
p
? œj ‰ j ‰ œ œ œœœ œ œ œ œ œ œ
j ‰ œj ‰ b b b œ œ œ œ œ œ œ œ œ œ œ œ œ n œ œ œ œ œ n œ œ œ œ
œ
œ bœ
bbb œ œ œ œ œ œ.
14
& œœœœ œ œœ ˙
3
˙ œœœ œœœœ ˙ œ œ nœ œ œ
? bb œœ œœ œœ œ n œ œœ œœ œ œœ œœ œœ œ œœ œœ œœœ n œœ œœ œœ œœ œœ
b œ œ œ œ œ œœœ œ œ œ
nœ œ
b b b œ œ3 œ œ œ œ œ œ œ j n œj n n n .. œ œ œ œ .
œ.
21 3 3 3
& œœœ œ
nœ œ œ œ œ œ. œ œ œ œœœ
3 3 3 3 3
œ œ
j j jF‰ Œ
? bb œœ œœ œœ œœ œœ œœ ‰œ n œœ ‰ œœ n œœ œœœ œ
b œ œ œ n œ œ nœ n n n .. œ œ œ œ œ œ œ
œ œ œ œ œ bœ
J
j
& œ œ œ œ. œ œ œ œ. œ œ œJ œ œ œ œ . œœ œœœœ
28
œœœœœœ ˙
3 3
œ œ
3 3 3 3
œœœœ
œœœ œœ œœ œœ œœ b œœ
3
œœœ œœœ œœœ œ œœœ œœœ œ œœ œœ œœ

3 3
? œœ œœ œœ œœ œ #œ
œ œ œ œ œ j‰ Œ œ
œ œ #œ œ œ œ
-œ 3 j j %
& œœœœ œ œ- œœœœ œ œ œœœœ j œj .. j ‰ œœœ ‰ œœ ‰ Œ
36 3 3 1. 2.
œœ ˙ œ œ œ œ œœ
nœ ‰ œj ‰ œj jf‰ Œ j
? œœ Œ !
œ œ œ
œ œœ œœ œœ œœœ œœœ œœœ œ œ œœ .. j‰ œ ‰ j
œ‰Œ
œ œ œ œœœœ œ
œ œ
œ
J
001
œ œ œ œ œ œ. œ œ œ œ œ œ.
2 œ
&b 4 J œ
œ œ #œ #œ œ œ œ œ
œ œ
œ œ œ # œœ œœ œœ # œœ œœ œœ œœ œœ œœ
? b 42 ‰ œ œœ œœ œœ
Piano
œ œ œ œ #œ œ œ œ œ œ œ œ
œ œ œ. œ œ œ œ œ œ. œ
œ œ œ œ œ #œ #œ œ œ œ
&b œ
5
œ œ
œ œ œœ # œœ œœ œœ # œœ œœ œœ œœ œœ œœ
? b œ œœ œœ œ œ œ œ œ
œ œ œ œ #œ œ œ œ
œ œ œ œ œ œ #œ nœ œ œ œ œ œ œ œ œ
œ œ œ œ ! œ
&b
9
œœ œœ œœ # œœ œœ œœ # œœ œœ œœ œœ œœ œœ
? œ œ œ œ œ œ œ
b œ œ œ œ œ #œ œ œ œ
œ œ œ œ œ œ œ œœœ œ œ œ œ œ œ
&b œ #œ œ œœ j
13
œ
œœ œœ œœ # œœ œœ œœ œœ # œœ œœ
?b œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ J
002
Andante
b œ œ nœ œ œ œ œ œ œ nœ œ œ œ œ œ œœ
& b b b 42 œJ œ œ œ œ œ œ œ œ œ œ œœœ
œ œ œ œœ œœ œœ œœ œœ œœ œ œœ
? b b b 42 ‰ œ œ œ œ nœ œ œ œ œ œ œœ
Piano sentimental
b œ œ
n œ œ
bb œ œœœ œœœœœœœ
&bb œœœ œ œ œ œ nœ œ œ œ œ œ œ œ œ œ œ œ nœ œ
5
? b b œ œœ œœ œœ œ
n œ œœ œœœ œœœ
œ
œœœ œœœ œœœ œ œœ œœ œœ
bb
bb b œ œ œrK œ œ œ n œ œ œ œ œ œ œ n œ œ œrK œ œ œ œœœœœœœ

& b œœœ œ œœœ œ
9
œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œ œ œ

? bb b œ œ œ œ œ œ œ
b
bb b b œ œ œr œ œ œ œ ‰ œJ n n n n b
K
œ
œ œ œ nœ œ œ œ œ œ œ œ œ œ œ nœ œ œ
13
&
? bb œ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œ œœ œœJ ‰ n n

bb œ œ nnb
œ œ œœ œ œ œ œ œ œ œœ œ œ œ œ
&b ‰ ‰
17
J J
? œœœœœœœœ œœœœœœœœ œœœœœœœœ œœœœœœœœ

b
œ œ œœ œ œ œ œ
2 002
b œ œ œ ‰ œ œ œ œœ ‰ J
21
& J
œœœ œœœ œœœœœœœœ œœœ œœœ œœœ œœœ
œ œ œœ œ œ œ
&b œ œ œ œ œ œ œ ‰ œJ
25
œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ
? œ œ œ œ œ œ œ œ ‰
b œ J
b œ œ nœ œ œ œ œ œ œ œ œ ! œ œ œ nœ œ œ œ œ œ œ œ œ ! œ
29
&
? œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œœœ œœœ œœœ

œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœœœœœœ
&b nœ œ ! nœ œ
33
œœ œœ œœ œœ œœ œœ œœœ œœœ œœœ œœ œœ œœ

?b œ œ œ œ œ œ œ œ œ œ
œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰ œ
37
& J
? œ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œ œœ œœJ ‰

b œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ j
œ
&b œ œ œ œ œ
41
? œ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œ œœœ œœœ

b œ œ œ J
003
œ
Andante
2 j œ . œ œ œ œ œ # œ œ œ . œ œ œ œ œ œ ‰ J
&b 4 œ ‰
J
p sentimental
? b 42 ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
œ. œ œ œ œ œ #œ œ œ. œœ œœ œ œ
&b ‰ J ‰ œ
5
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ
œ
b œ. œœ œœ œ œ ‰ œ ‰ J
œ œœœ
9
& J œ œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ œ
œ. œœ œœ œ œ œ œ œ œ œ œ œ œ ‰ j
&b ‰ J œ
13
?b œ œ œ œ œ œ œ œ œœœ œœœ œ œ œ œ œ œœœœœ ‰

œ œ œ #œ œ J
j j
2 003
œ œ
&b œ œœ œ œ œ œ ‰ œ œœ œ œ œ ‰ œ
17
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ
œ œœœ œ
&b œ œ ‰ œ œ œœœ ‰ J
21
œ J œ œ
b œ J
œ œœœ œ œ œ
œ œ œ œœ œ œ œ ‰ œ
&b ‰ J J
25
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ
œ œœœ œ œ œ œ œ œœœ
J œ œ
&b ‰ ‰
29
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ
b œ J
004
Allegretto
& 42 Œ Œ ‰ ! r œ œœœœ œ œ œ œ ! œ œœœœ œ œ œ !œ

œ œ
? 42 œ œœ œ œ # œœœœ œœœ ‰ œ œ œ œ œ œ # œœ œ œ œœ œ œ # œœ œ œ œœ œ œ œ œ œ œ œ
Piano
œ œ
œ œJ
œ œœœ œœ œ œ !œ œ œ œ œ œ #œ œ œ !œ
6
&
? œ œœ œ œ œœ œ œ œ # œœ œ œ œ œœ œ œ œ œ œ # œœ œ œ œ œ j ‰
œœœœ œ œ œ œ ! œ œ œœœ œ œ œ !œ
&œ
10
? œ œœ œ œ œœ œ œ # œœ œ œ œœ œ œ # œœ œ œ œœ œ œ œœ œ œ œœ œ
œ œœœ œœ œ œ !œ œ œ œ œ j
œ #œ œ œ ‰
14
&
? œ œœ œ œ œœ œ œ œ # œœ œ œ œ œœ œ œ œœ œ œ # œœ œ œ œœ œ j ‰
œ œ œ œ œ œ
005
œ œœœ œœœœ œ œ
Andantino
2 œ œœ œ œ œJ œ œ
&4 J J J œ J
? 42 ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
œ œ œ œ œ œ
œ œ œ œ œ œJ # œ œ œ œ #œ œ œ œj œ œ
5
& J J
? œ œ œ œ œ œ
œ œ œ œ œ œ #œ œ œ œ œ œ œ œ ‰ œJ
9
& J J #œ
? j
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœ ‰
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
&œ œ
13
#œ
?
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œj
&œ œ œ œ œ œ
17
#œ
? j
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœœœ
006
Andante
& b 42 ! œ ! ! j
œœ ‰ ‰ œj
œ œœœ
œ œ
p
Piano
j j j j j
? b 42 œ ‰ œ ‰ œ ‰ ‰ œ œ ‰ Œ
œ œ œ œ œ
&b œ œ œœœ œ œ ‰
œœ j
4

b
&b œ œ œ œœœ œ œ
œœ
!œ
8
œ œœ œ œ œ œ
œœœ œœœ œœœœ œœœ œœœœ œœœ œœœ œœœ
& b bœ œ œœ œ œ ! œ œ œ œœ !œ
12
œ œ œ œ œ œ
œ œ bœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
dolce
?
b œ œ
U
&b œ œ œœ œ œ œ œ œ œ œ œ
16
œ œ œ œ œ œ
U œ œ œ œ œ œ œœœœœ
? œ œ œ œ b œ œ œ œ œ œ œœ ! Œ œ
b œ J ‰
2 006
œœ
‰ j
20
œ œœ œ œ œ
œ
œœœ œœœ œœœœœœœœ œœœœ œœœ œœœ œœœ
?b œ œ œ œ œ
œœ
!œ
24
œ œœ œ œ œ œ

b
& b bœ œ œœ œ œ !œ œ œ œ !œ
28
œ œ bœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
dolce
?
b œ œ
U j
&b œ œ œœ œ œ œ œ œ œ œ œ
32
œ œ œ œ œ
U œ œ œ œ œ œ œœœœœ
? œ œ œ œ b œ œ œ œ œ œ œœ ! Œ œ œ
b J
007
Augusto Hylario
U % j
Andante expressivo
& b 42 œ
f p
œ #œ ggg œœœ
? b 42 gggg œœœ g œœœœ
Piano
gœ gg œ ‰
œ
œ œ œ œ !
&b œ œ œ œ œ œ œ œ œ œ œ ! œ
3
œ œ
œ œ œ œœ œœ œœ # œœœ œ œœ œœ œœ
? œ œœ œœ œœ œ œ œ œ œœ œ œ œ œ
b œ œ
œ œ ! œ #œ
&b œ œ œ œ œ œ œ œ œ œ œ œ œ ! œ
7
œ œ œ œœœ œœœ œœœ

? b œ œœ œœ œœ œœœ
# œ
œœœ œ œœ œœJ ‰
œ œ
&b œ
11
œ œ œ œ œ œ œ ! œ œ œ œœ œ œ œ œ ! œ
? œ b n œœœ œœœ
œ
œœ œœ œœ œœ œ œœ œœ œœ
œ œ ‰
b nœ œ œ J
œ ! b œœ
œ œ œ œ œ
&b œ œ œ œ œ œ œ œ ! œ œ œ œ œ œ œœ
15
œ œ œ
? b n œœœ œœœ œœœ œœ œœ œœ œ œœœ gg œœœ U

b nœ œ J ‰ œ gg
J
‰ Œ
œœ b œœ œœ œœ œœ œœ ! b œœ b œœ b œœ œœ œœ œœ b œœ b œœ œœ ! œœ
2 007
b b œœ œœ
19
&
? b œ b œœœ œœœ œœœ œ œ œ b œœœ œœœ œœœ

œœ œœ œœ
dolce
œ b œœ œœ œœ œ œ œ œ œ
b
b b œœ b œœ œœ œœ œœ œœ œœ œœ ! b œœ b œœ b œœ œœ œœ œœ b œœ b œœ œœ ! œœ
23
&
œœ œœ œœ œ œ œ œœ œœ
?
b œ œ œ œ
b œ b œœ œœ œœ b œœœ œœœ œœœ œ œ
b œ ‰
œ J
b b œœ œœ œœ œœ œœ b œœ œœ ‰ œj œ b œ œ œ œ œ œ œ ! b œ
27
& bœ bœ œ œ bœ œ bœ œ œ œ
? œœ œœ œœ b œœœ œœœ œœœ œœ œœœ œœ œœ œœ

b œ œ œ œ
b œ b œ b œ œ œ œ œ
b
œ
b œœ œœ œœ œœ œœ b œœ œœ ‰ œj j %
b œ b œœ œ œ œ b œ œœ
œ ‰
31
& bœ bœ œ bœ œ
? œœ œœ b œœ œœ œœ œœ œœœ œœ œœ
b œ b œ œ œ b
œ œ œ œ œ b œ œ œ œ
b ‰
œ
008
Andante
b
& b 42 œJ œ
. œ #œ œ œ œ œ nœ ‰. œ œ. œœœ œ œ œ
œ ‰.
R R
œ œ œ œ œ œ œœœ œ œ œœ œ œœœ œ œ œœœœœœœœ
express.
? b b 42 ‰ œ œ œ œ œ
Piano
b %
& b œ. œœœœœ œ œ ‰ œ œ œ œ œ #œ ‰ .. œJ
5
J œ œ
œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ œœœœœ
? bb œ œ œ œ œ œ œ J .. ‰
b œ. œ #œ œ œ œ œ nœ ‰. œ œ. œ œ œ œ œ œ œ ‰.
&b œ
9
R R
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b
b œœœœœ œ œ œ
& b œ. œ ‰ œ œ œ œ #œ œ œ ‰
13
J œ J
œ œ œ œ œœœœœœœ œœœ œœœ œœœœœœœœ
? bb œ œ œ œ œ œ œ
œ #œ œ œ œ œ nœ
2 008
b . œœœœœ œ
&b œ ‰.œ œ œ. ‰.
17
R R
œ œ œ œ œœœœœœœ œœœœœœœœ œœœœœœœœ
? bb œ œ œ œ œ
b b œ. œœœœœ œ ‰ œ œ œ œ
œ œ œ œ #œ œ œ ‰ J
21
& J
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœœœ
? b œ œ œ œ œ œ œ œ œ œ
b J ‰
b œ œ œ œ œœ œ œ œ œ œ œ œ œ
&b œ œ
25
œœ œ # œ œ #œ œœ
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ
œ œ œ œ
b œœœœœœ %
b œœ œ œœœ œ œ œ œ œ #œ ‰ .. ‰ œ œ
29 D.C. ao
& œ œ œ œœ œœ ˙˙
˙
œ œ œ
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ .. ‰ œœ œœ ˙˙
b œ œ œ J œ œ ˙
009
Andante
b rK
& b b b 42 œ œ œ œ. œ œ œ œ œ œ œ œ œ œ œ nœ œ. œ œ œ œ nœ œ. œ œ œ œ œ
Kr
nœ œ nœ
Piano
p
? b b b 42 ! ‰
b œœœœœœ œœ œœœ œ œ œœ œ œœœœ œœ œ œ œœœœœœœœ
bb œ œ œ œ nœ. œœ ! %
b
& b œ œœœ œ j
œ œ ‰ .. ! œ œ œ
5
œ œ n œ
? bb b œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ .. œ œ œ œ
bb
Kr
& b b œ. œ œ œ œ œ œ œ œ nœ œ œ nœ œ. œ œrK œ n œ œ . œœœœœ
9
œ
œ œ nœ
? bb œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
bb œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b rK
& b bb œ . œ œ œ œ œ œ œ œ nœ œ œ nœ œ. œ œr œ n œ œ .
K
œœœœœ
13
œ
œ nœ
œ
? bb b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ œ œ
2 009
œ œ œ œ œ œ. œ nœ œ œ.
bbbb nœ. rK
œ œ
œœœ œœœ œœ œ
17
&
? bb œ œ œ œ
bb œ œ œ œ œœœœœœœœ œœœœœœœœ œœœœœœ œœ
bb œ. œ
rK
& b b œ œ œ œ nœ œ œ œ œ œ. œ œ œ œ n œ œ œ œ ! œ œ
21
œ
n œ œ
nœ œ œ .
? bb œ œ œ œ
bb œ œ œ œ œ œ œœœœœœ œœœœœ œ œœ œœœœœœ œœ
bb %
D.C.
& b b œ œ œ œ nœ œ œ œ. œ .
œ œ œ. œ œ œ n œ œ œ œ . !
25
n œ œ
nœ œ œ .
? bb œ œ œ œ . r
bb œ œ œ œ œ œ œœœœœœ œœœœ œ œ œœ œœœœ. œ
010
œ
Andante
œ œœœœ œœœ œ œ œ œœœœ œ œ
bb 2 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰ œœ
& 4 œ œ J
J J
œ œ œ
? b b 42 ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ œœœ œœœ
œ
Piano
œ œ œ
œ œœœœ œœœ œ
œ œ œ œ œ œ %U
bb œœ œ œœœœ œœœ
œ
œ
œ
œ œ œ œ œ œ ‰ œj
œ
œ
5
& J
p
œ œ œ
? bb œ œ œ œ œ œ œ œ œœ œœœ
œ
œ œ œ œ œ œ œ œ œj ‰
œ œ œ
b b œ œ œœœœ œ œ œ œ œ œ œ œ œ œ œ ‰ œj
9
& œ J
? b œœœ œœœ œœœœœ œœ œœœ œœœ œœœ œœœ

b œœœ œ œ œ
&b œ œ œ œ œ œ
œœœœ œ ‰ J
13
œ J œ
f
œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œ j
œ ‰
œ œ œ
2 010
bb œ œ œ œ œ œ œ œ œ œ œœ œœ œ œ œœœœ
œ œ ‰ œJ
17
&
? bb œ œ œ œ œ œ œ œœœ œœœ œœœ

œ œ œ œœ œœœ œœ œ œ œ œ
œœ œ
œ œ œ œ
bb œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
21
&
? b œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ œ œ œ œ œ œ œ
œ œ œ œ
œœœœ œ
b œ œ ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ %‰
&b
24
? b œœœ œœœ œ œ œ œ œ œ œœœ œœœ j

b œ œ œ œ œ œ œœ œœœ œœ œ œ œ œ
œœ œ
011
Andante
# j . j œ œ œ œ œ œ ‰ œj
& # 42 œ œ œ œ œ œ œ œ ‰ œ œ.
œ œ œ œ œ œ œ œœ
gracioso
? # # 42 ‰ œ œ œ œ œœ œ œ œ œ œœ œ œ œ œ œœ œ
Piano
œ œ œ œ
## œ . œ œ œ œ œ. j j %
œ œ œ œ œ. œœ œ j ‰ œœœ ‰ œœ ‰
5
& œ œ œ œœ
œ œ
? # # œ œ œ œ œœ œ œ œœ œ œ œ œœ œœœ ‰ œœœ ‰ œœœ ‰
œ œ œ œ œ œ J J J
## . ‰ j œ . j œ œ œ œ œ œ ‰ œj
& . œ œ œ œ œ œ œ ‰ œ œ.
œ œ œœ œ œ œ œ œ œœ
? # # .. Œ œ œ œ œ œœ œ œ œ œœ œœ œ œ œ œœ œ œ œ œ œ
# . ‰ œj œ . ‰ œj
& # œ œœ œœ œ œ œ œ œ œ œ
14
œ
œ œ œ œœœ œ œœœ œ œœ
? # # œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ œ œ
2 011
## œ . ‰ œj œ . œ œ œ œ œ œ ‰ œJ
œœ œœ œ
18
& œ
œ œ œœ œ œ œ œ œ œœ
? ## œ œ œ œ œœ œ œ œ œœ œœ œ œ œ œœ œœ œ œ œ
22
## œ . œ œ œ œ œ. œ œ œ œ œ. œœ œœ œ ‰ œJ
& œ
œ œ œ œœ
? # # œ œ œ œ œœ œœ œœ
œ œ œ œœ œ œ œ œ œ
## œ . œ œ œ œ œ. %
œ œ œ œ œ. œ œ œ œ j ‰ .. ggg œœœ ggg œœœ Œ
26
& œ gg œ gg œ
œ œ
? # # œ œ œ œ œœ œ œ œ œœ
œ œ œ œœœ œ œœœ .. œ œ Œ
œ œ œ œ œ œ
012
Andante
# r œœ œ œœ œ
& 42 œr œ œ œ œ œ œ œ œ ‰ . œ œ œ ‰. r
œ
œ œ œ œ œ œ œœœ œ œ œœœ œœœ œœœœœœœ
? # 42 ! œ œ œ œ œ
Piano
œ œ !
# r
& œœ œ œœ œ œ ‰. œ œœ œ œœ œ œ œ ‰ œœ .. n œœ
5
œ
?# œ œ œ œ œ œ œ
# œ œœ œ œ œ œ œ œ
& # œœ œ œ œ œœ œ œœ ‰ œœ .. œœ œœ œ œ œœ ‰ œœ .. n œœ
9
œ œ œ
œ œ œ œ œ œ œ œ œœœ œœœ œœœœœœœœ
?# œ œ œ œ œ œ œ œ œ œ
# œ œœ j ‰ ‰.
& # œœ œ œ œ œœ œœ œœ ‰ œœ .. œœ œœ œœ œœ œ œ
13
œ œ œ R
œœœ œœœ œœœœœœœœ œœœ œœœ œœœœœœœœ
?# œ œ œ œ
œ œ œ œ œ œ
# œ œ œœ œ œ œ œ !œ œ œ œœ
2 012
œ œ œ œ !œ
17
&
œ œ œ œ œ œ
21
# œ œ œœ œ œ œ œ !œ œ œ œœ œ œ œ œ œ #œ
&
?# œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
25
# œ nœ œœ œ œ œ ‰ œ #œ œ œ œ œ œ œ œ ‰ œ #œ
&
œœœ œœœ œœœ œœ œ œ œ œ œ œœœ œœ œ
? # #œ œ œ œ œ œ œ œ œ œ
# œ nœ œœ œœ œœ
œ œ œ œ #œ œ œ œ œ œ
D.C.
‰ œ œ œ ‰
29
& J
œœœ œœœ œ œ œ œ œ œ œ œ œ œ
? # #œ œ œ œ œ œ œ œ œœœœœ ‰
J
013
Reynaldo Varella
ggg œœœœ œ œ œ œ œ œ œ œœœ % U r

Andante
b
&bb 42 gg œ œ ! œ
J
f p
24 ˙˙˙ œœ .. U
Piano
? bb œ. ‰
b
œ œ.
bbb œ. œœ œ œ œ œ œ œ œ œ. nœ œ bœ œ
œ !œ
3
& #œ
? b œ œœ œœ œœ œœœ œœœ œœœ œœ œœ œœ

œ œ œ œ œ œœ œœ œœ
bb œ
œ œ.
bb
& b œ. œ œ œ œ œ œœœ œ œ. nœ œ bœ œ œ !œ
7
#œ
? b b œ œœ œœ œœ œ
œœœ œœœ œœœ œ œ œ
œ œœ œœ œœ
œ œ œ
œ œ œ œ
b
b nœ nœ œ œ œ œ !œ œ œ
& b b œ. œ nœ œ. œ œ !œ œ
r
11
œœ œœœ œœ œœ œœ
? b
b b nœ œ œ œ œœ œœ œœ œ
n œ œœ œ œœ
œ
œ œ œ œ
U
œ. œ œ.
r
b œ œ
œ œ œ. œ bœ. j‰
& b b œ.
15
œ œ œ œ œ œ
? b œ œœœ œUœœ œ œœ œœ œœ œœ œœœ œ œœ œœJ ‰

bb œ œ œ œ
014
Andante
& b 42 Œ ! œ #œ œ œ ! œ #œ œ ! œ #œ œ œ œ œ ‰ œj
#œ œ œ
dolce
? b 42 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œœœ œœœ ‰
Piano
œ J
œ œ œ œ œ #œ œ j j
&b ‰ œ œ œœ œ œ œ ‰ œ
6
? œœœœœœœœ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ
œ œ œ œ #œ œ ‰ j j
&b œ œ œ œœ œ œ œ ‰ œ
10
?b œ œœ œœ œœ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ j
&b œ œ œœ œ ‰ œœ œ œ œ ‰ œ
14

b
œ œ œ œ œ œ œ ‰ œ œ œ œœ œ œ œ
D.C.
b J‰ œ œ œœ Œ
18
& J œœ œœ œ
œ œ
? œœœœœœœœ œ œ œœœ œœœœœœœœ œœœœœ ‰ œ œ œ Œ
b œ J œ œ œ
015
### 2
Andante
& 4 œœœ œ œ œ œ œ œ œ œ œ œ œ œœ
Kr
œ œ œ œ œœ
œ
f
? # # # 42 ‰ ! œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œœ
Piano
œ
### œ œœ
œ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ œœ
5
&
? ### œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
# œ œ œ œ
œ
### œ œ œ œ œ # œ œ # œ œœœœ œ
œ œ œ œ œ #œ œ œ œ œ œ œ œ
9
&
? ### œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
# œ œ œ
# œ œ ‰
œ œ œ J
### œ œ œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œœ œ œ
Kr
œ œ œ
13
& œ
œœœ
? ### œœ œœ œœ œœ œœ œœ œœ œœ œœœ œœœ œœœ

œ
2 015
### œœœ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ! œ œ #œ
17
&
? ### œœ œœ œœ œœ œœ œœ œœ œœ œœ œœœ œœœ

# œ
J
‰
œ
###
œ œ ! œ œ #œ œ ! #œ œœœœ œœ œ œ
œ œ #œ œ œ œ ! œ œ #œ
œ
21
&
? ### œœœ œœœ œœœ œœ œœ œœœ œœ œœœ œœœ

œ œ
# œ ‰
J œ œ
# J ‰
œ œ œ
### %
œ œ ! œœœ #œ œ ! #œ œœœœ œœ œ œ r! ‰ !
25
& œ œ #œ œ œ œ œ
? ### œœœ œœœ œœœ œœ œœ œœ œœœ œœœ œœœ œœœ

œ œ œ œ
# œ œ R
! ‰ !
œ œ
016
œ œ
œ œœ œ œ œœ
## 2 œœœ œ!œ œ œ œ!œ œ ! œœ œ
Allegro vivo
œ
& 4 J
p
Piano
? ## 42 ! ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œJ
## . j œ ! œ œ œ œ œ œ ‰ j
& . œ œ œ œ œ œ œ œœœœ œ œ œ
? # # .. ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
# œœ j
& # œ œ œœ œ œœœœœœ!
œ œ œ œ œ ‰ œ
9
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
## œ œ œœ œ œœ œœœ!œ œ j
œ œœ ‰ œ
13
& œ œ œ œ
? ## œœœ œœœ œœœ œœœ œœœ œœœ œœœœœœœœ

œ œ œ œ œ œ
## œ œ œœ œ œœœœœœ!œ œ .
œ œœ œ œ œ ‰ . ‰
17
&
? ## œœœ œœœ œœœ œœœ œœœ œœœ œœœœœ . ‰

œ œ œ œ œ œ J .
017
#### 2 !
Andantino
& # 4 œ œ œ œJ ‰ ! œ
p
? # # # # 42 ˙ œ œ œ
Piano
# œ œ ‰ ˙ œ
˙ œ œ ˙
J
# # # # œ œ œ œJ ‰ ! œ œ ‰ Œ ‰ ! œR
& # œ œ œ J
4
œ
œ œ œ œ œ
? #### œ œ œ œ
# œ œ ‰ ˙˙ œ œœ œœ ‰ œ œ œ œ œ œ
œ J
J
#### œ œ œ œ ‰ œ
& # œœ œ œ œ œ‰ œ
8
œ œ œ œ œœ
œ
? #### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
# œ œ œ œ œ
#### œ œ œ œ
& # œœ œ œ œ œ‰ œ
12
œœ œ œ œ œœ œ
œ œ
? #### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ
œ œ œ œ
# œ œ œ
2 017
####
# œ œ œ œ œ œ˙
œœœœ œ
16
& œ œœ œœœ œ œ œœ œœ œ œ ‰ œ
J
? #### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
# œ œ œ œ œ œ œ œ
####
# œ œ œ œ œ œ˙ œœœœ œ
œ œœ œœ œœ œ œ œ œr ! ‰
20
&
œ
? #### œ œ œ œ œœœ œœ œœœ œœœ œ
# œœ œœ œ œ
œ œ œ œ œ œ œr
!‰
018
Moderato
# 2 ! U
& 4 ! ! !
œ œ œ œ œ œ
p
? # 42 œ œ œ œ œ œ œ œ Ur ! ‰
Piano
œ ! ‰ œ ! ‰ œ ! ‰ œ
R R R
# œ.
œ œ œ œ œ œ. œ œ œ œ #œ œ. œ œ œ œ #œ œ.
3
& œœœœœ
? # œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ
# œ. r
œ œ œ œ œ œ. œ œ œ œ œ. œ œœœ œ œ # œœ œœ œœ
7
& #œ #œ
? # œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ ! ‰
R
# œœ œœ œœ œœ œœ ! œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœr œ.
& œœ ! œœ œ œœ # œœ œœ œœ
11
œ
R
œœœ œœœ œœœ œœœ œœœ œœœœœ
3
?# œ œ œ œ!‰ œ œ !‰
R R
2 018
# œ!œ œœ œœ œœ œœ œœ œœ œœ œ œ œ œ . r! j
œœœœ œ
15
& œ œ œ #œ œ #œ œ #œ œ
R œ
œœœ œœœ œœœ œœœ œœ œ œœœ
?# œ œ œ œ! ‰ œ œ!‰
R œ œ R
# !
œ œ œ œ œ œ. r
œ œ œ œ # œ œœ œ œ œ œ j
19
& œ œœœ œ #œ
R œ
? # œ œœ œœ œœ œœ œœ œœ œœ œœ
œ
œœ œœ
œ œ œ
# ‰ œ # œ œ œ ‰
# r! ‰
œ œ œ œ œ. œ œ œ œ
23
& œ œœœ œ #œ œ #œ œ œ
? # œ œœ œœ œœ œœ œœ œœ œœ œœ œ œœ œœœ ! ‰
œ œ
# œ ‰ œ œ
# œ œ
R
019
Andantino
b
& b b 68 ! ! ! Œ ‰ Œ œ
J
œ œ œ œ
Piano
? b b 68 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ ‰ Œ ‰ œ ‰ Œ ‰
œ ‰ Œ ‰ œ ‰ Œ ‰
b b œ œ œ œ œ œ j ‰ œ œ œ œ œ œ œ j‰ œ
b œ œ
5
& œ œ œ œ
œ
? b œœ œ œ œ œœ œ œœ œœ œœ œœ œ œœ œœ œ œ œ œ
œœ œ œ œ
bb J J J J J J J J
b œ œ œ j‰ œ œ œ œ œ œ œœ œ j
&bb œ œ œ œ œ œ ‰ œ
9
œ
? b œœ œ œ œ œœ œ œœ œœ œœ œœ œ œœ œœ œœ œœ œ œ œ
bb J J J J J J J J
b œ œœ œ œ œ œ
r
j‰ œ
&bb œ j‰ œ œ œ œ œ
13
J œ œ œ J œ œ
œ œ œ œ œ
? b b œœ œ œœ œœ œ œœ œ œœ œœ œœ œœ œ œœ œœ œœ œœ œ œœ œœ œœ
b J J J J J J J J
b œ œœ œ œ œ œ
r
jŒ
&bb œ j‰ œ œ œ œ œ ‰
17
J œ œ œ J œ œ
œ œ œ œ œ œ jœ
? b b œœ œ œœ œœ œ œ œ œ œ œ œ œ œ
œœ œ œ œ œœ œ œ œ œœ œœ œ œ
b J J J J J J J J
020
Gracioso
## 2 j œ . r
& 4 œ #œ œ œ œ œ nœ œ #œ œ œ œnœ œ ‰ œ
p
œ œ œ œ œ œ œœœ œœœ œœ œ œœœ œœœ
? # # 42 ‰ œ œ œ œ œ œ
Piano
œ œ œ œ œ
## œ . œœ œ. r
œ #œ œ œ œ œœ œœ ‰. œ
5
& œ
œœœ œœ œ œœœ œœœ œœœ œœœ œœœ œœœ
? ## œ œ œ œ œ œ œ
œ
## œ . œœ œ.
œ #œ œ œ œœ ‰ ‰
9
& œœ œ
D.C.
œ
œœœ œœ œ œœœ œœœ œœœ œœœ œœœ
? ## œ œ œ œ œ œ œ ‰
œ J
021
j
Moderato
# ‰ œj œ œ œ œJ œJ œ œ œ œ œ œœ œœ œœ œœ œœ œœ œ œœ œœj ‰ œj
& # 42 Œ œ
œ œœ œœ œ œœ œœ œ œ œ œ œœœ œœœ œœ œœ œœ
? # # 42 œœœ œœ ‰ œj œœ œ j œœJ œœJ œœJ j œœJ j
Piano
œœ œ J J œ œ œ œ
J J J J J J œ œ J J J
œ œ œ œ
## œ œ œ œ œ œ œ œ œ j j
œ œœ œœ œœ œœ œœ œœ œ œœ œœ ‰ ! œr
5
& J J œ
œ œœ œœ œ œ œœœ œ œ œ œœ œœ œœ
? # # œj œœ œ œ j œœJ œœJ J j œœJ œœJ œœJ œ
j œ œ œ
J J J œ œ J J J
œ œ œ œ
## r j j œ œ œ ‰ ! œr
œ œ œ œJ œJ œ œ œ ‰ ! œ œ œ œ œ œ
9
&
? ## j œœ œœ œœ j œœœ œœœ œœœ j œœœ œœœ œœœ œœ œœ œœ

œ œ œ œ œ œ J J J
œ J J J J J J J J J J
## ! œr j
œ œ œ œJ œJ œ œ œ ‰ œ œ œj j ‰
13
& œ œ œ œ
? # # j œœœ œœ
œ
œœ
œ j œœœ œœœ œœœ j œœœ œœœ œœœ œ
j œœœ œœœ ‰
œ J J J œ J J J œ J J J J J
œ
022
Andante
j œ œ œ œ œ œ œ œ œ ‰ œj
& 42 œ œ œ œ œ œ œ œ œ ‰ J
dolce
? 42 ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
œ œ œ œ
œœ œ œ œ œœœ j
&œœœ œ ‰ J œ œœ œ œ ‰
5
?œœœœœœœœ œœœ œœœ œœœ œœœ œœœœœœœœ

œ œ œ œ
œœ œ j
&œœœ œ œ œœœ ‰
9
œ œ œœ œ œ œ
?œœœœœœœœ œœœœœœœœ œœœ œœœ œœœœœœœœ

œ œ
&œœœ œ œœ œ œ œœœ ‰ ‰
13
œ œ
D.C.
œœ œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œj ‰
œ œ
023
Moderato
œ œ œ ‰ œ
& 42 œj œ œ œ œ J
œ œ œœ œ œœœ ‰ j
œ
Piano f
? 42 ‰ œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ œœœœœœœœ
œ œ #œ œ # œ œ
œœ œ œ %
&œ œ œ ‰ œ œ œ œœ œ œœœ ‰ j
5
J œ
p
? œœœ œœœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ ‰
œ œ #œ #œ œ œ œœ
œ œ J
œœ œ œ œ œ œœœ ‰ j
&œ œ œ ‰ œ œ œœ
9
J œ
? œœœ œœœ œœœ œœœ œœœ œœœ œœœœœœœœ

œ œ #œ œ # œ œ
œœ œ œ œ œ œœœ
&œ œ œ ‰ œ œ
‰ œ œœ
13
J J
F
? œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ
œ œ #œ œ # œ œ œ œ
2 023
œ œœ ‰ #œ œ œ œ œ œ #œ œ œ ‰ œ
17
&œ œ œ œ J
?
œœœœœœœœ œœœœœœœœ œ œ œ œ œ œ œ œ œœœœœœœœ
œ œœ ‰ #œ œ œ œ œ œ #œ œ œ
21
&œ œ œ œ œ
? œœœ œœœ œœœ œœœ œ œ œ œ œ œ œ œ

œ œ œ œ #œ œ œ œ œ œ œ œ
&œ ‰ j ‰ j
œ œ #œ œ œ
25
œ œ œ #œ œ œ œ œ œ
? œœœ œœœ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ #œ œ œ œ œ œ œ œ œ œ œ œ
%
&œ ‰ j ‰ ‰
œ #œ #œ œ œ
29
œ œ œ #œ œ œ œ œ
? œœœ œœœ œ œ œ œ œ œ œ œ œ œ œœ ‰
œ œ #œ œ œ œ œ œ œ œ œœ
œ œ J
024
Andantino
b .
& b 42 œJ œ œ œ #œ œ œ œ œ œ œ œ œ œ. œ œ œ
r r
œ œ
œ œ
p
Piano
? b b 42 ‰ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ
%œ
bb œ œ œ #œ œ
. œ œ œ œ œ œ œ
r r
œœ ‰ J
5
œ
& œ œ
F
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? bb œ œ œ œ œ œ œ œ œ œ œ ‰
œ œ œ œ
bb œ œ œ #œ
œ. œ œ œ œ œ œ œ œ. œ
r r
œ
9
& œ œœ
œ œ
? b œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
b œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ
bb œ œ œ #œ œ
. œ œ œ œ œ œ œ ‰ œJ
r r
œ œ
13
œ
& œ œ
? bb œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ
2 024
bb œ œ œ #œ œ
. œ œ œ œ œ œ œ œ. œœ œ
r r
œ
17
& œ œ œ
? bb œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ
bb œ. œ œ œ œ œ œ
r
œ œ œ #œ
r
œ œ ‰ j
21
œ
& œ œ
œ
œ
? b œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
b œ œ œ œ œ œ œ œ œ œ œ œ ‰
œ œ œ
b
& b œ. œ œ œ #œ œ. œœ
25
œ œ œ œ œ. œ œ #œ œ. œ
? b œœ œœ œœ # œœœ œœ œœ # œœœ œœ œœ œœ œœ œœ
œ
œ œ œ
b %
b
& œ. ‰ ‰
29
œ œ œ œ œ. œ œ #œ œ. œ œ œ #œ œ
? b œœ œœ œœ # œœœ œœ œœ # œœœ œœ œœ œœ œœ
œ
œ œ ‰
025
U % œ
Andante
# ‰ gg œœ gg œœœ ‰ ggg œœœ ‰ ‰ œj
& 42 ‰ gg œœœ ggg œœœ ‰ gg œ g
‰ ggg œœœ g ggg œœ
‰ ! R
g g
p g
j j j
œ œ j ‰ œ œ ‰
Piano
? # 42 ‰ œ œ œ œ
# œ œ œ œ #œ œ nœ œ ! œ œ œ œ œ œ œ œ œ ! œ
5
&
?# œœ œœ œœœ œœœ œœœ œœœ œœ œœ

œ œ œ ‰ œ œ œ ‰
œ
œ œ ‰ œ œ œ ‰
# ! œ œ œ œ œ œ œ œ œ ! œ
& œ œ œ œ #œ œ nœ œ
9
œ
?# œ œ ‰ œœ œœ ‰ œœ œœ ‰ œ œ ‰
œ œ œ
œ
# r
& œœ œœ œœ œ œ œ œ ! œ œœ œ œ œ œ œ œ ‰. œ
13
œ œ œ œ œ
?# œœ œœ œœ œœ œœœ œœœ œœ œœ
œ œ œ ‰ œ ‰
œ
œ œ ‰ œ œ œ ‰
# %
& œœ œœ œœ œ œ œ œ ! œ œ œ œœ œ œ œ ‰ ‰
17
œ œ œ œ
?# œœ œœ œœ œœ œœœ œœœ œœ œœ
œ œ œ ‰ œ ‰
œ
œ œ ‰ œ œ œ ‰
026
Andantino
. œ. œ œ œ œ œ œ ‰ œ
& b 42 œJ œ œ œ œ œ œ # œ ‰ œJ J
f
? b 42 ‰ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
b œ. œœœ œ œ
œ ‰ œJ œ œ œ # œ œ œ œj ‰
D.C.
œ.
5
&
b œ J
027
Moderato
# 2 j j U
& 4 œœœœ œœœœ œœœ œ œ ! ! œ œ œ
f
?# 2 œ œ œ œ ! Œ
Piano
4 œ œ œ œ
# j j œ ! œ œ œ œ œ œ œ œ œ jœ jœ
œ œ œ œ œ œ œ œ œ œ œ œ #œ œ
2
& œ œœœ œœœ œ œ œ
?# œœ œœ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ
œ œ œœœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ
# j j
œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ‰ # œ œJ œ œ œ œJ œ œ œ œ œ œ œ œ œ œ ! ! œ œ œ
6
&
œœ œœ œœ œœ œœ œœ œœ œœ œ # œ œ œ œ œ œ œœ œœ œœ œœ œœ œœ
?# œ œ œœœ œ œœ œœ œœ œ œœ œ # œœ œ # œœ œ œœ œ œœ œœ œœ œ œ
œ œ œ œ œ
#
œ œ œ œ œ œ #œ œ œ œ #œ œ œ œ œ nœ œ œ œ œ œ œ œ œ œ ! ‰ ! œ
10
& J J
œœœ
#œ R
? # œ œœ œœ œœ œ œœ œœ œœ # œœœ œœœ œœœ œœœ œœœ œ œœ œœ œœ œœœ œœœ œœœ œ œœ œœ œœ œ œœ œœ œœ

œ œ œ œ
# œ œœœœ œœœœ œ œ
# œ œJ œ œ # œ œJ œ œ œ œ œ œ n œ œ œ œ œ œ œ œ # œ œ œ ! Œ
14
& J J
œ œ œ œ œ
? # # œ œœ œ œœ œ œœ œœ # œœ œ œœ œ œœ œ œœœ œœœ œœœ œ œœ œœ œœ œ œœœ œœœ œœœ œ œœ œœ œœ œ œœ œœœ !
028
# r U! % œ
Moderato
& 42 œœ œœ œœ J
œ œ œ
? # 42 œ r U!
Piano
œ œ œ ‰
œ
# œ #œ œ œ œ nœ œ œ œœœ œ œ œ
2
& œ œ œ œ
?# œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ

# œ œ
#œ œ œ œ œœœœœ œœ œœ œœ œœ œ œœ œœ œ œ œ ‰ œj
6
&
œ œ œ
?# œ œ œ œ œ œ œ œ œ œ œ bœ œ œ œ œ œ œ œ œ œ œ œ œ
#œ œ œ œ œ œ œ œ
#
‰ j ‰ œ
10
& œ œ œœ œ œ œ œ œ œ œœ œ œ œ J
?# œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ

# %
œ œœœœ œ œ ‰ ‰
œ œ bœ œ œ
14
& œ œ ! œ œ œ
œœ
?# œœœ œœœ œ œ œ œ œ œ œ œ œ œœœ œœœ œœ œœœ œœœ
œ œ œ œ œ œ J ‰
029
Andante
b œ œ œ œ œ œ œ ! œ ! œœ
& b 42 œ œ œ œœœœ œ #œ
œœ œ œ œ
F
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œœœ œ œ
? b b 42 ‰ œ œ œ œ #œ œ nœ œ
Piano
b œ œ œ!œ! œ r
œ œ œ œ œ œ œ œ œ #œ œ œ œ œ ! ‰ ‰ œ œ
&b
5
œœœ œœœ œ œ œ œ œ œ œ œ œ œ œ œ
? b œ œ #œ œ œ œ œ œ œ œ œ œ œ œ
b
r
bb œ ! œœ ! œ ! œ ! œ ! ‰
! œ œ œ œ ! œœ ! œœ ! # œœ ! œœ ! ‰ ! œ œ œ
œ œ #œ
9
&
R
œ œ œ œ œœœœœœœœ œœœœœœ œœ œœœœœœœœ
? bb œ œ œ œ
œœ œ œ # œ œ œ œœ n œœ œœ œœœ œ œ œœ œœ œœ œ œ œœ œ
b
œ œ #œ
œ œ œ
&b
13
œ œ œ œœœ œœœ œ œ œ œ œ œ œ œ œ œ œ
? b œ œ œ #œ #œ œ œ œ #œ œ œ œ
b
œ œœ œœ œœ œœ œœ # œœœ œ œ œœ œœ œœ œœ # œœ œœœ œ
bb
17
œ œ #œ œ #œ nœ
&
œ œ œ œ œ œ œ œ œœœœœœœœ œœœœœœœœ
? b œ œ œ œ œ œ œ œ
b
bb œ œ # œ œ œœ œœ œœ œ œ
œ œ œ œ œ œ œ œ œ œ œœœ œ
2 029
#œ œ œ œ œ ! œ !
21
&
? bb œœœœœœœœ œœœœœœœœ œœœœœœ œœ œœœœœœœœ
œ œ œ œ œ# œ œœ œ œ œ œ œ
b œ œ œ œ #œ ‰ œ œ œJ
& b ‰ œJ J ‰ œJ œ ‰
25
? b b œJ ‰ œ œœœ œ œœœœœœœ œœœ œœ œœœ œ œœœœ œœœ
œœ œ œ œ œ œ œ œ œ œ œ œ œ# œ œ œ œ œ
b œ œ œ œ œœ # œ R ! ‰‰! R
&b ‰ J J ‰ ‰ ? r!
29
œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ
? b œ œ œ œ œ œ œ œ œR ! ‰Œ œ œ œ œ œ œ œ ‰
b J
? b œ œ œ œ œœ j
b œ ! œ ! œ ! œ ! œR ! ‰ & ! œ œ #œ œ œ œ œ ‰
‰ ? r!
34
œ
œ œœœœœœœ œœœœœ œœœ œœ œœ œœœœ œœœœœœœœ
? b
b
? b œ œ œ œ œœœ j
b œ ! œ ! œ ! œ ! œR ! ‰ & ! œ #œ œ œ œ œ ‰ Œ
38
œ œ œ œ œ œ œ œ œ œ œ œ œœœ
? b œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰
b J
030
% Andante œ œ
2 j # œ œ œ œ œ œ œ
&b 4 œ œ œ œ œ œ
r
œ
œ œ œ. œ
p
œ # œœ œ œœ œ œ #œ œœ œœ œœ
Piano
? b 42 ‰ œ J
J
œ œ œ
œ œ œ- . ! œ-j .
rit.
œ œ œ ‰ . œR
a tempo
&b œ # œ r
! Œ
5
œ
J
p
# œœ œ œœ œ œ œœ . #œ. œœ œœ
?b œ œ œ œ .. ! œœ .. ! œ œ
J J
œ œ œ œ œ œ
&b œ. œ œ œ œ œ œ
9
œ œ œ
œ œœ œ œ # œœœ œœ œœ
? œ œ œ œ œ œ œœœ œ œ œ
b J J œ œ J J
œ œœ
&b œ œ œ ! r œ œ œ ‰. r
r
œ
13
J œ œ œ
œ
œ œ
œ œ œœ œœ œœ # œœœ
?b œ œ œ œ œ J œ œœœ ‰
J œ œ J œ œ
2 030
œ œ
&b œ œ œ œ œ. œ œ œ œ œ #œ . J
17 cresc.
œ œ
œœ œ œœ œ j
œ n œ bœœ œ œ œœ œ
#œ # œœœ # œ n œœœ
?b r
J œ œ œ
œ J œ
œ %
&b œ œ œ œ œ. ! œ œ œ œ œ œ ‰ ‰
21
#œ œ
p Uu
œœ œ j œœœ œœ œœ # œœœ
?b œ œ n œ # œœ œ œ ! œ œœ
œ œ ‰
R œ œ J œ
031
%j œ œ œ œ!œ œ œ œœ œ œ œ ‰ j
Andante
## 2 ‰ œ œ œ ® œ œ
& 4 œœ œœ œœ œ œ
œ œ œ
p
œœ œœ œœ œ œœ œœ ‰ œœ œœ ˙œ œœ œœ œœ
Piano
? # # 42 œ ‰ ‰ œ œ œ
œœ
œ
## œ œ œ ®œ œ œ œ ‰ œJ œ œ œ œ œ œ œ ‰ œ
7
& J
? ## œ œœ œœ œœ œ œœ œœ ‰ œœ œœ œ œ
œ œ œ œ œ ‰
# œ œ œ ‰ œJ
& # œ œ œ œ ! œ œ œ œ œ
11
œ œ œ
? # # œ œœ œœ œœ ˙œ œœ œœ œœ œœ œœ œ œœ œœ ‰
œ œ
# œ œ œ œ
& # œ œ œ ‰ j œ œ œ œ œ œ œ ‰ J
15
? # # œ œœ œœ œœ œ œœ œœ œœ œœ œ œ
œ œ œ
œ œ œ ‰
œ
# œ œ œ %
& # œ œ œ ‰ j œ œ œ
œ œ œ œ ‰ ‰
19
? # # œ œœ œœ œœ œ œœ œœ œœ œœ œ œœ œ œ
œ œ œ ‰
œ
032
Andantino
# œœ !œ œ œ œ œœ œ
& 42 œj œ œ œ œ œ œ. œ !œ
p
œœ œœ œœœ œœœ ‰ œœœ œ œœ œœ
Piano
? # 42 ‰ œ œ ‰ œœ œ œ ‰
œ œ œ œ œ
# œ. #œ œ œ œ œ. œ j
œ œ œ œ œ œ œ œ ! œr
5
&
f
?# œ œœ œœ œœ œœ œœ œœœ œœœ œœœ œœ œœ
˙œ œ ˙œ ˙œ œ œ œ ‰
# œ œœ œœ ! œœ œœ .. # # œœ œœ n œœ œœ
& œœ œœ œœ œœ œœ œœ .. œœ œœ ! œœ œœ œœ œ œ œ œœ
9
œœ œœ œ œ œ œ œœ œœ œ œ œ
?# œ œ ‰ œœ œœ ‰ œœ œœ ‰ œ œ ‰ ˙œ œ œ œ
œ œ œ œ
# j
& œœ .. œœ œœ œœ œœ œœ ‰
14
œœ œœ œ
D.C.
œœ
œ
? # ˙œ œœ œœ œœ œœœ œœœ œœœ œœ
˙œ œ œ œ ‰
033
Andante
& 42 ! œ ! œ b œœ ! œœ ! œ ‰ œj
œœ œœ œ œ
bœ œ œ
œ
œ
œ œœ œœ œ
p œ
? 42 œœ œ œ bœ œ œ œ œ œ œ œ
Piano
œœ œ œœ œ œœ œ œ œ
&œ!œ œœ œœ œ!œ œ!œ œ!œ œœœœ œ œ!œ

3
œ
F
?œœ œ œ œ œ œœœ œœœ œœœ œ œœ œ œ œ
œœ œ œ œ œ œ œ œ œ œ
&œ!œ œœœ œ œ!œ œ!œ œ!œ œœœœ œ œ!œ bbb

7
œ œ œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ bb
œ b
b œ œœ œ œ œ œ!
&bb œ œ œ !œ œ œ œ œ #œ œ
11
œœ
œ œ œ œ œ nœ œ œ œ œ œ œ
? bb œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ œ
bb œ œœ œ œ! j‰
D.C.
& b œ œ œ œ œ œ œ nœ œ
15
œ œ
? bb œ œ œ œ œ œ œ œ œ œ œ œ œ nœ œ œ œ œ œ j
b œ œ œ œ œ œ œ œ œ œ ‰
œ
034
Allegretto
b .
& b 42 œj œ œ œ œ œ œ œ ‰ œj œ . œ œ œ œ œ œ ‰ œj
Piano
? b b 42 ‰ œ œœ œ œœ œ œ œ œ œ œ œœ œœ œœ
œ œ œ œ œ œœ œ œœ
b .
&b œ œ œ œœ œ œ ‰ j œ. œ œ œ œ œ œ ‰ œ
5
œ J
? b b œ œœ œœ œœœ œœœ œœœ œœœ œœœ œ œœ œœœ ‰

œ œ œ œ
b ‰ œJ
& b œ. œ œ œ œ ‰ œJ œ.
9
? b œœ œœ œ œœ œœ œœ œœœ œœœ œœœ œ œ œ

b œ nœ œ œ œ œ œ œ
b
& b œ. œ œ œ œ ‰ œJ œ. œ œ œ ‰ ‰
13
œ œ
D.C.
œ œ
? bb œ œœ n œ œœœ œ œ œ
œ œ œ œ œœœ œœœ œœœ œ œœ œœ ‰
œ
035
œ œ . œ
Moderato
œ œ œ œ. œ œ œ œ œ œœœ œ œ œ œ
#
& 42 œJ œ . œ œ œ œ œ . œ œ œ œ œ œœœ œ œ œ
J
œ
J
œ œ œ œ œ œ
? # 42 ‰ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ # œ œ œj ‰
Piano
# œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
j
œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ
5
& œ œ œ J
?# œ œ œ œ œ œ
œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ
œ œ ‰
#œ œ œ J
# œ. œ œ œ œ œ. œ œ œœ œ œœœ œ œ œ œ
9
& J J
? # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ
œ
# œ œ œ œ œ.
& œ. œ œ œ œ œ œœœ œ œ œ œ
13
J J
œ œ œ œ œ œ œœœ œ œ œ œ œœœ
?# œœœœœœœœ œ #œ œ œ œ œ #œ œ
# œ.
2 035
œ œ œ œ œ
œ œ œ œ. œ œ. œ œœœ œ !œ
17
&
?# œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ
œ
#œ œ œ
#
œœœ œ œœ œœ œ j j
21
? # #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
#
œœœ œ œœ œœ œ j j
25
? # #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
#
œœœ œ œœ j ‰
29
& œ œ #œ œ œœ œ œ œ
? # #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ‰
œ œ J
036
Allegretto
b
& b b 42 œ
j œ œ œœ œ œ œ. œ œ œ œ œ œ œ
j
œ œ
j
œ œœ œ œ œœ œ œ n œœ œ œ œœ œ
Piano
? b b 42 ‰ œ œ œ
b œ
b b œ œ œ œ œ œ j j
& b œ œ œ
5
œ œ œœ œ œ n œœ œ œœ œ
? bb œ œ œ œ
œ
œ
œ
b œ
b j j
& b b œ œ œ œ œ œj œ j œ œ
9
? b œ œ œœ œ
œ œ œ œ
œ œ n œœ œ œ œ œ œ
œ
bb œ
b
& b b œ œ œ œ œ œj œ j j ‰
13
œ œ œ œ œœ œ œ n œœ œ œ œ
? bb œ œ œ ‰
b œ œ œ
037
Andante
œ œ œ œ œ. œ œ œ
& 42 J ‰ œJ œ œ œ œ. œ œ œ ‰ œ
J
œ œ #œ œ œœ
! œ #œ œ
Piano
? 42 ‰ ! œ œ !œ œ œ
œ œ œ !
& œ œ œ œ. œ œ œ ‰ œ œ œ ‰ œJ
5
J œ œ #œ #œ œ
œœ œ œ
?
œ ! œ œ ! œ #œ œ ! œ #œ œ œ ! œ
#œ œ
œ œ œ œ œ ‰ œ œ œ œ œ #œ œ ‰ œJ
9
& #œ œ J
œ œ œ œ œ
?
œ ! œ œ ! œ #œ œ ! œ œ œ ! œ
œ
œ œ œ œ ‰ œ œ œ œ œ #œ Œ
œ œ
13
& #œ œ J
œ œ œ œ
?
œ ! œ ! œ #œ œ ! œ œ
œ Œ
œ
038
Allegretto
#
& 42 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œœœœœœœœ
? # 42 ‰ ! œ œ œ œ œ
Piano
œ œ œ
# œ œ œœ œ œ œ œ œ œ œœ œ œ œ œœ œ
5
& œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœœœ
?# œ œ œ œ œ œ œ œ J ‰
# œ #œ œ œ œ œ œ œ œ œ œ
& œ œ œœ œ œ #œ œ
9
? # œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œœœœœœœœ œœœœœœœœ
œ
# #œ œ œ œ r !‰
œœ œ œ #œ œ œ œ
D.C.
& œ œ
13
#œ œ œ
? # œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ !‰
œ œ R
039
Andante
œ œ. œœ œœ œ œ œ. œ œ œ œ œ œ ‰ œJ
r
& 42 J ‰ œJ
œ r
œ
œ œ #œ ! œ #œ œ œ œ
Piano
? 42 ‰ ! œ œ !œ œ ! œ
œ œ œ
œ. œ œ œ œ œ œ œ. ‰ œj
r
‰ œJ œ œ
r
œ
œ œ œ œ
5
œ
&
? œ œ œ #œ œ ! œ #œ œ ‰
œ ! œ !œ œ œ ! œ
œ
œœ œ j ‰ œj
&œ œ œ œ ‰ œ #œ. œ œ œ œ œ œ
9
? ! œ œ œ ! œ #œ œ ! œ #œ œ ! œ œ œ
œ œ œ œ
œ œ j
&œ œœ œ œ ‰ œ #œ. œ œ œ œ œ œ ‰ ‰
13
? ! œœ œ ! œ #œ œ ! œ #œ œ œ œœœ ‰
œ œ œ œ. œ
J
040
%
Andante
b
& b b 42 ‰ œ œ ‰ ‰ œ œ ‰ ‰ œœ œœ ‰ ‰
œœ œœ
! œ
R
œœ œœ œœ œœ œœ œœ œ œ
j j
Piano
? b b 42 œ . œ. j œ œ Œ ‰
b œ. œ œ œ. œ
œ œ. œ œ. œ
b œ œ œ œ œ
&bb œ œ œ œ J œ œ œ œ œ œ œ œ !œ
5
œœ œœ œœ œ œ œ œœœ œœœ œœœ œœ œœ œœ

? b œ œ œ œ œ œœ œœ œœ œ œ œ œ
bb œ
b œ œ œ œ œ ! œr
& b b œ œ œ œ œJ œ œ œ œ œ œ
9
J
? bb œ œ œ œ œ œœ œœ œœ œœœ œœœ œœœ œ œ œ ‰
b œ
n œ œ œ ‰ œj
& b b œ nœ œ œ
13
J J
n œœœ œœœ œœœ œœ œœ œœ n œœœœ œœœ œœœ œœ œœ œœ

? b œ œ œ œ œ œ œ œ œ
bb œ œ œ
%
bbb n œ œ œ œ œ œ œ œ œ œ œ œ
nœ œ œ ‰
17
& œ J J J
n œœœ œœœ œœœ œœ œœ œœ n œœœœ œœœ œœœ œœ œœ
? b œ œ œ œ œ œ œ œ ‰
bb œ œ œ
041
Andante
& b 42 œJ œ œ œ œ œ œ œ. ‰ œ
J
nœ œ œ œ bœ œ œ ‰ œJ
œ œ œ œ œœ œœ œœ œœ œœ œ œ œ
Piano
? b 42 ‰
nœ œ œ œ bœ j
& b œ œ œ œ œ œ œ. ‰ œ œ œ ‰ œ
5
J
œ œ œ œ œœ œœ œœ œœ œœ œ œ
?b œ œ œ œ œ œ œ œ œ œ œ ‰
œ œ œ œ œ œ œ œ œ œ #œ œ ‰ nœ
& b œ #œ #œ
9
J J J
œœ œœ œœ
? œ # œœœ # œ œœœ œ œ œ œ # œœ œœ œœ œœ œœ ‰
#
& b œ œ œ œ œ œj œ œ nœ œ œ bœ ‰ ‰
13
J œ œ œ
œ œ œ œ œœ œœ œœ œœœ œœœ œ œ
?b œ œ œ œ œ œ œ œ œ ‰
042
Andante
j œ œ œ œ œ œ œ œ ‰ œj
& 42 œ œ œ œ œ œ œ œ ‰ œ
r
œ
J
? 42 ‰ œ œ œ œ œ œ œ œ œœœœœœœœ œœœ œœœ œœœ œœœ

Piano
œ œ œ œ
‰ j ‰ j
& œ
5
œ œœœœ œ œ œ œœ œœ œ œ
œ œ
? #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
‰ j ‰ j
& œ
9
œ œœœœ œ œ œ œœ œœ œ œ œ œ
œ œ
&œœ œœ œ œœœœœœœœ
j
œœœœœœœœ
r
œœœœœœœœ
r
13
œ œ œ
?œœœœœœœœ œœœœœœœœ œœœ œœœ œ œ œ œ

œ œ œ œ
œ œ
2 042
U
‰ j ‰ j
& œœœœœœœœ œ
r
17
œ œ œ œœ œœœ œ œ œ
U
œ œ
U
‰ j ‰ ‰
& œœ
21
œœœ œ œ œ œ œœ œœœ œ œ
U
? #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ œ ‰
œ œ J
043
Andante
j
& 42 ‰ œœœ œœœ ‰ ‰ œœœ œœœ ‰ ‰ œœœ œœœ ‰ ‰ œœ œœ
œ œ œ
œ. j j
œ œ. œ œ ..
Piano
? 42 œ . j œ ‰ ‰
œ œ. œ œ œ œ
œ
œ œ œ œ œ œ
& œ J œ œ œ œ ‰ œj
5
J œ œ œ
? œj œœœ œœ
œ
œœ
œ œ œ œ
œ œœ œœ œœ œœœ œœœ œœœ œœ œœ œœ
œ œ œ œ
œ
œ
œ œ œ œ œ œ
&œ J œ œ œ œ ‰ œj
9
J œ œ œ
?œ œ œ œ œ œœ œœ œœ œœœ œœœ œœœ œ œ œ ‰
œ
œ œ œ
&œ œ. œ œ œ œ œ.
13
œ #œ œ œ œ œ
œœ œœ œœ
? œ œœœ œœœ œœ
œ œ œœœ œœœ œœœ œœ œœ œœ
œ œ œ œ
œ
œ œ œ
&œ œ. œ œ j ‰
17
œ
œ œ œ
? œ œœœ œœœ œœ
œ œ œ œ œ œœœ œœœ œœœ œ œœ œœ ‰
œ
044
Andante
b b b 2 Œ ‰ œr œ n œ œ œr œ n œ œ œr œ n œ œ œ œ œ œ œr œ œ œ œr œ œ ! œr œ n œ
& b 4 œ œ nœ
R
f
? b b b 42 œœœ œœœ ‰ œœ œœ œœ œœ œœ œœ œœ œœ œœ
Piano
b œ œ œ œ œ œ œ ‰
œ œ
b j
& b bb œ œ nœ œ œ nœ œ œ œ œ œ œœœ œ
r r r
œ œ œ œ œ œ
r
6
œ œ œ
? bb b œœ œœ œœ œœ œœ œœ œœ œœ œœ
b œ œ œ œ œ œ ‰
œ œ
bb nœ
& b b œ œ œ nœ œ œ œ bœ ‰ œ œ œ œ œ œ œ ‰ œœ
10 3
œ œ nœ œ œ
? b b n œ œœœ œœœ œœœ œ œ œ

n œ œœ œœ œœ œ
œœ œœœ œ œ
œ œ œ ‰
bb œ
b
& b bb œ œ œ n œ œ n œ œ œ bœ ‰ œ œ œ œ œ œ œ Œ
14 3
œ œ nœ œ œ
? b b b n œ œœœ œœœ œœœ œ œ œ

n œ œœ œœ œœ œ œœ
œ
œœœ œ œœ œœ ‰
b
045
œ œ œœœœ œ œ œ œ œ
‰ J œ œ œ œ nœ œ
Andante
b œ
& b b b 42 J
œ
‰ J
œœ œœ œœ œœ œ œ œœ œœ
œ œ ‰ œ œœ œœ
Piano
? b b b 42 ‰
b œ œ œ ‰ œ œ
‰ œ œ œ ‰
n
œ œ œœ œœ œ œ œœ œœ œ œ œ œ œ œ œ %
b
& b bb ‰ j
5
œ
œœœ œœœ œœ œœ œœœ œ œœœ œœ œœ
? bb b œ œ œ œ œ
b œ œ œ ‰
bb œ œ œ œ œ œ œ œ ‰ j
&bb œ œ œ œ œ œ œ ‰ œJ
9
œ
œœ
? bb œ œ œ œœ œœ œœ œœ œœ œœ
bb œ œ œ œ œ œ œ
b
& b bb œ œ œ œ œ œ œ œ œ ! œ œ œ œ. œ œ. œ œ œ ‰ œ
13
J
œœ œœ œœ œœœ œœœ œœœ n œœœ œœœ œ œ
? bb b œ œ œ œ œ œ nœ œ œ œ œ ‰
b œ œ
2 045
b bbb œ œ œ œ œ œ œ œ ! j œ œ œ œ œ ‰ œJ
17
& œ nœ œ œ œœ
œ œ œœ
? bb œ œ œ œ œœ œœ œœ œœ œœ œœ
œ œ œ œ
n œ œœ œœ œœ
bb
bb œ œ œ œ œ œ œ œ ! %
&bb œ œ œ œ ‰ ‰
21
œ nœ œ œ nœ œ œ
œ œ œœ
? bb œ œ œ œ œœ œœ œœ œœ
œ nœ
œœ
œ
œœ
œ œ œœ œœ ‰
bb
046
# 2 œj œ . œœ œœ œœ œœ œœ .. œ ! œ œ . œœ œœ œœ œœ œœ .. œœ œ ! œ
Andante
& 4 œ œ. œ œœ œ œ . œ œ
p
œ œ œ œ œ œ œœœ œ œ œœœ œœœ œœœœœœœœ
? # 42 ‰ œ œ œ œ œ
Piano
œ œ
# œ. œœ œœ œœ œœ œœ .. œœ œœ ! œœ œœ .. œœ œœ œœ œœ œœ .. œœ œœ ! œœ
& œ.
5

# j j
& # œœ .. œœ œœ œœ œœ œœ œœ œœ œœ n œ œœ .. œœ œœ œœ œœ
9
œœ œ
?# œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œ œ
# œœ œœ œœ œœ œœ œœ œœ œœ n œ œ j
& # œœ .. œ œ œ œ .. œœ œ œ ‰
13
œœœ œ œœ œ œœ œœ œœ œ œœœ œœœ œœœœœ

?# œ œ œ œ J ‰
047
Andante
& 42 œ œ œ œ œ œ œ œ œ œ œ œ . œ œ œ œ œ œ œœœ
œ œ œ
œ œœ œœ œœ œœœ œœœ œœœ œ œ
Piano
? 42 ! ‰ œ œ œ œ œ œ œR ! ‰
œ
& œ œ œ œ œ œ œ œ œ. œ œ œ
5
‰
œ œ œ œ œ œ
? œ œœ œœ œœ œ œœ œœ œœ œœœ œœœ œœœ œ
œœ œœ ! ‰
œ R
j
& œ. œ œ œ #œ œ # œ ! œr œ . j ! r
9
œœ œœ œœ œ œœ œœ œ œœ œœ œœ œœ œœ
? # œœœ œœ œœ # œœœ œœ œœ
œ œ œ œ
œ œ œ œ œ œ
j rD.C.
& œ. œ œ œ #œ œ # œ ! œr œ . œ œ œ œ œ œ !‰
13
?
œœ
œ
œœ
œ
œœ
œ # œœœœ œœœ
œ
œœœ
œ # œœœœ œœœ
œ
œœœ
œ
œœ œœ
œ œ !‰
œ œ œ œ
048
b 2 œ. œ œ œ œ
Allegro
b
& 4 œ œ œ œ. œ
f
œœ œœ œœ œœ
œ œœ œœ œœ
Piano
? b b 42 Œ J ‰ Œ
œ
b œ œ œ œ œ j ‰ œ œ
& b œ œ œ #œ ‰ j ‰
r r r r
3
J
#œ œ #œ œ
œ œ
p
? b œœœ œœœ œœœ # œœœ ‰ Œ ! !
b J
b œ œœœ œ œ
&b œ œ œ œ œ œ œ œœœœ œ
r r
7
#œ œ #œ
œ
œ
œ
œœœ œœœ œœœ œœœ œœœ œœœ
3
œœ œœ
3
? bb œ œ œ œ #œ
œ œ œ
œ œ
bb œ œ œ œœ œ œ œ œœ œ œ #œ œ
œ
œ œ œ œ œ
11
& œ œ œ œ œ œ #œ œ
f
œ œ œœ
? b œ œœ # œ œœ œœ œ œœ œ œœ œ # œœœ œœ œœ
b œ œ
b œœœœ œ œ œ œ œœœœ
& b œ œ œ œ #œ œ œ œ œ œ #œ œ œ œ œ
15
? b œ œœœ # œœœ œœ œ œ œœœ # œœœ œœ œœ œœ

b œ œ œ œœ œ œ œ œ
œ. œ
2 048
b œ
& b œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ
19
p
œœœ œœ œœ œ œœ œœœ œœ œœ
? bb œ œ œ
# œ œ œ œ œ œ
# Œ
b œ œ œ œ œ œ œ. œ œ œ
&b œ œ œ œ #œ
r r
23
#œ œ
œœ œ œœ œœ œ
? bb œ œ œœ œ œ œ œ œ œœ # œœœ Œ
œ
b œ œœœ
œ œ œ œ œœ œ œ # œ œœ œ œ œ œ œ
&b œ
27
œ œ œ #œ œ
3
f
œœ œœ œ œ œ œœ œœ
? b
b œ œ œ œ œ œœ œ œœ œ œ œ #œ œ œ
œœ
# œœ œœ
bb œ œ œ œ
œ #œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ #œ œ
31
&
? b œ œœœ # œœœ œœ œ œœ œœœ # œœœ œœ œœ œœ

œ œœœ œœœ œœ
b # œœ œ œ œœ œœ œœ
&b Œ œœ œœ œœ Œ
35
f
? b # œœœ œœ œœ œœœ Œ œœ œœ œœ Œ
b œœ œœ œœ
œ œ
049
œ œœ
Andante
b b 2 œ # œ œ œ œ œ. œ œ. œ œ ‰ œ
& 4 J
f
? b b 42 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œJ ‰
Piano
b j œœœ œ œ œ ‰ œ
&b œ œœœ œ œ œ ‰ œ œ
5
J
œ œ œ œ œ œœ œœœ œ œ œ œ œ œ œœœœœ
? bb œ œ œ œ # œ œ #œ œ J ‰
b j œœœ œ œ œ ‰ œ
&b œ œœœ œ œ œ ‰ œ œ
9
J
œ œ œ œ œœœ œœœ œ œ œ œ œ œ œœœœœ
? bb œ œ œ œ # œ œ #œ œ J ‰
b œ #œ œ œ œ œ œ œ œ
‰ œJ
&b œ œ œ œ œ œ #œ œ
13
œ œ œœœœœ
? bb n œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ J ‰
b b œ #œ œ œ œ œ œ œ œ
œ œ œ œ œ œ #œ œ ‰ ‰
17
&
œ œ œœœœœ
? bb n œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ J ‰
050
Moderato
2 œ œ œ œ œ œ œ œ œ œ œ œ
3
& 4 œ #œ œ #œ nœ œ œ œ œ œ œ œ #œ œ œ #œ œ #œ œ œ œ œ œ œ œ œ œ #œ œ
3 3 3 3
œœ œœ œœ œœ
3
œœ œ œ œœ œœ œœ
Piano
? 42 ‰ œ # œ œ œ œ ‰ œ
œ œ
% œœ œœœ
& œ œ œ œ œ #œ œ nœ œ œ œ œœ ! r
œ œ œ œ œ œ #œ œ œ #œ œ œ œ œ œ œ #œ œ œ œ œ œ r! ‰
5 3 3
œ
3
œœ œ
œ ƒ
rit.
œ œ
3
? œ # œ œœ œ œœ œœ ‰ # œœœ œœ œ œ j‰
œ œ œ œœœ
œ
œ œ œ œ œj
3
‰ œ œ œ œ œj ‰ ‰
3
‰
r
& œ œ œ œ œ œ œ œ œ œ œJ ! œr œ œ œ œ œ œ œ œ œ œ œJ ‰ ! œ
10
‰
p
œ œ œ
? œœœ œœœ œœœ œ œœ œœ œœ
cresc.
œœ œœ œœ
3 3
œ œ
j j
‰ ‰ œ ‰ ‰ œ œ
œ œ œ œ œ œ œ œ œ œ œ‰ œ œ! œr œ œ œ œ œ Uœ œ œ œ œ œ œ œ ! r
œU
14 r r
& J J ‰ œ
#œ œ œœ œœ œœ œ œ œ œœ œœ œœ
dim.
? œ œœ œ œœ œ œ œ œ œ # œœ œœ œœ œ œ œ
œ œ œ œ œ
2 050
& œ œ œ œ œ œ œ œ œ œ œJ ‰ ! r œ œ œ œ œ œ œœœœ œ ‰! #œ
18
œ J R
œœ œœ œœ
œ # œ # œ œ œœ
œ œ œ œ œ œ œœ œœ
3
œ
3
cresc.
? œ œ œ œ œ œ œ œ
#
œ œ œ #œ #œ œ œ œ œ œ œ
‰ ‰ œŒ œ œ œ œj
3
œ œ œ# œ œ Uœ U j
J œ œ #œ œ œ œ œ œ œ œ œ œ œ
3
œ œ œœœœ œ
22
&
U f
? œ # œœœ œ œœœ # œœ œœ
rit. 3
œ œ œ œ œ
œ œ œœ œœœ œœ œ œ œœ
œ # œ # # # œ œ œ œœœœœ œ œ œ œ œ œ œ œ œ œ
‰ œ œ œ œ œ
26
& œœ œ œ
f accel. 3
œœ œ œ œ œ œ œ œ œ
œ œ œœ œ œœ œœ œœ œ œ œ œœœ œ œœ œœ œœ
accel.
? œ‰ Œ
a tempo
###
a tempo accel. 3
œ œ
J
### œ œ œ œ œ U n œ U
œ U œœ
œ
œ œ #œ œ œ œ œ œ œ œ œ œ J œ œ œ œ œ œ œ œ œ œ œ œ œ ‰ ‰ œ œ œ œ ‰ œ‰
31 1, 2. 3.
& J
3
#œ
# œœœ
3
œœ œœ œœœœ œ œ œœ œœ j
rit.
œ œ œ œ œ œœ œœ ‰
3
? ### œ œ œ œ n œœ‰
œ œ œœ
Ao sinal duas vezes
051
b
& b b b 42
œ œœœ œ œ œ œ œ œ œœœ œ œ œ œ œ
Piano !
? b b b 42 œ œ œ œ
b œ œ œ œ
bb
&bb œ œ œœ
2
œ œ œ œ œ œœœ œ œœœ œ œœ œ œ œœœ œ

œ œ œ œœœ œ œœœ œ
? bb œ œ œ œ œ œ œ œ œ
b
& b bb œ œ œ œ œ j j
4
œ œœ
œ œ
œœ œœ r œœ
? bb b " œ " œ " œ " œ " œJ J œ
b œ J
œ œ œ
U r
bbbb œ. œ œ. œœ œœœœ j œœ
œ " œœ
6
& œ œ œ œ œ
œ
œ œ
œ > œ
U
œ
? bb œ œœ œœ œœ
bb r " œJ œ " œ " œ " œ "
œ
œ œ
bb j j
&bb œ œ œ œ œ
8
œ œœ
œ œ
œœ œœ r œ
? bb œ œ œ
bb œ " " œ " " œ " œJ J œ
J
œ œ œ
U r r r
2 051
b œ œ.
& b bb œœ
. œœ œœœœ œ
œœ " œœœ " œœœ "
10
œ > œ œ œ œ œ œ œ
U
œ
? bb b œ gg œœ gg œœ g
g
œœ
b r " œJ œ " gg œ " gg œ " gg œ "
œ œ
œ
r r r
b œ. œ œ œ œ œ >œ œ œ œ œ œ " œ " œ "
& b bb œœ œ œœ
12
œœ œœ œœ
‰ R " ‰ R "
œ
? b b b œ " gggg œœ " ggg œœœ ggg œœœœ ggg œœœœ ggg œœœœ
b œ " g " œ " g " g " g "
œ œ œ
r r r
bbbb œ . œœœ œ œ œ œ œ œ œ œ œ œœœ " œœ
œ " œœ
œ "
14
& œ œ œ œ
>
œ œ g œœ g œœ ggg œœœ
? bb œ œ " ggg œ " ggg œ
b b œ " œ " œr " œR " œ " g "
œ œ
b b œ. œ œ >œ œ œ œ œ œœr " œœ

œ œ œ " gg œœœ
&bb gg œœ gg œœ "
16
gg gg gg œ
œ
œ œ
œ g œœœ g œœœ ggœœœœ
? bb g g
bb œ " œ " œ " œ " œ " ggœ " ggœ " gg "
œ œ œ
j j
bb j œ œ
& b b œ. œ œ œ œ œ œœ œœœ œœ ‰ " œ œ œ
œ œ œ.
18
œœ J œœ J œ
J p
œœ œœ J g
œœ
rit. 3
? bb œ
œ " r " œœ " œ
bb œ " œ R œ " " ‰ Œ
œ œ
œ
œ œ œj
051 3
b œ œ œ œ œ œj œ œ œ
& b bb œ . œœ œ œ.
21
œ
œ œ œ
œ œ J J œ J J
? bb b œ œ
b œ
œ
b
& b
b b œ. œ œ œj œ œ œ œ. œ œ œj œœ
œ œ œ
œ
23
œ œœ œ
J œ œ J
œ J J
? bb œ œ
bb
œ œ
bb
& b b œ. œ œ œj œœ œ œœ œ œ. œ œ œj œ œ œ
25
œ œ
œ œ
œ œ J J œ J J
? bb œ œ
bb
œ œ
U
b b b œ. œ œ œj œ œœ œ œ. œœœ œœ ..
& b
27
œ œœ œœ œ. œ
œ J J œ J
? bb œ œ "
bb œ
œ
4 051
b
& b bb œ œ œ œ œ j j
29
œ œœ
œ œ
œœ œœ r œœ
? bb b " œ " œ " œ " œ " œJ J œ
b œ J
œ œ œ
bb U
œ. r r r
b œ. œ œœ œœœœ œœ " œ œœ " œ œœ "
& b œ
31
œ
œ
> œ œ œ œ œ œ œ
U
œ gœ gœ gg œœœ
? bb
bb r " œJ œ
œ " ggg œœ " ggg œœ " gg "
œ
œ œ
bb j j
&bb œ œ œ œ œ
33
œ œœ
œ œ
œœ œœ r œ
? bb œ œ œ
bb œ " " œ " " œ " œJ J œ
J
œ œ œ
bb U
œ. r
b œ. œ œœ œœœœ j œœ
œ " œœ
& b
35
œ
œ
> œ œ œ œ œ
œ
œ œ
U
œ
? bb œ gg œœ gg œœ g
g
œœ
bb r " œJ œ " gg œ " gg œ " gg œ "
œ
œ œ
œ œ œ œ œ œ œ œ. œ.
051 5
b œ.
& b bb œ œ
œ
37
ggg œ " ‰ gg œ ggg œœœ " ggœœœ œ

‰ R gg R " ‰ gR " ggg œœ "
R R
œ gg œœ
? b b b œ " gggg œœ " gg œ " ggœœœ " ggœœœ " gggœœ
œ
b œ " œ " g g "
œ œ œ
b b œ. r
b
& b œ œ œJ œ. œ œœœœ j œœ
œ " œœ
39
> œ œ œ œ œ
œ
œ œ
œœ œœ gœ gœ gg œœœ
? bb
b b œ " œ " œr " œR " œ " ggg œœ " ggg œœ " gg "
œ œ
b b œ. œ œ œ œ œ >œ œ œ œ œ
&bb œ ggg œœœ " gg œœ "
41
J g gg œ
? bb œœ
œ
œœ ggœœœœ ggœœœœ ggœœœœ
bb œ " " œ " œ " œ " gg " gg " gg "
œ œ œ
j j
œ œj
b bb
& b œ. œ œ œ. œ œ œJ œ œ œ œ œj œœ œœ ‰ Œ
43
œœœ œ œ
! J g J
œœœ œœœ œœ ‰ Œ
3
? bb
b b œ " œ " œr " œR " œ " œ " j
œ œ œ œ.
!
052
U
Muito Vagamente
" œœœœœ
rit.
& b 42 ! ! ! !
R
œœ
j j #
" œ ‰ œ " œ ‰ œj " œ ‰ œ " œ ‰ œj U
Piano
? b 42 œ œ œ œ œ œœ œ œ œ œ œ "
œ œ b œ œ œ. œ œ œ
œ bœ œ œ œ œœ œ œ œ œ œ
.
%
& b œ. œ œ œ œ œJ œ œ œ. œ œ œ œ œœ .. œœ œ
6
?b œœ œœ œ œ œ œ œ œ œœ
œ œ œ œœ œ œœ
œ œ œœ œ œœ œ œ œ
œ
œ
& b œ. œ œ œ œ œJ œ œ œ. œœ œœ œ œ œ œ œ œ
10
? œ œœ œ œ œ œ œ œœ
b œ œœ œ œœ
œ œœ
œœ
œ œœ œ œ œ
œ œ œ œ œ
U
. œ. nœ œ œ œ œ œ
&b œ œ œ œ œ œ œ œœœœ
14
œœ
œ œ œ œ nœ U œ
?b
œ œ #œ œœ œ œ œ œœ œ œ # œœœ
˙
œ
œ œ #œ œ
2 052
œ œ œ œ œ œ œ œ U œ. œ œ
&b "œ œœœ œ œ œ
18
œœ
# rit.
? œ œ # œœœ œ œ œ œ œ œœ œ
b œ œ nœ œ bœ œœ œ œ ˙
˙
œ
U
œ %
rit.
&b ! ! ! " œ œ œR
22
œ
œœ
j
" œ ‰ œj " œ‰ œ " œ ‰ œj U
Fim
?b œ œ œœ œ œ œ œ œœ "
œ œ œ bœ œ œ œ œ
œ. œ œ œ. œ œ œ œ
053
j j
& 42 œ œ œ œ j
œ œ œ œœ j ‰ ‰ j
˙ œ œ
F
œ œœ œ œœ œ œ
Piano
? 42 ‰ œ œ œ œ œ œ œ
œ œ œ œ
œ
j
& .. j
5
œ œ œ œ œ œ
? .. œ œœ œœ œ œœ # œ b œœ œ œœ œœ œ œ œ œ
œ œ
j j ‰ œj
&œ œ œ œ œ œ
9
œ œ ˙ œ
?œ œœ œœ œ œœ œ œ
& œJ œ œ œ œ œ œ œ œœ œœ
13 3
J œ œ œ œ
œœ œœ œ œ
?
œœ œ œœ
œ
œœ
œ
œœ
j j ‰
17
&œ œ œ œ œ œ œ œ ˙ œ j
œ
? œ œœ œœ œœ œœ œ œœ œ œœ œ œ œ œ
œ œ œ
2 053
j
21
& j œ œ œ œ œ œ
? œ œœ œœ œœ b œœ
œ #œ œ œœ œ œ
œ œ œ œ œ
j j ‰ œj
œ œ œ œ œ œ
25
& œ œ ˙ œ
?œ œœ œœ œœ œœ œ œ
& œJ œ œ œ œ œ œ œ œœ œœ
29 3
J œ œ œ œ
? œ œœ œ œœ œ œ
œ œ œ œ œœ œ œœ
œ
œœ œœ
œ œ
œœ œœ
j j ‰ j .. œœ œœ Œ
33 1. 2.
&œ œ œœ œ œ œœ ˙ œ œ œ
? œ œœ œœ œœ œœ œ œ
œ œ œ œ œ œ œ œ .. œ
œ
œ Œ
œ œ œ œ œ
054
Moderato
#
& 42 ‰ œ œ œ œ ‰ œ œ œ œ ‰ œ œ œ ‰
œ œ œ œ œ œœ œœœœ
f
? # 42 œ œ ggg œœœ ggg œœœ œ gg œœœ gg œœœ œœ g œœœ g œœœ œœ g œœ œ œ
Piano
œ gg gg œ g
g g g œ g
g g
œ g g œ œ g g
œ œ
5
# ‰ œ gœ g œ ‰ œ œ œ ‰ œ œ œ ‰ œgœ œ
& gg œœœ ggg œœœ ggg œœœ gg œœ gg œœ gg œœ gg œœ ggg œœ gg œœ gg œœ ggg œœ gg œœ ggg œœ gg œœ
gg g g gg g g g g g g g g p
?# œœœœœœœœ œœœœœœœ œœœœœœœ œœœœœœœ œ œ ‰
œ œ œ œ
œ
# œ. œ . œ j œ. œ œ œ œ œ
œ œ.
10
& œ œ œ œ. J
?# ggg œœœ gg œœœ gg œœœ ggg œœœ ggg œœœ œ

ggg œœ ggg œœœ œ
ggg œœ ggg œœœ gg œœœ gg œœœ
œ g gg gg œ g g g œ g g œ g gg gg
œ
# œ œ. œœ j œ. œ œ œ œ
& œ. œ œ. !œ
14
œ œ. R
?# gg œœœ ggg œœœ ggg œœœ gg œœœ gg œœœ œ

gg œœ gg œœœ œ
gg œœ gg œœœ ggg œœœ ggg œœœ
œ gg g g œ gg gg gg œ gg œ
gg œ gg g g
2 054
# .
& œ œ œ. œ œ œJ œ œ œ œ œ œ œ œ œ. ‰
18
?# gg # œœœ gggœœœ ggg œœœ ggg œœœ gg œœœ ggg œœœ gg œœœ ggg œœœœ ggg œœœœ ggg œœœ gg œœœ ggg œœœ
g gg œ g g gg
œ gg g g œ œ œg g
#
& ! œ œ œ œ j ! œR œ œ œ œ œ œ
22
œ #œ œ œ œ œ œ œ œ œ
f
?# ggg # œœœ gggœœœ ggg œœœ gg œœœ ggg œœœ gg œœœ gg œœœ gg œœœœ gg œœœœ œ gg œœœ ggg œœœ
g g g g gg œ gg gg œ œ gg g
œg œ œ
# ‰
œ œ œ ‰ œ œ œ œ œ
‰ ‰
26
& œ gg œœœ ggg œœœ

œ œ œ œ œ œ œ gg g
? # œ œ ggg œœœ ggg œœœ œ gg œœ gg œœœ gg œœœ œ gg œœ gg œœ œ œ œ œ ‰
œ g g œ g g œ œ
œ œ
055
j . œœ œ œ œœ .. œ
& b 42 œ œ œ œ #œ œ œ œ #œ œ. J
3
œœ œœ œœ # œœœ œœ œœ
3
# œœœ œœœœ
Piano
? b 42 ‰ œ œ œ #œ œ œ œ œ œ œ œœ œœ ‰
œ œ J
œœ .. œ œ œ œ œ œ œ œ œ œ œ œ œ. œ
& b .. œ
. œ œ J
5
œ œœ œœ œœ œœ œœ œœœ œœœ
3
? b .. œ œœ œ œ œ œ œ œ œ œ œ œ
œ œ
œœ .. œ œ œ œ # œ œ œ # œœ œœ œœ œ œ œ œ
& b #œ. œ œ . J
9
œœ œœ
3
# œœœ œœœœ # œœœ œœœœ # œœœ œ

3
? #œ
b œ œ œ œ œ œ œœœ œ œ
œ
œ
œ œ
œœ .. œ œ œ œ U
. œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ
&b œ œ
13
œ œœ œœ œœ œœœ œœœ œœœ

3
? b œ œœ œ œ œ œ œ œ
# œ Œ
œ œ œ œ
œ œ œ œ Uœ œ œ
‰ œj
2 055
œ œ œ œ #œ œ
&b œ œ
17
œœœ U œœœ # œœœ œœœœ œœ œœ œœ

?b ! œ #œ œ œœ # œœœ œœ ‰
œ œ J
œœ œ œœ ..
&b œ
. œœ œ œ # œœ œ œ œ #œ œ. œ
21
J
3
œ œœ œœ œœ œœœ
3
# œœœ œœœœ # œœœ

? b œ œœ œ #œ œ œ œ œ œ œ # œ
œ œ œ œ
œœ .. œœ œ. œ œ œ œ
œ. œ œ œ œ J œ. œ
&b J
25
3 3
œœ œœ œœ œ œœœ œœœ œœœ œ

? œ œ œ œ œ œœ œ œ œ # œœœ
b œ œ œ œ #œ
œ. œ œ œ œ œœ .. œ œ. œ œ œ œ œœ .. j
&b #œ. J #œ. œ
29
3 3
œœœ œœœ # œœœ œœœœ œœœ œœœ # œœœ œœœœ

?
œ. œ œ œ œ œœ œ œœ .. œ œ œ # œ œ1.. œ œ2. > œj

&b # œœ J .. œ œœ ‰
33
3 >
œ œœ œœ # œœ œœ œœ œœ œ œœ
3
? œ œœ # œœœ œœœœ
b œ
œ #œ œ œ œ œ œ œœ œœ œœ
œ J ‰ . œœ # œœœœ œœ ‰
.
J
056
œ œ œ œ œ œ œœœ œ œ œ œ œ œ œ œ œ œ œU
# # 2 œ œ œ œ œ œ œ œ œ œ œ œ œœ œœ œœ œœ
# œ œ œ œ œ œ œ œ œ œ ! ! r nnn
& 4 œ
!œ R œ
œœ œœ œ œ œ œ
œ œœœ œ œœœ œœ U
Fine
# œ œ œ œœ œœ œœ nnn
Piano
? # # 42 ! ‰ œ œ œœ ! ‰
œ œ œ œ
œ œ. œ œ
& œ. #œ œ œ œ œ œ. œ œ. œ œ. œ œ œ œ œ œ œ œ œ
5
‰ œ
œ # œ œœ œ œ œ
œ œ œ œ œ œœœ œ œœ œ œ œ
?œ œ œ œ œ
œ œ œ œ œJ œ œœ œ œ œ œ œ œ !
& œ. œ.
9
œ
œœ œœ œ œœ œœ œœ œœ œœ œœ œœ œ
? #œ œ œ # ‰˙ # œ œ œœ œ
œ
f
œ œ. œ #œ œ
& œ. œ. œ
œ œ. œœ œ .
13
œ. J œ
œœ œœ œ œ œ œœ œœ
? œœ œ #œ œ œ œ œ œ œ œ œ
œœ œœ œœ
œ
2 056
###
&œ!œœœœœœ œœœœœœœœ œ œ #œ œ œ œ œ œ œ œ #œ œ œ
17
R œœœ
œ œ œ œ œ œ
œ œ œ # œ œ œ œœ œ #œ œ œ
rit.
? œ œœ œœ œœ œ œœ œœ œœ ###
### . œ œ œ œ œ. œ. œ œ œ
.˙ œ œœœ œœœœœœœœ œ œ œ˙ œ œ œ œ œ œ œ
21
& ˙ œ œ
œœ œœ œ œ œ œ œœ œœ œœ œ œœ œœœ
Vagaroso
#
? # # .. œ œ œ œ œ œ
U œ œ œ
### œ
œ œ œ œ œ œ œ œœœœ œ œ œœ œœ œœ
3 3
œ œœ œœ œœ .. œ œ œ œ œ
25 1. 2.
& ˙ œ œ
U R rit. R
? ### œ œ œ œ œ . œ
D.C. al Fine
œ ! ‰ . œ ! ‰
057
%-
# # # 2 œ. œ œ œœ- œ. œ # œ n œ. œœ- œœ- œ. œ œ œ. œœ- œœ- œ œ œ œ. œœ- œœ- œ. œ œ
Allegretto
& 4 œ. œœœ œ œ œ œ œ œ œ
F
? # # # 42 ! ‰
Piano
œ
. . .
# # # œ. œœ- œœ- œ. œ # œ n œ. œœ-
œ œ œœ- œ. œ œ œ œœ- œ œ œœ- œ œœœ œ U nn
œœ ‰ n
5
& œ œ œ œ J
œœ œœ œœ œ œ œ œ œ œ œ. œ. U n n
Fine
? ### œ œ œ œœ œœ œœ œœ œœ œœ
œ œ œ œ. ‰ n
œ œ œœœœ œ œ
Meno
& œ. œ œ œ œ œ œ œ œ ‰ ‰
9
J
p
. . œ. g œœ
œ . . œ. œœ
œ . . œ. g œœ
œ . œ. œ. œ
.
? . œ œ ! g œ .
gg ! . œ œ ! g œ . .
gg ! œ œ œ ! g œ .
gg ! . œ œ ! œœ !
œ œ œ
œ
œ. œ œ œœœœ œœœ ‰ œ œ œ œ œ #œ j
œ. œ ‰
13
&
. œ. œ . œ. œ. œ. œ. œœ œ œœ œ # œœ œ œ. œ. œ. .
.
? œ œ. # œ œ
! œœ ! œ. œ. œ œœœ œ œ œ. # œ œ. œ. .
œ œ. œ œ œ
.
œ. œœœœœ œ œ ‰
& œ. œœœœœ œœœ ‰
17
J
. . œ. g œœ
œ . . œ. œœ
œ œ œ ! œœœ . . œ. œœ
œ
? . œ. œ ! œ
gg ! œ. œ ! gg œ œ œ ! gg œ
œ g œ. g ! œœ ! œ. œ. g !
œ
œ. œ œœœ œ j ‰ œ. U ###
2 057
œ œ
poco più
œ œ œ œ œ œ œœ œ œ œ! ! œ
21
& J
f œ œ œ . œ
œ œ ! œœ ! œ œ œ œ œ œ œ œ œ. # œ œ œ œ œ ! œœ !
œ . œ. œ.
? œœœ
dim.
œ œ œ œ #œ # # # œ œ œ. œ œ œ.
œ œ œ
œ
### j œ.
‰ œœœ œ œ œ Œ
25
& œ. #œ œ œ.œ #œ œ œ
œ. . œ. . œ. . œ œ. . .
œ .
œ .
œ .
œ
# œ
? ## œ œ œ œ œ œ
œ œ œ œ œ œ œ. œ œ œ. œ œ œ œ œ œ œ.
.
### œ . œ œ œ œj
œ ‰ œ. #œ œ nœ œ œ œ ‰ ! œr
29
& œ œ
œ. . œ. . œ. . œ. . œ. . œ. . œ œ. œ
.
œ
? # # # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ. œ œ.
33
### j ‰ œ. œ #œ nœ œ j ‰
& œ.
#œ œ œ . œ #œ œ œ œ œ œ
.
œ .
œ .
œ .
œ œ. . œ. . ..
œ ..
œ
# œ
? ## œ œ œ œ. œ . œ . œ . # œ œ
œ œ œ œ œœ œ œœ œ
œ œ œ œ œ œ œ œ #œ œ œ œ
### œ . œ œ œ œ œ œ. œœœ œ œ r œ. œ œ
œ œ ! œ
37
& J œ R œ œ
f
œ .
œ .
œ .
œ .
œ .
œ .
œ
? # # # œ œ . œ. œ n œ œ œ. œ œ œ œ. œ œ œ. œ œ œ. œ œ œ. œ œ œ. r !‰
D.S. al Fine
œ œ allarg.
058
Fox Medium
>
r œ bœ œ nb b b b
& b 42 œœ
bœ
œ œ
nœ œ œ œ œ b œœœ œœ ! ‰
œ œ
F R p
œ œœ œœ
Piano
? b 42 #œ œ œ œ r! ‰ nb b b b
œ
bb b . œ œ œ n œ b œ œ œ n œ b œ œ n œ œ œ œ b œ n œ b œ œ œ n œ b œ œ œ n œ b œ œ œ œ œ
& b .
3
œœ œœ nœ œœœ n œœœ œœœ œ

? b b b .. œ œ œ œœ nœ œ œ œ
b œ œ œ
œ œ œ œ œ œ œ œ œ. œœ œ œœœœ œ œ nœ
bb œ nœ œ œ R œ
&bb
7
œœ œœ n œœœ œœœ nœ œ
3 3
? bb œ œ œ œ œ œœ œ œœ œ œ œ
R ! ‰
bb
bb b b œ œ œ n œ b œ œ œ n œ b œ œ n œ œ œ œ b œ n œ b œ œ œ n œ b œ œ œ n œ b œ œ œ œ œ œ œ œ
& b
11
œœ œœ nœ œœ œœ
? bb b œ œ œ œœ œœœ nœ
n œœœ œœœ œ œ œ
b œ œ œ
b œ œ n œ b œ œ n œ b œ œ œ œ œ œ ~~ œ œ œ n œ œj ‰
œ œ œ
2 058
b nnnnb
& b bb œ œ nœ œ
15
œ œ œ œ œ n œœ
œœœ œœœ œ œ n œj
? bb b n œ œ
b œ n œ œœ ‰ œ nnnnb
rœ œ œ
& b œœ œ b œœœ œ n œ n œœœ œ œ œ œ nœ œœ œœ
19
œœ
œ œ œœ ‰ œ œ œ ! ‰
R
J
œ œ œœ œœ # œ r
? b œ œ #œ œ œ
œ œœ œ œ œ œ ! ‰
œ œ œ œ œ œ œ œ œ œœœ œ œ b œ œ
œœ
œ ! œ œ bœ œ
&b ! œœ
23
œ
‰ œj ‰ œj ‰ œj œ œœ
? œ œ bœ œ œ œœ Œ
œ œ œ œj ‰ œ œ œ
b J ‰
> > >
œ œ nœ œ j
& b œœœ œ œ œ œœ œ n œ œœœ .. œ
b œœ
27
n œœ œœ
œ œœ ‰ .
J
? œ œœ # œ œ œœ œ œœ # œ œ
b œ œœ œ œ b œ
j ‰ >œ r œ b œ œ .. j j œœ
& b œ œ b œ œ œ œ œ œ œŒ b œœ œœ œœœ
œ‰
œ
œœœ ‰ œœ ‰ Œ
31 1. 2.
n œ œ œ œœ œœ ! ‰ œ
bœ œ J
j R S Sj
˙
? b ˙˙˙ œ
œ ‰#œ œ œœ œ œœ œ œ r ! ‰ .. œ ‰ œj ‰ œ‰Œ
b œ œ J œ
059
.. œ œ œ >œ œ œ >œ
Moderato
& b 42 Œ ‰ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ . œ œ œ œ œ
p > >
? b 42 œ œ œ œ œ œ œ .. # >œ œ œ œ >œ œ œ œ >œ œ œ>œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
Piano
œ œ œ #œ œ œ
>œ
& b œœ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ œ œ œ ‰ œ
6
> #œ œ J
? b #œ œ œ œ œœœ œœœœ œœœ œœœœ œœ œ œœœœ œœœ

œ œ œ œ
œ j
&b œ ‰ œ œ #œ œ œ œ œ ‰ œ
10
œœœ œ œ J
? œ œ œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ œ
>œ j
b œ œ œ ‰ œj ‰ œ n##
14
& œ œ œœœ œ œ
>
œ œ œ œ œ œ œ œ>œ>œ
dim.
?b œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ n##
## œ j
2 059
œœœ œ j œœœ œ
œ œ. œ
18
& œ. œ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ
## œ œ n
œœ œ œ œ œœœ œ ‰ œ œ .. n b
22
& œ. J œ
? # # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ n œ œ .. n n
œ œ œ œ b
>œ >œ
b œ œ œ œ œ œ œ œ œ œ œ œ œ œ. œ œ œ œ
26
& œœ œœ œ
F> >
? # >œ œ œ œ > œ œ œ >œ œ œ œ > œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ
b œ œ œ
>œ
& b œœ œ œ œ œ œœ œœœ œœ œ œ œ œ œ œ œ Œ
30
> #œ œ
? # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ Œ
b œ œ œ
060
2 œœ .. œœœ œœ œœ œ œ #œ œ œ
b
& 4 œ. œ œ œœ .. œœ .. ‰
J J
f
œ œœ œ œœ # œœ œ
3
? b 42 œ œœ œ œœ œ # œ œœ œœœ œœœ
Piano
œ œ œ #œ
œ œ œ œ œ
œ
&b j œj œ j ‰ !
5
œ œ œ œ. œ
3
œ. #œ œ œ œ œ œ
? œ œ # œœ œœ œœ œ œœ œ œœœ j
b œœ œ # œ œ œ #œ œ
œ œ œ ‰
&b œ !œ œ œ !œ œ œ œ œ. œ !
9
œ œ œ
œ œ œ
œœ
3
? b œ œœ œœ œ œœ œ œœ œ œ
œ œ œ œ
œ œ œ
œ
œ œ œ œ œ
&b œ !œ œœ ! œ œ œ œ œ œ œ. œ !
13
œ
œ œ œ
œœ œœ œ œœ œ œœ œœ œ œœ œ
3
? œ œ
œ
b œ œ œ œ œ œ
œ œ
U
U œj
& b .. œ ! œ œ !œ œ ! œ œ œ œ.
17
œ œ œ œ ‰ g œœœ gg œœœ
U gg J g J
? .. œ œœ œœ
œ
œœ œ œœ œ œœ œ œœ g œœ .. ‰
b œ œ œ ggg œœ ..
œ œ g
U
n##
2 060
& b œ œ œœ œ œ œ j ‰ ! œ .. j ‰ ‰ # œj
21 1. 2.
#œ œ œ ˙ œ œœ ˙ œ
j œœ œ œ œ
? b œ œœ œœ œ œœ œœ œ
œ œ
œ
œ œ œ œ
œ œ ‰ .. œ # œ
œ œ #œ n##
œ œ œ œ œ
œ œ œ œ œ #œ
A
## . j j
& . œœ œœ œœ œœ n œ œœ
œ œœ œœ œ œ œ œ . œœ .. j
27
œœ .. œ. œ
œ œœ œ œ
3
? # # .. œ œœ œ n œ # œœ œ œ
œ œ n œœ œ œœ œœ œ œœ
œ nœ œ œ
œ
## j
& œœ œœ œœ œœ œœœ œœœ œœœ œ œ œ œ œ œ œ . ‰ œj .. œ œ œ ‰
31 1. 2.
3 3
œ œ. œ.
? # # œ œœ œœ œ œ œœ œ œœ œœ œ œœ . œœ œ œœ œ œœ œœ
œ œœ œ œ œ œ . œ
œ œ œ œ œ
## . j j
& . nœ.
37
œ œ nœ œ œ œœœ .
3
œ œ.
œ
? # # .. œ n œœ œœ n œ œ œ œœ œ œœ
œ
œœ œœ
œ
œœ
œ œ nœ œ œ
## fi
j œj œ œœ #œ œ œ. œ. ‰ .. j œ
41 . 1. 2.
& œ nœ
bœ œ œ œ œ.
3
œ. œ
œ œœ œ œœ
? # # œ œœ œ œœ
œ
œœ œœ
œ œ # œœ œ œ œœ œ .. œ œœ œ œœ œ n œœ œ œœ
œ #œ
## U fi j
j œ
Coro ao A e 3ª vez
# œ œj œ j ‰ ! n œ j ‰ n œœ ‰
47 3.
& nœ
bœ œ œ œ œ.
3
œ. œ œœ œ œ
œ j j j
? # # œ # œœ œœ œœ œœ œ n œœ œœ œ n œœ œ ‰ œ ‰ œ ‰
œ œ œ œ
œ œ œ œ œ œ
Do % ao fi
061
U r
r
& 42 œ œ œ œ œ b œœ œ œ œ œ œœ œœ œ œœ b œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œ œ
œ œœ œ
œ #œ œ œ œ œœœ
U œœ
œœœ œ œœ b n œœœœ ! ‰ œœœ
a tempo
œœœ
Piano rit.
? 42 ! ‰ #œ œ œ œ œ r!‰
œ R œ
#œ œ œ œ œ œ œ
% ! œ œ œ U ! œ
œ
œ
œœœ œœœœœœœœ œ
œ œ
œœœ
5
& œ œ œ œ œ œ
œ œ œ œ œ U
œ œœœ œœœ œœœ œœ œœ œœ
? œ œœ œ b œ œ œ #œ œ œ œ
œ œ b œ œ œ œ œ œ œ
œ œ œ
U
œœœœœœœœ ! œ œ œ œ
œ œ œ œ œœœœ œ
œœœœ œœœœ
9
& œ
U
? œ œœ œœœ œœœ œœœ œœœ œœ œœ
œ œ bœ œ œ œ œ œ!‰
œ œ bœ œ œ œ œ œ R
œ œ
! œ œ œ !œ œ œ
œ œ
œœ œœœœœœœœ œ œ œ #œ œ
13
& œœœ œ œ œ
œœ œ œ œ
œ œœœ œœœ œ œœœ # œœœ œœœ
? œ œœ œ b œ œ œ œœ #œ
œ bœ œ œ œ œ œ
œ œ œ #œ œ
"r "
2 061
œ
& œ œ œ # œ œ œ œ œ œ œ n œ œr n œ œ œ œ œ œ rœ œ
œ œ œr œ œ œ
17
œ œ
n œœ œœ œœ bUœœ œ
? #œ œ œ œ œ œ # œœ ! ‰ # œœœ ! ‰ n œœœ ! ‰ œ œ r! ‰
#œ œ œ R R R
œ œ
œ œ œ œ
& .. œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œ œ œœ œœ œœ œœ œœ œœ œœ œœ œœ
œ œ œ œ
21
œœ œ œ œ
? .. œ œœœ # œ œœœ œ œœ œœ œ
œ œœ œœœ œ
œœ œœ
œ œ œ
U
r .
& # œœ œœ œœ b œœ œœ œœ œœ œœ œœ œœ œ œœ b œœœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œ œ .
25 1.
n Uœ œœ œ
a tempo
œœœ œœœ œœ b œœœ ! ‰

œ œ œ œœ œ œœ œ œœ ..
rit.
? #œ œ
œ R
% œ
r œ
œ œ œ œœ œœ
.
& œœ œœ œœ œœ œœ œœ œœ œœ ‰
29
œ œ œ œ œ œ J
?œ œœ œœœ œ j
FIM
œ œ r ! ‰ œ œ ‰
œ œ œ œ œ œ
062
Lento
& b 42 ! œœœ œœ œœœ œœœ .. œœœ œœœ œœœ œœœœœ œœ œ œ œ œœœœ œœ œœ # œ œ œ œ œœ œœ œœœ œœ œœœ œœœ
œœœ œ œœ œ œ œœœ œ œ œ œ
" ß #œ
œœ œ œ # œ œœœ œ œœœœ
? b 42 œ œœœ .. œ œ œ œœ œœ œœœ œ œœœ œœœ
Piano
#œ œ ‰
œ œ œ # œ œ œ
œ œ
r
& b œœœ œœœœ œœœœ œœœ œ œœœ œ œœ œ œœœ œœœ œœ œœœ œ œ œ œœœœ œ j !œ
5
œ œœ
œœœ œ œœ œ # œœœœ œœœ
? b œ œœœ œ œ œœ œ œ #œ œ œœ œr œ œœ ‰
œ œ œ œ œJ
œ œ #œ œ
Lento
& b œ. œ œ œ œ œ œ #œ. œ œ œ œ œ œ œ œ œ !œ
9
J R
p
œ œ œ œ
? œ œ œ œ œ œœœ œœ œœ
#œ œ œ œ œ #œ œ
# œœœ œ œœœ
œ œ
œ œ œ œ
b œ œ œ
œ œ œ #œ œ œ
œ œ œœœœ œ œ œ œ œ #œ œ œ œ. œœ
&b ‰
13
J
œ œ # œœœ !œ œ ! j
? b œ œ œœ œ œ œœ œ
œœ
œ
œœ
œœ
œ œ œ
œ œ œ œ œ œ ! œ ! œ ! œ‰
œ œ œ œ œ
œ
œ. œ
2 062
b œ #œ œ œ œ œœ œ œ œ œ œ œJ œ œ
17
& J J
p
# œœœ œœ œœ œœœ œœœ œœ œœ
? b œ œœœ œ œ œ œ œ œ œœœ œ œ œ œ œœ
œ œ œ œ
j j U œ œœœœœœœœ
&b œ œ œœ œ # œ
21
œ œ
œ œ œ œ
U ‰ J
? # œ œ œœœ œ œ œœœ œ
œœ
œ
œœ œœœ # œœœ " œœœ œœœ œœœ
b œ œ œ #œ #œ œ œ
#œ œ œ œ œ #œ œ
^^ ^ ^^
! œœœ œ œ œ ^^
œ # œ œ œ œ œœ œ œ œ n # # œ œ . œ œ . œ œ . œ
& b ‰ œœ œœ œœœ œ œ œ œ œ œ œ œ œ œ
25
œ
J
? œ œ œ œ # œ # œœœœ œœœœ œœœœ # œœœ œœœœ œœœœ
dolce
œ # œ œ œ
b œ n# œ
œ #œ œ œ
## ˙ œœ œ . œ œ . œ œ . œ œ œ .
œ œ œ œ. œ œ. œ ˙
29
& Œ ‰ J J
? # # ‰ œœœ œ œœœ œ œœœ œ œœ œœ œœ œœ

œ
œœ
œ
œœ
œ ‰ œœœ œ œœœ œ œœœ œ
œ œ œ
œ œ œ
## . œ œ. œ œ. œ œ. œ œ. œ ˙ œ œ œ œ œ œ
062 3
œ œ . œ œ œ œ. œ œ. œ œ. œ
œ œ . œœ
33
&
‰ ‰ Œ
œ œ œ œœœ œœœ œœœ œœœ œœœ œœœœ
? # # œ œœœ œœœ œœœ œ œœ œœ œœ œ œ œ œ œ œ œ
œ œ œ œ
U 1.j
# ˙ j œ œ œ œ œ œ œ ‰
& # œœ . œœ . œ œœœœœœœœ œ œ ! œ œ œ b œœ œœ ..
37
‰ ! . RÔ R n œ
‰ ‰ ! œœ œœœ
œ œ œ œ œ œ œœ œœ œœ
? # # œ œœ œœ œœ œ œœ œœ œœ œ
œœœ œœœ œœœ œ œ œ œ n œœœ œœœ œœ .
œ .
œ œ œ œ
œ œ
œ œ œ #œ œ. œ
Piu Lento
## œ œ nn
‰ b ‰ œœœœœœ #œ.
42 2.
& J
^ ^
? # # œ n œœ œœ œœ ! ! n n œ œœ œ œœ ! ^ ! #œ œ #œ œ
œ b œ œ^ œ œ œ ! œ^ ! œ œ œ œ ! œ^ !
œ œ œ
œ
U
œ œ œœ œœ œ œ œ œ œ ‰
3
œœœœœœœœ œ
&b œ. œœ ! œœœ œœ œœœ œœœ
46
Œ
^ ^ ^ "
? œœ œ œœ ! ^ ! œœ œ œœ ! ^ ! œœ œ œœ ! ^ ! œ œœœ œ œœœ
b œ œ # œ œ œ œ
œ # œ œ œ œ
j j j
œ œœ œœ œ œœ œ œ œœ œœ ‰
& b œœ œœ œœ œœ œœ ‰ g œœ ‰ g œœœ
50
œ œ œœ œœ œ œœ œœ œœ œœ gg œœ gg # œœ œœ œœ
gg gg p
? ! œœ œœœ œ œœœ
b œ œ œ ! œ #œ œ œ œ # œ œ ‰
œ œ œ œ œ #œ œ œ œ œ œ
œ
063
% Moderatoœ {q = 75}
œ œ œœ œ œ œr œ œ œ œ œ œr œ œ .œ . œ. . œ. . . . . œœ^ œœ^ U
r
œ œ œ œœ !
& 42 œ ! œœ ! œ œ œ œ œ œgg œ œ ‰ bbb
Jgg
ƒ com animação œœ œœ œœ œœ œœ œœ U
? 42 œ œ œ œ œ œ œ œ œ œ œ œ œ^ ^ ‰
Piano
œ œ œ œ œ œœ bbb
b œ œ œ nœ œ œ ‰ j
Não depressa, bem ritmado
&bb ! œ œ œ œ œ œ #œ
6
œ œ œ œ
F
œœ œ œœœ œœœ œœ n œœœ œœ œœ j
? bb œ œ œ œ œ œ œ
# j œ
b œ nœ œ œ œ œ œ
œ
b
&bb œ œ œ œ Œ ! œ œ œ œ œ œ œ
10
œ œ
nœ
? b œ œœœ œ n œœœ œ œœ œœ j
œ ‰ Œ
œ œ
œ œœ œœ œj
bb œ œ œ
œ œ
b œ œ œ œ ‰ j œ œ œ œ
a tempo
&bb ! œ œ œ œ
rit.
œ œ œ œ !œ œ
14
œœ œœ œ
? b b œ œœœ œœœ œ œ œ ‰ œœ
œ œ #œ
œœ œ n œœœ œ œœ
b œ J œ #œ œ œ
œ
œ œ œ
2 063
b nnn
&bb œ ! œ œ œ œ
18
œ œ œ œ nœ
œ œ
œ œœ f
? b b œ n œœ œ œ œ n œœœ nœ
b œ œ œ
Œ Œ !
œ œ nnn
& œ. œ œ #œ. œœ œ œ œ œ œ œ œ œ œ‰ . œœ œ # œ‰. œœ œ œ œ œ œ œ ! œ

21
‰ œ ‰ œ
J J J J j F
? œœ #œ œ œ n œœ œœœ œ # œ œ œ n œœ œœ ‰
œ R œ œ œ œ
J ‰ œJ ‰ œ J ‰ œJ ‰
œ. œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ !
#œ œ
25
& œ œ n œ œ
œ œ
œœ œœ œ œ f
?œ œ œ œ
#œ œ
œœ œ
n œœœ œœ œ œ
œœ œœ œ œ
œ œ œ œ œ Rœ
œ
& œ‰ . œœ œ # œ . œ œ œ œ ! œ
œ œ‰ . œœ œ # œ‰.
29
œœ œ œ œ œ œœ œ œ œ œ
J ‰ J œ œ J J j F
? œœ #œ œ œ n œœ œœœ œ œ n œ œœ ‰
œ R œ œ œœ #œ
œ œ
J ‰ œJ ‰ œ J ‰ œJ ‰
œ œ œ
& œ . œ œ œ œ # œ # œ œœ œ œœ œœ œœ œœ œœ œœ œ œ # œ n œ œ ! œ œ
33
œ œ œ œ œ
œœ # œœ f ^ ^ƒ
?œ œ œ # œœ œœ œœœ œœ œœœ
œ œ œ œ œ ‰
œ œ œ œ œ J œ
œ œ œ
œ œ œ œ œ œ œ œ œ œ 063œ œ œ œ œ œ œ œ œ œ
! œ œ œ œ œ œ œ œœ œ œ
3
& œ œ œ œ œ œ œ œœ ‰ j
37
œ
œ
œœ œœ œ œœ œœ œœ œœ œœ f
œ œ œ œ œ ‰ œ
marcato
? œ J œ œ œ œ œ j
œ œ œ œ
Œ ‰ Œ ! œ œ œ
œœœ œ œ œ
œ. œ œ j œ œ œ‰ œœ œ œ‰
41
& œ œ œœ œœ ! ‰
œ ‰ œJ ‰ œ œ œ
J ƒ
? œ œ jœ œ œ jœ. j jœ œ œ œ œ œ
œ
œ
œ ‰
œ
œ
œ œ œ
j
J
œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œ œ œ œ œ
œ œ ‰ j
& œ œ œ œ œ œ œ œ ! œ œ œ œ œ œ œ œ œ
45
œ œ
œœ œœ œ œœ œœ œœ œœ
? œ œ œ œ œ œ ‰ œ œ œœœ œ œ œ ‰
œ J œ J
œ œ
Ÿ œ^ %
œœœ œ œ œ œ œ œ œ œ #œ nœ œœœ^ œœ ‰
œ œ œ œ #œ œ œj
49
& œ. œ
Z^ Z^
? # œ œœ # œœœ # œœœ œœœ œœ œœœ œœ œ
œ œ ‰ œ œ œ œ ‰
œ œ J œ
D.C.
œ œœ r
œœ œ œ œr œ œ œ œ œœ œ œr œ œ œ. . œ. . . . ^ ^
œ
œ œ œ
œœ ! œ œ œ œ œ. œ œ. œ
œ.gggœœ œœ U‰
&œ ! !
53
Jg
ƒ œœ œœ œœ œ œ
œœ œœ œœ œœ œœ œœ œœ œœ œœ ^ ^ U
?œ œ œ œ œ œ œ œ œ œ œœ‰
œ œ œ œ œ
œ
064
#
& # 42 œ œ œ œ œ œ œ œ ˙ œ œ œ œ
f
n ˙˙˙
#˙
œœ
œ ‰ œœj ‰ # œœj n œjœ ‰ Œ
Piano
? # # 42 ‰ Œ Œ œ œ œ œ œ œ
# j œ œ ‰ œj j j j nœ œ œ nn
& # œ ‰ œ ‰ œœ ‰ œœ œœ b
5
? # # œœœ œœœ j
‰ œ œ ‰ œj ‰ j
œ ‰ j ‰ j ‰ Œ
nn
b
J œ œ œ œ
œ œ œ
%
&b œ œ œ œ˙ #œ œ ˙˙ œ œ œ œ
9
‰ # œœ ‰ œœ
J J
œ # œœœ œœœ œœœ œœœ j
? b œœœ ‰ Œ œ œ ‰ # œj ‰ j
J œ œ ‰ Œ
#œ œ
œ #œ œ j
&b œ œœ œ # œ œ
13
‰̇ œœ œœ œœ
œ œ œ
œœ
œ ‰̇ œœœ ‰ # œœœ œœ
J J J
? œœ ‰ # œœJ ‰ œ Œ œ ‰ œ ‰ j
œ ‰ Œ
b J œ œ
œ œ
œ ˙ œœ
&b ˙ œ œ œ ˙
17
Œ̇ œ œ œ œŒ ˙ œ œ
b œJ ‰ Œ œ œ œ œ œ œ
j ‰ j œ œ œ
2 064
& b œœ œ œ œ œ
3
˙ œ
21
œœ œ œ œ œ #œ
J ‰ ˙ œ
? b œ œœœ œ œœ
œ œœ
#œ œ œœ œ
#œ j ‰ Œ
œ œ œ
j j j
œ œ œ j‰ œ œ œ œ œ œjœ # œœ œ œj ‰ ‰ j n # #
& b œ œ œ œœ .. œœ œ œ œ œ œ
25
œ J œ #œ
J J J
œ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ #œ œœœ # œœœœ ‰ Œ
? b œœœ ‰ Œ œ œ œ œ # œ œœ n##
J œ œ œ œ J
## . j œ
& . œœœ ‰ œœ ˙˙ ˙˙ œœ œ œ
31
j j
œœœ œ ‰ œœ ‰ œœ œœ œœ œœ œœ
? # # .. # œ Œ œ œœ œœ œœ œœ œ œ œ
#œ œ œ ‰ œ ‰ œ
## œ œ j
œ œ ˙ œ œ œ œ
35
& ˙ ˙ œ̇ œœ ‰ œœ œœ
J
œœ œœ œœ œœ >œ
? # # œ œœ œœ œ œ œ œ œ œ ˙
œ œ œ ˙
# œ j
& # ‰ œ œ‰ œ ˙˙ œœ œ œ œ ˙˙
39
# œœ œœ
J J
œ œ œ œ œ œ œœœ œœœ œœœ œœœ
? ## n ˙ œ œœ œœ œœ œœ œ œœ œ œœ
n˙ œ œ
## œ j
064 3
‰ j ˙ œ ‰ œ
43
& œ
œ
j j j j
œœœ œœœ ‰ œœ ‰ œœ ‰ œœ ‰ œœ
? ## œ œ œ œ œ œ nœ œ
œ
# ˙ œ œ œ œ
& # ˙ œœ œ œ
46
˙ œ
œœœ œœœ œœ œœ œœ œœ
? ## œ œ œ œ œ ‰ Œ
œ œ
œ œ œ ‰ Œ
J J
## œ
œ œ œ
50
& ˙ ˙ œ œ
œ œœ œœ œœ n ˙˙˙
? # # œ œœ œ œ œ œ œœ
œ ‰ Œ #˙
œ œ J
## œ œ. j
œ œ œ œœœ
54
& ˙ œ œ œ
œœ j j j ‰
? ## œ ‰ œœ ‰ # œœ œœ œ Œ œ
Œ œ œ œ œ œ œ ‰ œ
J
# # ˙1. j j j j j jfi n œ
œœ ‰ ‰ œ .. œ œ ‰ œ ‰ œ ‰ œ œ œ œ
& ˙˙
58 2.
œ œ œ œœ œœ œœ œœ
? ## œ œ œ œ œ ‰ Œ .. œ ‰ œœ ‰ œ œ j
J œ œ ‰ œ ‰ œ
œ
‰ Œ
D.S. al Fine
065
%
>œ >œ >œ >œœ œ . >
Allegro animato
#### 2 œœ ! œœ .. œœ ! œœ œœ œ n œ œ
œœ ! œœ œœ œ # œœ n œ .
& 4 œ œœ œœ œ œœ ! œœ œœ œ. œ. œ
Ra tempo J J >
F ƒ œ !ggœ>œœ ggœœœ gg œœœ . œ. œ ggg œœœ
rit.
? # # # # 42 . œ. œ. œ.
œ . œ. gg œœœ œ
Piano
!‰ ! œ. œ g g g .
!œ #œ gœ !œ. g
œ.
ƒ
. œ. œ. œr œ œr . œ. . .
œ œ œ. œ. œ
r
. .
œ r . ..
. œrit. >œ >œ >œ
# œ. œ
œ œ œ œ œ
œ œ
œ œ
œ œ . . .
œ œ œ œ œ œ . . .
œ œ œ œ œ œ
œ œ œ .
# œ œœ œœ œ œ. ® U
& ##
5
œ
fF
g œœ g œœ g œœœ œœœ g œœœ œœœ pœUF
? # # # # œ ggg œ œ ggg œ œ gg œ gg
g œ gg œ gg
g
œ œ. œ. RÔ ® ! ‰
œ œ. .
f F
>œœ œ ! œ œ >
#### œœ ! œœ ... œœ œœ œœ œ n œ œ œ œ
œ œ œ œ œ! œ œœ œ # œœ n œ œ ! œ œ . œ. œ.
9
& R J J œ >œ œ œ
ƒ . œ. œ. œ. œ !ggœ>œœ ggœœœ gg œœœ œ .œ œ
œ. œ. ggg œœ ! œ. œ. œ gggg œœ
a tempo
? #### ! œ œ œ. œ g g g ! œ # œ.
.
ƒ
. . >œœ >œ
# # # # œ. œ œ œ. œ. # œ. œ. œ. œœ r
œ œ ‰ œ # œ œ œ œ œ œœ œ œœ œœ œœ
œ œ U r n nn#
n
!
13
& œ œ
J
ƒ
g
? # # # # # œ gg
# œœœ > g œœœ gœg œœœ f >œœ >œœ œ œ œ Fine
>gg œ > gg œ >œ # œ œ œ pœ U nnnn#
œ gg gg Œ œg œ g J ‰
f ƒ
.
. . . . r
2 065
. . . r œ œœ
! œ œœ œœ ! ‰ ! œœ œœ œ ! ‰
Cantabile suavissimo
# œ . œ
& œ . œ œ # œ œ œ œœ œ œ ! œR œ
œ #œ œ œ œ
17
J J ! R
œœœ œœ œœœ g œœœ g œœœ
?# gg œœœ œ
ggg # œœ gg œœœ œ
œ
gg œ ‰ > g
g œ gg g
g # œ g
g g g
œg œ g œ g g œg œ g g g
# œ. j r >œœ U œœ œœ
œœ œœ œœ œœ # œœ œ ! œ œ. œ œ g
g #
œœ œ œ œœ gg œ œ œ œ œ
21
& œ. œ œ œ.
ƒ U
#œ ggg œœœ > gg# œœœ gg œœœ gg œœœ # œœœ > g œœœ # œœœ œœœ
cresc.
? # # œ ggg œœ œ
cresc.
g
# œ gg gg
œg g g œ œ œ
œ œ
œ œ œ œ œ . œ
# g œœœ ... # œ œ œ œ œ œ
& ggg œ . œ œ ! œ # œœ . œ œ œ œ !œ
25
g
œœ
œ >œ
Ï œœ g œœ œ F p
gg œœ gg œœœ gg œœœ gg œœœ ggg# œœœ ggg œœœ ggg œœœ ggg œœœ gg œœœ ggg œœœ
a tempo dim.
? # œ œ ggg œ g g g g œ œ g gg g
œ œ
Ï
a tempo F p
œ #œ œ œ œ U >œ œ
# œ œ œ œ œ œ œ œ ˘> gg œœœœ œ ! n# # # #
g
& œ œ #œ œ œ ! œ œ # œœ œ œ œ
29
J J R œœ g œœ
J
œœ ggg œœœ gg œœœ œ gg œœœ gg œœœ r # œœœ œœœ j
g
? # œ gg œ g
g œ œ n# # # #
œ œ g gg gg œ œ œ œ # œ œg
œ
‰
œ fl
>
065 3
œ
# # # # œ œœ œœ œ œœ œœ œœ ! œœ œœ œœ œ œœ ‹ œ œ œ
Piu mosso
œ œ œ
œ ‹œ œ œ
!
33
& J œ
F R
g œœ g œœ g œœ g œœ g œœ g œœ œ œœœ g œœœ
# œ œ g œ
? # # # œ œ œ œ gg œggg œ œ œ g œ
œ œ œ œ gg œggg œ œ œ g œ œ g
œ œ œ œ gg œggg œ œ œ œ œ œ gg œgg
#œ œ œ œ
# # # # œ œ- ‹ œ- œ- - œ œœ ! # œ œ œ œœ œ# œ œ œ œ œ
œ œ œœ œ œ œ# œ œ œ œ !R
37
& R J
J
œ œ œ œ œ œ œ gg œœœ gg œœœ œ œ œ œ œ œ œ œ œ
? # # # # œ œ œ œ œ œ ggg œœ œggg œœ œ œ œ œ œ œ ggg œœ œggg # œœ œ œ œ œ-g œg œ œ œ œœ œœ
œ œ œ œ œ
œ
# # # # œ œœ
œœœ œœ œœ œœ n œ # œœ œœ œ
œ œ œœ œœ œœ ! œœ
J ! n œR
41
&
g œœ œœ g œœ # œœœ œ œ œ œ œ gg # œœœ g œœœ œ œ œ œ œ g œœ g n œœ

? # # # # œ œ œœ œ œ ggg œ œgggg œ œ œ œ œœ œ œ ggg œ n œ # œ œ œ œg gg œ œ
œ
œ œgg œ gg œ
g nœ
œ œ œ
%
nœ >
œ >
#### n œ œ œœ j
œ ! œr œœ œœ œœ >œœ œ U® œ >œ
œ œœ œ œ œœ œ
45
& œ
nœ œ œ RÔ rit.
n œ
œ œ
œ œ
œ œ œ œœ g œœœ œœœ œ œ œ
? # # # # ggg œ n œggg œ ggg œ œ œ œ œ œ œgggg œœ œggg œœ œ g œ U
rit.
œ œ
a tempo
g g
œ œ œ œ œg œg g œ œ K
œ. œ r® ! ‰
œ œ œœ œ .œ
066
b 2 œ œ œ œ œ œ œ œ œ
& b 4 œ œœœ œ œœœ œ
Piano
? b 2 œ œ œ œ œ œ
b 4 œ œ
b b œœœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ ! ! œ #œ
3
& œ œ
œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ
? b œ œ œ œ œ œ œ œ œ
b œ œ œ
b .. j j j j œ #œ œ œ
b œ œ b œ nœ bœ œ œ œ #œ œ œ œ #œ œ œ
! ! œ ! ! œ #œ
6
& œ œ œ #œ
œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? b b .. œ œœ œ œ œœœ œ œ œ œ œœœ œ
œ
b b j j ! ! j j œ œ #œ œ œ!!œ
& œ œ œ #œ œ œ bœ nœ bœ œ œ œ #œ œ œ œ #œ œ œ
10
œ œœ œœ œœ œœ œœ œ œ œ œœ œœ œœ
? b b œ œœ œ œ œœœ œ
œœ œœ œœ
œ œ œ œ
œ
b œ œ œ œ œœœœ j j œ
&b J œ œ œ œ!!
14
J œ œ œ œ #œ œ # œ # œ n œ œ œ œ bœ œ
œ œ œ œœ œœ œœ œœœ œœœ œœœ œœ œœ œœ
? b n œ
œ œ
œ œ
b œ œ œ
œ
b œ œ œ œ œœœœ
2 066
b œ œ œ œ œ #œ œ #œ œ œ œ #œ œ œ œ œ œ nœ œ œ!‰
18
& J J R
? bb n œœœ œœ œœ
œ œ
œœ œœ
œ œ
œœ
œ
œœ
œœ
œœ
œœ
œœ
œœ œœ œœ œ
œ œ œ œ œ œ # œœ
b œœœ œ œ œœœœœ œ œ- œ œ œr œ œ .
&b ‰ œ œ #œ œ
22
œ œ #œ œ #œ
n œœ œœ œ œ œœœ œœœ
? b b ggg œœ ‰ Œ
accel. a tempo
#œ œ œ # œœ œ œ œ
ggg œ #œ œ œ
J
b b ‰ œœœœœœ œ œœ œœ œœœ œ œœ
œ #œ œ œ œ œ œ œ œ œ œ
26
& œ œ
? b œ œ œ œ œ œ # œœœ œœœ œ œ œ
b œ œ œ œ
œ
bb œ œœ
& œ œœœ œ œ œ œ œ n œ œ. œ œ œœœœ œ ‰ ‰ ! œ # œ ..
30 1.
œ #œ œ œ #œ J
œœœ œœœ œœœ œ œ œ œ œ œœ œœ œœ
? b œ œ
#œ œ œ œ œ
#œ n œœ œ œ œ ..
b œ nœ œ
œ œ
œœ
bb œ œ œ œ œ œœ œœ ‰
34 2.
& œ #œ œ œ #œ
? b œ # œœœ œœœ œœœ œœ

œ œ ‰
b œ
œ
067
Muito devagar
#
& 42 Œ ‰ œj œ œ œœœœ œ
œ œœ œ. œ œ œ œ œ œ œ œ œ œ œ
j
3
œœ
p J
? # 42 œœœ œœœ œœœ œœ œœ œœ œœœ œœœ œœœ œ œœ œ œœ œ œœ œœ œœ
Piano
œ œ œ ˙ œ
˙ ˙
#
œ œœ œ œœ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ
6 3
&
œ œ œ œ. .
?# œœ œœ œœ œœœ œœœ œœœ œ œœ œœ œœ œœ œœ

œ œ œ ˙ œ œ
˙
# # œ Uœœ œœ œ œ œ œœœœœœœ œœœœœœœœ œ j r œ

Fine
& # œ œ œ
10
œ œ œ R
U pr
ƒ œ. f
œœ .. œœœ œœœ œœœ œ œ œ œœ œœ œœ œœ
?#
œ œ œ œ ˙ œœ œœ œœ œ œ œ œ
˙ œ œ œ œ
˙ J R R
# œ œ œ œœœ œ œ œ œœ
œ œ œ œ œ. œ
14
& # œ œ œ œ œ œ
œ
œ œœ œ . œœ œœ
cresc.
? # # œ . œœ œœ
œ ˙ œœ œœ œœ œ. œ œ œ
J J J
2 067
U
œ. œ œ nœ
# œ œ œ œ œ j œ œ œ
œ œ œ
18
& œ œ
U
f œœ .. f
œœ œœ ˙ # œœœ œœœ œœœ
? # œ. œ
J #˙
# œ
œ œ. œ œ œ œ œ œ œ œ œ œ.
21
& œ œ # œ
œ
p
œœ œœ œœ
# œœœ œœœ œœœ œœ œœ œœ
D.C. al Fine
?#
˙ œ œ œ ˙ ˙ œ œ œ
068
# # # # 2 ‰ œj œ œ j
‰gg œœœœ ‰ œœ œœ œœ
‰ gg œœœ
j
‰gg œœœœ ‰ œœ œœ !
‰ gg œœœ
& 4 œ œ œœ gg Jgg J œ gg Jgg J
p œ
œ œœ œœœ œœ œ œœœ œœ œœœ œœ œ œœ œ
? # # # # 42 œ œ œ œ œ œ
Piano
œ
œ œ œ œ
œ œ œ
œ œ œ
U
#### j j œœ œ œœ œ œ œœ œ œ
œ œ
œ œœ œœ
6
& œ
? #### œ œœ œœœ œ œœœœœ œœœ œœœ œ œœœœœ

œ œ œœ œ œ œœ
#### œ œ œ
œ œ œ œ œ œœ œœ J œ œœ œœ œ œ œœ œœ ..
10
&
J œ
œ œ œ œœœ œœœœœ
? #### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ
U
œ œ œ œ
#### j œ œ
œœ œ # œ œ
j
œ œ œœ
œ œœ
14
& J J
f œœ
#
? ## # œ œ œ œ œ œ œ œ œ œ œ œ œ œœ
œœ ‰ œ œ œ œ œ œ œ œ #œ œ œ œ
œ œ
J
#### œ œ œœ œ œœœ œ œ œœ œœ œ œ œœ ! œ n n n n
œ
18
& œ œ œ œ œ œ œ
? #### œ œ nœ œ œ œœœœœ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ nnnn
rit.
œ œ œ
œ #œ
œ œ œ œ œ œj œ
2 068
œ œ œ œ
œ œ œ œJ œ
22
& J J
p
œœ œœ œœ œœœ œœ œœ œ
?œ œ œ œ œ œ œ œ œœ œœ œœ
a tempo
œ œ œ œ
œ
U
œ
œ #œ œ œ œœ ####
rit.
& œ œ œ œ œJ œ J œ # # œœœ œœ !
26
œ œ
œœ f œ j ‰
œœ œœ œœœ œœ
?œ œ œ J ‰ œ œ œ # # œœœ œœ ####
œ œ #œ
œ ‰
J
^ j U
œ
#### œ œ
œ œ j œ œ œ
œ œ œœ
œ œ œ œ œ œœ œœ œ # œœœ œ
30
& J J
f œœ
œ œ œ œ
a tempo
? #### œœœœœ œœ œœ œœ œ œ œœœ œœ

œ œ œ œ # œ œ
œ œ
#### œ œ œœ œ œœœ œ œ œœ œœ œœœ œ œœ œœ g œœ

œ
34
& œ œ œ gg œ.
? #### œ œ œ nœ œ œ œ œ œ œœ œœœ œœœ œœœœœœœ

rit.
œ #œ œ œ œ œ œ œ
#### j œ j j j U
œ
‰ ggg œœœ ‰ ggg œœœ ‰ œœ œœœ œœœ ‰ g œœ ‰
œœœ œœ
38
& gg gg J ggg n œ œ œœ
œ
"
U
? #### œ œ œ œ œ œ œ œ œ œ œ œ nœ œ œ œ œ œ œ
morr. poco a poco
069
>œœ n œ %œœ œœ œœ >œ

Tempo de Marcha
œ œ œ œ œ >œ j
œ œœ œœ œ
& b 42 œ n œJ œ œ œ œ
J œ œœ œ œ œ œ
J J
ƒ
œ œœœ ‰ œj >œ œœœ œœœ œ
œ œ ‰ œj >œ
œœ œœ
Piano
? b 42 ‰ Œ œ
œ œ œ œ
> œ œ
j j
œ j >œ œ
œ ˙ œ. œœ œœœ nœ
& b œœœ œœœ œœœ œ œ. œœ œ nœ
5
œ œ J ‰̇ J
j
? œ œœœ ‰ œ >œ œœœ
œ
œœœ œ œ œ œ j
œ ‰ Œ
b œ œ œ œ
> > >œ >œ œ
œœœ œœœ œœœ >œ j
œ œ œœ œœ >œ œ
&b œ œ œ œ œ œœ
9
J J œ œ œ œ
! J
œ j >œ œœœ œœœ œ œ œœ
? b œ œœ ‰ œ œ œ œ ‰ œj œ œ
œ œ œ œ œ
> œ
>œ >œ
j >œ œ œ œœ œœœ
œ
& b œœœ œœœ œœœ œ œ ‰ œJ ‰ J ‰ ‰ j nb b b b
13
œ œ J œ œ
J F
ƒ
j
? œ œœœ ‰ œ œœ œœ
cresc.
j
FIM
b œ œ œ œ œ œ
J ‰ œ ‰ j ‰ " nb b b b
œ œ œ œ
œ
bb j j ‰ ‰
& b b œ. œ œ œ œ œ j
17
‰̇ n œœ .. œœ œ
œ
J
? b b œ œœ œ œœ œ œœ œ œœ ‰ nœ nœ bœ œ #œ nœ #œ
bb
j j ‰ ‰ œj
2 069
b
& b bb œ . œ œ œ œ œ œœ
21
‰̇ œœ .. œ
J
œ œ œœ œœ
? bb b n œ œ œ œ nœ œ ‰ œ œ bœ œ #œ œ #œ
b
j j
b b b œ. n œ n n œ˙ n œ œ œ b n œ˙ . nœ œ˙ n œ n œ œ
& b n˙
25
j j
j œœ œœ j n œœ œœ nœ œœ n œœ œ
? b b b n n œ˙ b œ˙ œ œ nœ œ œ
b œ
b b nœ. j j ‰ ‰
&bb ˙ œ œ œ nœ œ nœ. œœ j
29
‰̇ œ . nœ
J œ
œ n œœ œœ œ n œœ œœ
nœ œ nœ œ bœ
? bb J J J J ‰ œ œ
bb
bb j j ‰ ‰
& b b œ. œ œ œ œ œ j
33
‰̇ n œœ .. œœ œ
œ
J
? b b œ œœ œ œœ œ œœ œ œœ ‰ nœ nœ bœ œ #œ nœ #œ
bb
bb j j ‰ ‰ j
& b b œ. œ œ œ œ œ œœ
37
‰̇ œœ .. œ œ
J
? b b n œ œœ œ
œœ œ
nœ œ œ œ
œ
‰
œ œ bœ œ #œ œ œ
bb
j j
069 3
b
& b bb œœ . œ n œœ œ œ œ œ. œ œ œ œ œ
41
œ
j J
j œœ œœ œ œ œ œœ œ œœ œœ œœ
? b b b n n œ˙ nœ œ œ œ œ
b œ
b ‰ j j
& b bb œ . j
œ nœ œ œ œ
n œœ œœ œœ . nnnnb
45
˙ œ œ nœ nœ
J
œ œœ p
? b b b œj œ œ œœ œœ œ œ
œ nœ œ œ œ ‰ Œ
b J œ J nnnnb
& b .. ‰̇ œœ œ‰ œŒ œ œ œ œ œ
49
œœ ‰ œœ ‰̇ œœ ‰ œœ œœ ‰ Œ
œ J J J
J J J
? .. œ ‰ œ ‰ œ ‰ œ ‰ œ ‰ œ ‰ œ ‰ œ ‰
b œ œ œ œ œ œ
j - - -
œœ œ œ œ ˙ œœœ œœœ œœ
& b ‰̇ œœ œ œ œ
53
œœ ‰ œœ œ ‰ Œ ‰ ‰
J J J J J J ‰ ‰ ‰
? œ œ œ œ œ œ
b œ ‰ ‰ œ ‰ œ ‰ ‰ œ ‰ j
œ œ ‰ Œ
j j
œ. œ .
œ nœ œ œ œ
& b ‰ œœ œœ ‰ œœ œœ œ œ
j
œ
57
‰ œ œ ‰
#œ
‰̇ ‰
? ˙ #˙ nœ n >œ b >œ
b ˙ #˙ nœ ‰ nœ bœ ‰ Œ
J J
4 069
&b œ œ œ œ
61
œ j
œŒ
‰̇ œœœ œœœ œœœ ‰ œ œ œœ œœ œœ œœ œ œ >œ
#œ
J J œ œ œ œ > >
F
> > > >
?b ‰ œœ
œ œ bœ œ ‰ œ ‰ œ ‰ œ ‰ j ‰ œ
œ
˙ œ œœ œ œ œ ˙ œœ œœ œœ œ œ œ
œœ œœ
&b œ œœ œ œ œ œ œ œ
65
J
‰̇ > J ‰̇ >
œ œœ œ œ œ œœ œœ
?b œ œ œ œ œ œ œ œ
œ ‰ œ
J œ
˙ œ œœ œ œ œ ˙ œ. œœ
œ œœ œœ œ œ œ ˙ œ. œœ œ œ œ
& b ‰̇
69
J ‰
œœ œœ œ œœ œœ
? œ œ ‰ œ œ œ œ œ ‰ Œ
b œ J œ J
bœ œ œ bœ bœ bœ bœ nœ
&b œ œ œ œ
73
œ œ #œ œ
f
ggg œœœœ œœœœ
j œœ œœ œœœ
? b œj ‰ g J ‰ Œ
b œ J J ‰ Œ
bœ
j j œœœ n %
œ
. j œ nœ
& b œ‰ . œœ œ œ. œ œ œœ œ œ œ. .
77 1. 2.
œ. ‰œ ‰œ . œ ‰œ ˙‰ œ œœ J
œ .. œ . . œ
j f ƒ
? œ. œ j ˙ ‰ œ ‰ Œ .. œ ‰ Œ
b œ. œ œ. œ ‰
œ. œ œ œ œ œ
J J œ œ
J J
071
# # 2 j . œj œ
œ œœ œ
j ‰ j
& 4 œ . ˙˙ œ œ œœ œœ œ œ œ œ œœ
‰̇ œœ ‰ œœ ‰ œœ ‰ œœ
J J J J
œ œœ œ œœ œ
? # # 42 ‰ .. œ œœ œ œ œœ œ œ œ œ #œ œ œ
Piano
œ
œ œ
# # œj j ‰ j
& ˙ œ œ œ œœ œ œ œœ œœœ
5
œ œ
Œ ‰̇ œœJ ‰ œœJ ‰ J ‰ J
œ œœœ œ œœœ œ
? # # œ œœ œ œ œœ œ œ œ œ #œ œ œ
œ œ
9
# # œj œ j
œ œ œœœ œ œ œ œ œ œ
‰
œœ
j
& ˙˙ œ ‰̇ œœœ ‰ œœœ ‰ œœ ‰ œœ
J J J J
œœ œœ œœ œœ œ
? ## œ œ
œ
œ œ œ œ œ œ œ œ œ #œ œ œ
# # œj j ‰
& ˙ œ œ œ œœ œ œ œœ
13
Œ œ œ ‰̇ œœJ ‰ œœJ ‰ J Œ œ œ
? # # œ œœœ
œœ œ œœœ œ œ œ
œ
œ œ œœ œ œ ‰ # œœœ
œ œ J >
# # œj œ j j
2 071
œ œ˙ œ œ œ œ œ œ #œ. j
3
œœ .
17
& ˙ ˙ . œœ
J
? ## œœ œœ œœ œœ # œœœ œœ
œ # œœœ œ œœœ
œ œ œ œ œ œ œ #œ
œ œ œ œ
# j ‰
& # œœ œ Œ œ œ œœ œ œœ . œ œ˙ œ œ œ .
21 3
œœ Œ œ œ
? ## œœ
œ
œœ
œ
œœ
œ
œœ
œ # œœœ œœœ # œœœ
œ œ œ œ #œ œ #œ
œ œ
# # œj œ j j j
œ œ˙ œ œ œ œ œ œ #œ.
3
œœ .
25
& ˙ ˙ Œ œ
? ## œœ
œ
œœ
œ
œœ
œ
œœ
œ # œœœ œœœ # œœœ
œ œ œ œ œ #œ œ #œ œ
## j j ‰ j 2. ##
œœ œ .. œ œ # œœœ n œœ œ # #
29 1.
& œœ œ Œ œ œ œœ œ‰ Œ # œ œ œ̇ œœ œ Œ œ # œœ œ
J
# œœœ œœœ f
? # # œœœ œœœ œœ œœ œœ œœ . ####
œ œ œ œ œ œ œ œ œœ ‰ œœ . œ œ #œ #œ
J œ #œ #œ
# # # # . œœj >œ j >

œœ œ œœ œ œœœ œœ
‰ j
œœœ
. œ œœ # œ œ œ ‰̇ œœ œœ
35
& œ œ
J J J J J
œ
œ œœ œ
œœ
? # # # # .. œ œœ œ œœ
œ œœ œ # œœ œ œ œ œ
œ œ
# # # # œj œ œ œ j ‰ j 2. j‰
071 3
œ œ œ œœ œœ œœ œ œœ œœ œœœ œœ œœœ .. œœ œœ # œ œœ œ œœ œœ
39 1.
&
J J J J ‰̇ J J J ‰̇ J ‰
? #### œœ œœ œœ œœ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ .. œ œ œ œ œ œ œ œ
# # # # œj œ j ‰
œ
œœ œœ œ œœ œ œ œ œ œ #œ.
3
œœœ œœ
45
& œœ œœ œœ œ œ œœ œœ ˙˙ œœ œœ
J J
? # # # # œ œœœ œœ
œ œ œœ
œ œœ
œ œ
# œœœ œœœ #œ
œ œœ œœœ
œ œ œ œ
#### j j œœ œ œ œ œ œ .
œ œœ œœ
49 3
& œ œœ œœ œœ œœ œœ œœ œœ
J J J ˙
? #### œœœ œœœ œœœ œœœ œ œœœ > œœœ >œ œœœ œœ
œ œ
#œ œ #œ œ œ œ œ
>
# # # # œj œ j ‰
œ
œœ œœ œ œœ œ œ œ œ œ #œ.
3
œœœ œœ
53
& œœ œœ œœ œ œ œœ œœ ˙˙ œœ œœ
J J
? # # # # œ œœœ
œœ œ œœ # œœœ
œ œ œœ œ œ œœœ # œœœ œœœ
œ œ œ #œ œ
j
œ
j ‰ œœ ‰
#### j j
œ œœ œœ œ œœ œ ‹ œ œ ˙œ œœœ
œ œœ œ
57
& œ œœ œœœ œœ
J J J J
? # # # # # œ œœœ œœœ # œ œœœ œœœ œœœ œ œ œ j ‰ j ‰
FIM
œ œ
070
Lento
#
& # 42 œ œ œ # œ œ œ œ œ j
œœ ‰ Œ !
œœ œ
œ bœ
‰ œ œ œ œ œ œ œ œœœ. œœ. œœœ œ œœ.œ œœ. œœ.œ
Piano
? # # 42 ‰ Œ
œ . œ
## . ‰ œ
& . œ œ œ œ œ œ œ œ œœœ . # œœ ‰
œ œ œ œ œ œ ˙ œœœ n # œœœ # œœœ
4
œœ
œ œœ œœ œœ
? # # .. œ œœ œ œ œ œ œ œ œ œ œ
œ œ
## ‰ œ j
œ œ œ œœ . œ œ ‰ œ œ œ œ œ œ œ œ œ œ.
8
& œ œ œ œ œ œ
œ œœ
? # # œœœ œ œœ œ œ œ œ œ œ œœœ œ œœœ œœœ œœœ
œ œ œ œ . . .
œ
# ‰ œ#œ#œ œ œ œ œœœœ œ.
& # ‰ œ #œ #œ œ œ œ œœœœ œ.
3
12
œ œ œ œœœ œœœ œœœ # œœœ œœœ œœœ œœ œœ œœ

rit. a tempo
? # # œ œœ
# œœ œœ
3
#œ œ œ œ œ
œ
#œ œ
## ‰ œ œ œ œ œ œ
2 070
œ œ œ
œœ œ œ œ œ œ œ œ œ œ "œœœ
16
& œ
œ œ œ œ œ œ œœ œœ œœ œœ
? # # œ œœ œœ œœ œ œœ œœ œœ œ œ œ œ œ œ ‰
œ œ œ œ
# œ œ œœœ œ
& # œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ. œ
20
œ œ
œ œ œœ œœ œœ œœ œœœ œœœ
? # # œ œœ œœ ‰ œ œ œ ‰ #œ œ œ ‰ œ ‰
œ œ #œ œ
# # œ œ œ œ œ œ œ œ œ 1. ..
œ œœ œ œ œœœœ œ Œ !
24
& œ
œ œ œ œ
? # # œ œœ œœ ‰ œ œœ œœ ‰ œ œœœ œœœ œ œœ
œ.
œœ
œ.
œœ œ œœœ b œœœ œœœ .
œ. .
œ œ œ œ œ
œ
## œ œ " œ œ œœ œ œ œ œ ‰ n b œœ n œœ œœ ‰ œœ n œœ œœ ‰ n b œœ n œœ œœ
œ
29 2.
& œœ œœ œœ œ œ œ œ œ œ œ œ œ
? # # œ œœ j j j
rit.
œ œœ ‰ Œ œ ‰ Œ œ ‰ Œ œ ‰ Œ
œ œœ J œ œ œ
œœ
# œœ
& # ‰ œœœ n œœœ œœœ ‰ œœ b œœ œœ Œ J ‰ j ‰ Œ
34
œ œ œ ˙ œ
œ
? # # œj ‰ Œ j j
animando
œ ‰ Œ œ ‰ œ ‰ ˙ œ ‰ Œ
œ œ œ J J
072
b 2 ! œ œœgg œœœ ! ‰ œœ œœ œg œœ
b
& 4 g œœ ggœœ gggœgg R
œ
œ g
g
!g œ ggœ gggœ gg œ ! ‰
œ ! œœœ ! # œœœ
œ œ gœœ ggœœggg œ !
ggg œœ ggœ R œ œ œ# œ
r
gg œ gg œ ggg œœ ggœ g œ
g g g
œœ œ œ œ œ œœ œœ
r
? b b 42 œ œœœœ
Piano legato
œ œœœœ œ œ œ r! ‰
œ
"
j ! œr œ .
bb œ . œ œœ. œ œ œœ œ œ œœ œ œœ œ j
5
&
3
œJ # œ . œ œ œœœœ œ œ
? b gg œœœ ggg œœœ œ gg œœœ ggg œœœ œœœ œœœ œ

b œ gg œ gg g œ œ # œ œ œ œœ œ œ œ
œ œ œ œ œ
œ œ
j ! œr . œ
bb œ. œ œœ. œ œ œœ œ œ œœ œ œ œ œ œœ .. œ œ # n œœœ
9
& œ J œ # œ. œ œ œ œ
J
? bb gg œœœ gg œœœ œ ggg œœœ gg œœœ œœ œœ
gg œœœ gg œœœ ‰
g œ g gg œ œ œ
# œ œ g
œ gg œ œ œ œ œ œ gg
œ
j j j
bb œœœ œœœ œœœ œ œ œ œœ œœ œœ œ œ œœ œœ œœœ œœ œœ œ
# n œœ
13
& œ œ œ œ #œ œ
J
? b œ ggg œœœ gg œœœ ggg œœœ ggg œœœ œœ œœ œ
ggg œœœ ggg œœœ ‰
b g g ‰ œ g ‰ œ # œJ œ œ g
œ œ œ
J
U
œ
b b œœœ œœœ œœœ œ œœ œœœ œœ œ œ
œœ œœ œœ œ œœ j bœ œœœ #
n n œœ œœ œ
œœ # œœ œœ
17
& œ J
? b œ ggg œœœ ggg œœœ œœ œœœ œœ # œœ œœ # œœ œ œ œ œ œ œœ œ œ œ œ

nn # œ œ œ œ œ œ œ œ
rit. a tempo
b g ‰ œ œ œ
œ œ
# œ œ œ œ œj j
2 072
œ j j
œ œ œ œ œ œ
21 3
& œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œœ œ
? # œ œ œ œ œ œ œ # œœ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œœ
3
œ œ œ
# j œœ œœ œœ
œœ œ œœ œ œ œj œœ œ œœ œœ # œœ œœ
3
œ
25
& œ œ
3
#œ nœ ‰ Œ œ œ J œ J œ œ œ
J
œœ œœœ œœœ œœœ œœ œœœ œœœ œ œœ œœ œœ œœ
? # œœ œ œœ b œœ œœ œœ œ œœ œ œ œ œ œ œ œ
œ œ œ œ œ œ
# j j
& œ œ œ œ œ œ œ œ œ œœ œ œ œj j
29 3
œ œ
œ œ œ œ œœ œœ œœ œœ œ œœ œ œ œœ œœ
? # œ œ œ œ œ œ œ # œœ œ œ # œ œœ œœ
3
U
# œœ ! œ œœ œœ œ œœ œœ œgœ ggg œœœ b œ œ nbb
3
œ œ œ œ œœ R œ œ
33
& œ œ œ œ g œœ ggœœ ggœ g œ

3
œ œ gg œ gg
g
U rœ rœ rœ
œ œ œ œ œœœ œœ œ
?# œ œ œ n œ œ œ œ œœœ ‰
œ œ œ # œ œ r! ‰ nb
#œ œ œ
œ œ œ œ b
œ œ œ
bb ! œ œœ gg œœœ ! ‰ ! œ gœœ gœœœ ggg œœœ ! ‰ ! r œœ !

# œœœ œœ
œ ! ! r! ‰
gg œœœ gggœœ gggœ gg R gg œœ ggœ gg g R œ œ œ# œ
37
& œ œ œ œ
g g g
g g
"g #
œ œ œ œ œ
r
? b œ œ œ œ œ œ
œœœ œ œœœœ œ ! œ ! œr ! ‰
pizz.
b
#
073
& 42 œj .. œ . œ œ œ œ œ œ j œ. œœœ œ œ œ œ
J œ
œ œ œœ œœ # œœœ œœ œœ œœ
.. œ œœ œ œœ
Piano
? 42 ‰ œ œ ‰ œ
#œ J œ œ œ œ œ ‰
J
U
œ #
œ. œ œ œ œ œ j ##
a tempo
œ œ. œ œ. #œ œ ‰
5
& J J œ
œœ #Uœœ œœ œœ œœ # œœœ œœ œœ
rit.
? œ œ ‰ ###
œ œ œ
œ
œ J œ
œ
œ œ œ œ ‰
J
### œ. œ œ œ œ œ œ
9
& œ œ
œœ œœ œ œ œœ œœ œœ œœ
? ### œ œœ ‰
œ œ œ œ œ œ J œ
œ
œ
œ œ œ œ ‰
J
### œ œ œ. œ œ œ œ œ œœ œ
œ. œ œœ œ œ œ ‰ J
13
&
œœ œœ œœ œœ œœœ œœœ œœ œœ
? ### œ œ œ œ œ ‰ #œ œ œ J ‰
œ œ J
### œ . œ œ œ œ œ œ œ
2 073
œ œ œ œ # œ œ œ œ œ œ œ œ œ œ œ ‰ œJ
17
&
œ œ œœ œœ œœœ œœœ œ œ
? # # # œ œœ œ œœ œ J ‰ œ #œ œ œ œJ ‰
### œ . œ œ œ œ œ œ œ œ œ œ œ j . œœ œœ j n n n
œ œ œ œ #œ œ œœ œ œ œ œ . œ œ œ œ
21 1. 2.
&
œœ œœ œ
? ### œ œ œ œœ œœJ ‰ œœœ
# œ
œœ œ œœœ œœ ‰ .. œ œ ‰ n n n
œ œ œ J œ
œ œ œ œ œ œ j œ. œ œ œ œ œ œ œ
& œ.
26
J œ
œœ œœ œœ œœ œ œ œœ œœ
? # œœ œœ
œ œ œ œ
#œ
œ œ ‰
J œ œ œ œ œ ‰
J
U
œ œœ œœ
œ. œ œ œ œ œ
a tempo
œ œ. œ œ. #œ œ œ œ
30
& J J
œœ #Uœœ œœ œœ œœ # œœœ
rit.
? œ œ œ œ ‰ œ œ
œ œ œ J œ œ œ œ
074
œ œ œ œœ ! œ œœœ œœ
& 42 ! œ œ œ œœ œ œœ œ ! œ # œ œ œœ œ œœ œ ! œ# œ œ œœ œ œ œ œ œ œ
f
œœ œœ # œœœ œœœ œœœ # œœœ œœœ œœ œœ œœ
Piano
? 42 œ œ
œ œ œ
œ œ œ œ œ
& .. œ . œ œ œ œ œ. #œ œ œ ! œ #œ nœ j ‰
5
œ #œ œ œ œœ œœ œœ œ œ
p
? .. œœ œœ œ # œœœ œœœ #œ œœœ œ œœ œ
œ œ œ œ œ œœ œ œ
œ œ ###
& œ œ œ œ œ œ œœ œœ . # œ. œœ œœ . œ. œœ œœ n œ œ œ œ
9
œ # œ œ œœœ œœœ œœœ # œ

? œœœ œ œœ œ # œœœ œœœ œœœ œ
# œœœ œœ ###
œ œ #œ ˙
### œœœ ! U!
! œr œ œ œ œœ œœ œ œ œ œ
13
& œ . œ œ œœ œœ œœ œœ œœ œ œ œ œ
? ### œœ œœ œœ œ œ œ œ œ œœ œœ U
œ œ œ œ œ œœ œœ œœ œ œœ œ œœ œ œ œ ! ‰
R
###
œ œ œ œ œ œœ œœ œ. œ. œœ œœ œ. œ. œœ œœ œ œ œ œ œ ‰ ! œr
17
&
œ œ
? ### œœ œœ œœ œ
œ œ
œ œ
œ œ
œ œœ œœ œœ œœ
œ
### ^
2 074
œ œ œ œ œ œœ œœ œ. œ. œœ œ
œœ . œ. œœ œœ œ œ œ œ œ œ. œ
21
& œ œ
œ
œ œ. . . . .
? ### œœ œœ œœ œ œ œœœ œ œœœ
œ œ œ œ œ œœ œœ œ œœ œ œœ œœ œœ
œ
### ^ ^ ^
œœ
25
& œ œ
œ. œ. . . œ. œ. œ. œ. œ œ œ œ œ. œ. œ . œ. œ. œ. œ. œ. œ œ œ >œ >œ
>
? ### œœ œœ œ œ œ
œ œœ œœ œœ œœ œœœ œ œ
œ œ
œ
œ œ œ œ œ >œ >œ
>
###
œ œ. œ œ œ œ !œœœ œ œ œ œ œ
29
& œ > œ œ œ œ œ
œ œ œ >
f > >
? # # # œ œœ œœ œœ œ
œœ œœ œœ
œœ
œ œ
œœ œ œœ > > >
### œœ œ œ œ !œ œ œ œ œ œ
33
& œ œ œ. œ œ
œ œ œ
? # # # œ œœ œœ œœ œ œ
œ œ œ
œœ œ œœœ œœ œœ œœ
### œœ^
j‰ Œ .. j ‰ œœ ! ‰
œœœœœ
37 1, 2. 3.
& œ œœ œœœœœ œ œœœœœœ œ R

œ œ œ
? ### œœ œœ œ œ œ œœ œœœ œœ œœ œœ . œœ^
œ œ œ œ œœ œœ œœ œ œ
œ œ nœ œœ. œ œ œœ ! ‰
R
075
# # # 2 œ œ œ œ% œ œ œ œ œ œ
Allegretto
œ œ œ œ
& 4 œ œ œ œœ œ œœ œœœ œ œ œ œœ œ
f œœœ œœœ œœœ œœœ
? # # # 42 ! ‰
Piano
œ œ œ œ œ J ‰ œ œ œ œ ‰
œ J
### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œœœ U® œ œ œ
Lento
œ œ œ œ œ
5
&
RÔ
p
œœ œœ œ
œ œ
œ œœ œ
œœ KrU® ! ‰
? ### œ œ œ œ œ œ
FIM
œ ‰ œ J ‰ œ œ œ
J œ œ
### œ œœœ œ œ œ œ j œ
œ œ œ œ œ œ œ œ œj œ œ œ
9
& œœ œ œ œ
? ### œ œ œ œ œ
J
‰
### œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œj œ œ œj œ œ œ
13
&
œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœ œœ

? ### œ œ œ œ œ œ œ ‰
œ œ œ J
# # # œr œ œ œ U® œ œ œ œ œ œ œ œ
2 075
œ œ œ
œ œ œ œ œ œ œ œœ œ
17
&
p
œœ œœ œœ œœ œœ œœœ œœ œœ
? ### œ œ
œ œ œ œ œ J ‰ œ œ œ œ œ
J
‰
œ
# # # œr œ œ œU œ œ œ œ œ œ
œ œœ œ œ œ ®œ œ œ œ œ œ ®œ. œ. œ.
21
&
p
œœ œ œ œ œ œœœ œœ œœ
? ### œ œ œœ œ œœ œœ ‰ œ œ
espress.
œ œ ‰
œ J œ œ J
œ
### œ
œ œ œ œ œ œ œ œ œ. œ. œ. œ œ œœœœ œ œ œ ®œ. œ. œ.
25
&
"
? # # # œ # œ œ œ œ œœœ œ œ œ œ œ œœœ œ # œ œ œ œœ
œ œ œœœœ œ œœœ
œ œ
29
### œ œ œ œ œ #œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ œ
&
f
? ### # œ œ œ œœ œ œœ œœ œ œ œ œ œœ
œ œ œ œ œ œ œ #œ
œ
œ œ
œ œ œ œ œ
%
### œ œ œ . . . œ œ
œœœ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œ
33
&
p f
? ### œ œ œ œ œ œ œ œ œ ‰
œ œ œ œ œ œ J
076
Devagar
œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ
‰ ‰ œ œ œ œ œœ œ œ
& 68 J œ œ
J œ
> ‰
p
^ ^ ^ ^ ^
Piano
& 68 œ œ œ œ œ œ œœ œœ œ œœ œœ
œ . œ œ œ œ œœ œJ œœ œ œ œ J œ
J J
>
œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ
œ œ œ œ‰ œ œ œ œ œœ œœ
J
3
& J J Œ
^ ^ ^ ^
& œœ œœ œ œœœ œœ œœ œœ œœ œœ
œ œ J œ œ œ œ
J J J
œœ œœ œœ œœ œ œ œ œ œœ œœ œœ œœ œœ œœ # # œœ n œœ œœ œ œœ œœ
œ œ œ œ œ
5
&
^ ^ ^ ^
& œ œœ œœœ œ œœœ œœ œœ œœ œ œœœ œœ
J J œ œ J
J
œ œ œœ
œœ œ œ
œœ œ œ œ œ œœ œ œ œ œ
! R œ œ œ œ œ ! œ œ
7
&
rit.
^ ^ œœ œœ ^ ^
a tempo
&œ œœ œœ œœ œœ
œ œ œ œœœ œœœ œ œ
J J œ œ J
J
œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ
œ œœ œ œ œ œ
2 076
œ œ œ œ œ
9
& J J
> ‰
^ ^ ^ ^
& œœ œœ œ œœœ œœ œœ œœ œœ œœ
œ œ J œ œ œ œ
J J J
>
œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ
œ œ œ œ‰ œ œ œ œ œœ œœ
J
11
& J J Œ
^ ^ ^ ^
J J œ œ J
J
œœ œœ œœ œœ œ œ œ œ œœ œœ œœ œœ œœ œœ # # œœ n œœ œœ œ œœ n œœ
œ œ œ œ œ
13
&
^ ^ ^ ^
J J œ œ J
J
œ œœ
œ œ œœ œœ œ œ œ œ œœ
! R œ œ œ œ
15
&
f
r
rit.
^ ^ œœ œœ ^ ^
a tempo
&œ œœ œœ œ œ œœ œœ œœ
œ J œ œœ œ œ
J œ
J
077
œ
Andante
b .
& b 42 œrœr œ œ œ œ œ œ œ œœœœr œ œ œ
œ œ .œ# œ n œr œ .# œ œ œ œ œ œ# œ œ # œœœ œœœ n œr
œœ œ œ œ œ œœ
F F
œ œ œ œœ œ œ œ œ
œœ œ # œœ œ œœ œœ œ r! ‰
œ # œœ
Piano
? b b 42 ‰ ! œ
œ œ œ œ œ œœ
œ
%
b . œ œ . œ # œ œ œ œ œ . œ œ œ œr
& b œ œ œ œ œ œœ œ œr œ œ œ
5
œœ
œ œœ œœ œ œ œœ
r
œ œ
œ œ g œœ œ œ œ g œœ
? bb œ gg œœ g œ gg œ œ œ œ g
œ œ ggg œœ ‰
œ
gg œ ggg œ gg œJ ‰ œ # œ J
œ œ œ
"œ fi
œ œœ
bbœ n œ œ œ œ œ gg œœ ggg œœœgggg œœ œœ
œ œœ ! ! r œ œ œ b œ . œ . œ œ . œ n œ nn#
9
& œ œ œ gg œ œ œ #œ œ œ œ
n œ œ œœœ œ œ œ œ œœ œœ œ
? b œ œ œ !‰ œ œœœ œ œ œR ! ‰ #
œ nn
?
b œ & R œ œ œ œ
œ œ #œ œ
# œ. œ œ œœœ œ œ œ œ œ œ #œ
œ œ. œ œ. œ
r
14
& œ. œ œœ œ
œœ œœœ œ œœ
?# œ J ‰ œœ œ œ œ œ ‰ œœ œ ‰
J
œœ œ
œœ œ œ œ J ‰
œ œœ J œ
œ
# œœœœ œœ œ œ œ œ
2 077
œœœœ
œ œ œ œ œ œ œ œ œ œ œ œ œ œœ
18
& œœœ œ
œœ œœ œ œ œœ œ œ œœ œ œ œ œ ‰
?# œ œ ‰ œœ œ J ‰ œœ œœ œœ ‰ œœ
œ J œœ œœ J J
œ
# œ. œ œ œœœœ œ œœ œ œ œ œ œ œ #œ
. œ . œ
r
22
& œ. œ œ
œ
œœ œœœ œœ œœœ
?# œ J ‰ œœ œ œ œ œ ‰ œ J ‰
œœ œ
œœ œ œ œ J ‰
œ œœ J œ
œ
U
# œœœœ nbb
œ œ œ n œ œ œ œ œ œœ .. œ œ . œ œ . œ œ . œ œ . œ œ . œ œr œ
26
& œ œ n œr b œ œ
œœ œœ œ œ
œ œ œ œ
œ œ ‰ œ œ œ J ‰ œœ œœ œ œ œœ
D.S. al O
?# ‰ nb
œ J œœœ œ bœ J b
œ
bb fi œ œ .œ œ r œ œ . œ œ œ œœ
œ œ œ œ œ œ œ œ # œ œœ n œ œ œ œ . # œ œ œ # œ œ # œœœ œœ ‰
œ
30
& œ œœ œœ œ r r
J
œ
#
rit.
? b œ œ œ
œ
b œ œœ R! ‰ œ œœ œ œ #œ œœ œ # œœ œ œœ œ j‰
dim.
œ œ œ œ œ
œ œ œ
078
#
& 42 œ œ œJ œ œœ œ œ œœ œ œ
J œ #œ œ œ œœ
F
Piano
œœœ œœœ œœœ œœ œœ œœ œœ œœ
? # 42 Œ bœ œ
œ œ œ œ
œ
œ œ œ œ œ
# œ œ
œ œ
5
& J œ œ ˙ ˙ œ
p
?# œ n œœ œœœ œ œ
œ œ œ œœ
œœ
œ
œœ
œ
œ œ œ œ œ Œ
#
& .. œ œ œ œ œ œœ œ œ œ
9 3
œ ˙ œ
œ œ œœ œœ œœ œœ œœ
? # .. œ œ œ œ œ
œ œ #œ œ Œ
œ œ #œ œ
# œ œ
& œ œ œ.
13
œ œ œ œ œ
3
œ ˙ œ
?# œœ œœ œœ œœ œœ œœ œœ œ
œ œ œ œ œ œ œ œ œ œ #œ Œ
œ œ œ œ #œ œ
# œ œ œ œ
& œ œ œ œ œ œ œ ˙ œ
17
œœ œ œœ œœ œœ œœ œœ
œ œ œœ
3
?# œ œ œ œ œ œ œ œ #œ œ Œ
œ œ œ #œ œ
# œœœ
2 078
œ œ œ
21
& œ œ ˙ ˙ œ
œœ œœ œœœ œœœ œœœ œœœ œœœ
3
?# œ œ œ œ œ œ #œ nœ œ Œ
# œ œ
& œ œ œ œ œ œ œ #œ
25 3
œ ˙ œ
œ œ # œœœ œœœ œœœ œœœ # œœœ # œœœ œœœ
? # œ œ n œ œœ œ œ nœ œ
r
œ œ œ
# œ U
œ
œ œ œ œ ˙ ˙ œ œ œ
29
&
œœ U œœœ œœœ œœœ œœœ œœœ œ œ œœ œœ
3
?# œ # œ œ œ œ œ œ œœ œ
œ œ œ
# œ œ œ œ œ. œ œ œJ œ œ #œ œ.
33
& J œ œ
bœ œœœ œœ œœ
? # b œ œœœ œœ œœ
œ œ
# œœ
œ
œœ
# œ œ .. œœ
œ œ œ Œ
37 1. 2.
& J œœ ˙ ˙ œ ˙
œ œ œ œ œ œ f
? # œ n œœœ œ œœ œ œ œ œœ œ œœ œœ Œ .. œœ œ œ Œ
œ œ œ
œ
079
œ œ œ œ œ ®œ œ ®œ œ œ œ œ œ œ œ
œ œ ® ® œ œ # œ œ
& 42 ‰ ® ® œœ ®
œ
! ® # œœ ® œœœ ®
œœ œ
"
œ œœœ œœ ! œœœ œœœœ
? 42 ! œœœ ! œœ ! # œ # !
Piano
œ ! œ ! ! ! œ ! ! !
œ œ œ
œ ® œ œ ® œ œ œ #œ œ œ œ œ œ œ œ ®œ œ ®œ œ œ œ œ
® ® ! œ ® œ ® œœ ® ® ®
4
& œ
œœ œœœ ! œœœ f œœ œœ # œœ
? œ ! #œ ! œœœ œ ! œœœ
! œ ! ! ! œ ! ! œ ! œ !
œ œ œ
œ #œ œ œ œ œ œ œ œ nœ ### !
& ® œœ ® œœœ ® œœœ ® # œœœ ® œ ® # œ œr ! ‰ ! ! œœ ‰ œ
7
œ œœ œ œœ œœ œœœ œ œ œœœ œ
R f .
œ ! œœ œ œ
? œœ ! ! ! ! ‰ œœ ® ® œ
# ### œ œ œœœ œ œœœœ œ œœ œœ œ œœœ œœœ
œ œ œ
J
*
## j œ #œ œ œ œ œ ˙ œ œ œ œ œ ‰ œ œ œ œ œ œ
& # œ œœ œ œœœ .œœœ œœœ œœœ ‰
11
. œ œ œ œ
.
œœ œœ œœ œœ œœ !
? ### ! !
œ
!
œ œ œ œ
! !
œ œ œœœœ œ œœ œœ œ œœ œœœœ
œ œ œœ œ œ
### œ ˙ œœ œœ œœ œœ œœ ˙ œ œ œ œ œ œœ
œ ‹œ œ œ œ #œ œ œ œ œ œ
15
& œ œ œœœ
! œ œ!œ!œ ‰ œ œ
. . ƒ . .
? # # # œœ œœ œ œœ œœ œ œœœ œœœ œœ œ œœ œœ œœœ œœœ œ œœ œœ œœ œœ œœ
œœ œ œ œ œœ œ œœœœ
œ œ
### ˙ œ œ œ œ œ œ #œ œ œ œ œ ˙ œ œ œ œ œ œ œ œ œ œ
2 079
œ
˙ œ œ œ œ œ œ
! ‰ œ # œ œ œ œ ˙ œ œ œ œ œœ ! œœ ‰ œ œ œ œ œ ## œœ
19
& ! ! ! !
. . .œ
œ œ œœœœ œ œœ œœ œœœ œœœ œ œ # œ œœ œœ œœ. œœ œœ
? ### œ
œ œ œ #œ œ #œ œ
œ œ œ œ œ œ
œ œ
### œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ œ œ. œ œ œ œ œ œ œ œ œ œ œ œ œœ
23
& œ œ œ œ œ œ œ œ
. . .
œ œ œ œ ƒ
? # # # œœ œ œ œ œœ œœ # œ œ œ œœ
œœ œ œ œ œœ œ œ œ
a tempo œ
œœ œœ œœ
rit.
œ œ œ
## œ œ nnn
& # œœœ œœ œ œœ œœ œ œœ œœ œ œœ œœ œœœ œœ œ œœ œ œ œ œœ œ œœ œœœ œœœ œœ œ œœ œœ œ œœ œœ œ œœ œœœ
26
œ
œœ œœ œœ œ œ œ œ œ œ
? ### œ œ œ œœ œœ œœ œœ œœ œœ nnn
œ œ œ
œœ œ œœ œœ œœ œœ œ œœ œœ œœ œœ œ œœ œœ œœ œ
& œœ œ œ œ œ œ œœ
29
œ œ œ œ œ
"
? œ œ œ œ œ œ
œ œ
œ œ j r
& œœœ # œœ œ œœ œ œ œ œœ œ œœ œœœ œ œ œ œ œ œ œ #œ œœ ‰ ‰ ! œ
31
œœ œ œ œ œ œ œ œ œ œ œœ œ
œ œ œ œ œ œ œœ œœ œœ
? # œœ œœ œœ # œœ œœ œœ
#œ œ œ œ œ œ
Lento
œ œ œ œ œ . œœ œ œœ œœ œ œœ œœ œ œœ œœœ
079 3
œ œ
& œ œ œ œ œ œ! . # œœ œ œœ œ œ œ œœ œ œœœ œœœœ œ
34
! œ
? œœœ œœ
œ # œœœ œœ
œ
œœ
œ # œœœ œœ
œ œ
œœ
œ
œœ
œ
œœ
œ
œ œ # œ #œ œ
œ œ œ œ œ œ. œ ! œ œ
& œœœ œ #œ œ j œ
38
J œ œ œœ œ œ
œœ œœ œœ œœ œœ œœ # œœœ œœ œœ œœ œœ œœ
? œ œ œ œ
œ œ œ œ
œ
œ œ œ
œ œ
& œ œ œ œ œ œ . # œœ œ œœ œ œ œ œœ œ œœœ œœœœ œ œ œ œ œ œ . œœ œ œœ œœ œ œœ œœ œ œœ œœœ
42
! ! œ
œ œœ œ œ œ œ œ œœ œœ œœ
? œœ œ # œœ œœ œœ # œœ œœ œ œ œ
œ œ #œ #œ œ œ
œ œ œ œ œ œ œ. œ ! œ œ œ œ. œ
œ #œ œ œ œ ! œ . œ
46
&œ J œ œ.
œœ œœ œœ œœ œœ œœ œ œ œ œœ œœœ ! œœœ
? œ œ œ œ # œœ œœ œœ
œ œ œ œ
œ œ ! œ ! !
. .
& œ œ œ . œ œ . œ œ . œ œ . œ œ . œ# œ . œ œ œ œ . œ œ . œ œ . œ# œ . œ œ . œœ
50
œœ œœ œœ œœœ œœœ # œœœœ œœ œœœ ! œœœ œœ œœ œ

? ! ! œ # œ ! œ ! ! # œ ! ! ! ! # œ! ! ! œ !œ ! œœ !
œ œ œ œ
œ
4 079
j œ œ œ œ œ
&œ œ œ #œ œ
54
œœ
œ œ œ œ œœœ œ œ #œ J
œ œœ œœ œœœ
? œœ œ œœ
œ #œ # œœœ œœ œœ œœ
œ ! œ ! œœ œ #œ œ
œ œ œ œ
œ œ
œœœ œ œ œ
& œ œ œ œ œ œ œ #œ œ œ œ #œ œ #œ œ œ œ œ ® œ ® œœ ®
58
œ
œ
œ œ œœ # œ
œ œœ # œœ œœ " œœ œœœ !
? œ œœ œ œ œ œ œ œ ! œ ! !
œ #œ œ œ œ
œ ®œ œ ®œ œ œ œ œ œ #œ œ œ œ œ ® œ œ ® œ œ ® œ # œ ®œ œ œ œ
® ® œ œ
! œ ®œ# œœ ® œœœ ® ! œ ®œ ® œœ ®
62
& œœ œ œ
œœœ œœ ! œœœ œœœœ œœœ œœœ ! œœœ f œ
? #œ # ! # œ ! œœœ ! œœ ! !
œ! ! ! œ! ! ! œ! ! !
œ œ œ
œ
°
œ ®œ œ ®œ œ œœ œ œ >œ
® ® ® # œ œœœ ® œ œœœ ® œ œœ ® œ œ ® n œ œ ® r ! ‰ Œ ggg œœœ ! ‰
g
66
& œ œœ # œœœœ #œ œ gR
œ R ƒ
œ œ # œ ! œœ œ œ
? œ ! œœ ! œœ ! œœ ! œœ ! ! ! ! ‰ œœ ® ® œ
# œ œ r
œ œ! œ ! œ !‰
œ
>
080
Vagaroso %
œ œ
& b 42 œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ œ œ œ
œœ œœ œœ #œ œ œ # œœœ œœœ œœ œœ œœ
œ œœ œœ œœ
Piano
? b 42 ! ‰ œ œ œ œ #œ œ œ œ œ
œ
œœ œœœœ œ œ j
b œ œ œ œ œ œ œ œ œ œ œ œ œ ! r n##
5
& œœ #œ œ œ #œ
œœ œœ œœ œ œœœ œœœ œ
œ œœ œœ
Fine
?b œ œ œ œ œ œ j! ! n##
œ œ œ œ
#
œ œ
##
œ œ œ œ œ œj œ œ œ œ œ œ œ œ œ œ ! œ
9
&
œ œ œœ œ œ œ œ œ œœ œœ œœ
? # # œ œœ œœ œ œ œœ œœ œœ œ œœ œ œœ œ œ œ œ
œ œ
# j œ œ œ œ œ œ œ œ
& # œ œ œ œ œ œ œ œ ! œ
13
œ œ œ œ œœœ œœœ œœœ œœ œœ œœ œœ œœ

? # # œ œœ œœ œœ œ œ œ œ œ œ œ
œ œ œ
2 080
## œ Uœ œ œ œ œ U
œ œ œ œ œ
œ œ . . œ œ œ œ œ œ œ œ œ œ !œ
17
&
œœ œœ Uœœ œœ œœ U
? ## œ œ œ œ œ œ œ œœ œ œœ œœœ œ œ œ œ
œ œ œ
%
## œ Uœ œ œœ œ U
œœ œ œ
œ œ . . œœœ œœ œœœœ œ bœ
œ œ œ
21
&
œ œ Uœ œœ œœ Uœœ œœ œœœ
? # # œ œ œ œœ
œ œ œ œ œ œ œ œ œ œ r! ‰
œ œ œ œ œ
081
Tempo de fado vagaroso

œ œ r
& b 42 œ œ œ œ œ œœœ œœœ œ œ œ œ œ œœœ œœœ œ œ œ œ œ # œœœ œœœ œœœ œœœ œœ œœ œœœ œœœ !œ
p
œœ œœ œœ # œœ œœ œœ f œœ œœ œœ
œ œ œ
Piano
? b 42 ! ‰ œ
œ œ œ œ œ œ œ œ œ œ œœ œ œœ œœ œœ
œ œ
r
& b .. œ . œ œ œ œ œ œ œ œ œ œ j !œ
5
#œ œ œ œ œ œ œ œ
? b .. œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ
œ œ œ
œ œ
& b œ. œ œ œ œ œ j ! œR
9
œ œ œ œ œ #œ œ œ œ œ œ œ œ
F
? œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœœ œ
b œ œ œ œ
œ œ
.
& b œ #œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œJ œ ! œR
13
œ œœ œœ œœ # œœœ œœœ œœœœ œœ œœ œœ

? b # œ œœœ œ œœ œ œ œ œ œ œœ œœ œœ œ
#œ œ œ œ œ œ
2 081
Estribilho
b œ. œ œ œ n##
# œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œœ
17
& œœ œœ
f
œœ œœ œœ œœ
? # œ œœ œ œœ
b #œ œ œ œ œ œœœ œ # œœœœ # œœœ œœœ œœœ #
n#
œ œ œ
Mais Depressa Vagaroso

U
#
& # œœœ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œ œ œ œ b œ œ
21
œ œ œ œ œ œ œ œ œ œœ œœ œœ
p
? ## œœœ œœœ œœœ œœœ œœœ œœœ œœœœ œœœœ U
œ œ œœœ ! ‰
œ
œ
œ œ
œ œ R
œœ
## œœ œœ œœ œœœ œœœ œœ œœ r . 2.œ œ œ œœ
œ œggg œœœ œœ œ œ œ œ gg œ œ œ œ
œœ œ œ n œ œggg œœ ! œ . œœ œœ œœ J‰
25 1.
& gg œ n œgg œœ œœœ nœ œ

g g
œœ œœ œœ f œœ œœ œœ
? # # b œœœ œœ œœ
œ œ œ n œœ
œ œœ œœ
œ œ œ œ œ œ n œœ œœ œœ .. œ j
œ ‰
œ œ œ œ œ
082
#
& 42 œ œ œ œ œr œ œ œ œ œ œ œr œ œ
œ œ œ œ F
Piano
? # 42 œ œ œ œ œ œ
œ œ
#
& œ œ œ œ œ œ. œ œ œ œ œ œ œ œ œœœœœ œ
3
œ œ œ œ œ œR œ œ J
?# œœ œœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ

œ œ œ
# r!
œ œ œœ œ œœ œ œ œœ œ œ œ
7
& œœ
œœ œ œ œ œ œ œœ œ œ
r
œ
œœ œ œœ œœ œœ œœ œœ œœ œœ œœ
?# œ œ œœ œœ œœ œœ œœ œœ œ œ œ
œ œ œ œ œ
#
11

œ œ œ
# r!
œ œ œœ œ œ ! ! œ
15
& œœ
œœ œ œ œ œ œ œ œ œ œœ œ œ œ œ
œœ œœ œœœ œœœ œœœ œœœ œœœ œ œ
?#
œ œ œ œ œ œ œ œ œ œ œœ œœ œ
œ œ œ œ
# œ
2 082
j j
œ œ œ œ œ œ #œ œœ œ # œ˙ œ œ œ œ œj œ œ œ œœ
19
& œ
! œ œ œ !œ œ œ
? # œ œœœ œœœ œœœ œ # œœœ œœ
œ
œœ
œ
œœ œœ
œ œ œ
#
œœ œ œ œ
œ œ œ œ
# j j œggœœœ ! !
œ
23
& œœ œ œ œœ œ œ œ #œ
œ œœ œ œ œ œ œ œ #œ œ œ ggg œ œ
œœ œœ œœ œœ œœ œœ œœ œœ œœœ r
?# œ œ œ œ # œ œ œ œ œ
# œ œ œ ggg œ ! ‰
œ œ gg œ
œ œ œ œ
#
27

œ œ œ
# œ œ œ
& œ œ œ œ œ œ œ œ œ œr ! œ œ œ œ œ œ
31
f
œœ œ œ œ œ
œœœ œœœ œœœ œ
œœœ œ
œœœ œœ œœ œœ
?# œ œ œœ œ œ œ
œ œ œ œ œ
r œœ œœ
# œ œ œœ œ œ œ œœ
œ œ œ œ œœ œ œ œœ
& œ œ œœ œ œ !œ ! ! R ! ‰
35
?# œœ œœ œœœ œœœ œœœ œœœ œœœ œ ! œ ! r! ‰

œ œ œ œ
œ œ
083
Allegretto Moderato
b œ œ œ œ.
& b b b 42 r œ œ œ œ œ œœœ
œ.
!
œ œ
œ œ œ œ œ œ j
? b b b 42 œr œ œ
Piano Com sentimento
œ œ ‰
b
œ
bb œ œœ œ œ œ œ !
&bb œ œ œ œ œ œ ! œ nœ
3
œ œ
? b b œ œ œ œœ œœ œ n œ œ œœ œœ
œ n œ œ œ œœ œœ
œ œ œ œœ œœ
bb œ œ œ
b œ
& b bb œ œ œ œ œ œ ! œ œ œ œ œ œ !œ
7
œ
œ nœ
œ œ œ œœ œœ œ œ œœ œœ œ œœ œœ
? bb b œ œ œ œ œ œ nœ œ n œ œ œ
b œ œ œ
bb b b œ œ œ œ œ œ œ œ ‰ œ œ . n œ n œ œ œ œœ œ œ œœ œœ ! œ
11
& œ œ. nœ nœ œ œ œ œ
œ œœ œœ œœ œ f œ
cresc.
? b b œœœ œœ œ œ œ n œ
œ œ
œ
œœ œœ
œ œ ‰
bb œ œ œ œ œ œ J
œ. U
œœ œ œ œ œ œ
r
bbbb œ . œ œ ! œ œ. œ œœ œœ œ œ œ œ œ J ‰
15
& nœ.
œ œœ œ œœ œœ
dim.
? b b œ œ œ œ œœ œ œ œ n œ œ œ j‰
bb œ œœ œ œ œ
084
rit. a tempo
# nœ.
& # 42 œœ œœ œœ œœ œœ œœ œœ œœ # œœ œœ # œ- . # œœ n œœ œœ œœ œœ œœ œœ ! ! œ
œ
f R
œ œ œ- œœ œœ
# œ œ
Piano
? # # 42 œ œ œ œ œ œ ! ! œ
## r" nnb
! œ gg œœœ œ œ œ
3 3 3
& œœ œ œ œ œ œœ ! ! œœ œœ œ œ œ ggg œœœ ggg œ œ œ œ

J gg
F
? ## œœ œœ ! ! œ œœ ! œ œ œ nnb
œ œ œ
œ
R
& b ..
5
œœ œ œ œ œ œ œr œ # œ œ œ œ
r r
3
œ #œ œ œ œ
bœ œ r
œœ œ n œ # œ
œ
œ œ
? .. œœ œœ #œ œ œ œ œ
b œ œ œ #œ
œ
& b œr œ œ œ nœ
7
œ œ œ
r
œ œœ # œ
r
#œ œ œ œ
r
œ œ œ
?b œœ œœ œ œ œR ! ‰
œ œ œ œ
2 084
& b # n œœ # œ œ œ œœ œ œ œ œ œ œ œ œ
9
œ
# # œ˙
?b œ # œœ œœ œ œœ œœ
œ œ nœ #œ
& b #œ œ œ œ œ œ œ œ œ œ
11
# ˙ œ #œ œ œ œ œ
œ œ œ nœ #œ
? b #œ œœ œœ œ nœ œ
œ nœ œ œ œ œ
& b œ œœ œ œr œ œ œr œ œ œr œ # œ
13
œ œ œ
r
3
œ #œ œ œ
œ œ œ œœ œ n œ # œ
? œœ œœ #œ œ œ œ œ #œ
b œ œ œ
œ
& b œ#r œœ œ œœ œœ œœ œœ # œœ
15
œ œ œ
r
œ # œœ
r
œ œ
. œ œ
r
œ
? œœ œœ œ
b œ œ œ œ œ R ! ‰
œ
œ.
& b œ˙
r
œ œ œ
r
œ œ œ œ œ œœ
r
17
œ
œ
œ œ
œ œ
œ œ
R
? œœ œœ œœ œœ œœ œ œ
b œ œ œ
##
084 3
& b œœ ®œ œ œ ®œ œ œ ! ! n
19 3 3
#œ œ œ #œ œ œ #œ
J p
œœ œœ œœœ œœ œœ
? b #œ œ ! ‰ n##
œ œ
œ
R
#
& # œ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œ ! !
21
œ œœ œ œ
? ## œ œœ œœ œ œœ œœ œ œ œ
œ
##
œ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œ ! ! œœ œœ
23
& œ œœ
? ## œ œœ œœ œœ œœ œ œ œ
œ œ
## œ œ œ œ œ œ œ œ # n œœ .. œ n œœ œœ œœ œœ œ œœ ! ! œ
25
& œ œ œ œ œ œ #œ œ - #œ œ œ
R
-
? ## œ œœ #œ œ œ œ
œœ œœ
! ! œ
œ œ
##
! !
27 3
& œœ œ œ œ œ œœ ! ! œ œ œ
3
œ œœœœ œ œœ
œ
R
? ## œœ œœ ! ! œ œ œ !
œ œ œ œ œ
œ
##
4 084
œ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œ ! !

29
& œœ œ œ
œ
? ## œ œœ œœ œ œœ œœ œ œ œ
œ
#
& # œ œœœ œœœ œœœ œœœ œœœ œœœ œœ œœ œœ œœ œœ œœ œœ œœ ! ! œœ œœ
31
œœ
? ## œ œœ œœ œœ œœ œ œ œ
œ œ
## œ œ œ œ œ œ œ œ # n œœ .. œ n œœ œœ œœ œœ œ œœ ! ! œ
33
& œ œ œ œ œ œ #œ œ - #œ œ œ
R
-
? ## œ œœ #œ œ œ œ
œœ œœ
! ! œ
œ œ
## ! ! œ ! œ gg œœœ œ œ œ ..
35 1. 3 3
& œœ œ œ œ œ œœ œ œœ œ œ œ ggg œœœ ggg œ œ œ œ

J gg
? ## œœ œœ ! ! œ œœ œ œ
œ œ ! œ œ ..
œ
r
œ
ggg œœœ ! ‰
3
## œœ œ œ œ œ
! ! œœ gg
37 2. 3
& œœ œ œ œ œ œœ
? ## œœ œœ ! ! œ œœ gg œœ ! ‰
œ œ œ ggg œ
gR
085
œ œ œ œ œ œ œ. œ œ œ œ‰Œ ˙ % b
& b 42 J J J n ˙˙˙˙ nb b
F 3
œœ œœ œœ œœœ œœœ
Piano
? b 42 ‰ Œ œœœ b
œ #œ œ œ œ œ œ #œ œ œ œ n œ nb b
b œ œ œ. œœ œœœ
& b b œ. nœ. ˙
5
p 3
œ œ œ œœ œœ œ
? b b œ œœ œœ œœ œ œ œ # n œœ œœ œœ œ # n œœ œ
b œ œ #œ
b nœ œ œ. œ nœ j
& b b œ. œ œ œ nœ
9
? b n n œœœ œœœ œœœ n œœœ œœœ œœ œœ œœ

bb œ œ nœ œ œ œœœ œ œ ‰ Œ
J
b œ œ œ œœœœ œ œ œ
& b b œJ œ œ œ
13
J J J J J J
3
F
? b b œ œœ œœ œœ œ œ œ œœ œœ œœ œœ œ
œ œ œœ
b œ œ bœ œ œ œ œ œ
œ
œ
2 085
b
&bb œœœœ œ œ œ œ œ
17
œ ‰̇ œœ .. ‰̇ n œœ ..
3 p
? bb œ œœ œœ œœ œœ œœ œ œ œ
b œ œ nœ œ nœ œ
œ œ
b œ œ œ. œœ œœœ
& b b œ. nœ. ˙
21
œ œ œ œœ œœ œ
? b b œ œœ œœ œœ œ œ œ # n œœ œœ œœ nœ
œ#œ œ
b œ œ #œ
b nœ œ œ. œ nœ
& b b œ.
25
˙ ˙
? b n n œœœ œœœ œœœ n œœœ œœœ œœ œœ œœ

bb œ œ nœ œ œ œœœ œ œ œ œ nœ
b j j œ œ j œ œ œ œ œ œ
&bb œ œ œ œ œ.
29
? b œ œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
œ œ œ
œ j
bbb œ œ œ nœ œ œ œ œ nœ ˙˙ œœ n œ n nnb
33
& n˙ œ œ œ
3
? b œ # n œœœ œœœ œœœ n n œœœ œœœ

bb œ nœ œ bœ œ
nœ j‰ Œ nnnb
œ œ
085 3
œ
& b .. œ œ œ . œ œ œ œ
37
œ œ œ œ œ œ
œ œ
œ
3
? b .. œ œ œ œ œ œ œ œ œ j ‰ Œ
œ œ œ œ œ œ
œ
& b œ œ œ. œ œ œ œ œ œ œ œ œ œ œ œ œ
41
?b œœœ œœœ œœœ œœœ œœœ œœ œœ œœ

œ œ œ œ
œ j ‰ j ‰
œ œ œ œ œ bœ
j ˙ œ œ
&b ˙ œ œ œ œ œ
45
? # œœ œœ œœ # œœ œœ œœ œœ œœ
j
b
œ
œ
œ #œ œ œ œ œ œ œ ‰ Œ
fi
œ j
& b œ œ œ. œ œ j j œ œ ..
49 1.
J œ ‰̇ œœ ‰ œœ œœœ
œ ‰ Œ
Jf
3
? œ #œ œ œ œ œ œ ‰ Œ ..
b œ œ
J
j % fi j j
œ œœ
&b œ œ ‰ œœœ œœ ‰ Œ
53 2.
n n n ˙˙˙˙ J
S
S
? œ ‰ œ ‰ œ ‰ Œ
b œ œ œ #œ œ œ nœ nœ J J œ
J
086
œœ œ œ œ œ œ œ œ œœœœœœ
Moderato
2 œ œ œ # œ #œ œ
&b 4 ‰
f 3 3
œœ œœ
3 3
œœ œœ # œœœ œœœ
Piano
? b 42 ! œ œ œ œ œœ œœ œ ‰ Œ
œ œ œ œ œ œ J
j
œ œ œ œ œ œ œ œ # œ œœ
nœ j% # #
rit.
œ #œ #œ
&b nœ ‰ ‰ œ n
5
œ œ #œ J
? œœ œœ œœœ œœœ #
b œ œ œ œ #œ œ œ ‰ Œ n#
# j Uœ
a tempo
& # œ œ œ œ œ j
9
J œ j œ
œ œ
? # # œ œœ œœ œ œœ n bUœœœ œœœ œœœ œœœ œœœ
œ œ œ œ œ
# j
& # œ j j j ‰ ‰ œ
13
œ œ œ œ
˙ œ
? # # œ œœœ œœœ œ œœ œ
œ œœ œ œ œ œ œ œ
œ œ œ
## j U œ j
œ œ œ œ œ
17
& J œ j œ
œ œ
? # # œ œœ œœ œœ n bUœœ œœœ œœœ œœœ œœœ
##
2 086
œ j j
21
& œ
œ œ œ ˙ œ œ
? # # œ œœœ œœœ œ œœ œ
œœœ œ œœ œœ œœ œœ œ œ
œ
##
j ‰ œ bœ œ œ œ nœ
25
& œ œ œ œ
3
nœ œ œ œ
nœ œ b n œœœ œ n œœ œœ œœ œœ
? # # n œ b œœ œ œœ nœ œ œœ nœ œ œ
##
j ‰ œ œ
29
& œ œ nœ œ œ nœ œ œ
3
#œ œ œ œ
? # # # œ œœœ œœœ œ
# œ œœ œœœ œ n œœ œœ œœ œœ
œ œ
œ œ
##
rit.
j ‰ œ œ œ œ œ bœ nœ œ ‰ nœ nœ œ ‰ nœ
33
& nœ nœ nœ
? # # n œ b n œœœ œ œœœ nœ
b n œœœ œœ
œ œ n œ n œœ n œ œœ n œ n œœ n œ œœ
# nn
a tempo
& # bœ ‰ bœ ‰ ˙˙ b
37
œ œ nœ œ œ n œ n œ œ œœ # œœ ˙
? # # œ b n œœ œ œœ œ b n œœ œ œœ nœ ‰ Œ n
J œ œ œ œ nb
> j j
086 3
.. œ œ j
&b ‰ œ œ œ œ œ œ œ œ œ œ
41
J J J J J J œ œ
? b œ œœ œœ .. œ œœ j >œ œœ œœ œ œœ j # œ œœœ œœœ

œ J œ œ J œ œ
> >
j j j j j j j j
&b œ œ œ œ œ œ œ
46
œ œ œ œ œ
? b # œj œœœ j >œ œœ œœ j œœ j > œœ œœ

œ œ #œ œ œ œ œ
> >
j j j j j j
&b œ œ œ œ œ
50
œ œ œ J J œ œ œ œ
? œ
œœ j > œœ œ œœ j œœ j > # n œœœ œœ
œ
b J œ œ œ œ nœ #œ
> >
fi 1.
j j j j j j œ ..
&b œ œ œ œ œ œ œj œ
54
œ œ œ J
# # œœœœ > n # œœœ œœ n # œœœ œœ œœ œœ
? œ œ œ œ œ ..
b œj œ
j nœ #œ #œ œ œ
>
j
œ œ œ #œ œ nœ jfi # # j j
œœ ‰ Œ
&b œ nœ ‰ ‰ œ n œœœœ ‰
58 2.
œ œ #œ J œ œœ
> >
? œ # j j
b ‰ Œ n# œ œ ‰ œœ ‰ Œ
% fi > œ
>
087
Fado Marcha
# 2 j œ # œj œ œ œ œ
& 4 œ # œj n œ œ # œ œ . œ
#œ œ J œ œ œ
3
œ
f
# œœœ œœœ œœœ # œœœ œœœ n # œœœœœœœœœ # œœœ œœœ
3
Piano
? # 42 Œ ‰
#œ œ # œ œ œ œ #œ
œ
# j œœ %n
œ j ‰ œ œ ‰ œœ ‰ Œ bb
& œœœœ œ œœ
5
œœ œ
3
œ œ J œ J
ß ß
?# œ n œœ œœ œœ œœ œ j nb
œ œ œ œj ‰ œ ‰ œj ‰ Œ b
œ œ
œ
b j
& b œ. œ œ.
9
œ œ œ œ œœœ ˙
3
p œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ
? bb œ œ œ œ œ œ
œ œ
b j
& b œ.
13
œ #œ
3
œ œ œ œ. œœœ ˙
? b œ # œœ œœ œœ œ œ
œ #œ œ œ
œœ œœ œœ
œ œ œ
b œ œ œ œ
b j
&b j
17
j j œ œ œ œ œ.
3
œ œ œ œ œ ˙
? b j ‰ Œ ˙
˙˙ # œœœ œœœ œœœ # œœœ œœ
b œ œ œ #œ œ œ
2 087
bb j j œ œ œ œ.
21
& j j
3
œ œ #œ ˙
œ œ œ
˙ œœ œœ œœ œœ œ
? bb j ‰ Œ # ˙˙ œ œ œ œ œ œ œ
#
œ œ
b j
& b œ. œ œ.
25
œ œ œ œ œœœ ˙
3
? bb œ œ œ œ œ œ
œ œ
b j
& b œ.
29
œ #œ œ ˙ ˙
œ œ
? b œ # œœ œœ œœ œ # œœ œ
œœ œœ
œ
œœ œ
b œ œ œ œ nœ
b j j œ
& b œj œ œ œ œ œ œ œ œ œ œ œ
33
œ J J
? b œ œœ œœ œœ œœ œœ œœ œœ œœ
b œ œ
œ œ œ œ œ
b j j - n#
& b œ #œ œ œ œ œ œ #œ œœ œ # œj n
37
‰̇ n œœ ‰ œœ nœ
J
? b # œœ œœ # œœ œœ
nœ œ nœ j #
b œ #œ
œ œ œ ‰ Œ nn
œ
# . 3
087 3
& . œ œ œ œ. j j j
41
œ œ œ œ œ œ œ. j
3
œ œ œ
œœ œœ œ œœ œœ œœ œœ
? # .. œ œ œ œ œ
œœ
œ œ œ œ
#
j
45
& œœœ . j
œ œ œ œ œ œ œ œ. œ œ
3 3
œ #œ
œœ œœ œ œ œœ
?#
œ œ œ œ
œ œ œ œ œ œ #œ nœ
œ œ
# œ œ #œ œœœœ œ œ
49
& nœ œ œ œ. œ œ
3
œ #œ œ J
œ œœ œ œ
n # œœœ œœœ œœœ # œœœ œœ
3
?# #œ œ œ œ œ
# œ #œ œ
#œ œ #œ œ œ œ
#
& œœœœ j œ j ..
53 1. 3
œ œ œ œ œ œ. j
3
œ œ œ œ œ #œ
f
?# œ n œœ œœ œœ œœ œœ œœ œœ œœ
œ œ œ œ j ‰ Œ ..
œ œ
# j >œ %
œ œ j ‰ œ œ ‰ œ
œ ‰ Œ
& œœœœ œœ
57 2. 3
œ œ J œœ œ œ œ
> J
œœ œœ œœ œœ œœ ß ß
?# œ œ ‰ œ ‰j j ‰ Œ
œ œ J œ
œ œ
œ
088
%
& b 42 Œ ‰ j ..
œ œ œ œ œ # œ˙ œ œ œ œ
œ cresc.
!
? b 42 ˙>
r>
.. œ œ œ œ # œ œ œ œ # œr >œ
Piano
œ #œ
Í œ œ
rall.
accel.
œ
& b œœ . œ . œ œ. œœ
3
# œœ œ. J
œ f œœ œœ œœ œ
? œ œ œ œ œ œ œ œ œ
b œ œ œ
œ
a tempo
œ œ œ œ œ
&b ˙ œ œ
5
œ ˙ œ œ
p
œ œ r œ œ œ œ
œ
#œ
?b œ œ
r
œ œ #œ
œ œ
" r
rall.
& b œœ . œ # œœ .
j œ
7
œ œœ œœ
J p
œ #œ œ "
dim.
? œ œ œ œ
Fim
b œ œ œ œ œ œœ "
œ œ œ
!
2 088
rall. a tempo
œ œ œ œ
& b œœœ # œœ
n bn œœ œ #œ œœ œ œ #œ œ œ œ œ
9 accel.
J
fœ-j >
S‰ œ
j p‰ œœ
j j
‰ # œœ n œœ ‰ n œœœ
?b œ œ œ #œ
>
r
a tempo
& b nœ œ œ œ œ œ œ œ
rall.
œ. " œœ
11
j p
œ j œœ œœ œ œ
? œ ‰ œ ‰ # œ
œœ œ œ œ œ œ œœ
b œ
& b œ. œ œ œ œ œ œ
13
?b œ œ œ œ œ œ œ œ
U
n##
rit.
œ #œ œ
a tempo
&b œ œ j ‰
15
œ œ œ œ œ
œ œ #œœœ œ œœ œœ
? b # œj œœ ggg œœ # œœ œ œ
gg œ œ n##
g œ
>
088 3
# # ˙˙ ˙˙ ˙˙ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ
˙ ˙ ˙
17
&
p >
S œœ
œ œ œ œœœ œ œ œœ œ
? ## œ œ œ œ œ
œ œ
# # ˙˙ ˙ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ œœ œ ˙˙˙ ˙˙
˙
˙ ˙
19
&
S> œ
ƒ
? ## œ œœ œœ œœ
œ œ
# # b ˙˙˙ b˙
˙˙
˙ ˙ n ˙˙˙ ˙˙˙
˙ ˙
21
&
p j j
œœœ n œœ
? ## ‰ b œœœ
œ œ œœœœ ‰ œœ
nœ œ œœ
œ
˙ ˙
> U
j
# # # ˙˙˙ ˙˙˙ n ˙˙˙ .. œ . œœœ œ
. j
˙.
23
& #˙ ˙ œ œ
S j ƒ j œœ p
‰ œ œœ œ ‰ œœ œœœ œœ "
? ## n œ # œœ œ œœ œ
œ œ "
œ œ
u
4 088
œ. œœ b œœ n œœ b œœ1... œœ œœ
# # œœ .. # œœ œœ œ œœ
œ œ bœ nœ bœ. œ œœ œ œ œ
25
&
j j j j
œœ œœ œ
? # # # ‰œ n œœœ ‰ ‰œ ‰ œœ
cresc.
œ bœ œ
œ œ
œ œ
n œœ U
œ œ œœ
# # n œœ œœœœ n œœœ œœ ..
"
27
&
j j !r œ
f n œœ
dim.
œœ p n œœ œœ œœ !
‰ ‰
U " " ‰
? ## œ œ œ œ œ œ œ ..
œ œ œ " œ " œ " ‰
œ œ
œ
R
U
œ
œœ n œœ œ œœ
# # bb œœœ ... œœ œ œœ œœ
œ n œœ œœœœ œ
œ œœ œ œ
29 2.
&
j j f j ƒj
œœ œœœ n œœ œ
? # # ‰œ bœ ‰
œ œ ‰œ œ ‰ œœ
œ œ œ œ
œ
u
œ œœ
D.S. ao Fim
31
# # n œœœ œœ "
&
p r ! œ
" n œœ œœ œœ " ‰
? ## œ œ œ œ
" œ " " ‰
œ œ œ
œ
R
!
089
& 42 œj .. œ ! ‰ # œj œ # œ œ ! œ ‰ j
œ œœœ œ p
œ œ œ
œ œ
? 42 ‰ .. œ œ œœœ # œ # œ œ œ œ
Piano
œ œœ œ J ‰ œ œ
œ œ œœ œ œ J ‰
œ
œ
&œ ! ‰ # œj œ #œ ! œ ‰ œj
5
œ œ œœ œ œ œ œ
? œœ # œ œJ ‰ # œœœ œœœ œœœ œœ j ‰

œ œ œœ
& œ œ œœ !œ ‰ ‰ œj
9
! j
œ œ œ œ œ œ œ œ œ ! œ œ
p
? œ œœ œœ œœ œ œ œ œœ œœ œ œœ œœJ ‰
œ œ œ
œ œ ! œ œ œ œ !œ œ ‰ œJ
13
& œ !
p œœ œœ
Sentimental
? œ œœ œœ œœ œ œ œ œ
œ œ œ œœ œ œ ‰
œ œ J
œœ!œ œœ œ‰
2 089
œ œ œ œ œ œ œ ! #œ œ ‰ œJ
17
& J
œœœ œœœ œ œ œ œ pœ œ œ œ œ œ œ œ œ œ œ œ œ œ
?œ œ œ œ œ œ œ œ
œ œ œ œ !œ œ œ œ œ ‰
œ œ œ œ ! ‰ œ
#œ œ
21
& J J
S
œœ œ œ œ œ œ œ œ œ pœ œ œ œ œ œ œ œ œœœ
?œ œ œ œ œ œ œ œ ‰
J
œ œ
& œ œ œ œ ! œ œ œ œ œ œ ! œ œ ‰ œ
25
œ J
œœ œœ œœ p œœœ œœœ œœœ S
œ œ œ
? œ œ œ œ œ œ œ œ ‰ œ œ œ œ œ œ œj ‰
œ œ J œ
œ œ ! œ œ œ œ œ ! œ œ1. œ œœ
& œ œ œœ!œ œ ‰ j .. œ œ
29
œ
p f
œ
? œ œ œ œ œ œœœ ‰ œ œœœ œœœ œœœ œ œ ‰ .
œ J . œ œ
Fim
œœ J œ œ
090
%n
# # # 2 . œ. œ œ œ. œ œ œ œ œr œ œ œr œ œ. . . . . œ
Allegretto Moderato
œ œ œ œ œ U
œ œ. œ œ. œ œ œ ggg œœ ! n n
& 4 œ œ œ œ œ œ. œ. gg œ
.. f
p
Piano
# œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ fœ Up n
? # # 42 ‰ ! œ œ œ œ œ œ œ œ œ œ œ œgg œ ! ‰ n n
œ œ ggg œ
R
œ j œ œ œ œ œ ! j
& œ œ œ œ œJ œ œ œ œ œ
5
œœ
J J R œœ
j j jm.d. j œ
œœ œœ
?
œœ
œ # œœœ œœ
œ
œœ
œ
œœ
œ
‰ œœ ‰ œœ ‰
œ œ œ œ #œ
œ œ œ œ
œ
œ j œœ œ
& œ œ œ œ œJ œ œ œ œ œ œ. œ œ œ œ
9
œœ
j j r
œœ œ œœ ‰ œœ ‰ œœ œœ œœ ! ‰
? œ # œœ œ œ œ œœœ œœ œ
œ œ œ œ œ
œ ! ‰
R
>j
& œœ . # œ œ œ œ >œj # œ j œ
j
13
œ. œœ œ œœ œ œœ ! ‰ œ
œ .
. œœ œ œœ . œ ‰
p # œ œ œ # œ œ .
J R J
? œœ œœœ œ œ œ œœœ œœ œœœ œ œ œ œœœ
2 090
###
& œ˙˙ œ œ œ # œ œ œœœ ...
œœ # œœ œ œ # œ n œœ œ œ œ œ .
œœ
17
œœ œœ ! œœ ..
J #œ œ
R
? œ œ œ œ œ œœœ œ œ œœœ œ œ œ œ œ ###
œ œ œ œ œ œ œ œ #œ œ œ œ œ œ
### œ . œ œ. œ œ. œ œ œ œ œ œ. œ
œ œ. œ œ œ #œ œ œ.
21
&
œœ œœ œœ œ œ œ œœ œ
com sentimento
? ###
œ œ œ œ œ œ œœ œ œ œ œ œ œ œœ
œ œ œ œ
# # # œ œ- œ œ . œ œ œ œ œ
a tempo
nœ œ œ œ œœ œ
rall.
œ œ
25
œœ
& J ! œœ ! œœ œœ ! ‰
R
œ œœ œœ œœ œ œ
? ### œœ œ #
œ œ œ œ œœ œ œ œ ! œ ! r! ‰
œ œ œ œ
# # # œ œ. œ œ œ œ œr œ œ r
œ œ. œ. . œ œ œ. œ. œ. . œ œ g œœ U %
œ œ
œ œ œ. œ. . œ œ œ gg œœ !
œ g j
29
& gR œ
. f p p
œœ œœ œœ œœ œœ œœ œœ œœ œœ œœ œ œœ
? ###
Fim
œ œ œ œ œ œ œ œ œ œ œ œ g
g ! ‰
œ œ g
gg R œ
u
091
%
# 2 œ œ œ œ œ œ œœ œ œ . œ
Fado Marcha, Não Depressa
j nb
& 4 œ œ œ œ œ œ œJ œ . œ ‰ Œ ‰ bœ œ b
J
f jF
œ œ œ œ œ œ œ ‰ Œ
Piano
? # 42 ‰ œ œ œ œ nbb
œ >œ >œ œ œ œ
> > >œ >œ œ > >œ >œ >œ >
>
bb ‰ œ œ œj œ . œ œ œ œ œ œ j
5 3 3
& œ œ œ œ.
Œ ‰ j ˙ Œ
? bb œ œ œ
!
œ œ œ œ œ œ
bb œ œ j
œ œ. ˙ œ œ œ
9 3
& œ œ #œ œ
3
J J
œ j
#œ # n ˙œœ # œœ̇ #œ ‰ Œ
œ
? b
b œ n œ n œœ œ nœ #œ J ‰ Œ
>
bb œ œ ‰ #œ œ b˙ œ œ #œ œ œ œ # ˙˙˙
13
&
3 3
Œ ˙ œ œ
? bb ! œ œ Œ
œ
œ œ œ œ œ
œ œ
gg# ˙˙˙ j
2 091
b j ggg ˙
&b j œ œ
17
œ œ.
œ œ œ œ œ œ
3 3
˙ œ
œ Œ
? bb ˙ œ œ œ œ̇ œ œ œ
#œ œ œ œ œ œ œ œ œ œ ‰ Œ
#œ œ J
bb ‰ n œ œ œj œ œ œ œ œ œ j
21 3 3
& œ œ œ. œ œ.
Œ œ œ ˙ Œ œ œ
? bb ! œ œ œ œ œ œ
bb œ bœ œ œ œ œ
j
œ œ œ
25
& ˙ ˙
3
3
? b n˙ œ̇œ ˙
b nœ œ̇ œœ œ
œ œ œ nœ
nœ œ œ > œ œ nœ œ ‰ Œ
J
> > > >
‰
bb œ˙ œ œ. œ œ œ œ j œ.
29 3 3
& ˙ œ #œ œ J œ œ #œ
? b !
b œ œ œ nœ
œ œ œ œ œ
" œ nœ œ œ œ œ œ j
b œ #
&b # œ œ œ œ œ œ œ n ˙˙˙ nn
33
œœ
œ nœ œ œ
? b n˙ #œ #
b œ nœ œ nn
œ œ œ #œ nœ œ
> > >
j j ˙
# . ‰ œ œ œ ‰ œ œ œ
091 3
& . ˙ œ̇ œ œ œ œ œ œ ˙˙ œ œ œ œ œ œ œ
36
˙˙
? # ..
œ œ œ œ
j œ j
# ‰ œ œ œ #œ nœ œ œ
j
œ œ œ
œœ œ œ œ
40
& ˙˙˙ œ œ œ œ œ œ n ˙˙
˙ ˙ œ - - -
J
?# œ #œ 3
3
œ œ œ œ ‰ Œ
œ œ œ œ œ J
œ œ œ
b˙
rall.
# œœ U #U #
a tempo
˙˙ œ œ nœ bœ œ #œ j n œj
44
& ˙ œ œ œ œ œ œ
œ
œœ j ggg œœ ggg # # œœ
3
?# ‰ œ œ b˙
3
œ œ œœ Œ gg œ
J g gg œ
œ
j j
œ œ œ œ fi
œ œ > > >
# " œ œ "
œ j jœ œ œ .
œ œ œ œ œ œ .
1.
.
48
& œ œ œ ‰ œ
œ œ œ œ œ
œ œ œ
J J œ
n
?# n œœ g
gg œœ œ œ
œ œ œ œ ..
3
œ œ g
g
gg œ œ œ œ
œ g > œ
>
% fi j >œ
# j j œœ œœ ‰ Œ
& œ œ. œ ‰ Œ œ œ
52 2.
œœ œ
> J
j
? # ggg œœ œ œ œ œ œ œ ‰ j j
ggg œ œ œ œœ œ œ œ œ ‰ œ ‰ Œ
J > >œ >œ > > > > >
œ
œ
>
092
œœ œ œ œœœœ œ œ œ œ
œ œœœœ œ
Allº Moderato
2 œ œ œ œ œ
&b 4 ‰
œœ œœ œ œœ œœ f
œœœ
3 3
œ œœ
ligeiro
œ œ œ
Piano
? b 42 œ œ œ œ œ œ #œ œ œ
œ j j
b .
. œ ‰ ‰ œ ‰ œ
5
& œ
œ œ œœ œ œ œœ œ. J
p j
? b .. œ j œ j œ
J ‰ ‰ œ œ
‰ ‰ œ œ œ œ œ
œ œœœœ œ
œœœ œ
& b œ œ œ œ œœœ œ œ œ œœœœ œ œ œ
8
œ. J
3
œ œ œ
3
? œ œ œœ œœ œ
b œ œœ œ œ
œ œ œ œ œ œ œ œ œ œœ œœ œœ œ
œ œ œ œ œ
œ œ œ œ œ nœ
& b œ œ œ œ œ nœ œ œ œ nœ œ #œ œ œ œ nœ œ œ œ œ. œ
14
?b œ œœ œ œ œœ œ œ œ #œ œœ
œ
œœ
œ n œ
# œœœ œœœ œœ Œ
œœ
2 092
&b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ. œ
20
œ œ œ œ œ
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœœ œœœ œœœ œœœ Œ

b œ œ œ œ œ
&b œ œ œ œ œ œ œœœ œ
26
? œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œœœ œ œœ œœ
œ
œœ
œ
b œ œ
œ
&b œ œœœœ œ œ œ œ œ œ œ
30
œ œœœœ œ œ
3
œ
œœ œœ
3
?b œœ œœ œ
œ œ œ œ œ œ œ œ #œ œ œ Œ
œ
œ œ œœœ œ œ œ
& b œœ œœœ œ œœ œœœ œ œœ œœ œ œ œœœ œ œ œœœœ œ œ œœ œ
34 3
? b œ œœ œœ œ œœ œœ œ œœ œœ œ œœ œœ œ œœ œœ œœœ Œ
œ œ œ œ œ
œ
092 3
œ œ œ œ œ œ
& b œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ œœœœ œ œ œ
40
3
? œ œœ œœ œ œœ œ œœ œœ œœ œ œœ œœ œ œœ œœ œœœ Œ
b œ œ œ œ œ
œœœœ œœœ œœœœ œœœ œœœœ œœœ œœœœ œœœ

&b œ œ œ œ
46
? œ œœ œœ œ œœ œœ œ œœ œœ œ œœ œœ
b œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œ œ œœ œ œ œ œ œœœ œ
œ œ œ œ œ œ
&b
50
œœ œœ
?b œ œœ œ
œœ œ
œœ
œ
œœ œ œœ œ
œœ œ œ
œ œ
œ œœœœ œ œ œ œ œ œœœ œ œœ
œ œœœœ œ œ
&b œ œ œ œ .. J ‰ Œ
54
ƒ
œ œœœ œœ œœ
3 3
j
? œ œœ œ œ œ œ œ # œ œ œ .. œJ ‰ Œ œ ‰ Œ
b œ
093
bœ œ œ œ œ œ œ œ œ
r r
œn œ œ œ œ nœ œ
r r
2 œ œ œ œ œ œ
r
œ
&b 4
œ
f
œ œ œ œ
œ ! œœ ! œ ! œœ ! œ ! œœ ! œ ! œœ !
Piano
? b 42 ! ‰
œ œ
œ œ
r
œ œ œr œ œ œ ‰
3
œ
œ œ
&b œ ! ! œ œ œ œ nœ œ
3
œ œ œ ˙
S p
œ œ œœ œœ œœ œœ œœ œœ
œ œ g
? b œ ! œ ! œ ! œ ! œ ! ggg œœ ggg œœ ggg œœ œ
! gg ! gg œ œ œ ! œ ! œ ! ! œ
œ œ g
3
& b .. œJ œ. œ !
6
œ œ œ œ nœ œ
p
? b .. œ œœœ œœœ œ œ œ œ œœœ
! ! œ œ ! œœ ! œ ! !
œ œ œ œ
f
! r ! ‰ ‰
œ
r
œœ œ
&b œ œ œ œ œ œ œ #œ œ
8
œ
œ ˙
. . p
œœœ œœ œœ
? œ ! ! œ œ œ ! œ ! œ ! !
b œ œ. œ. œ
f
2 093
&b œ œ œ œ ! œ œ.
10
R œ œ. œ œ
? œ # œœœ œœœ ! œœœ # œœœ œœœ

b ! ! ! œ ! ! œ ! !
. ‰
&b œ œ . œ œ œ œ nœ œ
12
J œ œ œ œœ˙
S p
œ
? œ ! œœ ! œ œ œœœ. œœœ œœœ
b œ œœœ ! ! ! !
œ. œ.
œ . . ‰
&b J œ. œ! œ œ œ œ #œ œ
14
œ œ œ œ n œ œ œœ œ œ œ ˙
. œœ œœ
? œ ! œœœ ! œœœ œ œ. œ. œ ! œœœ œ œœ
! œ ! œœ ! œ ! œ ! œ œ œ
œ ! ! œ ! !
b œ œœœ œ œ œ
œ œ
! r
&b œœœ œ ! œ œ. œ œ œ. œ œ
18
R œ œ.
œœ œœ
? œ ! # œœœ ! œœœ ! œœœ ! œ ! # œœœ ! ! œœœ ! œ ! œ ! œ ! œ.
b œ
œ
093 3
& b œ. œ œ œ œ œœœ œ ! œr œ . œ œ n##

21
> œ œ #œ œ œ œ #œ
>
? œ ! œœ ! œœ ! œœ ! œ ! œ ! œ ! œœ ! œ ! # œœœ ! ! œœœ ! œ ! # œœ ! œœ ! ‰ #
b œ œ R n#
œ
## ! . œ œ œ . œ œ œ œœ . œ œ œ œœ œ œ
r ! ‰
œœ œ # œ œ
œ œ œ
25
& œ œ œ
œœ œœ œ œ œ
? ## œ ! œ ! œ ! œ ! œ ! œœ ! œœ ! œœ !
œ
r
# # !œ . œ œ œ . œ œ œ œœ œ œ œ œ œ œ œ ! ‰
3
œ œ œœ œ œ #œ œ
27
& œ œ œ œ
œœœ œœœ œœ œœ œœ
? ## œ ! ! ! ! œ ! œ ! œ ! œ !
œ
r
# # n !œ . œ œ œ. œœœ œ
3
œ œ œ œ #œ œ œRÔ ! # œ œ œ œ œ œ
œ #œ œ œ n œ œ
29
&
œœœ # œœœ œœ U
? ## # œ ! ! ! œ ! œœ ! ‰ Œ
œ R
. . . .
4 093
r
œœ œœ œœ œœ œœ œ œ
# # !œ œœ œœ œœ ! œ œ ! ‰
œ #œ œ
31
& œ œ œ œ œ œ œ œ
œ
œœ œœ œœ f
? ## ! œ ! ! œœœ ! œ ! œ ! œR ! ‰
œ œ
## n œ œ œ œr ! œ œ œ œ
œ #œ œ œ nœ œ #œ œ
33
&
œœ # œœœ œœœ
? ## # œ ! œ ! œ ! œ ! œ ! ‰ Œ
R
## œ nnb
œ œ
35 1.
& œ œ œ œ #œ œœ œœ
j
nœ
f
œœ œœœ
? ## ! œ ! ! ! œœ œ nœ nn
œ œ œ b
nœ
‰ r
& b œ œœ˙ œ œ œ œ n œ œ .. n # # ! œœœ ! œœœ ! ‰
37 2.
œ > >
.
œ œ œ œœ > r
œ œ œ
? œ œ œ œ ! œ ! œ ! ! œ .. # # œœ ! œ ! œ ! ‰
b n
œ
094
### 2 œ œ œ œ œœœœœ œ œœ
œœ œ œœœ œœœ ®œ œ œ
& 4
F
œœ œœ œœ œœœ œœœ œœœ
? # # # 42 ! ®
Piano
œ œ œ œ œ
### œ œ œ
œ ®œ œ œ œ œ œ
œ œ œ œ. œ
3
& œ. œ œ.
? ### œ œ œ œ œ œ
œ œ
### œ œ.
œ œ œ. œ œ.
5
& œ. œ œ. œ œ. œ œ.
œ
? ### œ œ œ œ œ œ œ
œ
7
###
& ˙ œ ! œ. œ œ. œ œ. œ
? ### œ œ œ œ œ œ œ
œ
###
œ œ. œ
9 3
& œœœœ œ. œ œ. œ œ.
? ### œ œ œ œ œ œ
œ œ
### œ . nnn
2 094
œ œ œ œ
œ œ. œ œ. œ œ. œ œ œ
11
& œ
œœœ œœœ œœœ œœ œœ œœ
? ### œ nnn œ œ œ
œ
mosso
&œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ œ œ #œ
13
œ œ
?œ
œœ
œ
œœ
œ
œœ
œ œ # œœœ œœœ œœœ
œ œ œ œ œ œ ® œ œ #œ
œ œ œ
15
& œ #œ œ œ œ œ œ
œ œœ œœ œœ œœ œœ
# œœ
menos
? œ œ œ œ œ
œ œ
%
œ # œ œ œJ œ œ .
& œJ œ œ! œ #œ œ. œ
17
œ œ œ œœœœ
3
R œ
? œ œœœ œœœ œœœ œ # # œœ
œ
œœ
œ
œœ
œ
œ œ
j U
rit.
œœœœ j
a tempo
&œ œ ! œœ ! œ œ ! œ œ ® œ #œ
21
œ ! œ
œ œ œ #œ œ # œ œ œ
œœ œœ
? œ œœœ œœœ œœœ œ œ œ # œ œœ œœ œœ
œ œ œ œ # œœ œœ œœ
œ
094 3
œ #œ œ œ œ œ œ œ œ œ œ #œ œ œ œ œ œ ! r
a tempo
& œJ œ j
25 apress.
œ œ œ œ œ
œœ œœ œœ # œœœ œœœ œœœ # œœœ œœœ œœœ œœ œœ œœ
? œ œ œ œ
œ œ œ œ œ œ
più vivo
œ U ##
&œ œ œœ œ œ œ œ œ ®œ œ œ #
29
œ ! œ œ œ œ œ # œ
œ œ œ œ œ œ œœ œœ œœ œ œœœ œœœ
? œ œœ œœ œœ œ # œœ œœ œœ œ œ œ œ œ # œœ ###
### œœœ œœœ œ

œ œ ! œ œ œ œ œ œ œ œ #œ œ
33
& œ œ œ œ œ œ œ œ œ œ nœ
menos
? ### œœ œœ œœ œ œ œ œ œ œ
### . œ œ #œ œ œ #œ œ. œ œ œ
rall. a tempo
. œ n œ œ œ œ œ # œ
37
& nœ œ œ œ œ œ œœ
? # # # .. n œ n œœ œœ œœ # œœ œœ œœ
vivo
œ œ œ œ œ œ
### j U %
œ œ œ œ œ n œ .. œ œ œ ® œ
41 1. 2.
& œ œ
Fim
? ### œ œ œ .. œ œ œ
œ œ J J
095
Moderato
& 42 ! œ œ œ œ œ œ œ .. œ œ # œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ # œ œ œ œ œ œ œ ‰ œj
œ
œœ œœ œœ
œ # œ œ # œ œœœ
Piano
? 42 œ œ .
.œ œ œ œ œ œ œ œ
œ œ œ
&œ œ œ œ œ œ œ œ œ œ œ œ œ j
5
œ œ œ
3
J J œ
œœœ œœœ œœœ œœ œœ œœ œœœ œœœ œœœ œœ œœ œœ
3
œ œ œ œ
&œ œ œ œ œ j œ œ œ œ œ
9 3 3
œ. œ œ œ œ
œ œ œœ œ œœ œœ œœ œœ œœ # œœœ œ
? œ œœ œœ œ œ #œ œ œ œ œ œ œ
&œ œ œ œ œ œ œ œ œ œ œ œ œ j
13
œ œ œ
3
J J œ
œœœ œœœ œœœ œœ œœ œœ œœœ œœœ œœœ œœ œœ œœ
3
œ œ œ œ
j #œ œ œ œ œ œ . ###
& œ œ œ œ œ œ. œ œ œ œ œ œ œ .
17 3 1.
œ
œ # œœ œœ œœ
3
? œ œœ œœ œœ # œœ œœ œœ œ œ œ .. # # #
œ
###
2 095
œ ‰ j j j j
21 2.
&
œ œ œ œ œ œ œ œœ
œ œ œ œ œ œ œ œœ œœ œœ œœ œœ .. œœ
3 3
? ###
œœœœœœœœ œœœœœœœœ œœœœœœœœ œœœœœœœœ œœœœœœœœ
### œ œœ œ œj œ
j
œ œ #œ œ œ
3
œœ œœ œ œ œ œœ j
26
& œ œ œ œ œ œ œœ .. œ
œœœœœœœ œœœœœœœœ œœœ

3
? ### œ œ œœœ œœœ

œ œ œœ œ œ
### j j
j
30
& œ œœ œœ œ œ œ œœ
œ œ œ œ œœ œœ œœ œœ œœ .. œœ
3 3
? ###
œœœœœœœœ œœœœœœœœ œœœœœœœœ œœœœœœœœ
### œ j j
œ œ œ œ œ œ œ #œ œ œ nœ œ œ œ nœ
3
œœ œœ œ œ œ œœ
34
& œ œ œ œ œ œ œœ œ œ
œœœ œœœœœœœ œœœ œ

3
? ### œ œ œ œ œ œ œ œ
œ œ œ œ œ œ
###
œ œœ Œ
38
& œ œ œ œ œ nœ œ œ
? ### œ œ œ nœ œ Œ
œ œ
096
## 2 œ œ œ œ œ% œ œ
& 4 ® ® œ œ ® œœœ ® œ œ œ. œ
p
Piano
œœ œœ
? # # 42 ® ‰ œ ! œ ! ! œ !
œ
œ
œ
2
# # œ ® œ œœ ® œœ œ ® œ œ ® œ œ œ ® œ œ œ ® œ œ œ ® œ œ œ # œ œ
&
œœ œœ œœ œœ
? # # œ ! œœ ! œœ ! œ ! œ !
3
R ‰ œ ! œ !
œ
# # n œœœ œœœ œœœ œœ œ œ œœ

® ® ® ® œ œ œ ® œœœ œ œ
4
&
œœ œœ œœ œœ
œ ! œR !
3
? ## œ ! ‰ œ ! œ ! œ !
œ !
œ œ
## œ œœœ œœœ œ œ œ œ œ
œœ œ œ œ r nn
! œ œ œœ ! œœœ ! œœ œœ ! ‰ b
6
& œ œ
œœ œœ œœ !‰
œ œ
Fim
? ## ! œ ! œ ! ‰ ! œ ! œœ ! nnb
œ R œ ! œ
œ œ r
œ œ
R
j !
2 096
œ
& b œœœ œœœœ #œ œ œ
3
œœ œœ œ œ œ œœ œœœ œœœ
9 3
œœ
p
? b œ ! œœ ! ! œœ ! œ ! œœ ! ! œœ ! œ ! œœ ! ! œœ ! œ ! œœ ! ! œœ !
œ œ œ œ
œ j !
& b œœœ œ œ œ . œ œœ #œ œ œ
3
œœ œ œ œ œœ œœœ œœœ
13 3
œœ
? b œ ! œœ ! ! œœ ! œ ! œœ ! ! œœ ! œ ! œœ ! ! œœ ! œ ! œœ ! ! œœ !
œ œ œ œ
œ j
#œ œ # ˙˙
3
& b n œœ œ œ œ œ # œœ œœ œ œ œ œœ
17 3
? n œœ œ ! œœ œœ
# œ !
œœ
œœ ! ! œœ ! œ ! # œœ ! œœ ! œœ !
b œ ! ! ! œ ! ! ! œ ! œ
> > j3
œ # œœ œ ® œ n œ œ ®œ œ
œ œœ œ . # œ œœ
œ
&b n œœ œ œ œ œ # œœ œ œœ œœ
21
œ #œ
R!
n œœ œ ! œœ œœ œœ # œœœ œœœ œœ œœ
? #œ ! ! œ ! ! œ ! ! œ ! œ !#œ ! œ !
D.S. ao fim
b œ ! ! ! œ ! œ œ œ R ‰
œ
097
# # 2 œ œ . œœ œ œ œœ n œœ œœ œœ
Allegretto
& 4 œ .œ œœ œ œ œœœ œ œ œ œ
œ œ œ œ œ œœ œœ œ œœ
œœ œ œ œœ œ
Piano
? # # 42 ‰ ! .. œ œ œ œ #œ œ œ
œ
œ #œ œ œ
3 3
3 3
# # œœ œ œ œ œ
œ #œ œ œ œœ œ œœ œ œ œ œ œ œ œ
3
& œ œœ œœ œ œ #œ
œ
œœ œœ œœ œœ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ
œ œ œ œ œ œ œ œ œ #œ
œ
## œœ œœ œœ >œ
œœ # œœœ œœ œœ # œœ œ ‰
5
& œ œ œ œ œ œ J
> > > > > > >
>œ
? ## œ œ œ œ œ œ œ œ œ ! œ
> > > œ œ > >
> >
2 097
## ! ! ! ! ‰ œ œ œ œ #œ œ
œœ œœ
7 3 3
& œ œ
œ œ
? ## œ ! œ œ ! œ œ œ œ œœœœ œœ
œ œ œ œ œœ
## œ œ Œ ‰ œ
œ œ œ #œ œ
9
&
œ œ œ œ
? ## œœ œœ œ œ œ œœ œœ œ
œ œ œ œ œ œ œ œ œœ œ
3 3
# œ
3
& # œ ! ® œ® œ ! œ œœœœœ œœ
11
œ œ
œœ nœ #œ œ œ œ œ œ œœ
? ## œ œ œ r
œ œ œ œ. ! œ
## ® ® !
œ œ œ œœœœœ œœ
13
& œ œ œ œ œ
œ œ œ œ œ œ œ œ œ œ œ œœ
? ## œ œ œ œ œ. ! œr
œ œ
097 3
##
® œ œ. ® œ œ. ‰ œ œ œ œ #œ œ
15
& œ œ
œ œ. œ œ.
? ## œ œ œ œ œ œ œ. ! œr
## œ ® ! œ œ œ œ œ œ œ œ
17
& œ œ œ. œ œ œ. œ
œ œ. œ œ. œ œ œ œ œ
? ## œ œ œ ! r
œ œ œ œ. œ
##
! ! ! œ
19
&
bœ œ nœ œ œ nœ œ œ œ œ œ œ œ œ œ
œ œ œ n # œœ œœ œœ
œ œ œ
? ##
bœ. œ nœ. œ œ œ
## U j
œ œ œ ® œ
œœ ‰
21
& œ œ œ œ œ œ œ œ œ œœœœœ œ
œ œ
? ## œ U
œ œ œ bœ nœ nœ ® œ œ œ j ‰
œ
4 097
# œ œ œ #œ œ j
& # œ. nœ œ.
23
œ œ œ œ œ œ
? ## œ œ œ œ œ œ œ œ œ œ œ œ
œ #œ
œ
œ
## œ #œ j
œ.
25
& œ œ œ œ œ.
? ## œ # œ œ œ œ œ œ œ œ œ œ
œ
## œ œ œ œ œ. œ œ ! œ
27
& œ. œ
œ œ œ œ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ
œ œ
# œ.
& # œ œ œ œ
29
œ œ
? ## œ œ œ œ œ œ #œ œ œ. œ œ œ
œ œ œ #œ œ œ
# œ œ.
097 5
& # œ. œ nœ œ #œ œ J
31
œ œ
? ## # œ œ œ #œ œ œ œ œ œ œ œ œ
œ. œ œ œ œ œ
## œ œ.
œ. œ œ œ #œ œ
33
& J
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ
35
## œ . œ œ œ œ œ
& œ œ J œ J
œ œ œ œ
? ## œ œ œ œ œ œ œ œ œ
## œ œ œ œ
#œ œ. ‰
37
&
œ œ œ œ
? ## œ #œ
œ œ œ œ œ œ œ ! œ
œ œ œ
##
639 097
! ! ! œ œ œ œ œ œ œ œ
3 3
& œ œœ œ œœ œ œ œ œ œ œ
œ œ œ œ
? ## œ œ œ œ œ ! r
œ œ. œ
#
& # ! ! ! œ œ œ œ œ œ œ œ
41 3 3
? ## œ œ œ œ œ œ ! r
œ. œ
##
! ! ! œ œ œ œ n # œœ œ œ
43
& nœ œ œ
bœ œ œ œ nœ œ œ œ œ œ œ œ > >œ >œ
? ## œ nœ. œ œ œ
bœ. œ œ œ
## U
œ œ œ œ œ ® ! œ œ œ ..
45
& œ œ œ œ œ œ œ
œ œ œ
f
? ## œ U
œ œ œ bœ nœ nœ ® œ œ œ Œ ..
œœ n œœ œ œ
097 7
# # œœ œœ œ œ œ œœ
& œ œ œ œ œ œœ œœ
47
œ œ œ œ œ œ œœ œœ œ œ œœ
? ## œ œ œ œœ œœ
œ œ œ #œ œ œ
œ #œ œ œ
3 3
3 3
# # œœ œ #œ œ œ œ œ œ œ œ œ
œœ œœ œœ œ œ œ œ œ œ œ
49
& œ œœ œ œ #œ
œœ œœ œœ œœ
? ## œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ #œ
œ œ œ œ œ œ œ œ #œ
## œœ œœ œœ >œ
œœ # œœœ œœœ œœœ # œœœ œ ‰
51
& œ œ œ J
>
? ## œ œ œ œ œ œ j
œ œ œ œ œ œ œ ‰
œ œ œ œ œ > œ
œ >
098
^
2 œœ. œœ. œ^ œ œœ œœ. œœ œœ. j
& 4 œ #œ œ œ œ œ. œ œ
œœ # œ œ œ # œœœ. œ œ œ œ œœœ œœ ‰ Œ
J œ
F œ œœ
? 42 j œ # œ
œœ œœ # œœ œœ œœ œœ
Piano
œ œ œ œ œ œ
œ œ œ œ
%
‰ œj ‰ j Œ
5
& œ œ
œ. œ œ. œ œ œ œ œ œ œ œ
p
? œœ œœ # œœ œœ # œœ œœ œœ œœ
œ œ œ #œ œ
œ œ œ
‰ œj
9
& œ œ. œ œ. œ œ
œ. œ œ. œ œ œ œ œ ˙
p
? œœ œœ #œ œ # œœ œœ ˙œ̇
œ œ œ œ œ œ #œ œ œ œ œ
^ ^ rall.
U
& œ. œ œ. œ œ œ œ. œ œ. œ œ œ
13
œ. œ œ. œ œ œ œ œœ
#œ
œœ œœœ œœœ œœœ œœ
? œ œœœ œœ
œ œ œœ
œ œ œ #œ œ œ œ
œ
#
! r ##
2 a tempo 098
‰ œj
17
& œ œ. œ œ. œ œ œ
œ. œ œ. œ œ œ œ œ. œ
œœ œ # œœ œœ # œœ œœ ‰
? œ œ œ œœœ .. ###
œ œ #œ œ œ œ ‰
J
### .
. œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ œ. ! r
21
& œ
? # # # .. œ œ œ œ œ œ œ œ
œ œ œ œ œ œ #œ œ
###
œœœ œ œ œ œœœ œ œ œ œ œ ! œr
25
& œœ œ œ œ.
œœ œ œ œ œ œœ œ
? ### œ œœ œœ œœ œœ œ œœ
œ œ nœ œ œ œ œ œ
œ
29
### œ œ œ œ œ œ œ œ œ œ œ œ œ œ ! œr
& œ œ œ œ œ.
? ### œ œ œ œ œ œ œ
œ œ œ œ œ œ #œ œ œ
%
### ^ ^j
œ œ œ œ œ œ œ œ œ œ œ œ œ. r .
œœ œœœ œ ! œ . œ œ œœ ‰
33 1. 2.
& œœ œ
œœ œ œœ œ # œœœ n œœ œœ .. ‰ ^ Fim
? ### œ n œœ œ œ œœ œ œ œ . œ ^j
œ œ œ œ œ œ œ œ . œ œ ‰
099
Allegretto Agitato
b
& b b 42 ! œœ ! œœ ! œœ ! œœ ! œœ ! œœ ! œœ ! œœ
œ œ nœ œ œ œ nœ œ
Piano
? b b 42 œ ! œ ! œ ! œ ! ! œ ! œ ! œ !
b œ
b œ œ nœ
& b b .. œ . œ œ œ œ œ œ œ œ œœ œ œ œ œ̇
3
œ œœ œ œ œ
J
œ œ œ œ œ œœ n œœ œœ
? b b .. œ œ œ œ œ n œœœ œœœ œ œ œ œ œ œœ œ œœ œ œœ œ œœ œ n œœ œ œœ œ œœ œ nœ œ œ
b œ œ œ
b
& b b œ œ œ œ œ œ œ œ ˙! œœ ! œœ ! œœ ! œœ ˙! œœ ! œœ ! œ ! œ
7
nœ œ
? b œ œœ œ œœ œ n œœœ œœœ œ ! œ ! œ ! œ ! ! œ ! œ ! œ !
bb œ œ
b œ œ nœ
& b b œ. œ œ œ œ œ œ œ œ œ œœ œ
10
œ̇ œ œœ œ œ œ
J
œ œ œ œ œ œœ n œœ œœ
? b b œ œ œ œ œ n œœœ œœœ œ œ œ œ œ œœ œ œœ œ œœ œ œœ œ n œœ œ œœ œ œœ œ nœ œ œ
b œ œ œ
2 099
b n
& b b œ œ œ œ œ œ œ œ ˙ œ œ œœ œœ ˙ œ œ œ œ n n
14
! œ ! œ ! ! ! œ ! œ ! nœ ! œ
œ œ
? b b œ œ œ œ œ n œœœ œœœ œ ! œ ! œ ! œ ! ! œ ! œ ! œ ! n
b œ œ nn
! œ œ ®œ œ ®œ œ ! œ œ ®œ œ ®œ ˙ !œ œ œ œ
17
&œ œ œ œ
!œ!œ!œ!œœ œ œ
? œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ ! œ ! œ ! œ ! œ œœ œœœ œ œœœ œœœ
œ œ
! œ œ
21
&œ œ œ œ œ œ œ œ œ œ œ œ
? œ œœ œ œœ œ œœœ œœœ œ œœœ œœ œœœ œœœ

œ œ œ
œœ
œ
œœ œœœ œœœ
œ œ
œ œ œ œ œ
! œ œ ®œ œ ®œ œ ! œ œ ®œ œ ®œ ˙ !œ œ œ œ
24
& œ œ œ œ
!œ!œ!œ!œœ œ œ
? œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ œœœ œ ! œ ! œ ! œ ! œ œœ œœœ œ œœœ œœœ
œ œ
099 3
! œ œ œ œ bbb
28
&œ œ œ œ œ œ œ œ œ œ
? œ œœ œ œœ œ œœœ œœœ œ œœœ œœœ œ œœœ œœœ œ œœœ œœœ œ œœœ œœœ b
œ œ œ œ œ bb
bb ! ! ! ! ! ! ! . ! n œ œ œ œ œ ! nœ œ œ œ
& b œ .
31 1. 2.
œœœ œœœ œœ œœ œœœ œœœ œœ œœ œ œ œ

nœ œ nœ œ
? b œ ! œ ! œ ! œ ! ! œ ! œ ! œ ! .. œ œœ œ œœ œ n œœœ œœœ œ œœ œ œœ œ n œœœ œœœ
bb œ œ œ
b œ œ ! œ ! œ ! œ j
b
& b n œœ œœ ‰ n œœœ ‰ j‰ Œ
35
œœ œœ œœ nœ J ˙ œ
œ j‰ Œ
? bb œ ! n œ ! œ ! n œ ! œ
j
‰ œ ‰ n ˙˙ ˙ ˙
œ
n œœ
b nœ œ nœ J
œ ˙ ˙ œ ‰ Œ
J
100
^ ^ ^
Tempo de Fado Marcha
& 42 œ œ œ œ œ œ œ. œ œ œ œ œ œ œœœœ œ œ œ
3 3
˙
f
œœ œœ œœ œœ b œœ n œœ œœ œœ
3 3
œ
Piano
? 42 Œ ‰ œ œ #œ œ bœ
œ œ
& œœœœ œœœœ ‰ j ‰ b œj ‰ œj b b b

5 3 3
œ œ œ œ œ œ b œœœ œœ
œ œœ œœ
? œ œœœ œœœ œœœ œœœ œœœ œœœ j

œ ‰ œj j ‰ œj ‰
œ œ œ ‰
bœ bbb
%
b
&bb
9
œ œ œ œ œ œ œ ˙ œ œ œ œ œ œ #œ ˙
p
? b b œ œœ œœ
œ
œœ œœ œœ
œ
œœ œœ œ œ œ
nœ œ œ œ œ
b œ œ œ
b
&bb œ œ œ œ œ œ œ ˙
13
œ œ œ œ œ œ œ ˙
? b œœ œ œœ œœ œœ œœ œ œœ œœ œ œœ œœ œœ œ
b b nœ œ nœ œ
bbb œ œ œ œ œ.
17
& œ œ nœ œ œ œ
3
œ œ œ œ œ œ œ ˙
? b œ
bb œ œ œœ œœ œœ œ n œœ œœ œ
nœ œ œ
2 100
b nnn
& b b œ œ œ œ œ œ œ œ œ œ. j n œj
21 3 3 3
œ œ œ œ œ œ œ œ œ.
3 3 3
nœ œ œ
‰ j ‰ j jF‰ Œ
? bb œœ œœ œœ œœ œœ œœ œ n œœ œœ n œœ
b œ œ œ œ n œ œ nœ nnn
œ œ œ
J
& .. œ œ œ œ . œ œ œ œ.
26 3 3
3 3 3 3
? .. œ œœ œœ œœ œœ œœ œœ œœ œ
œ œ œ œ œ
œ bœ œ œ #œ
j
& œ œ œ œ. œ œ œ
30
œ œ œ œ œ œ ˙ J
3 3
œœœ œœœ œœœ œœœ œœœ œœ œœ œœ

3
? œ œ œ œ œ œ
œ œ œ j ‰ Œ
œ
œ
& œ œ œ. œ œ œ œ œ œ œœœœ œ œ- œ-
34 3 3
œœ œœ œœ œœ # œ b œœ n œœ
3 3
? œ œ Œ !
œ œ
j %
j œj .. j ‰ j‰ œ ‰ Œ
& œœœœ œœœœ
38 1. 2.
œœœœ œœœ
3 3
œœ œœ ˙ œ œ œ
‰ j j jf‰ Œ j
? œ œœœ œœœ œœœ œœœ œœœ œœœ œ œœ ‰ œœ œœ .. œj ‰ œ ‰ œj ‰ Œ
œ œ œ œœ œ œ
œ œ
J
General Statistics on Fado
Fado Edition Date Corpus Tonality Quality Tonality B Tonality Final Quality final Form
1 1840 0 Dm minor Dm Dm minor AA'BB'

2 1845 0F Major F F Major I1I2 AA B CC' D O
3 1845 0 Dm minor Dm Dm minor AABBCCDD
4 1848 0 Am minor Am Am minor I ABAB
5 1864 0 Am minor Am Am minor I AABB'
6 1890 0F Major F F Major I AABC AABC
7 1895 0 Dm minor F Bbm minor I AA'BB' CCDD
8 1868 0 Gm minor Gm Gm minor I(AB) ABAB CD O(AB)
9 1860 0 Fm minor Fm Fm minor I(Ab1c2) AAB CC O(Ab1c2)
10 1885 0 Bb Major Bb Bb Major I(AA) AA BB'B' O(AA)'
11 1852 0D Major D D Major I(AA) AAA BB'
12 1890 0G Major G G Major AABBCCDD
13 1887 0 Eb Major Bb Eb Major I AA BC
14 1892 0 Dm minor F F Major I AA BC
15 1870 0E Major E E Major ABC ABDD
16 1894 0D Major D D Major I AABB AABB
17 1895 0B Major B B Major I AABB'
18 1852 0 Em minor Em Em minor I AA'BB'CC'
19 1875 0 Eb Major Eb Eb Major I AABB
20 1895 0D Major D D Major ABB ABB
21 1895 0D Major D D Major AABB
22 1870 0C Major C C Major AABB
23 1850 0 Am minor Am Am minor I(AA) AABBCC
24 1850 0 Gm minor Gm Gm minor I(AA') AA'AA' BB
25 1894 0G Major G G Major I AABB AABB
26 1849 0 Dm minor Dm Dm minor ABAB
27 1870 0G Major Em Em minor I (a1a2a1b1) (c1c2d1c2)
28 1875 0G Major G G Major I ABCD ABCD
29 1870 0 Gm minor Gm Gm minor (a1a2b1b2) (c1c1'c2c2'c2''c2'''c2''b2) (d1d1'd1b2) (d2b2d2b2)
30 1898 0 Dm minor Dm Dm minor AA'BB'CD AA'BB'CD
31 1870 0D Major D D Major I AA'BB'B' AA'BB'B'
32 1871 0G Major G G Major ABA'B' ABA'B'
33 1879 0C Major Cm Cm minor I AABB X AABB
34 1898 0 Bb Major Bb Bb Major AABB AABB
35 1872 0 Em minor Em Em minor I(AA) AABCCC
36 1898 0 Cm minor Cm Cm minor AABB
37 1872 0 Am minor Am Am minor ABCC
38 1898 0G Major D G Major AABB AABB
39 1898 0 Am minor Am Am minor AABB
40 1898 0 Eb Major Bb Bb Major I AABB AABB
41 1898 0F Major Dm F Major AABA'
42 1850 0C Major C C Major ABB A'B'B''
43 1896 0C Major C C Major I AABB'
44 1866 0 Eb Major Eb Eb Major AABB
45 1850 0 Ab Major Fm Fm minor I(A'C') ABCC ABCC
46 1880 0G Major G G Major AABB
47 1898 0C Major Am Am minor AABB AABB
48 1858 0 Gm minor Gm Gm minor I(D) ABCC D(a1b2)C O
49 1898 0 Gm minor Gm Gm minor I AABB
50 1918 1 Am minor A A Major I ABA'B' CD X ABA'B' CD X ABA'B'CD
51 1927 1 Ab Major Ab Ab Major I AABC X AABC
52 1927 1F Major F F Major I AABC X AABC O
53 1940 1C Major C C Major I ABAB ABAB
54 1927 1G Major G G Major I AABB' O
55 1953 1 Dm minor Dm Dm minor I ABCD ABCD
56 1920 1 Am minor A A Major I AABC (d1d2d3) O
57 1925 1 Am minor A A Major I ABAC DEDE O
58 1960 1 Fm minor F F Major I ABAC DEDF ABAC DEDF
59 1900 1 Dm minor D Dm minor AA'BCDE AA'BCDE AA'
60 1946 1F Major Dm Dm minor I ABB' CC'DD' CD ABB' CC'DD' O
61 1947 1C Major C C Major I ABA'C DEDE ABA'C DEDE O
62 1920 1 Dm minor Dm Dm minor I ABCD Z (e1e2e3) O
63 1914 1 Cm minor Cm Cm minor I AB Z CDCD' EFEG I AB Z CDCD' EFEG O
64 1937 1 Dm minor Dm D Major I ABCDE FGH IJ FGHIJ ABCDE FGHIJ O
65 1916 1E Major Em E Major ABAC DEFG HIJK ABAC
66 1940 1 Bb Major Gm Gm minor I AABB' CDE AABB' CDE'
67 1927 1G Major Em G Major AAB CD AAB
68 1925 1E Major C E Major I ABA'C DE A'C O
69 1934 1F Major F F Major AA' BCBD EFE'F' EFE'F' AA'
70 1925 1D Major D D Major I ABCD EF ABCD EF O
71 1936 1D Major Bm C#m minor AABB' AABB'' (t2) AABB
72 1918 1 Gm minor G Gm minor I AA'BB' CDCD' O
73 1909 1 Am minor F#m Am minor ABCC'DD' ABCC'DD' AB
74 1916 1 Am minor A A Major I ABCDD EFF'E ABCDD EFF'E
75 1920 1A Major E A Major I AABB (c1c1c2c2) D O
76 1915 1C Major C C Major AABC AABC
77 1910 1 Gm minor G Gm minor I AB CDCE AB O
78 1934 1G Major Am G Major I ABCD ABCD
79 1924 1A Major Am Am minor I ABA' Z CDCD Z EE' I
80 1920 1D Major D D Major I AABB O
81 1942 1 Dm minor D D Major I AABB'C I AABB'C
82 1933 1G Major Em G Major I ABAB CD AB B'
83 1905 1 Fm minor Fm Fm minor I AA'BC
84 1950 1 Dm minor D D Major I ABAC DEDE ABAC DEDE
85 1946 1 Cm minor Eb F Major I ABAC DE ABAC DE O
86 1941 1D Major Dm D Major I AABC DEDE Z AABC DEDE O
87 1943 1 Gm minor G G Major I ABA'C DEDE ABA'C DEDE O
88 1930 1 Dm minor D Dm minor I ABCDE I ABCDE O
89 1900 1 Am minor C C Major AABB' CCDD' AABB' CCDD'
90 1915 1 Am minor C A Major I ABCDEF O ABCDEF O
91 1967 1 Gm minor Gm Gm minor I ABA'C DEDE ABA'C DEDE O
92 1905 1F Major Am F Major I AZBZCZCZ DZEZFFZ AZBZCZCZ DZEZFFZ O
93 1927 1F Major Dm D Major I ABAB CDEE' ABAB CDEE' O
94 1927 1 Am minor Am Am minor IZ ABCB' DEFF' BCB' DEFF' O
95 1911 1C Major Am A Major I ABAB' I ABAB' CDCD O
96 1920 1 Dm minor D D Major I AABB' O
97 1920 1D Major Bm D Major I ABA'C DEFG A'C I ABA'C DEFG A'C O
98 1945 1A Major C A Major I ABCB DEDF DEDF ABCB DEDF DEDF
99 1911 1 Cm minor C C Major I ABAB CDCD Z ABAB CDCD O
100 1945 1 Cm minor C C Major I ABCDCD ABCDCD
1
Huron
Fado Edition Simplified Form Prog01 Prog02 Prog03 Prog04 Ostinati
Contour
1 AABB i|V7|V7|i i|V7|V7|i convex corridinho

2 AAB CCD I|V7|V7|I ii|V7|I-V7|I I|V7|V7|I I|V7|V7|I convex corridinho
3 AABB CCDD i|V7|V7|i i|iv|V7|i i|V7|V7|i i|V7|V7|i convex alberti
4 ABAB i|V7|V7|i i|V7|i-V7|i convex hybrid
5 AABB i|V7|V7|i i|V7|V7|i desc alberti
6 AABC I|V7|V7|I i|V|ii|V ii-ii0|I|V7|I convex alberti
7 AABB CCDD i|iv|V7|i #iv07|I|ii-V7|I i|V7|V7|i i|vii07|V7|i convex corridinho
8 ABAB CD i|V7|V7|i i|V7|i-V7|i i|V7|V7|i i|ii|i-V7|i desc alberti
9 (Ab1c2) AAB CC (Ab1c2) i|V7|V7|i i|V7|i-V7|i i|V7|i-V7|i convex alberti
10 AABBB I|ii|V7|I I|V7|V7|I convex arpeggio
11 AAABB I|V7|V7|I I|IV|V7|I convex hybrid
12 AABB CCDD I|V7|V7|I V7/ii|ii|V7|I I|V7|V7|I V7/ii|ii|V7|I convex alberti
13 AABC I|V7|V7|I V7/ii|ii|V7|I V7|I|V7|I convex corridinho
14 AABC I|V7|V7|I I|V7|V7|I I|vi-ii|I-V7|I asc alberti
15 ABC ABDD IV|IV|IV|I IV|IV|V7|I I|V7|I-V7|I I|V7|I-V7|I convex corridinho
16 AABB I|V7|V7|I I|vii0|V7|I convex alberti
17 AABB I|V7|V7|I I|ii|V7|I desc arpeggio
18 AABB CC i|V7|V7|i i|i|V7|i i|V7|V7|i convex alberti
19 AABB I|V7|V7|I I|V7|V7|I desc fingered
20 ABB I|V7|V7|I I|V7|V7|I convex alberti
21 AABB I|V7|V7|I IV|I|V7|I convex corridinho
22 AABB I|V7|V7|I I|V7|V7|I convex alberti
23 AABBCC i|V7|V7|i i|i|V7|i i|V7|V7|i asc alberti
24 AAAABB i|V7|V7|i i|V7|V7|i desc corridinho
25 AABB I|V7|V7|I I|ii|V7|I desc chula
26 ABAB i|V7|V7|i i|V7|V7|i desc alberti
27 (a1a2a1b1) (c1c2d1c2) I|V7|V7|I III|V7/III-V7/V|V7|i IV|V|i-V7|i vii07|V7|i-V7|i convex corridinho
28 ABCD I|V7|V7|I V7/ii|ii|V7|I I|V7|V7|I I|V7|I-ii-V7|I desc alberti
29 (a1a2b1b2) (c1c1c2c2c2c2c2b2) (d1d1d1b2) (d2b2d2b2) i|V7|vii07-V7/III|III III|V7|i|i i|V7|V7|i i|V7|V7|i desc alberti
30 AABB CD V7|i|V7|i i|ii|V7|i III|III|V7/III-V7/V|V i-V7/iv|iv|i-V7|i convex baião
31 AABBB I|V7|V7|I I|V7|V7|I convex corridinho
32 ABAB I|V7|V7|I I|ii|V7|I convex corridinho
33 AABB I|V7|V7|I iv|i|V7|i convex arpeggio
34 AABB I|V7|V7|I I-V7/ii|ii|V7|I asc corridinho
35 AABCCC i|iv|iv|V7 i-V7/iv|ii0|V7|i V7|i|V7|i desc alberti
36 AABB i|iv|V7|i iv|i|V7|i convex hybrid
37 ABCC i|V7|V7|i i|V7|V7|i i|V7|V7|i desc habanera
38 AABB I|V7|V7|I I-V7/ii|V7|V7|I desc alberti
39 AABB i|V7|V7|i i|V7|V7|i desc habanera
40 AABB I|V7|V7|I V7|I|V7|I convex corridinho
41 AABA I|V7|V7|I V7|i|V7/V|V desc corridinho
42 ABB I|V7|V7|I V7/ii|ii|V7|I convex alberti
43 AABB I|V7|V7|I I|ii|V7|I convex corridinho
44 AABB IV|IV|IV|I V7|V7|I-V7|I desc corridinho
45 ABCC I|V7|V7|I III|V7/III|V7|i iv|i|V7|i convex corridinho
46 AABB I|V7|V7|I V7/ii|ii|V7|I convex alberti
47 AABB I|V7|V7|I i|V7|V7|i asc corridinho
48 ABCC D(a1b2)C i|i|i|V7 V7|i|i-V7|i ii07-V7|i|ii07-V7|i V7/III|III|iv-VI|V7 desc march
49 AABB i|V7|V7|i V7/iv|iv|V7|i convex alberti
50 ABAB CD i|V7|V7|i V7|iv|V7|i I|V7|V7|I I-V7/ii|ii-ii07|V7|I convex corridinho
51 AABC I|V7|V7|I I|V7|V7|I I|V7|V7|I convex ballad
52 AABC I|V7|V7|I V7/ii|ii|V7/iii|iii ii-V7/iii|I|V7|I desc hybrid
53 ABAB I-#i07|V7|ii-V7|I vi-V|V7|V7|I convex march
54 AABB I|V7|V7|I V7/vi|vi|V7|I desc ballad
55 ABCD i|i-iv|V7|V7-i i|V7/iv-iv|V7|i-V7-i i-V7|i-V7-i-V7/iv|iv|V7/III-III-V7/iv iv-V7|iv-V7|i-V7|i-V7-i desc march
56 AABC (d1d2d3) i-V7|i|ii0|i V7|i|V7|V7 ii0|i|V7|I I|I|V7|V7 convex corridinho
57 ABAC DEDE i|i|iv|i V7|i|iv|V7 V7/III|III|V7|I I|I|I|V7 desc fingered
58 ABAC DEDF i|V7|V7|i i|V7|V7|i V7/iv|iv|i-V7|i I-V7/ii|V7|V7|I convex march
59 AABCDE AABCDE AA V7|i|V7|i i|V7|V7|i i-VII|vi|V7|i-I I|I|I|V7 desc alberti
60 ABB CCDD CD vi|ii|V7|I i|iv|i-V7|i I-i0|V7|V7|I i|V7|V7|I convex march
61 ABAC DEDE I|V7|V7|I I|V7|V7|I V7/V|IV-#i07|V7/V-V7|I I-V7/ii|ii|V7|I asc march
62 ABCD (e1e2e3) i-bVII7|V7|V7|i V7/III|III|iv-V7|i ii-V7/vi|vi|V7|I V7|i|V7|i asc fingered
63 AB CDCD EFEG i|V7/V|V7|i i|V7|V7|i i-bVI|V7|i-bVI|V7-i i|V7/V|V7|i convex march
64 ABCDE FGH J FGHJ ABCDE FGHJ i|V7|V7|i i|V7|V7|i (V7/V)/V|V7/V|i-V7|i concave march
65 ABAC DEFG HJK ABAC i|V7|V7|i iv-V7|i|i-V7|i i|bVII|bVII|bVI bVI|V7|V7|i convex ballad
66 AABB CDE V7/ii|ii|ii|V7 V7|I|I|I i07|V7|V7|V7 IV|ii|ii|V7 convex corridinho
67 AAB CD AAB I|V7|V7|I I|V7|V7|I V7/ii|ii|V7|I i-V7|i|iv-V7/iv|iv convex corridinho
68 ABAC DE AC iv|i|V7|i I|I|I|V7 V7|V7|V7|I I|I-V7/ii|V7/ii|ii-iv convex arpeggio
69 AABCBD EFEF EFEF AA I|V7|V7|I V7/ii|ii|V7|I i|V7|V7|i iv|i|V7|i convex march
70 ABCD EF I|V7|V7|I vi|V7|V7|I V7|i|V7|i V7/iv|iv|V7|i asc corridinho
71 AABB I-V|I|IV-V|I V|V|V|I IV-#iv07|I|I-V7|I I|V7|V7|I asc march
72 AABB CDCD V7|I|V7|I I|V7|V7|I I-#iv07|V7|IV|I-V7 I|IV|i-V7|I convex ballad
73 ABCCDD ABCCDD AB I|I|I|V7 V7|V7|ii-V7|I V7/vi|IV|V7/ii|ii ii|I|V7|I convex march
74 ABCDD EFFE I|ii|V7|I i|V7|iv|V7 I-V7/ii|ii|V7|I i|V7|V7|i convex corridinho
75 AABB (c1c1c2c2) D i|iv|V7|I ii0|i|V7|i I|ii-V7/ii-ii|ii-V7|I I-i07|ii-V7|V7|I desc corridinho
76 AABC i|V7|V7|i i-V7/iv|iv|i-V7|i I/III|V7/III|V7/III|I/III iv|i|V7|i convex waltz
77 AB CDCE AB i|V7|V7|i i|V7|V7|i I|V7|V7|I I|V7|V7|I convex ballad
78 ABCD I|vii|V7|I I|IV|I-V7|I V7/ii|ii|V7|I V7|I|V7|I desc march
79 ABA CDCD EE I|V7|V7|I I|V7|V7|I I|ii|V7|I convex corridinho
80 AABB i|iv|V7|i V7/iv|iv|V7|i I|V7|V7|I I|V7|V7|I convex corridinho
81 AABBC I|I-V7|I-V7|V7 ii-V7|V7|V7-I|I bVII-V7|V7|V7-i|i ii0-I|I-V/V|V7-I|I concave corridinho
82 ABAB CD ABB V7|V7|I|I V7|V7|I|I i|V7|V7|i i|iv|V7|i convex corridinho
83 AABC I|IV|V7|I IV|I|V7|I convex hybrid
84 ABAC DEDE i|V7|V7|i V7/iv|iv|V7|i I|V7|V7|I concave march
85 ABAC DE I|V7|V7|I I|V7|V7|I i|V7|V7|i iv|V7|V7|i convex march
86 AABC DEDE i|V7|V7|i V7/III|III|V7|i iv|i|V7|i desc march
87 ABAC DEDE i|V7|V7|i I|V7|V7|I iv|i|V7|i I|V7|V7|I concave march
88 ABCDE i|V7/V|V7|i I|V7|V7|I iv|i|V7/V-V7|V7/IV I|V7|V7|I asc arpeggio
89 AABB CCDD I-i07|V7|V7|I V7/III|III|V7|i V7|I|V7|I i|V7|V7|i asc march
90 ABCDEF i|iv|V7|i i|V7|i-V7|i iv|i|V7|I I|ii|V7|I convex march
91 ABAC DEDE i07-ii07|i-V7/iv|iv-V7|i ii7|i|V7|I I|V7|V7|I iv|i|V7/V|V7 concave fox
92 ABCC DEFF i|V7|V7|i I|V7|V7|I i|iv|V7|i I|V7|V7|I asc march
93 ABAB CDEE i-V7|i|iv-i|V7 vi-V7/vi|vi-IV|I-V7|I V7|i|V7|i iv|i|V7|i desc march
94 ABCB DEFF BCB DEFF i|i|i-V7|V7 V7-iv|iv-V7|V7-i|V7-i ii-i|i-V7/V|V7|I I|I|I|ii asc corridinho
95 ABAB ABAB CDCD I|I|V7|I VI|VI|V7|i V7|I|V7|I V7|i|V7|i asc corridinho
96 AABB V7|V7|I |vi V7|V7|i|III iv|i|V7|I I|V7|V7|I concave march
97 ABAC DEFG AC i|V7|V7|i i|V7|V7|i III|iv-V7|i|V7 iv|V7|iv|V7 asc march
98 ABCB DEDF DEDF vi|I|IV|I iv|V7|iv|V7 I|V7|V7|I I|ii|V7|I asc march
99 ABAB CDCD i|V7|V7|i IV|I|V7|I asc march
100 ABCDCD I|ii|V7|I I|ii|V7|I bVI|I|ii|I i|ii07|V7|i convex march
2
Average
Average Time
Variability of Variability of
Average Note Variability of Maximum Minimum Staccato Average Time Between
Fado Edition Ostinati 2 Note Density Time Between Time Between Initial Tempo
Duration Note Duration Note Duration Note Duration Incidence Between Attacks Attacks For
Attacks Attacks For
Each Voice
Each Voice
1 7,84 0,40 0,11 0,69 0,11 0,00 0,31 0,11 0,31 0,11 64,00
2 alberti 7,64 0,35 0,12 0,95 0,12 0,00 0,25 0,06 0,25 0,06 64,00
3 6,11 0,28 0,12 0,95 0,22 0,00 0,24 0,03 0,24 0,03 64,00
4 6,82 0,30 0,22 1,38 0,22 0,00 0,24 0,06 0,24 0,06 64,00
5 5,40 0,33 0,18 1,06 0,25 0,00 0,27 0,04 0,27 0,04 56,00
6 5,75 0,35 0,27 2,41 0,20 0,00 0,26 0,21 0,26 0,21 64,00
7 7,49 0,46 0,20 2,50 0,13 0,00 0,36 0,18 0,36 0,18 64,00
8 6,41 0,29 0,20 1,87 0,19 0,00 0,25 0,10 0,25 0,10 120,00
9 7,50 0,24 0,11 0,74 0,05 0,01 0,23 0,06 0,23 0,06 64,00
10 6,71 0,29 0,17 1,90 0,19 0,00 0,25 0,13 0,25 0,13 64,00
11 6,26 0,36 0,15 1,11 0,06 0,00 0,30 0,16 0,30 0,16 64,00
12 7,17 0,28 0,11 0,53 0,10 0,00 0,23 0,11 0,23 0,11 64,00
13 5,68 0,50 0,33 3,49 0,12 0,00 0,37 0,39 0,37 0,39 64,00
14 6,11 0,31 0,18 1,38 0,22 0,00 0,25 0,09 0,25 0,09 120,00
15 7,67 0,40 0,14 1,17 0,12 0,00 0,29 0,13 0,29 0,13 64,00
16 6,64 0,26 0,11 1,38 0,22 0,00 0,24 0,03 0,24 0,03 64,00
17 5,95 0,36 0,49 3,28 0,22 0,00 0,25 0,09 0,25 0,09 64,00
18 corridinho 6,98 0,32 0,19 1,37 0,13 0,00 0,26 0,14 0,26 0,14 64,00
19 5,08 0,50 0,44 1,87 0,15 0,00 0,29 0,06 0,29 0,06 120,00
20 5,75 0,29 0,17 0,95 0,22 0,00 0,24 0,12 0,24 0,12 64,00
21 8,26 0,44 0,13 1,38 0,19 0,00 0,32 0,11 0,32 0,11 64,00
22 6,00 0,28 0,17 0,95 0,22 0,00 0,24 0,11 0,24 0,11 64,00
23 6,38 0,28 0,12 1,38 0,19 0,00 0,24 0,03 0,24 0,03 64,00
24 7,58 0,44 0,15 0,95 0,12 0,00 0,32 0,12 0,32 0,12 120,00
25 6,25 0,46 0,24 3,05 0,08 0,01 0,41 0,26 0,41 0,26 120,00
26 5,64 0,29 0,14 0,90 0,22 0,00 0,25 0,15 0,25 0,15 120,00
27 12,63 0,23 0,08 0,84 0,12 0,00 0,19 0,22 0,19 0,22 64,00
28 6,80 0,29 0,12 0,93 0,20 0,00 0,25 0,13 0,25 0,13 64,00
29 7,24 0,27 0,14 1,18 0,06 0,01 0,23 0,09 0,23 0,09 64,00
30 4,99 0,51 0,26 1,64 0,12 0,00 0,37 0,24 0,37 0,24 120,00
31 5,22 0,47 0,19 1,87 0,22 0,00 0,38 0,14 0,38 0,14 64,00
32 chula 6,83 0,48 0,27 1,87 0,22 0,00 0,40 0,18 0,40 0,18 64,00
33 6,59 0,28 0,15 0,95 0,20 0,00 0,24 0,10 0,24 0,10 64,00
34 5,71 0,44 0,11 0,95 0,22 0,00 0,38 0,17 0,38 0,17 64,00
35 6,89 0,30 0,16 0,95 0,20 0,00 0,24 0,03 0,24 0,03 64,00
36 5,07 0,40 0,14 1,23 0,20 0,00 0,32 0,11 0,32 0,11 120,00
37 4,03 0,40 0,14 0,95 0,22 0,00 0,37 0,12 0,37 0,12 120,00
38 6,61 0,27 0,14 1,03 0,19 0,00 0,22 0,07 0,22 0,07 120,00
39 4,29 0,38 0,13 0,71 0,12 0,00 0,35 0,15 0,35 0,15 64,00
40 7,28 0,46 0,16 1,41 0,22 0,00 0,37 0,13 0,37 0,13 64,00
41 5,87 0,44 0,09 0,93 0,22 0,00 0,36 0,12 0,36 0,12 64,00
42 6,14 0,28 0,20 2,02 0,12 0,00 0,25 0,17 0,25 0,17 64,00
43 6,78 0,47 0,17 1,41 0,22 0,00 0,37 0,12 0,37 0,12 120,00
44 7,10 0,38 0,13 0,95 0,12 0,00 0,29 0,12 0,29 0,12 120,00
45 6,91 0,43 0,12 0,95 0,09 0,01 0,34 0,13 0,34 0,13 64,00
46 8,32 0,32 0,20 1,38 0,09 0,00 0,23 0,04 0,23 0,04 120,00
47 7,30 0,43 0,13 1,16 0,19 0,00 0,31 0,14 0,31 0,14 120,00
48 11,21 0,28 0,12 0,56 0,07 0,04 0,22 0,13 0,22 0,13 120,00
49 6,03 0,30 0,16 1,17 0,22 0,00 0,25 0,06 0,25 0,06 120,00
50 6,83 0,44 0,40 2,39 0,05 0,08 0,27 0,29 0,27 0,29 108,00
51 arpeggio 8,26 0,30 0,32 2,40 0,05 0,11 0,19 0,26 0,19 0,26 64,00
52 5,08 0,45 0,37 3,13 0,11 0,00 0,31 0,27 0,31 0,27 64,00
53 7,46 0,31 0,25 1,67 0,09 0,02 0,24 0,07 0,24 0,07 108,00
54 7,74 0,34 0,22 1,41 0,06 0,11 0,15 0,10 0,15 0,10 64,00
55 7,34 0,45 0,27 2,38 0,19 0,00 0,32 0,19 0,32 0,19 72,00
56 7,69 0,41 0,27 2,04 0,03 0,02 0,28 0,21 0,28 0,21 64,00
57 7,79 0,32 0,26 1,88 0,09 0,23 0,22 0,16 0,22 0,16 64,00
58 6,44 0,46 0,29 1,95 0,07 0,00 0,30 0,16 0,30 0,16 64,00
59 6,72 0,28 0,17 1,41 0,22 0,00 0,23 0,00 0,23 0,00 72,00
60 9,05 0,31 0,27 1,88 0,09 0,05 0,23 0,19 0,23 0,19 108,00
61 8,32 0,42 0,32 2,04 0,15 0,00 0,27 0,24 0,27 0,24 64,00
62 corridinho 9,22 0,41 0,27 2,35 0,01 0,01 0,27 0,16 0,27 0,16 64,00
63 corridinho 9,08 0,32 0,14 1,20 0,06 0,05 0,24 0,15 0,24 0,15 74,00
64 10,20 0,32 0,30 2,11 0,02 0,00 0,24 0,09 0,24 0,09 128,00
65 8,78 0,36 0,32 1,96 0,03 0,11 0,17 0,19 0,17 0,19 120,00
66 march 8,42 0,38 0,18 1,63 0,08 0,00 0,24 0,12 0,24 0,12 64,00
67 7,07 0,57 0,58 3,57 0,07 0,02 0,31 0,26 0,31 0,26 64,00
68 march 6,55 0,44 0,29 2,09 0,08 0,00 0,27 0,17 0,27 0,17 64,00
69 8,86 0,40 0,27 2,14 0,07 0,00 0,27 0,06 0,27 0,06 108,00
70 chula 6,83 0,42 0,23 2,33 0,06 0,02 0,32 0,14 0,32 0,14 64,00
71 9,62 0,39 0,32 2,28 0,10 0,00 0,28 0,09 0,28 0,09 91,00
72 hybrid 9,10 0,37 0,28 2,09 0,07 0,04 0,20 0,18 0,20 0,18 64,00
73 5,76 0,46 0,22 1,58 0,22 0,00 0,34 0,15 0,34 0,15 64,00
74 7,83 0,37 0,16 1,87 0,09 0,05 0,29 0,17 0,29 0,17 64,00
75 hybrid 6,80 0,40 0,23 2,31 0,08 0,02 0,29 0,20 0,29 0,20 64,00
76 7,94 0,50 0,35 2,09 0,20 0,00 0,30 0,11 0,30 0,11 64,00
77 hybrid 7,70 0,35 0,24 2,03 0,03 0,05 0,22 0,17 0,22 0,17 64,00
78 7,40 0,36 0,40 3,27 0,08 0,01 0,29 0,17 0,29 0,17 91,00
79 9,79 0,32 0,22 3,05 0,02 0,02 0,25 0,13 0,25 0,13 64,00
80 6,94 0,47 0,29 2,04 0,09 0,00 0,32 0,21 0,32 0,21 64,00
81 10,22 0,38 0,21 2,05 0,07 0,00 0,26 0,16 0,26 0,16 64,00
82 9,11 0,40 0,18 2,34 0,09 0,01 0,27 0,12 0,27 0,12 64,00
83 march 6,32 0,40 0,23 1,56 0,12 0,00 0,28 0,14 0,28 0,14 64,00
84 10,53 0,29 0,23 1,87 0,05 0,05 0,19 0,11 0,19 0,11 64,00
85 10,94 0,23 0,21 1,88 0,06 0,05 0,19 0,06 0,19 0,06 128,00
86 8,73 0,30 0,20 1,58 0,06 0,01 0,26 0,13 0,26 0,13 128,00
87 10,08 0,23 0,18 1,88 0,06 0,09 0,19 0,09 0,19 0,09 128,00
88 march 13,46 0,30 0,39 2,46 0,03 0,46 0,14 0,19 0,14 0,19 64,00
89 alberti 5,83 0,38 0,17 0,95 0,19 0,00 0,30 0,11 0,30 0,11 64,00
90 alberti 7,36 0,41 0,30 1,95 0,05 0,05 0,25 0,16 0,25 0,16 64,00
91 4,73 0,71 0,59 5,00 0,10 0,00 0,35 0,22 0,35 0,22 72,00
92 hybrid 10,32 0,23 0,09 0,87 0,08 0,02 0,18 0,08 0,18 0,08 108,00
93 corridinho 8,53 0,25 0,20 1,87 0,03 0,09 0,25 0,17 0,25 0,17 64,00
94 7,82 0,41 0,25 2,21 0,06 0,01 0,28 0,19 0,28 0,19 64,00
95 alberti 10,88 0,27 0,12 0,94 0,13 0,00 0,19 0,08 0,19 0,08 95,00
96 8,00 0,34 0,34 1,87 0,03 0,13 0,26 0,17 0,26 0,17 64,00
97 arpeggio 8,69 0,24 0,21 1,41 0,06 0,09 0,18 0,15 0,18 0,15 64,00
98 8,91 0,31 0,24 2,20 0,07 0,03 0,22 0,15 0,22 0,15 91,00
99 9,90 0,27 0,30 3,75 0,06 0,03 0,23 0,08 0,23 0,08 64,00
100 8,25 0,29 0,21 1,64 0,09 0,03 0,21 0,13 0,21 0,13 91,00
3
Most Relative
Variation of Average Note To Relative
Compound Or Simple Changes of Overall Dynamic Variation of Common Most Common Pitch Strength of
Fado Edition Dynamics In Note Dynamics Strength of
Meter Meter Range Dynamics Pitch Class Prevalence Top Pitch
Each Voice Change Top Pitches
Prevalence Classes
1 0,00 0,00 20,00 6,64 6,64 5,84 0,19 0,28 0,67 0,75
2 0,00 0,00 25,00 6,09 6,09 5,30 0,21 0,34 0,49 0,46
3 0,00 0,00 17,00 5,09 5,09 3,70 0,29 0,42 0,33 0,44
4 0,00 0,00 25,00 6,52 6,52 6,92 0,29 0,37 0,31 0,50
5 0,00 0,00 13,00 5,09 5,09 3,41 0,26 0,37 0,34 0,42
6 0,00 0,00 29,00 5,83 5,83 4,19 0,27 0,35 0,37 0,50
7 0,00 1,00 63,00 14,25 14,25 9,96 0,13 0,20 0,65 0,83
8 0,00 1,00 24,00 5,55 5,55 3,69 0,23 0,40 0,39 0,39
9 0,00 1,00 17,00 4,40 4,40 3,80 0,27 0,39 0,29 0,42
10 0,00 1,00 53,00 14,61 14,61 4,88 0,11 0,23 0,72 0,88
11 0,00 1,00 30,00 6,41 6,41 5,40 0,13 0,23 0,64 0,99
12 0,00 1,00 29,00 6,29 6,29 4,93 0,21 0,31 0,37 0,41
13 0,00 1,00 55,00 7,44 7,44 5,60 0,17 0,27 0,64 0,63
14 0,00 0,00 23,00 5,66 5,66 4,06 0,24 0,33 0,42 0,48
15 0,00 1,00 19,00 6,09 6,09 5,23 0,14 0,26 0,81 0,72
16 0,00 1,00 17,00 5,40 5,40 3,27 0,24 0,36 0,36 0,45
17 0,00 0,00 29,00 5,66 5,66 4,43 0,11 0,21 0,79 0,93
18 0,00 0,00 30,00 6,40 6,40 6,35 0,21 0,32 0,56 0,63
19 1,00 0,00 19,00 6,24 6,24 6,91 0,31 0,48 0,52 0,36
20 0,00 1,00 13,00 5,07 5,07 3,45 0,17 0,32 1,00 0,86
21 0,00 0,00 30,00 7,42 7,42 7,50 0,17 0,25 0,59 0,83
22 0,00 1,00 13,00 5,01 5,01 3,41 0,33 0,43 0,35 0,51
23 0,00 1,00 48,00 12,11 12,11 3,83 0,33 0,43 0,29 0,38
24 0,00 1,00 46,00 10,70 10,70 5,91 0,20 0,32 0,52 0,61
25 0,00 1,00 29,00 7,27 7,27 6,74 0,13 0,29 0,75 0,52
26 0,00 1,00 17,00 4,89 4,89 3,47 0,32 0,37 0,33 0,59
27 0,00 0,00 39,00 6,36 6,36 5,51 0,13 0,21 0,73 0,89
28 0,00 1,00 30,00 5,99 5,99 4,78 0,13 0,24 0,95 0,94
29 0,00 0,00 28,00 5,96 5,96 4,40 0,26 0,41 0,48 0,36
30 0,00 1,00 30,00 7,46 7,46 7,31 0,13 0,26 0,94 0,86
31 0,00 1,00 26,00 6,63 6,63 7,08 0,25 0,31 0,51 0,61
32 0,00 1,00 56,00 17,19 17,19 7,87 0,13 0,22 0,89 0,83
33 0,00 0,00 43,00 11,88 11,88 5,06 0,15 0,31 0,91 0,75
34 0,00 1,00 20,00 6,82 6,82 7,04 0,19 0,27 0,65 0,63
35 0,00 1,00 29,00 5,62 5,62 4,56 0,20 0,38 0,57 0,43
36 0,00 1,00 20,00 5,96 5,96 6,61 0,16 0,27 0,61 0,77
37 0,00 0,00 17,00 5,73 5,73 5,43 0,24 0,34 0,40 0,60
38 0,00 1,00 13,00 5,05 5,05 3,58 0,18 0,31 0,91 0,84
39 0,00 1,00 20,00 5,69 5,69 5,27 0,23 0,37 0,52 0,55
40 0,00 1,00 25,00 6,72 6,72 5,87 0,15 0,24 0,68 0,69
41 0,00 1,00 20,00 6,87 6,87 6,68 0,17 0,30 0,74 0,56
42 0,00 1,00 13,00 4,95 4,95 3,19 0,20 0,34 0,52 0,45
43 0,00 1,00 25,00 6,77 6,77 6,69 0,18 0,28 0,54 0,65
44 0,00 0,00 46,00 8,24 8,24 7,28 0,14 0,23 0,80 0,96
45 0,00 1,00 25,00 7,07 7,07 5,95 0,13 0,18 0,77 0,97
46 0,00 1,00 29,00 6,30 6,30 6,63 0,19 0,29 0,67 0,64
47 0,00 1,00 20,00 6,71 6,71 6,03 0,14 0,20 0,79 0,72
48 0,00 0,00 63,00 16,98 16,98 7,63 0,10 0,22 0,85 0,95
49 0,00 1,00 21,00 5,54 5,54 3,84 0,17 0,27 0,89 0,73
50 0,00 0,00 75,00 16,65 16,65 6,78 0,10 0,20 0,99 1,00
51 0,00 0,00 51,00 7,92 7,92 5,62 0,12 0,29 0,96 0,69
52 0,00 1,00 34,00 7,71 7,71 4,42 0,12 0,30 0,77 0,47
53 0,00 0,00 31,00 6,96 6,96 8,45 0,17 0,33 0,60 0,40
54 0,00 0,00 55,00 18,19 18,19 5,59 0,16 0,30 0,82 0,65
55 0,00 0,00 33,00 7,51 7,51 6,54 0,11 0,22 0,83 0,99
56 0,00 1,00 44,00 10,71 10,71 7,54 0,10 0,25 0,85 0,68
57 0,00 1,00 56,00 13,56 13,56 7,57 0,14 0,28 0,60 0,63
58 0,00 0,00 62,00 9,04 9,04 6,86 0,11 0,29 0,82 0,48
59 0,00 0,00 71,00 12,93 12,93 11,24 0,29 0,49 0,42 0,25
60 0,00 0,00 41,00 9,67 9,67 9,26 0,13 0,23 0,68 0,70
61 0,00 0,00 25,00 7,18 7,18 8,04 0,08 0,20 0,91 0,74
62 0,00 0,00 91,00 12,42 12,42 8,29 0,11 0,22 0,71 0,83
63 0,00 0,00 57,00 15,05 15,05 7,67 0,12 0,25 0,67 0,65
64 0,00 1,00 42,00 7,37 7,37 7,91 0,12 0,21 0,85 0,78
65 0,00 1,00 82,00 17,28 17,28 7,82 0,10 0,22 0,78 0,72
66 0,00 0,00 30,00 6,97 6,97 5,69 0,09 0,16 0,93 1,00
67 0,00 1,00 74,00 17,51 17,51 7,05 0,12 0,20 0,68 0,82
68 0,00 0,00 89,00 19,75 19,75 6,86 0,09 0,24 0,98 0,82
69 0,00 1,00 88,00 19,85 19,85 9,73 0,10 0,24 0,54 0,70
70 0,00 0,00 51,00 8,10 8,10 7,49 0,10 0,19 0,92 0,90
71 0,00 0,00 58,00 12,15 12,15 7,45 0,09 0,18 0,89 0,73
72 0,00 0,00 50,00 8,17 8,17 7,15 0,09 0,21 0,80 0,85
73 0,00 0,00 20,00 6,98 6,98 6,53 0,10 0,20 0,94 0,87
74 0,00 0,00 79,00 21,72 21,72 7,32 0,16 0,27 0,60 0,58
75 0,00 1,00 68,00 17,04 17,04 6,70 0,12 0,21 0,74 0,88
76 1,00 0,00 75,00 9,20 9,20 9,66 0,15 0,25 0,67 0,73
77 0,00 0,00 73,00 8,97 8,97 5,95 0,12 0,29 0,75 0,68
78 0,00 0,00 56,00 10,22 10,22 6,88 0,10 0,22 0,85 0,75
79 0,00 0,00 91,00 26,50 26,50 7,70 0,10 0,27 0,97 0,70
80 0,00 1,00 25,00 6,74 6,74 6,58 0,13 0,24 0,98 0,93
81 0,00 0,00 85,00 17,33 17,33 7,33 0,13 0,23 0,74 1,00
82 0,00 0,00 62,00 11,84 11,84 6,43 0,12 0,19 0,83 0,99
83 0,00 0,00 51,00 10,94 10,94 7,52 0,13 0,26 0,68 0,56
84 0,00 0,00 64,00 13,44 13,44 6,87 0,10 0,21 0,88 0,70
85 0,00 0,00 51,00 13,25 13,25 7,00 0,12 0,20 0,60 0,78
86 0,00 0,00 47,00 8,46 8,46 7,60 0,13 0,26 0,63 0,58
87 0,00 0,00 70,00 15,17 15,17 7,71 0,11 0,20 0,90 0,74
88 0,00 1,00 102,00 21,99 21,99 10,81 0,11 0,23 0,73 0,85
89 0,00 0,00 58,00 8,52 8,52 6,32 0,11 0,25 1,00 0,70
90 0,00 1,00 86,00 11,89 11,89 7,26 0,15 0,29 0,44 0,65
91 0,00 0,00 61,00 10,03 10,03 8,09 0,10 0,25 0,74 0,78
92 0,00 0,00 66,00 9,03 9,03 6,86 0,10 0,29 0,89 0,64
93 0,00 0,00 73,00 15,48 15,48 6,58 0,10 0,20 0,72 0,68
94 0,00 1,00 20,00 6,23 6,23 5,33 0,12 0,18 0,84 0,88
95 0,00 0,00 49,00 6,72 6,72 6,29 0,12 0,26 0,99 0,60
96 0,00 0,00 33,00 7,40 7,40 7,25 0,11 0,27 0,55 0,65
97 0,00 0,00 58,00 11,82 11,82 4,67 0,08 0,17 0,93 0,85
98 0,00 0,00 61,00 9,62 9,62 7,18 0,09 0,21 0,91 0,97
99 0,00 0,00 75,00 9,77 9,77 7,73 0,15 0,30 0,74 0,73
100 0,00 0,00 70,00 13,66 13,66 6,79 0,14 0,24 0,63 0,63
4
Interval Interval
Number of Most Importance of
Between Between Pitch Class Primary Importance of Importance of Most Common Pitch
Fado Edition Common Pitch Variety Range Common Middle
Strongest Strongest Variety Register Bass Register High Register Class
Pitches Pitch Register
Pitches Pitch Classes
1 5,00 7,00 5,00 18,00 8,00 41,00 0,45 63,00 0,26 0,40 0,34 9,00
2 12,00 5,00 2,00 27,00 11,00 41,00 0,47 63,00 0,12 0,70 0,19 0,00
3 7,00 7,00 3,00 27,00 8,00 46,00 0,45 61,00 0,28 0,49 0,24 9,00
4 7,00 5,00 1,00 17,00 8,00 39,00 0,41 56,00 0,60 0,29 0,12 4,00
5 24,00 5,00 1,00 17,00 8,00 39,00 0,41 58,00 0,55 0,29 0,16 4,00
6 2,00 5,00 3,00 21,00 9,00 36,00 0,47 59,00 0,16 0,84 0,00 0,00
7 4,00 4,00 1,00 34,00 12,00 51,00 0,41 60,00 0,35 0,51 0,14 5,00
8 12,00 8,00 2,00 23,00 9,00 41,00 0,48 62,00 0,13 0,74 0,13 2,00
9 7,00 5,00 1,00 21,00 10,00 38,00 0,38 55,00 0,55 0,37 0,08 0,00
10 7,00 5,00 1,00 31,00 7,00 55,00 0,41 60,00 0,46 0,28 0,25 5,00
11 7,00 7,00 1,00 21,00 7,00 40,00 0,45 60,00 0,32 0,52 0,16 9,00
12 12,00 2,00 1,00 26,00 11,00 34,00 0,48 65,00 0,05 0,73 0,22 2,00
13 5,00 3,00 3,00 31,00 10,00 46,00 0,45 60,00 0,27 0,62 0,12 10,00
14 3,00 9,00 3,00 26,00 9,00 38,00 0,45 60,00 0,25 0,64 0,11 9,00
15 7,00 7,00 2,00 22,00 9,00 38,00 0,41 59,00 0,36 0,50 0,13 4,00
16 2,00 2,00 1,00 27,00 7,00 49,00 0,45 60,00 0,28 0,57 0,14 9,00
17 5,00 5,00 1,00 26,00 7,00 45,00 0,42 58,00 0,41 0,49 0,09 6,00
18 5,00 7,00 2,00 17,00 8,00 34,00 0,46 61,00 0,20 0,79 0,01 11,00
19 12,00 7,00 2,00 18,00 7,00 41,00 0,45 61,00 0,13 0,81 0,06 10,00
20 1,00 7,00 4,00 14,00 7,00 29,00 0,48 61,00 0,17 0,78 0,05 9,00
21 3,00 7,00 3,00 21,00 7,00 45,00 0,45 59,00 0,29 0,63 0,09 9,00
22 17,00 7,00 2,00 15,00 7,00 34,00 0,43 58,00 0,33 0,61 0,05 7,00
23 7,00 4,00 3,00 21,00 10,00 36,00 0,41 55,00 0,66 0,29 0,05 4,00
24 10,00 5,00 4,00 21,00 9,00 37,00 0,39 58,00 0,30 0,61 0,09 2,00
25 24,00 5,00 2,00 23,00 8,00 42,00 0,39 60,00 0,29 0,58 0,13 2,00
26 17,00 7,00 2,00 18,00 8,00 32,00 0,45 60,00 0,25 0,54 0,22 9,00
27 5,00 4,00 3,00 26,00 10,00 40,00 0,46 59,00 0,31 0,66 0,03 11,00
28 5,00 5,00 3,00 27,00 11,00 38,00 0,39 57,00 0,44 0,51 0,05 2,00
29 12,00 5,00 2,00 37,00 11,00 55,00 0,48 66,00 0,04 0,65 0,31 2,00
30 5,00 7,00 2,00 28,00 11,00 43,00 0,45 61,00 0,20 0,67 0,14 9,00
31 3,00 3,00 4,00 20,00 7,00 41,00 0,45 60,00 0,26 0,53 0,21 9,00
32 21,00 5,00 2,00 24,00 9,00 40,00 0,39 60,00 0,29 0,60 0,11 2,00
33 5,00 7,00 3,00 24,00 10,00 39,00 0,43 58,00 0,32 0,66 0,03 0,00
34 2,00 5,00 3,00 17,00 8,00 33,00 0,41 56,00 0,61 0,33 0,06 5,00
35 12,00 2,00 2,00 28,00 9,00 42,00 0,46 62,00 0,21 0,69 0,11 11,00
36 5,00 7,00 3,00 19,00 7,00 34,00 0,43 59,00 0,23 0,71 0,06 7,00
37 22,00 5,00 2,00 17,00 7,00 36,00 0,41 61,00 0,37 0,48 0,15 4,00
38 5,00 7,00 3,00 18,00 9,00 29,00 0,45 61,00 0,18 0,77 0,05 9,00
39 20,00 5,00 4,00 14,00 7,00 37,00 0,41 61,00 0,37 0,41 0,23 4,00
40 5,00 7,00 3,00 28,00 8,00 45,00 0,45 60,00 0,16 0,73 0,10 10,00
41 3,00 9,00 3,00 25,00 10,00 32,00 0,47 61,00 0,16 0,75 0,09 0,00
42 2,00 2,00 3,00 17,00 8,00 31,00 0,43 58,00 0,31 0,68 0,01 7,00
43 3,00 7,00 4,00 26,00 8,00 50,00 0,43 59,00 0,28 0,65 0,08 7,00
44 5,00 2,00 5,00 21,00 10,00 35,00 0,44 60,00 0,27 0,68 0,05 8,00
45 4,00 8,00 2,00 31,00 8,00 50,00 0,47 61,00 0,19 0,69 0,12 0,00
46 9,00 9,00 3,00 16,00 8,00 26,00 0,48 65,00 0,05 0,84 0,12 2,00
47 7,00 7,00 3,00 21,00 8,00 34,00 0,41 58,00 0,27 0,73 0,01 4,00
48 19,00 5,00 1,00 32,00 9,00 48,00 0,43 62,00 0,24 0,55 0,21 7,00
49 7,00 5,00 2,00 20,00 10,00 32,00 0,48 61,00 0,17 0,72 0,11 2,00
50 5,00 5,00 2,00 47,00 12,00 58,00 0,41 62,00 0,24 0,55 0,21 4,00
51 12,00 5,00 2,00 24,00 7,00 53,00 0,49 59,00 0,25 0,67 0,09 3,00
52 24,00 5,00 2,00 32,00 11,00 41,00 0,38 56,00 0,50 0,43 0,07 0,00
53 2,00 2,00 2,00 26,00 9,00 48,00 0,43 56,00 0,45 0,54 0,01 7,00
54 9,00 5,00 3,00 23,00 8,00 36,00 0,39 57,00 0,35 0,61 0,04 2,00
55 5,00 7,00 2,00 32,00 9,00 48,00 0,45 62,00 0,27 0,40 0,33 9,00
56 7,00 2,00 1,00 30,00 9,00 46,00 0,50 65,00 0,13 0,62 0,25 4,00
57 5,00 5,00 1,00 43,00 12,00 46,00 0,41 61,00 0,27 0,51 0,22 4,00
58 12,00 5,00 2,00 39,00 12,00 44,00 0,47 65,00 0,13 0,57 0,29 0,00
59 12,00 2,00 2,00 25,00 11,00 32,00 0,45 59,00 0,24 0,71 0,05 9,00
60 2,00 7,00 1,00 33,00 11,00 41,00 0,45 56,00 0,37 0,60 0,03 9,00
61 10,00 3,00 0,00 40,00 12,00 53,00 0,51 60,00 0,27 0,66 0,07 7,00
62 12,00 7,00 1,00 43,00 10,00 55,00 0,45 58,00 0,33 0,53 0,14 9,00
63 5,00 7,00 1,00 43,00 12,00 53,00 0,43 62,00 0,21 0,62 0,17 7,00
64 2,00 7,00 2,00 44,00 12,00 53,00 0,45 59,00 0,27 0,66 0,07 9,00
65 2,00 2,00 1,00 48,00 12,00 60,00 0,46 64,00 0,19 0,51 0,30 11,00
66 14,00 5,00 1,00 32,00 11,00 41,00 0,41 61,00 0,20 0,68 0,11 2,00
67 5,00 5,00 1,00 31,00 12,00 48,00 0,39 59,00 0,29 0,63 0,07 2,00
68 7,00 7,00 0,00 39,00 12,00 46,00 0,46 61,00 0,30 0,55 0,15 11,00
69 7,00 5,00 1,00 49,00 12,00 60,00 0,47 62,00 0,24 0,58 0,18 0,00
70 5,00 7,00 2,00 36,00 12,00 48,00 0,45 60,00 0,24 0,66 0,11 2,00
71 12,00 5,00 0,00 41,00 12,00 50,00 0,42 59,00 0,31 0,62 0,07 6,00
72 5,00 5,00 1,00 40,00 12,00 50,00 0,48 61,00 0,21 0,69 0,10 2,00
73 5,00 5,00 2,00 32,00 10,00 41,00 0,41 61,00 0,26 0,59 0,15 9,00
74 12,00 5,00 2,00 27,00 10,00 41,00 0,41 57,00 0,40 0,58 0,01 4,00
75 16,00 8,00 1,00 29,00 10,00 48,00 0,57 62,00 0,23 0,53 0,24 1,00
76 10,00 2,00 2,00 24,00 9,00 41,00 0,52 72,00 0,00 0,49 0,51 7,00
77 12,00 5,00 1,00 40,00 12,00 65,00 0,48 60,00 0,26 0,62 0,12 2,00
78 3,00 9,00 1,00 35,00 11,00 44,00 0,48 58,00 0,31 0,65 0,04 2,00
79 12,00 5,00 2,00 46,00 12,00 55,00 0,50 63,00 0,20 0,61 0,20 4,00
80 5,00 7,00 2,00 31,00 10,00 41,00 0,48 61,00 0,21 0,66 0,13 9,00
81 7,00 7,00 2,00 30,00 9,00 48,00 0,45 59,00 0,31 0,62 0,06 2,00
82 9,00 5,00 2,00 35,00 8,00 60,00 0,39 58,00 0,33 0,62 0,04 7,00
83 2,00 8,00 1,00 31,00 9,00 44,00 0,47 60,00 0,26 0,61 0,12 0,00
84 2,00 7,00 1,00 35,00 11,00 48,00 0,52 61,00 0,20 0,74 0,06 9,00
85 5,00 7,00 1,00 43,00 11,00 52,00 0,47 57,00 0,37 0,54 0,09 0,00
86 5,00 2,00 1,00 40,00 11,00 45,00 0,45 57,00 0,36 0,60 0,04 9,00
87 5,00 5,00 2,00 37,00 12,00 48,00 0,43 57,00 0,31 0,68 0,01 7,00
88 7,00 7,00 1,00 48,00 12,00 67,00 0,58 68,00 0,12 0,40 0,48 9,00
89 9,00 3,00 2,00 28,00 8,00 48,00 0,43 60,00 0,26 0,63 0,11 4,00
90 5,00 5,00 1,00 39,00 12,00 45,00 0,41 60,00 0,33 0,51 0,15 4,00
91 24,00 5,00 1,00 50,00 12,00 63,00 0,52 59,00 0,36 0,51 0,13 7,00
92 12,00 5,00 1,00 33,00 10,00 53,00 0,38 61,00 0,29 0,53 0,18 0,00
93 12,00 7,00 1,00 44,00 12,00 50,00 0,45 62,00 0,23 0,60 0,16 9,00
94 3,00 5,00 4,00 31,00 12,00 37,00 0,48 61,00 0,15 0,77 0,08 4,00
95 12,00 5,00 2,00 32,00 11,00 40,00 0,50 59,00 0,30 0,68 0,02 4,00
96 17,00 7,00 1,00 29,00 9,00 48,00 0,45 65,00 0,20 0,45 0,36 9,00
97 5,00 7,00 0,00 45,00 12,00 50,00 0,45 59,00 0,30 0,60 0,10 9,00
98 10,00 5,00 0,00 40,00 12,00 46,00 0,41 57,00 0,34 0,62 0,04 4,00
99 12,00 7,00 3,00 31,00 10,00 43,00 0,43 57,00 0,34 0,65 0,01 7,00
100 2,00 7,00 2,00 33,00 12,00 38,00 0,43 57,00 0,34 0,66 0,00 7,00
5
Distance Most Relative

Most Number of
Average Between Most Common Strength of
Dominant Strong Tonal Common Common Amount of Repeated Chromatic Stepwise Melodic
Fado Edition Melodic Common Melodic Most
Spread Centres Melodic Melodic Arpeggiation Notes Motion Motion Thirds
Interval Melodic Interval Common
Interval Intervals
Intervals Prevalence Intervals
1 3,00 3,00 13,55 12,00 11,00 0,18 0,71 3,00 0,29 0,00 0,13 0,22 0,05
2 4,00 1,00 10,14 12,00 10,00 0,12 0,83 2,00 0,46 0,04 0,06 0,16 0,12
3 4,00 2,00 13,19 7,00 2,00 0,16 0,66 4,00 0,40 0,00 0,00 0,08 0,09
4 4,00 2,00 11,29 5,00 2,00 0,17 0,91 4,00 0,38 0,00 0,00 0,08 0,15
5 4,00 2,00 14,63 12,00 12,00 0,15 1,00 3,00 0,38 0,00 0,00 0,06 0,06
6 4,00 1,00 7,45 7,00 2,00 0,17 0,93 5,00 0,55 0,01 0,00 0,09 0,17
7 3,00 3,00 11,88 12,00 12,00 0,15 0,65 2,00 0,47 0,04 0,02 0,05 0,12
8 4,00 3,00 10,22 7,00 5,00 0,15 1,00 4,00 0,56 0,01 0,00 0,06 0,11
9 3,00 2,00 17,50 24,00 7,00 0,12 0,78 2,00 0,23 0,00 0,02 0,09 0,07
10 4,00 2,00 17,93 5,00 1,00 0,13 0,91 3,00 0,33 0,00 0,00 0,00 0,13
11 4,00 2,00 9,97 4,00 8,00 0,16 0,97 3,00 0,51 0,01 0,05 0,08 0,27
12 6,00 2,00 8,80 5,00 2,00 0,15 0,97 4,00 0,55 0,01 0,00 0,05 0,17
13 2,00 3,00 8,06 2,00 1,00 0,21 0,60 3,00 0,44 0,05 0,05 0,26 0,24
14 6,00 2,00 9,87 5,00 2,00 0,14 0,95 4,00 0,52 0,01 0,00 0,07 0,14
15 3,00 2,00 10,43 2,00 1,00 0,15 0,82 3,00 0,27 0,01 0,13 0,28 0,07
16 4,00 3,00 12,42 3,00 4,00 0,14 0,83 4,00 0,51 0,00 0,00 0,05 0,15
17 7,00 2,00 9,56 3,00 1,00 0,17 0,90 3,00 0,49 0,00 0,00 0,02 0,32
18 4,00 2,00 7,72 12,00 7,00 0,17 0,78 5,00 0,56 0,01 0,04 0,14 0,17
19 2,00 2,00 7,76 5,00 7,00 0,18 0,96 5,00 0,57 0,02 0,02 0,11 0,20
20 2,00 2,00 8,88 12,00 8,00 0,24 0,55 3,00 0,60 0,00 0,03 0,03 0,15
21 3,00 2,00 10,89 12,00 5,00 0,13 0,85 2,00 0,42 0,03 0,04 0,12 0,06
22 3,00 1,00 11,66 12,00 5,00 0,18 0,83 3,00 0,50 0,00 0,00 0,05 0,09
23 2,00 2,00 12,54 7,00 3,00 0,14 0,97 3,00 0,41 0,00 0,00 0,06 0,13
24 4,00 2,00 7,55 7,00 2,00 0,14 0,98 5,00 0,49 0,05 0,11 0,23 0,04
25 4,00 2,00 10,57 3,00 2,00 0,11 0,84 2,00 0,53 0,02 0,10 0,14 0,20
26 4,00 1,00 10,52 5,00 2,00 0,13 1,00 5,00 0,36 0,02 0,00 0,11 0,10
27 4,00 3,00 10,89 1,00 6,00 0,14 0,72 2,00 0,32 0,01 0,14 0,21 0,05
28 3,00 2,00 11,99 12,00 3,00 0,19 0,65 2,00 0,42 0,01 0,01 0,02 0,08
29 4,00 2,00 10,42 12,00 5,00 0,16 0,96 3,00 0,51 0,00 0,04 0,11 0,13
30 3,00 2,00 11,05 12,00 9,00 0,13 0,88 3,00 0,53 0,02 0,05 0,10 0,15
31 3,00 2,00 12,12 3,00 9,00 0,12 0,92 3,00 0,44 0,00 0,05 0,15 0,18
32 4,00 2,00 9,77 2,00 5,00 0,12 1,00 3,00 0,42 0,03 0,04 0,16 0,08
33 2,00 2,00 10,82 7,00 1,00 0,15 0,91 4,00 0,47 0,01 0,01 0,02 0,03
34 4,00 3,00 11,17 4,00 2,00 0,11 0,89 3,00 0,45 0,06 0,05 0,15 0,18
35 3,00 2,00 9,19 7,00 5,00 0,17 1,00 5,00 0,56 0,00 0,00 0,02 0,13
36 4,00 2,00 10,52 7,00 12,00 0,16 0,62 2,00 0,46 0,00 0,02 0,09 0,14
37 4,00 3,00 14,92 17,00 13,00 0,10 0,88 1,00 0,27 0,01 0,04 0,09 0,12
38 3,00 2,00 8,96 12,00 5,00 0,15 0,84 4,00 0,56 0,03 0,00 0,08 0,17
39 4,00 2,00 16,48 19,00 17,00 0,13 0,73 3,00 0,19 0,00 0,06 0,16 0,06
40 4,00 2,00 11,19 9,00 3,00 0,12 0,91 3,00 0,32 0,00 0,07 0,17 0,06
41 3,00 2,00 12,10 12,00 11,00 0,11 0,78 1,00 0,41 0,02 0,09 0,12 0,05
42 5,00 2,00 10,26 7,00 2,00 0,16 0,74 3,00 0,52 0,00 0,01 0,04 0,12
43 3,00 2,00 11,55 7,00 5,00 0,11 1,00 2,00 0,45 0,01 0,06 0,09 0,13
44 3,00 3,00 10,00 1,00 1,00 0,23 0,38 1,00 0,34 0,05 0,23 0,31 0,09
45 4,00 3,00 9,09 2,00 1,00 0,13 0,90 3,00 0,34 0,02 0,12 0,25 0,14
46 6,00 3,00 7,25 9,00 2,00 0,19 0,83 4,00 0,56 0,00 0,00 0,06 0,22
47 7,00 2,00 7,34 2,00 2,00 0,13 0,85 4,00 0,49 0,11 0,04 0,18 0,11
48 4,00 2,00 10,53 2,00 1,00 0,14 0,96 3,00 0,33 0,02 0,14 0,28 0,07
49 3,00 2,00 9,34 7,00 1,00 0,20 0,97 3,00 0,51 0,03 0,00 0,04 0,12
50 4,00 2,00 9,79 2,00 1,00 0,16 0,72 2,00 0,32 0,07 0,11 0,27 0,10
51 3,00 2,00 8,72 12,00 10,00 0,16 0,88 3,00 0,57 0,04 0,02 0,16 0,20
52 3,00 2,00 10,76 5,00 7,00 0,12 0,92 3,00 0,43 0,03 0,04 0,11 0,18
53 3,00 3,00 10,88 12,00 10,00 0,18 0,64 2,00 0,43 0,00 0,04 0,16 0,08
54 4,00 2,00 7,67 5,00 1,00 0,15 0,94 4,00 0,52 0,02 0,03 0,08 0,24
55 4,00 3,00 13,57 12,00 9,00 0,14 0,71 2,00 0,39 0,03 0,07 0,16 0,13
56 4,00 3,00 7,39 12,00 10,00 0,10 0,92 4,00 0,48 0,09 0,07 0,16 0,14
57 3,00 2,00 9,73 4,00 3,00 0,11 0,85 2,00 0,46 0,04 0,04 0,09 0,17
58 3,00 1,00 10,95 1,00 2,00 0,19 0,53 2,00 0,34 0,01 0,19 0,24 0,15
59 3,00 2,00 9,91 12,00 5,00 0,27 0,42 2,00 0,60 0,00 0,00 0,06 0,09
60 4,00 2,00 8,30 12,00 9,00 0,15 0,77 3,00 0,58 0,04 0,04 0,11 0,23
61 7,00 3,00 8,22 2,00 1,00 0,11 0,94 4,00 0,41 0,05 0,05 0,17 0,20
62 4,00 3,00 9,84 0,00 3,00 0,13 0,87 2,00 0,57 0,13 0,02 0,09 0,19
63 3,00 2,00 10,74 12,00 10,00 0,13 0,68 1,00 0,48 0,06 0,05 0,14 0,12
64 4,00 2,00 10,00 3,00 1,00 0,09 0,96 0,00 0,38 0,01 0,03 0,11 0,16
65 3,00 3,00 10,18 3,00 1,00 0,17 0,71 3,00 0,47 0,01 0,02 0,12 0,29
66 6,00 1,00 7,38 1,00 2,00 0,20 0,42 1,00 0,40 0,08 0,20 0,26 0,12
67 3,00 2,00 8,43 7,00 3,00 0,11 0,92 3,00 0,46 0,02 0,07 0,15 0,19
68 2,00 2,00 10,11 3,00 2,00 0,14 0,98 3,00 0,47 0,01 0,00 0,08 0,27
69 4,00 2,00 11,18 1,00 11,00 0,09 0,96 1,00 0,35 0,02 0,09 0,17 0,08
70 6,00 2,00 8,73 2,00 5,00 0,14 0,74 2,00 0,40 0,02 0,08 0,22 0,13
71 6,00 1,00 12,45 12,00 7,00 0,12 0,69 1,00 0,34 0,02 0,06 0,14 0,06
72 4,00 1,00 7,98 3,00 2,00 0,14 0,84 4,00 0,48 0,02 0,05 0,09 0,26
73 4,00 2,00 9,14 2,00 1,00 0,14 0,84 4,00 0,40 0,06 0,12 0,25 0,06
74 4,00 2,00 8,90 2,00 5,00 0,09 0,94 1,00 0,40 0,04 0,05 0,14 0,14
75 4,00 2,00 9,47 12,00 12,00 0,13 0,82 4,00 0,42 0,11 0,10 0,19 0,12
76 4,00 3,00 7,97 10,00 5,00 0,11 0,89 3,00 0,44 0,00 0,08 0,13 0,16
77 3,00 1,00 7,27 3,00 1,00 0,19 0,61 4,00 0,52 0,05 0,07 0,16 0,30
78 6,00 3,00 10,14 10,00 3,00 0,11 0,91 3,00 0,50 0,01 0,07 0,17 0,12
79 3,00 3,00 10,50 12,00 10,00 0,09 0,94 1,00 0,40 0,03 0,06 0,15 0,12
80 3,00 1,00 8,99 1,00 6,00 0,12 0,76 2,00 0,42 0,01 0,12 0,21 0,13
81 4,00 1,00 9,19 7,00 2,00 0,11 0,90 4,00 0,45 0,09 0,04 0,07 0,12
82 3,00 2,00 10,70 2,00 8,00 0,16 0,49 1,00 0,34 0,05 0,03 0,20 0,08
83 4,00 2,00 9,50 4,00 1,00 0,11 0,79 1,00 0,48 0,07 0,01 0,05 0,20
84 4,00 2,00 8,28 2,00 3,00 0,17 0,72 2,00 0,32 0,03 0,08 0,25 0,10
85 6,00 2,00 11,74 2,00 1,00 0,10 0,95 2,00 0,38 0,01 0,06 0,15 0,14
86 2,00 3,00 11,93 12,00 5,00 0,15 0,65 2,00 0,51 0,01 0,02 0,08 0,11
87 4,00 2,00 10,07 2,00 3,00 0,13 0,61 1,00 0,32 0,01 0,07 0,19 0,10
88 4,00 1,00 9,14 12,00 6,00 0,15 0,96 3,00 0,48 0,03 0,02 0,05 0,14
89 6,00 2,00 9,38 12,00 8,00 0,12 0,91 3,00 0,55 0,02 0,02 0,12 0,19
90 3,00 1,00 11,73 12,00 7,00 0,13 0,80 2,00 0,42 0,05 0,04 0,10 0,11
91 3,00 1,00 12,29 2,00 1,00 0,17 0,63 2,00 0,28 0,02 0,11 0,27 0,11
92 3,00 2,00 12,95 12,00 10,00 0,13 0,66 1,00 0,39 0,03 0,05 0,14 0,06
93 5,00 2,00 8,04 2,00 1,00 0,15 0,73 3,00 0,41 0,05 0,11 0,26 0,14
94 4,00 2,00 6,79 2,00 1,00 0,12 0,90 4,00 0,40 0,10 0,10 0,22 0,15
95 3,00 1,00 9,07 5,00 4,00 0,11 1,00 4,00 0,41 0,01 0,06 0,11 0,15
96 4,00 2,00 8,91 2,00 1,00 0,28 0,30 1,00 0,36 0,00 0,08 0,36 0,10
97 7,00 2,00 8,42 3,00 1,00 0,19 0,69 3,00 0,52 0,02 0,02 0,05 0,32
98 4,00 1,00 11,84 0,00 2,00 0,10 0,80 1,00 0,35 0,10 0,05 0,13 0,07
99 3,00 2,00 10,59 12,00 3,00 0,18 0,60 2,00 0,48 0,00 0,02 0,03 0,14
100 4,00 1,00 9,22 2,00 1,00 0,14 0,81 2,00 0,41 0,05 0,06 0,20 0,16
6
Melodies
Melodies Melodies Melodies Melodies Melodies
Melodies Corpus
Melodic Melodic Direction of Duration of Size of Corpus Corpus Corpus Corpus Corpus
Fado Edition Melodic Fifths Corpus Note Average Time
Tritones Octaves Motion Melodic Arcs Melodic Arcs Average Note Variability of Maximum Minimum Staccato
Density Between
Duration Note Duration Note Duration Note Duration Incidence
Attacks
1 0,00 0,00 0,18 0,86 3,84 51,56 2,90 0,31 0,13 0,69 0,11 0,00 0,33
2 0,06 0,00 0,12 0,75 1,99 20,84 2,74 0,32 0,13 0,95 0,22 0,00 0,35
3 0,16 0,00 0,10 0,90 5,30 69,74 2,07 0,40 0,15 0,95 0,22 0,00 0,47
4 0,15 0,00 0,04 0,65 2,45 27,40 1,87 0,47 0,37 1,38 0,22 0,00 0,50
5 0,10 0,00 0,15 0,92 5,96 50,23 1,98 0,46 0,25 1,06 0,25 0,00 0,48
6 0,17 0,02 0,12 0,82 2,75 20,60 1,97 0,46 0,35 2,41 0,22 0,00 0,49
7 0,06 0,06 0,15 0,63 1,61 19,81 2,12 0,43 0,21 1,38 0,22 0,00 0,47
8 0,15 0,02 0,15 0,87 3,88 40,02 2,48 0,34 0,17 0,95 0,19 0,00 0,39
9 0,07 0,00 0,04 0,88 4,24 74,10 3,42 0,27 0,16 0,74 0,09 0,04 0,29
10 0,02 0,02 0,10 0,88 4,44 79,46 2,55 0,35 0,19 0,93 0,19 0,00 0,38
11 0,03 0,00 0,16 0,93 7,60 75,17 2,18 0,39 0,17 0,74 0,20 0,00 0,44
12 0,14 0,04 0,06 0,82 2,72 23,89 2,26 0,36 0,12 0,53 0,11 0,00 0,43
13 0,01 0,00 0,05 0,70 2,43 20,57 2,24 0,41 0,30 1,40 0,09 0,01 0,42
14 0,13 0,00 0,12 0,85 3,35 33,20 2,03 0,41 0,16 1,38 0,22 0,00 0,46
15 0,13 0,02 0,02 0,62 1,41 14,76 2,94 0,32 0,20 1,17 0,09 0,01 0,33
16 0,12 0,00 0,11 0,87 3,86 48,07 2,58 0,32 0,11 0,48 0,22 0,00 0,37
17 0,04 0,04 0,04 0,88 4,38 41,72 2,19 0,40 0,20 1,17 0,22 0,00 0,43
18 0,11 0,02 0,17 0,76 2,56 19,78 2,78 0,33 0,22 1,17 0,13 0,00 0,34
19 0,09 0,00 0,17 0,80 2,86 20,32 2,64 0,32 0,14 0,53 0,12 0,00 0,36
20 0,09 0,03 0,24 0,81 2,67 23,56 1,88 0,44 0,26 0,95 0,22 0,00 0,46
21 0,11 0,06 0,13 0,61 1,81 20,17 2,16 0,39 0,23 1,38 0,19 0,00 0,43
22 0,15 0,00 0,18 0,87 3,98 46,35 2,06 0,41 0,24 0,95 0,22 0,00 0,46
23 0,14 0,00 0,09 0,86 3,59 44,89 2,11 0,39 0,17 1,38 0,19 0,00 0,46
24 0,14 0,02 0,13 0,77 2,24 17,63 2,39 0,39 0,26 0,95 0,07 0,07 0,40
25 0,09 0,05 0,05 0,67 1,98 21,29 2,25 0,37 0,19 0,84 0,20 0,00 0,41
26 0,13 0,02 0,07 0,83 3,10 33,00 2,00 0,42 0,18 0,90 0,22 0,00 0,45
27 0,10 0,00 0,04 0,71 3,48 38,19 4,82 0,20 0,10 0,61 0,10 0,00 0,19
28 0,04 0,02 0,19 0,91 5,68 68,50 2,35 0,37 0,16 0,93 0,20 0,00 0,40
29 0,15 0,00 0,16 0,83 3,33 34,72 2,49 0,30 0,18 1,18 0,09 0,02 0,39
30 0,11 0,01 0,13 0,71 1,96 22,01 1,88 0,43 0,28 1,64 0,07 0,02 0,50
31 0,03 0,00 0,11 0,72 1,80 21,74 1,90 0,43 0,18 0,95 0,22 0,00 0,50
32 0,12 0,04 0,09 0,81 3,06 29,65 2,13 0,42 0,20 0,98 0,22 0,00 0,45
33 0,15 0,04 0,13 0,85 3,45 37,40 2,06 0,41 0,21 1,38 0,20 0,00 0,45
34 0,06 0,00 0,09 0,87 3,82 45,46 2,00 0,41 0,19 0,95 0,22 0,00 0,47
35 0,17 0,00 0,17 0,90 4,96 45,62 2,20 0,43 0,21 0,95 0,20 0,00 0,44
36 0,16 0,04 0,07 0,72 1,80 18,78 2,25 0,41 0,21 1,23 0,20 0,00 0,41
37 0,01 0,00 0,06 0,83 2,96 43,12 2,00 0,41 0,16 0,95 0,22 0,00 0,47
38 0,13 0,02 0,15 0,74 1,98 18,24 2,25 0,41 0,17 1,03 0,20 0,00 0,41
39 0,01 0,00 0,02 0,82 3,07 49,78 2,19 0,37 0,15 0,71 0,09 0,06 0,43
40 0,02 0,09 0,11 0,70 2,17 24,17 2,06 0,43 0,23 1,15 0,22 0,00 0,46
41 0,04 0,00 0,11 0,67 1,72 21,24 2,06 0,41 0,15 0,93 0,22 0,00 0,46
42 0,16 0,00 0,11 0,85 3,29 33,61 2,36 0,35 0,22 1,48 0,09 0,04 0,40
43 0,11 0,06 0,11 0,68 1,80 20,97 1,97 0,46 0,25 1,38 0,22 0,00 0,48
44 0,02 0,00 0,08 0,66 1,71 17,92 3,13 0,28 0,15 0,95 0,09 0,08 0,30
45 0,03 0,03 0,11 0,70 1,89 17,43 2,42 0,36 0,17 0,93 0,09 0,03 0,39
46 0,16 0,00 0,08 0,79 2,46 17,71 2,29 0,41 0,25 1,38 0,09 0,01 0,41
47 0,05 0,04 0,11 0,77 2,16 17,84 2,29 0,39 0,22 1,16 0,19 0,00 0,41
48 0,02 0,03 0,09 0,77 2,35 22,91 4,75 0,20 0,10 0,56 0,09 0,09 0,20
49 0,20 0,02 0,05 0,89 4,58 43,74 2,13 0,40 0,17 0,95 0,22 0,00 0,44
50 0,04 0,02 0,02 0,59 2,39 25,22 2,22 0,35 0,31 1,67 0,07 0,22 0,43
51 0,06 0,03 0,16 0,67 2,08 18,69 2,63 0,31 0,35 1,88 0,05 0,26 0,37
52 0,05 0,04 0,11 0,86 3,67 39,85 2,11 0,43 0,31 1,64 0,19 0,00 0,45
53 0,06 0,04 0,18 0,63 1,60 17,32 2,44 0,38 0,44 1,67 0,09 0,07 0,39
54 0,09 0,06 0,09 0,77 2,37 18,13 1,94 0,45 0,37 1,41 0,11 0,00 0,48
55 0,04 0,01 0,14 0,70 1,84 25,58 2,09 0,45 0,37 2,38 0,19 0,00 0,46
56 0,09 0,05 0,10 0,66 2,03 16,25 2,55 0,37 0,29 1,63 0,10 0,03 0,37
57 0,09 0,05 0,07 0,75 2,74 27,74 2,02 0,44 0,34 1,88 0,09 0,04 0,48
58 0,03 0,00 0,05 0,70 2,06 22,58 2,84 0,32 0,24 1,41 0,07 0,02 0,34
59 0,11 0,02 0,27 0,91 5,74 52,38 1,81 0,48 0,29 1,41 0,22 0,00 0,52
60 0,04 0,07 0,15 0,60 1,76 15,20 2,56 0,34 0,41 2,00 0,09 0,17 0,38
61 0,07 0,07 0,03 0,62 1,69 14,48 2,92 0,31 0,28 1,48 0,22 0,00 0,33
62 0,08 0,01 0,08 0,74 2,56 28,89 2,15 0,45 0,46 2,35 0,06 0,06 0,45
63 0,05 0,03 0,13 0,70 1,89 21,36 3,08 0,28 0,17 1,20 0,06 0,02 0,32
64 0,08 0,08 0,04 0,59 1,92 19,23 2,06 0,46 0,49 2,11 0,15 0,00 0,48
65 0,03 0,03 0,07 0,69 2,24 23,07 1,86 0,48 0,38 2,02 0,12 0,00 0,52
66 0,08 0,04 0,07 0,71 2,49 19,71 2,89 0,29 0,27 1,56 0,09 0,01 0,32
67 0,11 0,04 0,09 0,63 2,24 19,11 2,54 0,37 0,44 3,22 0,07 0,03 0,37
68 0,05 0,01 0,06 0,83 3,42 34,49 1,87 0,49 0,30 2,09 0,20 0,00 0,51
69 0,06 0,02 0,09 0,65 2,02 22,80 2,08 0,45 0,36 1,38 0,09 0,01 0,46
70 0,11 0,02 0,08 0,60 2,87 25,27 2,58 0,35 0,34 1,87 0,06 0,06 0,38
71 0,03 0,01 0,12 0,70 1,96 24,73 2,20 0,43 0,47 1,96 0,10 0,00 0,45
72 0,05 0,05 0,08 0,78 2,42 19,35 1,98 0,47 0,31 2,09 0,07 0,02 0,48
73 0,09 0,02 0,10 0,72 2,21 20,90 2,05 0,43 0,31 1,45 0,22 0,00 0,47
74 0,09 0,07 0,06 0,69 1,83 16,86 2,73 0,28 0,21 1,18 0,09 0,16 0,35
75 0,03 0,04 0,13 0,76 2,20 22,27 2,52 0,35 0,20 0,78 0,08 0,11 0,38
76 0,10 0,09 0,04 0,71 1,73 13,45 2,78 0,34 0,18 1,01 0,20 0,00 0,33
77 0,05 0,01 0,06 0,70 2,45 18,63 3,31 0,28 0,22 1,41 0,05 0,09 0,29
78 0,10 0,04 0,07 0,51 2,78 28,22 1,93 0,49 0,73 3,27 0,10 0,01 0,50
79 0,07 0,05 0,09 0,66 1,83 19,61 2,02 0,46 0,50 3,75 0,12 0,00 0,48
80 0,09 0,04 0,09 0,64 1,82 16,34 2,00 0,42 0,37 1,45 0,09 0,03 0,47
81 0,11 0,06 0,10 0,63 1,64 16,49 2,77 0,33 0,24 2,05 0,19 0,00 0,34
82 0,04 0,05 0,04 0,71 2,15 23,60 2,43 0,38 0,30 1,64 0,09 0,03 0,41
83 0,06 0,05 0,04 0,77 2,50 24,43 2,00 0,42 0,27 1,45 0,09 0,02 0,46
84 0,06 0,02 0,04 0,59 2,06 17,44 4,18 0,21 0,19 1,17 0,06 0,18 0,23
85 0,08 0,04 0,06 0,60 2,85 33,62 3,07 0,31 0,35 1,88 0,06 0,18 0,32
86 0,10 0,00 0,15 0,76 2,61 31,12 2,55 0,36 0,24 1,59 0,06 0,05 0,39
87 0,07 0,07 0,06 0,61 2,61 26,21 3,14 0,31 0,32 1,88 0,06 0,26 0,31
88 0,07 0,14 0,15 0,70 1,69 15,97 5,03 0,18 0,37 4,97 0,05 0,70 0,17
89 0,09 0,04 0,12 0,77 2,72 25,74 2,02 0,39 0,21 0,95 0,19 0,00 0,48
90 0,07 0,02 0,13 0,78 2,61 31,88 2,28 0,44 0,31 1,63 0,12 0,00 0,42
91 0,06 0,01 0,02 0,58 2,07 25,81 1,73 0,57 0,60 3,74 0,19 0,00 0,56
92 0,08 0,00 0,13 0,81 3,41 45,63 4,97 0,19 0,10 0,87 0,08 0,01 0,19
93 0,04 0,04 0,08 0,69 2,32 19,49 2,85 0,30 0,28 1,63 0,05 0,13 0,34
94 0,04 0,07 0,07 0,67 2,14 16,14 3,12 0,29 0,28 2,11 0,06 0,04 0,32
95 0,07 0,02 0,11 0,74 2,12 19,12 2,98 0,32 0,18 0,94 0,17 0,00 0,32
96 0,05 0,01 0,08 0,68 1,88 16,74 1,73 0,55 0,42 1,87 0,10 0,02 0,55
97 0,05 0,03 0,07 0,75 3,04 24,92 2,10 0,45 0,34 1,41 0,20 0,00 0,45
98 0,03 0,02 0,07 0,73 1,91 24,94 3,40 0,26 0,22 1,80 0,08 0,11 0,29
99 0,08 0,03 0,18 0,80 2,46 26,04 2,32 0,39 0,55 3,75 0,13 0,00 0,42
100 0,03 0,06 0,09 0,54 2,27 21,72 3,17 0,31 0,33 1,62 0,09 0,05 0,31
7
Melodies
Melodies Melodies Melodies
Melodies Corpus Melodies Melodies Melodies Melodies Melodies
Corpus Melodies Corpus Corpus
Corpus Average Melodies Corpus Corpus Corpus Most Corpus Most Corpus
Average Time Corpus Average Note Relative
Fado Edition Variability of Variability of Corpus Initial Overall Variation of Common Common Relative
Between Variation of To Note Strength of
Time Between Time Between Tempo Dynamic Dynamics In Pitch Pitch Class Strength of
Attacks For Dynamics Dynamics Top Pitch
Attacks Attacks For Range Each Voice Prevalence Prevalence Top Pitches
Each Voice Change Classes
Each Voice
1 0,14 0,33 0,14 64,00 14,00 5,74 5,74 8,08 0,21 0,21 0,79 1,00
2 0,17 0,35 0,17 64,00 14,00 5,79 5,79 7,79 0,32 0,32 0,65 0,73
3 0,23 0,47 0,23 64,00 18,00 6,09 6,09 8,33 0,14 0,29 1,00 0,89
4 0,42 0,50 0,42 64,00 14,00 6,15 6,15 9,00 0,21 0,28 1,00 0,75
5 0,26 0,48 0,26 56,00 14,00 6,02 6,02 7,71 0,24 0,24 0,76 0,76
6 0,39 0,49 0,39 64,00 18,00 6,29 6,29 8,94 0,27 0,27 0,71 0,71
7 0,28 0,47 0,28 64,00 14,00 6,04 6,04 8,80 0,20 0,20 0,80 0,84
8 0,28 0,39 0,28 120,00 18,00 5,73 5,73 6,81 0,25 0,30 0,78 0,64
9 0,17 0,29 0,17 64,00 18,00 3,73 3,73 3,89 0,17 0,25 0,80 0,76
10 0,21 0,38 0,21 64,00 52,00 16,57 16,57 5,11 0,15 0,21 0,93 0,75
11 0,23 0,44 0,23 64,00 18,00 6,11 6,11 8,83 0,22 0,22 0,89 0,89
12 0,23 0,43 0,23 64,00 18,00 5,97 5,97 7,81 0,20 0,22 0,71 0,73
13 0,31 0,42 0,31 64,00 10,00 2,70 2,70 3,14 0,19 0,19 0,93 0,93
14 0,20 0,46 0,20 64,00 18,00 6,16 6,16 8,54 0,19 0,19 0,92 0,92
15 0,20 0,33 0,20 64,00 19,00 5,44 5,44 5,56 0,20 0,24 0,69 0,84
16 0,16 0,37 0,16 64,00 14,00 5,80 5,80 8,18 0,23 0,23 0,67 0,78
17 0,25 0,43 0,25 64,00 14,00 6,10 6,10 8,71 0,19 0,19 0,62 0,92
18 0,21 0,34 0,21 64,00 18,00 5,41 5,41 5,61 0,20 0,22 0,92 0,86
19 0,14 0,36 0,14 120,00 8,00 2,07 2,07 2,18 0,45 0,45 0,20 0,33
20 0,32 0,46 0,32 64,00 14,00 6,26 6,26 9,40 0,33 0,33 0,40 0,40
21 0,30 0,43 0,30 64,00 18,00 5,96 5,96 6,84 0,25 0,25 0,59 0,65
22 0,28 0,46 0,28 64,00 14,00 6,15 6,15 7,00 0,34 0,41 0,45 0,38
23 0,25 0,46 0,25 64,00 30,00 7,98 7,98 8,99 0,20 0,20 0,84 0,95
24 0,28 0,40 0,28 120,00 18,00 5,44 5,44 5,78 0,19 0,29 0,95 0,66
25 0,24 0,41 0,24 72,00 14,00 6,12 6,12 8,95 0,32 0,32 0,70 0,70
26 0,23 0,45 0,23 120,00 19,00 5,79 5,79 7,50 0,31 0,38 0,70 0,67
27 0,14 0,19 0,14 72,00 18,00 4,39 4,39 3,85 0,20 0,20 0,89 0,89
28 0,20 0,40 0,20 64,00 18,00 5,87 5,87 7,78 0,15 0,25 1,00 0,67
29 0,29 0,39 0,29 64,00 18,00 5,79 5,79 7,51 0,24 0,28 0,37 0,61
30 0,46 0,50 0,46 120,00 24,00 6,74 6,74 6,17 0,22 0,24 0,70 0,77
31 0,29 0,50 0,29 64,00 18,00 6,20 6,20 8,62 0,26 0,31 0,42 0,57
32 0,21 0,45 0,21 64,00 52,00 18,43 18,43 9,97 0,27 0,27 0,67 0,78
33 0,21 0,45 0,21 64,00 36,00 9,38 9,38 7,31 0,25 0,28 0,69 0,61
34 0,26 0,47 0,26 64,00 18,00 6,12 6,12 8,77 0,26 0,26 0,63 0,81
35 0,21 0,44 0,21 64,00 18,00 6,14 6,14 9,07 0,28 0,28 0,57 0,57
36 0,18 0,41 0,18 120,00 18,00 5,91 5,91 7,68 0,22 0,25 0,79 0,75
37 0,24 0,47 0,24 120,00 18,00 6,22 6,22 9,00 0,19 0,19 0,92 0,92
38 0,15 0,41 0,15 120,00 14,00 6,01 6,01 6,67 0,30 0,30 0,58 0,58
39 0,23 0,43 0,23 64,00 18,00 5,79 5,79 7,38 0,24 0,24 1,00 1,00
40 0,25 0,46 0,25 64,00 18,00 6,00 6,00 7,91 0,20 0,20 0,92 0,92
41 0,22 0,46 0,22 64,00 14,00 6,10 6,10 9,09 0,31 0,31 0,55 0,60
42 0,34 0,40 0,34 64,00 18,00 5,93 5,93 7,47 0,26 0,26 0,48 0,65
43 0,26 0,48 0,26 120,00 18,00 6,05 6,05 8,26 0,15 0,21 0,78 0,85
44 0,19 0,30 0,19 120,00 19,00 6,11 6,11 7,18 0,23 0,23 1,00 1,00
45 0,22 0,39 0,22 64,00 18,00 5,89 5,89 7,01 0,21 0,21 0,88 0,88
46 0,23 0,41 0,23 120,00 18,00 5,66 5,66 6,31 0,25 0,28 0,78 0,70
47 0,22 0,41 0,22 120,00 18,00 5,99 5,99 8,11 0,27 0,27 0,53 0,53
48 0,10 0,20 0,10 120,00 52,00 13,84 13,84 7,14 0,19 0,21 0,69 0,81
49 0,23 0,44 0,23 120,00 14,00 5,97 5,97 8,30 0,27 0,27 0,67 0,67
50 0,46 0,43 0,46 108,00 36,00 9,76 9,76 1,83 0,23 0,23 0,84 0,84
51 0,47 0,37 0,47 64,00 38,00 6,83 6,83 6,34 0,23 0,23 0,71 1,00
52 0,47 0,45 0,47 64,00 43,00 8,61 8,61 4,28 0,30 0,32 0,45 0,46
53 0,45 0,39 0,45 108,00 14,00 5,90 5,90 6,36 0,24 0,31 0,76 0,59
54 0,42 0,48 0,42 64,00 14,00 5,77 5,77 7,50 0,30 0,30 0,61 0,61
55 0,39 0,46 0,39 72,00 14,00 5,85 5,85 7,49 0,20 0,21 0,83 0,85
56 0,31 0,37 0,31 64,00 18,00 4,10 4,10 3,95 0,32 0,32 0,56 0,56
57 0,46 0,48 0,46 64,00 52,00 13,20 13,20 2,72 0,17 0,19 0,76 0,92
58 0,24 0,34 0,24 64,00 31,00 6,02 6,02 5,61 0,21 0,27 0,56 0,43
59 0,34 0,52 0,34 72,00 35,00 9,97 9,97 4,95 0,20 0,20 0,73 0,73
60 0,49 0,38 0,49 108,00 14,00 5,68 5,68 5,94 0,15 0,21 0,94 0,73
61 0,31 0,33 0,31 64,00 18,00 5,38 5,38 5,79 0,15 0,16 0,96 1,00
62 0,49 0,45 0,49 64,00 30,00 6,81 6,81 2,43 0,20 0,21 0,88 0,96
63 0,27 0,32 0,27 74,00 30,00 7,03 7,03 5,77 0,19 0,23 0,49 0,60
64 0,50 0,48 0,50 128,00 18,00 5,36 5,36 5,17 0,25 0,25 0,60 0,60
65 0,46 0,52 0,46 64,00 72,00 15,27 15,27 7,32 0,14 0,17 0,76 0,76
66 0,39 0,32 0,39 64,00 18,00 5,58 5,58 6,75 0,23 0,23 0,67 0,71
67 0,43 0,37 0,43 64,00 59,00 16,96 16,96 6,46 0,17 0,21 0,71 0,76
68 0,37 0,51 0,37 64,00 64,00 14,27 14,27 6,47 0,19 0,21 0,88 0,96
69 0,43 0,46 0,43 108,00 39,00 8,72 8,72 3,91 0,12 0,17 0,94 0,81
70 0,45 0,38 0,45 64,00 18,00 4,95 4,95 4,55 0,21 0,21 0,63 0,79
71 0,53 0,45 0,53 91,00 33,00 6,54 6,54 6,61 0,27 0,27 0,67 0,67
72 0,36 0,48 0,36 64,00 18,00 6,04 6,04 7,94 0,17 0,19 1,00 0,96
73 0,36 0,47 0,36 64,00 14,00 6,11 6,11 8,80 0,17 0,17 1,00 1,00
74 0,29 0,35 0,29 64,00 72,00 18,81 18,81 6,87 0,22 0,23 0,55 0,69
75 0,33 0,38 0,33 64,00 69,00 19,18 19,18 5,77 0,23 0,23 0,71 0,71
76 0,20 0,33 0,20 64,00 38,00 3,90 3,90 2,46 0,19 0,19 0,92 0,92
77 0,26 0,29 0,26 64,00 27,00 5,78 5,78 6,23 0,16 0,23 0,73 0,70
78 0,74 0,50 0,74 91,00 14,00 5,70 5,70 7,40 0,22 0,22 0,94 0,94
79 0,56 0,48 0,56 64,00 56,00 13,26 13,26 4,55 0,20 0,20 0,71 0,85
80 0,46 0,47 0,46 64,00 18,00 5,62 5,62 6,82 0,17 0,20 0,92 0,80
81 0,27 0,34 0,27 64,00 67,00 17,37 17,37 9,50 0,19 0,21 0,65 0,92
82 0,35 0,41 0,35 64,00 23,00 5,34 5,34 5,84 0,26 0,26 0,97 0,97
83 0,33 0,46 0,33 64,00 39,00 9,07 9,07 6,06 0,15 0,21 0,80 0,86
84 0,25 0,23 0,25 64,00 40,00 11,68 11,68 3,49 0,19 0,20 0,77 0,71
85 0,35 0,32 0,35 128,00 40,00 7,22 7,22 5,08 0,17 0,18 0,96 0,96
86 0,30 0,39 0,30 128,00 18,00 5,95 5,95 5,65 0,21 0,22 1,00 0,96
87 0,32 0,31 0,32 128,00 18,00 5,18 5,18 5,09 0,17 0,19 0,67 0,73
88 0,30 0,17 0,30 64,00 60,00 17,55 17,55 5,51 0,21 0,34 0,62 0,40
89 0,31 0,48 0,31 64,00 39,00 7,80 7,80 8,95 0,13 0,21 0,75 0,96
90 0,31 0,42 0,31 64,00 18,00 6,90 6,90 2,55 0,21 0,21 0,64 1,00
91 0,61 0,56 0,61 72,00 41,00 11,03 11,03 2,08 0,28 0,28 0,60 0,60
92 0,10 0,19 0,10 108,00 18,00 5,05 5,05 5,11 0,16 0,27 0,62 0,61
93 0,41 0,34 0,41 64,00 52,00 14,15 14,15 6,14 0,18 0,20 0,60 0,69
94 0,36 0,32 0,36 64,00 18,00 5,28 5,28 5,65 0,17 0,22 0,90 0,72
95 0,19 0,32 0,19 95,00 18,00 5,61 5,61 6,25 0,16 0,18 1,00 0,91
96 0,44 0,55 0,44 64,00 28,00 7,49 7,49 7,58 0,25 0,25 0,77 0,77
97 0,37 0,45 0,37 64,00 18,00 5,49 5,49 5,30 0,21 0,21 0,92 0,92
98 0,27 0,29 0,27 91,00 17,00 5,93 5,93 7,16 0,18 0,18 0,81 1,00
99 0,56 0,42 0,56 64,00 18,00 4,68 4,68 4,35 0,40 0,41 0,29 0,32
100 0,33 0,31 0,33 91,00 36,00 11,76 11,76 4,88 0,22 0,23 0,52 0,70
8
Melodies Melodies
Melodies Corpus Melodies Melodies Melodies Melodies
Melodies Corpus Corpus Melodies Melodies Corpus Melodies Corpus
Interval Between Melodies Corpus Most Corpus Corpus Corpus Most
Fado Edition Interval Between Number of Corpus Pitch Corpus Pitch Importance of Importance of High
Strongest Pitch Corpus Range Common Primary Importance of Common
Strongest Pitches Common Variety Class Variety Middle Register
Classes Pitch Register Bass Register Pitch Class
Pitches Register
1 4,00 5,00 5,00 11,00 8,00 24,00 0,59 72,00 0,00 0,52 0,48 4,00
2 5,00 7,00 3,00 10,00 8,00 14,00 0,52 68,00 0,00 0,96 0,04 7,00
3 12,00 5,00 4,00 19,00 8,00 27,00 0,50 71,00 0,00 0,56 0,44 4,00
4 4,00 4,00 4,00 9,00 8,00 15,00 0,56 72,00 0,00 0,55 0,45 4,00
5 4,00 4,00 5,00 8,00 8,00 11,00 0,59 72,00 0,00 0,57 0,43 4,00
6 2,00 2,00 5,00 9,00 8,00 12,00 0,56 73,00 0,00 0,44 0,56 0,00
7 4,00 1,00 4,00 13,00 9,00 17,00 0,63 77,00 0,00 0,20 0,80 8,00
8 4,00 4,00 3,00 13,00 9,00 19,00 0,59 72,00 0,00 0,46 0,54 4,00
9 5,00 5,00 4,00 16,00 10,00 20,00 0,59 72,00 0,00 0,55 0,45 4,00
10 5,00 7,00 4,00 14,00 7,00 24,00 0,62 76,00 0,00 0,25 0,75 7,00
11 4,00 11,00 5,00 10,00 7,00 16,00 0,55 68,00 0,00 0,93 0,07 11,00
12 2,00 2,00 4,00 16,00 11,00 22,00 0,62 78,00 0,00 0,14 0,86 7,00
13 2,00 2,00 4,00 15,00 10,00 18,00 0,52 67,00 0,00 0,89 0,11 7,00
14 3,00 3,00 5,00 10,00 8,00 14,00 0,52 67,00 0,00 0,97 0,03 7,00
15 2,00 7,00 6,00 12,00 9,00 15,00 0,56 73,00 0,00 0,50 0,50 7,00
16 2,00 7,00 6,00 9,00 7,00 14,00 0,52 67,00 0,00 0,98 0,03 7,00
17 4,00 4,00 4,00 12,00 7,00 19,00 0,54 69,00 0,00 0,76 0,24 9,00
18 5,00 7,00 5,00 9,00 8,00 15,00 0,59 72,00 0,00 0,59 0,41 4,00
19 5,00 7,00 4,00 9,00 7,00 13,00 0,52 66,00 0,00 1,00 0,00 7,00
20 5,00 7,00 4,00 8,00 7,00 12,00 0,52 66,00 0,00 1,00 0,00 7,00
21 3,00 9,00 4,00 11,00 7,00 17,00 0,54 68,00 0,00 0,91 0,09 9,00
22 1,00 2,00 5,00 9,00 6,00 17,00 0,56 70,00 0,00 0,84 0,16 0,00
23 3,00 9,00 4,00 13,00 10,00 16,00 0,54 68,00 0,00 0,88 0,12 9,00
24 7,00 5,00 5,00 11,00 9,00 13,00 0,54 70,00 0,00 0,67 0,33 4,00
25 3,00 3,00 3,00 8,00 7,00 12,00 0,52 64,00 0,00 1,00 0,00 7,00
26 2,00 2,00 3,00 11,00 8,00 16,00 0,54 66,00 0,00 1,00 0,00 9,00
27 4,00 4,00 5,00 12,00 10,00 15,00 0,56 74,00 0,00 0,41 0,59 0,00
28 9,00 3,00 7,00 13,00 10,00 17,00 0,52 73,00 0,00 0,48 0,52 7,00
29 5,00 5,00 1,00 25,00 11,00 52,00 0,59 75,00 0,06 0,22 0,72 4,00
30 5,00 7,00 4,00 13,00 9,00 19,00 0,50 65,00 0,00 0,98 0,02 4,00
31 1,00 2,00 4,00 11,00 7,00 17,00 0,56 69,00 0,00 0,80 0,20 0,00
32 3,00 3,00 6,00 10,00 8,00 16,00 0,50 64,00 0,00 1,00 0,00 4,00
33 5,00 7,00 5,00 13,00 9,00 16,00 0,56 68,00 0,00 0,91 0,09 0,00
34 2,00 2,00 5,00 8,00 7,00 12,00 0,58 71,00 0,00 0,58 0,42 2,00
35 5,00 2,00 5,00 9,00 8,00 13,00 0,50 61,00 0,00 1,00 0,00 4,00
36 2,00 9,00 6,00 10,00 6,00 17,00 0,48 64,00 0,00 0,97 0,03 2,00
37 5,00 2,00 6,00 8,00 7,00 12,00 0,58 71,00 0,00 0,69 0,31 2,00
38 2,00 2,00 5,00 9,00 7,00 13,00 0,58 74,00 0,00 0,24 0,76 2,00
39 2,00 2,00 5,00 7,00 7,00 9,00 0,56 72,00 0,00 0,56 0,44 0,00
40 2,00 2,00 5,00 9,00 8,00 12,00 0,52 68,00 0,00 0,91 0,09 7,00
41 7,00 7,00 4,00 11,00 9,00 13,00 0,52 65,00 0,00 1,00 0,00 7,00
42 2,00 7,00 6,00 9,00 7,00 14,00 0,52 67,00 0,00 0,98 0,02 7,00
43 2,00 2,00 4,00 15,00 8,00 22,00 0,52 70,00 0,00 0,70 0,30 7,00
44 2,00 2,00 5,00 11,00 10,00 12,00 0,58 73,00 0,00 0,33 0,67 2,00
45 2,00 2,00 6,00 9,00 8,00 14,00 0,58 73,00 0,00 0,45 0,55 2,00
46 3,00 3,00 4,00 10,00 8,00 14,00 0,52 64,00 0,00 1,00 0,00 7,00
47 5,00 5,00 4,00 11,00 8,00 15,00 0,54 67,00 0,00 0,97 0,03 9,00
48 4,00 5,00 6,00 14,00 9,00 20,00 0,59 73,00 0,00 0,47 0,53 4,00
49 1,00 11,00 5,00 9,00 9,00 11,00 0,56 72,00 0,00 0,61 0,39 0,00
50 2,00 2,00 5,00 12,00 9,00 15,00 0,58 71,00 0,00 0,57 0,43 2,00
51 2,00 5,00 4,00 10,00 7,00 16,00 0,58 73,00 0,00 0,35 0,65 2,00
52 2,00 3,00 4,00 13,00 8,00 25,00 0,52 65,00 0,04 0,96 0,00 7,00
53 2,00 2,00 5,00 10,00 7,00 17,00 0,52 65,00 0,00 1,00 0,00 7,00
54 8,00 4,00 5,00 8,00 8,00 11,00 0,62 75,00 0,00 0,37 0,63 7,00
55 2,00 5,00 5,00 13,00 9,00 19,00 0,54 72,00 0,00 0,59 0,41 9,00
56 1,00 11,00 3,00 9,00 7,00 14,00 0,55 70,00 0,00 0,86 0,14 11,00
57 2,00 5,00 4,00 18,00 12,00 19,00 0,54 70,00 0,00 0,68 0,32 4,00
58 2,00 2,00 2,00 21,00 12,00 24,00 0,50 67,00 0,01 0,84 0,16 4,00
59 2,00 2,00 3,00 14,00 11,00 15,00 0,59 75,00 0,00 0,21 0,79 4,00
60 5,00 7,00 4,00 15,00 10,00 17,00 0,55 72,00 0,00 0,55 0,45 4,00
61 2,00 11,00 5,00 14,00 10,00 17,00 0,54 67,00 0,00 0,95 0,05 0,00
62 2,00 2,00 4,00 15,00 10,00 17,00 0,54 67,00 0,00 0,95 0,05 9,00
63 1,00 7,00 2,00 18,00 11,00 22,00 0,50 64,00 0,03 0,91 0,06 4,00
64 5,00 5,00 4,00 16,00 11,00 18,00 0,50 66,00 0,00 0,98 0,02 4,00
65 2,00 2,00 2,00 22,00 11,00 24,00 0,59 73,00 0,00 0,42 0,58 4,00
66 7,00 5,00 4,00 15,00 10,00 18,00 0,54 71,00 0,00 0,64 0,36 9,00
67 1,00 4,00 5,00 16,00 10,00 20,00 0,55 74,00 0,00 0,46 0,54 11,00
68 5,00 7,00 3,00 16,00 10,00 22,00 0,52 68,00 0,00 0,89 0,11 7,00
69 2,00 5,00 4,00 24,00 11,00 29,00 0,58 74,00 0,00 0,40 0,60 4,00
70 5,00 5,00 6,00 14,00 11,00 16,00 0,54 68,00 0,00 0,88 0,12 9,00
71 2,00 2,00 5,00 9,00 9,00 10,00 0,54 67,00 0,00 1,00 0,00 9,00
72 2,00 5,00 5,00 18,00 11,00 19,00 0,54 72,00 0,00 0,51 0,49 9,00
73 5,00 7,00 7,00 12,00 10,00 15,00 0,54 71,00 0,00 0,62 0,38 2,00
74 4,00 7,00 5,00 15,00 10,00 20,00 0,50 65,00 0,00 0,98 0,02 4,00
75 2,00 2,00 5,00 10,00 8,00 14,00 0,59 74,00 0,00 0,40 0,60 4,00
76 2,00 2,00 5,00 12,00 8,00 18,00 0,52 67,00 0,00 0,94 0,06 7,00
77 10,00 2,00 5,00 15,00 11,00 16,00 0,58 71,00 0,00 0,53 0,47 4,00
78 2,00 2,00 5,00 10,00 8,00 14,00 0,58 74,00 0,00 0,26 0,74 2,00
79 2,00 5,00 5,00 17,00 12,00 17,00 0,58 72,00 0,00 0,51 0,49 2,00
80 3,00 1,00 6,00 11,00 7,00 17,00 0,50 68,00 0,00 0,84 0,16 4,00
81 2,00 5,00 5,00 13,00 9,00 19,00 0,59 74,00 0,00 0,34 0,66 4,00
82 2,00 2,00 5,00 9,00 8,00 12,00 0,58 72,00 0,00 0,56 0,44 2,00
83 2,00 7,00 5,00 14,00 9,00 20,00 0,59 74,00 0,00 0,36 0,64 4,00
84 2,00 2,00 4,00 18,00 11,00 21,00 0,59 74,00 0,00 0,31 0,69 4,00
85 2,00 2,00 5,00 14,00 10,00 17,00 0,55 67,00 0,00 0,98 0,02 9,00
86 2,00 2,00 4,00 15,00 11,00 16,00 0,51 64,00 0,00 1,00 0,00 7,00
87 1,00 2,00 4,00 17,00 12,00 17,00 0,54 66,00 0,00 0,93 0,07 9,00
88 12,00 1,00 3,00 22,00 10,00 31,00 0,59 70,00 0,00 0,55 0,45 4,00
89 4,00 5,00 2,00 18,00 8,00 29,00 0,59 69,00 0,00 0,71 0,29 4,00
90 3,00 5,00 5,00 14,00 11,00 15,00 0,54 70,00 0,00 0,74 0,26 4,00
91 2,00 2,00 5,00 14,00 12,00 13,00 0,54 68,00 0,00 0,97 0,03 9,00
92 7,00 7,00 5,00 17,00 9,00 24,00 0,52 67,00 0,00 0,75 0,25 7,00
93 3,00 3,00 2,00 22,00 12,00 26,00 0,50 65,00 0,02 0,84 0,14 4,00
94 5,00 7,00 4,00 21,00 12,00 23,00 0,52 71,00 0,00 0,70 0,30 7,00
95 2,00 2,00 5,00 14,00 11,00 15,00 0,50 67,00 0,00 0,89 0,11 4,00
96 5,00 5,00 6,00 8,00 7,00 14,00 0,50 66,00 0,00 0,98 0,02 4,00
97 2,00 2,00 4,00 12,00 10,00 14,00 0,54 68,00 0,00 0,92 0,08 9,00
98 5,00 5,00 4,00 13,00 10,00 16,00 0,50 65,00 0,00 0,93 0,07 4,00
99 2,00 5,00 3,00 14,00 9,00 19,00 0,50 64,00 0,00 0,99 0,01 4,00
100 5,00 2,00 6,00 14,00 11,00 15,00 0,55 73,00 0,00 0,47 0,53 11,00
9
Melodies Melodies
Melodies Melodies
Melodies Melodies Corpus Corpus
Melodies Melodies Corpus Most Corpus Melodies Melodies Melodies Melodies Melodies
Corpus Corpus Most Distance Relative
Corpus Corpus Common Number of Corpus Corpus Corpus Corpus Corpus
Fado Edition Average Common Between Most Strength of
Dominant Strong Tonal Melodic Common Amount of Repeated Chromatic Stepwise Melodic
Melodic Melodic Common Most
Spread Centres Interval Melodic Arpeggiation Notes Motion Motion Thirds
Interval Interval Melodic Common
Prevalence Intervals
Intervals Intervals
1 4,00 1,00 4,56 3,00 2,00 0,25 0,73 3,00 0,47 0,07 0,18 0,35 0,30
2 2,00 2,00 2,09 0,00 2,00 0,33 0,54 5,00 0,54 0,33 0,15 0,33 0,17
3 3,00 2,00 2,17 0,00 1,00 0,34 0,62 4,00 0,50 0,34 0,21 0,38 0,11
4 2,00 3,00 1,88 2,00 2,00 0,32 0,89 4,00 0,46 0,28 0,18 0,49 0,11
5 4,00 3,00 3,47 3,00 1,00 0,39 0,44 3,00 0,68 0,05 0,06 0,19 0,56
6 4,00 3,00 2,29 0,00 2,00 0,31 0,60 5,00 0,50 0,31 0,13 0,31 0,14
7 2,00 4,00 1,63 0,00 2,00 0,37 0,70 4,00 0,54 0,37 0,14 0,40 0,17
8 3,00 3,00 2,37 0,00 2,00 0,27 0,70 4,00 0,53 0,27 0,18 0,37 0,21
9 3,00 2,00 2,57 1,00 2,00 0,24 0,82 5,00 0,51 0,14 0,24 0,43 0,31
10 7,00 1,00 2,74 2,00 1,00 0,40 0,39 4,00 0,30 0,08 0,16 0,55 0,21
11 5,00 2,00 2,23 2,00 2,00 0,34 0,54 4,00 0,32 0,18 0,17 0,51 0,13
12 3,00 3,00 2,82 0,00 1,00 0,24 0,67 6,00 0,45 0,24 0,16 0,29 0,19
13 5,00 3,00 2,30 2,00 1,00 0,52 0,32 3,00 0,27 0,11 0,16 0,68 0,11
14 6,00 1,00 1,97 2,00 2,00 0,34 0,62 4,00 0,40 0,21 0,19 0,53 0,19
15 3,00 2,00 2,70 2,00 1,00 0,28 0,61 5,00 0,50 0,14 0,17 0,45 0,22
16 3,00 3,00 2,04 2,00 2,00 0,38 0,67 4,00 0,39 0,25 0,10 0,48 0,13
17 7,00 3,00 2,40 2,00 1,00 0,40 0,41 5,00 0,43 0,12 0,16 0,57 0,27
18 4,00 2,00 2,49 2,00 1,00 0,29 0,78 4,00 0,38 0,12 0,23 0,52 0,21
19 3,00 2,00 3,03 0,00 2,00 0,25 0,75 5,00 0,51 0,25 0,03 0,22 0,25
20 4,00 2,00 2,79 2,00 1,00 0,41 0,50 3,00 0,17 0,00 0,21 0,62 0,14
21 6,00 2,00 2,32 2,00 2,00 0,29 0,74 6,00 0,44 0,21 0,11 0,39 0,23
22 4,00 3,00 1,92 2,00 2,00 0,38 0,58 5,00 0,37 0,22 0,16 0,54 0,13
23 4,00 2,00 2,36 2,00 1,00 0,31 0,77 5,00 0,35 0,07 0,24 0,55 0,26
24 3,00 1,00 2,87 1,00 1,00 0,29 0,56 5,00 0,38 0,11 0,29 0,46 0,13
25 2,00 3,00 1,50 0,00 2,00 0,45 0,43 3,00 0,61 0,45 0,13 0,32 0,16
26 1,00 4,00 0,90 0,00 1,00 0,52 0,38 4,00 0,65 0,52 0,19 0,35 0,13
27 2,00 3,00 1,58 1,00 1,00 0,49 0,50 3,00 0,25 0,10 0,49 0,74 0,13
28 7,00 2,00 2,43 2,00 1,00 0,33 0,54 4,00 0,35 0,15 0,18 0,51 0,18
29 4,00 2,00 2,63 1,00 1,00 0,32 0,97 3,00 0,28 0,15 0,32 0,63 0,09
30 3,00 2,00 2,67 3,00 2,00 0,21 0,95 6,00 0,46 0,11 0,20 0,40 0,32
31 4,00 3,00 2,64 2,00 2,00 0,23 0,82 5,00 0,49 0,19 0,15 0,38 0,26
32 6,00 3,00 1,95 2,00 1,00 0,46 0,60 4,00 0,22 0,09 0,28 0,74 0,12
33 4,00 3,00 2,68 1,00 2,00 0,27 1,00 4,00 0,49 0,13 0,27 0,33 0,35
34 5,00 2,00 2,90 2,00 2,00 0,28 0,94 3,00 0,51 0,26 0,07 0,34 0,16
35 4,00 3,00 1,51 2,00 2,00 0,40 0,73 3,00 0,38 0,29 0,18 0,58 0,09
36 4,00 3,00 2,69 3,00 2,00 0,39 0,46 4,00 0,60 0,05 0,18 0,35 0,52
37 4,00 4,00 1,85 2,00 2,00 0,30 0,83 4,00 0,44 0,25 0,23 0,52 0,15
38 3,00 3,00 2,74 2,00 2,00 0,31 0,89 4,00 0,39 0,27 0,03 0,34 0,10
39 4,00 3,00 1,24 2,00 2,00 0,33 0,95 3,00 0,40 0,31 0,27 0,60 0,09
40 4,00 3,00 2,98 2,00 1,00 0,35 0,77 3,00 0,37 0,06 0,10 0,44 0,30
41 3,00 2,00 3,00 1,00 6,00 0,29 0,83 4,00 0,49 0,19 0,29 0,43 0,06
42 7,00 3,00 1,36 0,00 2,00 0,39 0,89 3,00 0,49 0,39 0,14 0,49 0,07
43 7,00 3,00 2,85 4,00 2,00 0,25 0,93 5,00 0,52 0,05 0,13 0,37 0,47
44 2,00 3,00 1,69 1,00 1,00 0,46 0,73 2,00 0,21 0,06 0,46 0,79 0,14
45 4,00 3,00 1,89 1,00 1,00 0,27 0,95 3,00 0,45 0,26 0,27 0,50 0,14
46 2,00 2,00 1,89 0,00 2,00 0,34 0,75 3,00 0,66 0,34 0,06 0,31 0,29
47 4,00 2,00 2,00 2,00 2,00 0,33 0,70 4,00 0,41 0,23 0,16 0,49 0,17
48 4,00 3,00 2,25 1,00 1,00 0,39 0,85 3,00 0,19 0,04 0,39 0,72 0,11
49 4,00 2,00 1,49 2,00 2,00 0,31 0,85 4,00 0,45 0,26 0,25 0,55 0,17
50 4,00 3,00 2,78 2,00 1,00 0,30 0,96 4,00 0,53 0,14 0,14 0,43 0,31
51 4,00 2,00 2,60 2,00 2,00 0,43 0,38 3,00 0,39 0,16 0,06 0,49 0,19
52 6,00 2,00 2,40 0,00 2,00 0,36 0,77 4,00 0,52 0,36 0,07 0,34 0,12
53 3,00 2,00 2,26 2,00 2,00 0,33 0,55 4,00 0,34 0,18 0,18 0,52 0,16
54 3,00 3,00 3,17 1,00 1,00 0,20 0,92 6,00 0,49 0,19 0,20 0,34 0,17
55 4,00 2,00 2,09 2,00 1,00 0,33 0,80 4,00 0,35 0,14 0,27 0,60 0,18
56 4,00 3,00 2,65 0,00 2,00 0,23 0,83 3,00 0,46 0,23 0,18 0,37 0,15
57 4,00 1,00 2,20 2,00 1,00 0,35 0,68 4,00 0,33 0,15 0,24 0,60 0,17
58 1,00 3,00 2,33 1,00 2,00 0,42 0,46 3,00 0,31 0,07 0,42 0,56 0,22
59 4,00 2,00 2,95 2,00 1,00 0,35 0,74 5,00 0,27 0,00 0,25 0,60 0,16
60 3,00 3,00 3,22 2,00 1,00 0,20 0,90 4,00 0,50 0,09 0,16 0,37 0,35
61 6,00 3,00 2,25 0,00 2,00 0,32 0,55 5,00 0,56 0,32 0,12 0,30 0,17
62 3,00 3,00 1,80 2,00 2,00 0,32 0,78 4,00 0,46 0,25 0,18 0,50 0,20
63 2,00 3,00 2,54 2,00 1,00 0,26 0,80 4,00 0,42 0,19 0,21 0,47 0,17
64 5,00 3,00 2,87 2,00 1,00 0,39 0,42 4,00 0,29 0,03 0,15 0,54 0,23
65 4,00 4,00 3,37 3,00 1,00 0,25 0,83 5,00 0,44 0,03 0,18 0,39 0,34
66 3,00 3,00 1,59 1,00 1,00 0,45 0,46 4,00 0,36 0,20 0,45 0,61 0,14
67 5,00 3,00 3,67 2,00 1,00 0,22 0,77 4,00 0,48 0,05 0,15 0,38 0,35
68 4,00 1,00 2,67 2,00 1,00 0,30 0,64 5,00 0,42 0,13 0,13 0,43 0,27
69 5,00 3,00 2,64 2,00 1,00 0,44 0,64 2,00 0,23 0,01 0,28 0,72 0,15
70 7,00 1,00 2,58 2,00 1,00 0,34 0,49 4,00 0,37 0,10 0,17 0,51 0,25
71 3,00 4,00 1,84 2,00 2,00 0,33 0,69 4,00 0,43 0,22 0,19 0,52 0,20
72 4,00 2,00 2,12 2,00 2,00 0,30 0,82 4,00 0,41 0,24 0,17 0,47 0,15
73 4,00 2,00 2,04 2,00 1,00 0,33 0,84 4,00 0,34 0,17 0,27 0,60 0,11
74 4,00 2,00 2,75 2,00 1,00 0,32 0,54 4,00 0,43 0,11 0,17 0,49 0,25
75 4,00 3,00 2,30 2,00 2,00 0,21 0,97 5,00 0,48 0,20 0,20 0,41 0,26
76 3,00 4,00 2,61 2,00 1,00 0,38 0,77 3,00 0,41 0,05 0,13 0,51 0,33
77 2,00 3,00 3,37 3,00 1,00 0,27 0,69 4,00 0,55 0,08 0,16 0,34 0,38
78 6,00 3,00 2,81 2,00 1,00 0,44 0,49 3,00 0,30 0,01 0,21 0,65 0,21
79 4,00 3,00 2,21 2,00 1,00 0,29 0,78 4,00 0,47 0,14 0,18 0,47 0,31
80 6,00 3,00 2,45 2,00 1,00 0,24 0,83 5,00 0,41 0,15 0,20 0,44 0,27
81 4,00 2,00 2,57 2,00 1,00 0,28 0,73 3,00 0,40 0,19 0,20 0,48 0,15
82 4,00 3,00 1,69 2,00 1,00 0,53 0,51 3,00 0,19 0,10 0,27 0,80 0,08
83 4,00 2,00 2,85 2,00 1,00 0,29 0,68 3,00 0,38 0,08 0,20 0,49 0,25
84 5,00 2,00 1,91 2,00 2,00 0,37 0,59 4,00 0,40 0,22 0,19 0,55 0,17
85 3,00 3,00 2,18 2,00 1,00 0,35 0,60 3,00 0,37 0,16 0,21 0,56 0,17
86 2,00 3,00 2,46 2,00 1,00 0,30 0,67 4,00 0,39 0,17 0,20 0,50 0,16
87 4,00 3,00 2,39 2,00 1,00 0,44 0,53 2,00 0,20 0,06 0,23 0,67 0,13
88 2,00 3,00 8,42 12,00 4,00 0,42 0,29 2,00 0,65 0,02 0,07 0,13 0,06
89 3,00 2,00 2,87 0,00 2,00 0,31 0,61 5,00 0,55 0,31 0,10 0,29 0,17
90 3,00 1,00 2,35 1,00 1,00 0,31 0,61 4,00 0,40 0,16 0,31 0,50 0,19
91 3,00 4,00 2,69 2,00 1,00 0,29 0,95 3,00 0,27 0,01 0,28 0,57 0,26
92 3,00 2,00 4,39 0,00 2,00 0,20 0,89 5,00 0,61 0,20 0,08 0,26 0,17
93 2,00 3,00 2,46 2,00 1,00 0,29 0,96 4,00 0,35 0,12 0,28 0,58 0,17
94 5,00 2,00 2,47 2,00 1,00 0,29 0,71 5,00 0,42 0,15 0,20 0,49 0,23
95 5,00 3,00 2,19 2,00 2,00 0,26 0,87 5,00 0,50 0,22 0,17 0,43 0,26
96 5,00 2,00 3,41 4,00 1,00 0,29 0,87 5,00 0,53 0,10 0,04 0,22 0,39
97 4,00 3,00 2,66 1,00 1,00 0,32 0,70 3,00 0,31 0,08 0,32 0,55 0,15
98 4,00 1,00 2,11 2,00 2,00 0,32 0,64 4,00 0,42 0,21 0,19 0,52 0,19
99 3,00 1,00 2,11 0,00 2,00 0,27 0,73 4,00 0,51 0,27 0,19 0,38 0,21
100 5,00 3,00 1,86 2,00 1,00 0,36 0,63 4,00 0,38 0,19 0,22 0,58 0,17
10

Melodies
Corpus Corpus Corpus Corpus Corpus Size Melodies Melodies Melodies Melodies Melodies Melodies Melodies
Fado Edition Corpus
Melodic Melodic Direction of Duration of of Melodic pitch 0 pitch 1 pitch 2 pitch 3 pitch 4 pitch 5 pitch 6
Melodic Fifths
Tritones Octaves Motion Melodic Arcs Arcs
1 0,00 0,03 0,06 0,54 1,57 7,38 0,21 0,07 0,00 0,01 0,07 0,21 0,00
2 0,04 0,00 0,00 0,49 1,38 4,20 0,32 0,00 0,09 0,00 0,17 0,23 0,00
3 0,04 0,00 0,01 0,37 1,77 5,60 0,29 0,10 0,00 0,02 0,03 0,25 0,00
4 0,07 0,00 0,00 0,49 2,93 7,57 0,28 0,03 0,00 0,03 0,03 0,07 0,00
5 0,08 0,05 0,00 0,45 1,38 4,89 0,24 0,02 0,00 0,01 0,09 0,13 0,00
6 0,05 0,03 0,00 0,39 1,74 5,80 0,27 0,00 0,19 0,08 0,05 0,16 0,00
7 0,00 0,00 0,01 0,44 1,74 4,43 0,20 0,09 0,09 0,07 0,13 0,00 0,04
8 0,02 0,01 0,03 0,45 1,93 6,21 0,30 0,07 0,00 0,03 0,04 0,09 0,00
9 0,04 0,00 0,02 0,46 2,63 7,80 0,25 0,08 0,00 0,03 0,09 0,19 0,00
10 0,00 0,00 0,01 0,35 2,26 6,46 0,21 0,00 0,13 0,00 0,12 0,15 0,00
11 0,00 0,00 0,00 0,37 2,03 5,21 0,22 0,19 0,00 0,10 0,00 0,08 0,10
12 0,02 0,07 0,00 0,51 1,68 6,23 0,22 0,01 0,14 0,01 0,09 0,14 0,01
13 0,03 0,00 0,00 0,54 2,24 5,66 0,19 0,05 0,18 0,00 0,14 0,04 0,00
14 0,00 0,00 0,00 0,53 2,72 6,50 0,19 0,03 0,14 0,00 0,17 0,10 0,00
15 0,14 0,02 0,00 0,52 1,64 5,11 0,24 0,00 0,09 0,00 0,11 0,20 0,01
16 0,01 0,05 0,00 0,42 2,19 5,56 0,23 0,00 0,15 0,00 0,15 0,18 0,00
17 0,00 0,00 0,00 0,37 2,95 7,20 0,19 0,00 0,12 0,12 0,00 0,10 0,00
18 0,05 0,03 0,00 0,47 2,00 5,63 0,22 0,05 0,00 0,02 0,07 0,16 0,00
19 0,02 0,00 0,00 0,47 2,13 8,26 0,45 0,00 0,03 0,00 0,15 0,15 0,00
20 0,03 0,00 0,00 0,31 2,64 6,36 0,33 0,00 0,13 0,00 0,13 0,13 0,00
21 0,00 0,02 0,00 0,58 1,86 5,43 0,25 0,00 0,10 0,16 0,00 0,12 0,00
22 0,02 0,00 0,00 0,39 2,58 5,79 0,41 0,00 0,16 0,00 0,00 0,06 0,00
23 0,02 0,00 0,00 0,52 2,23 5,45 0,20 0,00 0,12 0,19 0,00 0,12 0,02
24 0,14 0,02 0,00 0,49 1,49 4,74 0,29 0,09 0,00 0,04 0,04 0,19 0,00
25 0,00 0,00 0,00 0,41 1,70 4,40 0,32 0,00 0,10 0,00 0,00 0,08 0,00
26 0,00 0,00 0,00 0,27 2,14 3,86 0,38 0,00 0,25 0,03 0,00 0,06 0,00
27 0,01 0,00 0,00 0,48 2,07 3,64 0,20 0,01 0,13 0,05 0,18 0,13 0,00
28 0,01 0,06 0,00 0,52 1,85 5,27 0,25 0,01 0,10 0,00 0,11 0,12 0,00
29 0,03 0,01 0,01 0,47 2,06 6,30 0,28 0,06 0,01 0,02 0,07 0,17 0,01
30 0,02 0,03 0,00 0,53 1,88 5,58 0,24 0,10 0,00 0,08 0,03 0,17 0,00
31 0,03 0,00 0,01 0,46 1,90 6,10 0,31 0,00 0,18 0,00 0,15 0,09 0,00
32 0,00 0,00 0,00 0,47 2,57 5,30 0,27 0,12 0,03 0,21 0,00 0,09 0,00
33 0,02 0,00 0,00 0,49 1,62 4,94 0,28 0,00 0,09 0,08 0,00 0,03 0,02
34 0,07 0,00 0,02 0,49 1,50 5,73 0,26 0,00 0,21 0,08 0,00 0,15 0,00
35 0,00 0,03 0,00 0,41 3,09 6,35 0,28 0,10 0,00 0,04 0,03 0,15 0,00
36 0,03 0,02 0,00 0,39 1,79 4,91 0,25 0,00 0,17 0,14 0,00 0,00 0,00
37 0,05 0,03 0,00 0,35 1,84 4,48 0,19 0,00 0,15 0,00 0,02 0,00 0,13
38 0,02 0,11 0,00 0,44 1,55 5,83 0,30 0,00 0,17 0,13 0,10 0,13 0,00
39 0,00 0,00 0,00 0,46 2,71 4,71 0,24 0,00 0,24 0,00 0,18 0,03 0,00
40 0,00 0,03 0,00 0,47 1,55 4,92 0,20 0,00 0,19 0,00 0,16 0,14 0,00
41 0,24 0,02 0,00 0,43 1,55 5,70 0,31 0,06 0,09 0,00 0,14 0,19 0,00
42 0,03 0,00 0,00 0,35 2,03 4,40 0,26 0,00 0,13 0,00 0,12 0,17 0,00
43 0,00 0,02 0,00 0,46 2,85 8,50 0,21 0,03 0,18 0,00 0,10 0,13 0,00
44 0,01 0,00 0,00 0,43 1,80 3,20 0,23 0,06 0,23 0,10 0,04 0,06 0,00
45 0,05 0,00 0,00 0,53 2,50 6,32 0,21 0,00 0,15 0,16 0,00 0,03 0,04
46 0,00 0,00 0,00 0,37 2,19 6,05 0,28 0,00 0,20 0,00 0,03 0,04 0,03
47 0,01 0,03 0,00 0,48 1,80 4,50 0,27 0,00 0,14 0,08 0,00 0,10 0,00
48 0,02 0,00 0,02 0,43 1,91 4,45 0,21 0,08 0,00 0,01 0,11 0,17 0,00
49 0,02 0,00 0,00 0,50 1,92 3,84 0,27 0,00 0,12 0,03 0,12 0,06 0,00
50 0,04 0,00 0,02 0,44 1,71 5,41 0,23 0,00 0,18 0,06 0,00 0,05 0,01
51 0,01 0,02 0,02 0,42 1,94 5,96 0,23 0,00 0,17 0,12 0,00 0,23 0,00
52 0,03 0,00 0,01 0,68 2,24 7,38 0,32 0,00 0,14 0,00 0,09 0,11 0,00
53 0,00 0,00 0,00 0,49 1,82 4,87 0,31 0,00 0,18 0,00 0,11 0,09 0,00
54 0,12 0,10 0,00 0,44 1,66 6,41 0,30 0,02 0,03 0,00 0,18 0,13 0,00
55 0,01 0,00 0,02 0,50 2,24 5,43 0,21 0,00 0,17 0,13 0,01 0,12 0,00
56 0,06 0,05 0,01 0,55 1,71 5,86 0,32 0,18 0,00 0,10 0,00 0,14 0,06
57 0,00 0,01 0,02 0,39 1,77 4,40 0,19 0,05 0,09 0,02 0,04 0,18 0,01
58 0,02 0,01 0,00 0,40 1,86 4,65 0,27 0,10 0,12 0,08 0,05 0,09 0,02
59 0,11 0,04 0,00 0,49 1,96 5,64 0,20 0,05 0,09 0,02 0,09 0,13 0,00
60 0,02 0,02 0,02 0,45 2,32 8,10 0,21 0,08 0,04 0,06 0,10 0,14 0,00
61 0,07 0,02 0,00 0,49 1,89 6,20 0,16 0,00 0,11 0,00 0,14 0,09 0,01
62 0,01 0,00 0,00 0,44 1,71 4,02 0,21 0,00 0,20 0,04 0,06 0,05 0,00
63 0,03 0,01 0,02 0,52 2,70 8,35 0,23 0,09 0,09 0,03 0,08 0,09 0,00
64 0,01 0,02 0,00 0,47 1,94 5,63 0,25 0,06 0,12 0,02 0,09 0,15 0,00
65 0,02 0,05 0,01 0,41 1,68 5,80 0,17 0,11 0,07 0,02 0,13 0,12 0,02
66 0,01 0,00 0,00 0,58 3,09 6,15 0,23 0,08 0,12 0,04 0,00 0,06 0,02
67 0,02 0,02 0,05 0,44 1,82 7,00 0,21 0,13 0,00 0,12 0,00 0,08 0,09
68 0,01 0,03 0,02 0,44 1,64 4,96 0,21 0,05 0,08 0,02 0,08 0,21 0,00
69 0,03 0,00 0,01 0,58 2,41 6,36 0,17 0,03 0,11 0,01 0,09 0,14 0,00
70 0,03 0,00 0,00 0,55 1,94 5,47 0,21 0,00 0,13 0,13 0,01 0,11 0,01
71 0,00 0,00 0,00 0,49 1,64 3,76 0,27 0,00 0,18 0,04 0,00 0,04 0,01
72 0,02 0,02 0,01 0,42 2,37 6,34 0,19 0,00 0,18 0,05 0,11 0,10 0,01
73 0,06 0,00 0,00 0,48 2,29 5,58 0,17 0,00 0,14 0,03 0,04 0,02 0,09
74 0,05 0,01 0,01 0,39 2,27 6,87 0,23 0,04 0,09 0,02 0,14 0,15 0,00
75 0,02 0,03 0,00 0,53 1,71 4,87 0,23 0,10 0,00 0,08 0,00 0,12 0,00
76 0,03 0,07 0,00 0,45 2,34 6,41 0,19 0,00 0,17 0,00 0,14 0,09 0,00
77 0,03 0,02 0,03 0,36 1,89 6,95 0,23 0,04 0,09 0,03 0,08 0,09 0,00
78 0,08 0,00 0,00 0,34 1,84 5,21 0,22 0,00 0,21 0,12 0,00 0,12 0,00
79 0,02 0,01 0,00 0,49 1,88 4,83 0,20 0,11 0,02 0,11 0,02 0,17 0,05
80 0,00 0,00 0,00 0,47 2,37 6,63 0,20 0,16 0,00 0,16 0,00 0,13 0,00
81 0,04 0,03 0,02 0,41 1,78 5,57 0,21 0,07 0,05 0,00 0,11 0,19 0,00
82 0,00 0,01 0,00 0,48 1,94 3,64 0,26 0,00 0,14 0,05 0,00 0,03 0,01
83 0,03 0,03 0,02 0,45 2,07 6,14 0,21 0,11 0,02 0,05 0,05 0,15 0,00
84 0,01 0,01 0,00 0,41 1,53 3,70 0,20 0,02 0,11 0,01 0,09 0,14 0,00
85 0,03 0,00 0,01 0,50 2,09 5,39 0,18 0,00 0,17 0,13 0,04 0,08 0,00
86 0,02 0,02 0,02 0,46 1,75 5,15 0,22 0,07 0,05 0,06 0,07 0,05 0,00
87 0,01 0,03 0,00 0,48 2,05 5,17 0,19 0,01 0,14 0,08 0,04 0,09 0,01
88 0,06 0,00 0,42 0,49 1,12 9,62 0,34 0,14 0,06 0,01 0,11 0,11 0,00
89 0,02 0,02 0,05 0,54 1,91 7,84 0,21 0,05 0,00 0,15 0,05 0,20 0,00
90 0,03 0,03 0,01 0,45 1,93 5,39 0,21 0,05 0,03 0,02 0,08 0,21 0,00
91 0,00 0,01 0,00 0,43 1,81 4,88 0,28 0,01 0,17 0,10 0,03 0,05 0,01
92 0,04 0,00 0,18 0,46 1,58 8,60 0,27 0,00 0,03 0,00 0,12 0,17 0,00
93 0,02 0,02 0,02 0,46 1,89 5,28 0,20 0,05 0,12 0,14 0,05 0,11 0,01
94 0,03 0,01 0,00 0,46 1,92 5,55 0,22 0,06 0,10 0,01 0,14 0,16 0,00
95 0,01 0,00 0,00 0,48 1,84 5,12 0,18 0,10 0,16 0,08 0,01 0,15 0,02
96 0,04 0,00 0,00 0,57 1,59 5,97 0,25 0,00 0,12 0,00 0,13 0,19 0,00
97 0,08 0,03 0,00 0,40 1,84 5,23 0,21 0,02 0,19 0,13 0,00 0,10 0,00
98 0,01 0,03 0,01 0,44 2,17 5,74 0,18 0,10 0,08 0,03 0,06 0,18 0,00
99 0,04 0,00 0,00 0,46 1,71 4,81 0,41 0,05 0,00 0,02 0,09 0,13 0,00
100 0,02 0,01 0,00 0,50 2,10 4,78 0,23 0,11 0,04 0,09 0,00 0,12 0,07
11

Fado Edition
pitch 7 pitch 8 pitch 9 pitch 10 pitch 11
1 0,12 0,17 0,00 0,14 0,00

2 0,06 0,00 0,02 0,05 0,06
3 0,17 0,08 0,00 0,06 0,00
4 0,17 0,21 0,00 0,17 0,00
5 0,17 0,18 0,00 0,16 0,00
6 0,13 0,00 0,02 0,00 0,13
7 0,00 0,16 0,06 0,00 0,17
8 0,07 0,19 0,00 0,19 0,02
9 0,14 0,12 0,02 0,07 0,03
10 0,15 0,00 0,13 0,09 0,00
11 0,00 0,19 0,00 0,12 0,00
12 0,06 0,00 0,13 0,16 0,01
13 0,09 0,03 0,16 0,09 0,03
14 0,13 0,00 0,17 0,06 0,00
15 0,14 0,00 0,13 0,03 0,06
16 0,08 0,00 0,10 0,13 0,00
17 0,15 0,18 0,00 0,15 0,00
18 0,19 0,15 0,00 0,13 0,00
19 0,09 0,00 0,06 0,06 0,00
20 0,13 0,00 0,00 0,07 0,07
21 0,15 0,07 0,00 0,13 0,00
22 0,13 0,00 0,09 0,00 0,16
23 0,19 0,07 0,01 0,02 0,06
24 0,15 0,09 0,00 0,07 0,04
25 0,06 0,00 0,22 0,19 0,03
26 0,06 0,09 0,00 0,03 0,09
27 0,06 0,01 0,09 0,00 0,13
28 0,10 0,01 0,16 0,12 0,01
29 0,14 0,14 0,00 0,10 0,01
30 0,18 0,14 0,01 0,04 0,00
31 0,09 0,00 0,07 0,00 0,11
32 0,06 0,09 0,00 0,12 0,00
33 0,17 0,09 0,09 0,00 0,14
34 0,11 0,00 0,05 0,15 0,00
35 0,16 0,09 0,00 0,16 0,00
36 0,11 0,00 0,19 0,13 0,00
37 0,18 0,00 0,16 0,18 0,00
38 0,00 0,00 0,10 0,08 0,00
39 0,00 0,03 0,15 0,00 0,15
40 0,13 0,00 0,06 0,06 0,06
41 0,02 0,00 0,09 0,05 0,05
42 0,09 0,00 0,10 0,13 0,00
43 0,10 0,00 0,10 0,15 0,00
44 0,04 0,00 0,09 0,11 0,03
45 0,13 0,00 0,09 0,19 0,00
46 0,06 0,00 0,20 0,17 0,00
47 0,14 0,13 0,00 0,08 0,06
48 0,14 0,14 0,00 0,11 0,03
49 0,06 0,03 0,12 0,00 0,18
50 0,11 0,00 0,13 0,20 0,02
51 0,11 0,00 0,07 0,07 0,00
52 0,09 0,00 0,15 0,07 0,03
53 0,06 0,00 0,16 0,09 0,00
54 0,03 0,00 0,15 0,15 0,00
55 0,18 0,10 0,00 0,02 0,05
56 0,00 0,00 0,05 0,15 0,00
57 0,13 0,09 0,06 0,11 0,03
58 0,08 0,05 0,04 0,09 0,03
59 0,14 0,07 0,05 0,14 0,02
60 0,15 0,12 0,02 0,09 0,00
61 0,14 0,03 0,15 0,01 0,16
62 0,15 0,06 0,04 0,07 0,12
63 0,14 0,02 0,12 0,07 0,04
64 0,12 0,05 0,03 0,09 0,01
65 0,12 0,06 0,06 0,13 0,00
66 0,17 0,08 0,00 0,11 0,10
67 0,01 0,16 0,02 0,16 0,01
68 0,11 0,05 0,08 0,12 0,00
69 0,12 0,11 0,07 0,13 0,01
70 0,16 0,09 0,01 0,13 0,01
71 0,16 0,13 0,00 0,14 0,02
72 0,18 0,08 0,05 0,02 0,04
73 0,17 0,00 0,16 0,09 0,09
74 0,16 0,01 0,08 0,09 0,00
75 0,14 0,16 0,01 0,16 0,00
76 0,09 0,00 0,13 0,17 0,01
77 0,10 0,04 0,12 0,16 0,01
78 0,10 0,00 0,06 0,14 0,02
79 0,05 0,02 0,09 0,14 0,02
80 0,08 0,14 0,00 0,13 0,00
81 0,13 0,08 0,03 0,13 0,00
82 0,09 0,00 0,17 0,25 0,00
83 0,18 0,12 0,00 0,12 0,00
84 0,13 0,06 0,09 0,15 0,00
85 0,15 0,02 0,08 0,10 0,04
86 0,15 0,12 0,02 0,21 0,01
87 0,10 0,07 0,08 0,07 0,12
88 0,04 0,03 0,00 0,07 0,10
89 0,15 0,15 0,00 0,03 0,00
90 0,12 0,13 0,06 0,08 0,02
91 0,12 0,06 0,06 0,03 0,09
92 0,16 0,01 0,16 0,06 0,02
93 0,08 0,04 0,08 0,09 0,04
94 0,11 0,07 0,02 0,10 0,01
95 0,10 0,04 0,10 0,07 0,00
96 0,15 0,00 0,02 0,13 0,00
97 0,13 0,08 0,02 0,08 0,06
98 0,15 0,04 0,08 0,10 0,00
99 0,06 0,05 0,07 0,12 0,00
100 0,09 0,02 0,07 0,16 0,01
12

Tiago Videira - Instrumental Fado (Complete With Appendixes) PDF

Uploaded by

Copyright:

Available Formats

Tiago Videira - Instrumental Fado (Complete With Appendixes) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tiago Videira - Instrumental Fado (Complete With Appendixes) PDF

Uploaded by

Copyright:

Available Formats

Instrumental Fado:

A Generative Interactive System

Tiago Miguel Gonzaga Videira

Tese de Doutoramento em Media Digitais

Instrumental Fado - A Generative

A thesis submitted in fulfilment of the requirements

Faculdade de Ciências Sociais e Humanas

This work was funded by FCT – Fundação para a Ciência e

António Victorino d’Almeida

Instrumental Fado - A Generative Interactive System

KEYWORDS: fado, ethnomusicology, music information retrieval, style imita-

Instrumental Fado - A Generative Interactive System

O Fado é considerado, pela UNESCO, património imaterial da Humanidade

PALAVRAS-CHAVE: fado, etnomusicologia, musicologia computacional, com-

First, and most important I thank my family. My mother, father, grandmother,

I also thank my professors – a group of remarkable scholars who aided me in

List of Figures xix

List of Tables xxiii

I On the musics and sounds of fado 1

3.4.2 Test Implementation . . . . . . . . . . . . . . . . . . . . . . 43

4 On the concept of fado 51

5.18.3 Size of Melodic Arcs . . . . . . . . . . . . . . . . . . . . . . 171

7 Fado as Memory 207

8 The Creative Processes 211

10 Future Work 221

II Automating the Composition Process 223

12 State of the art 229

13 The model 249

14.1 Interactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256

15 Implementation of the Model 261

16 Templates for rendering the MIDI files 311

18 Future Work 323

A Supplemental Material 343

3.1 Complementing Fieldwork . . . . . . . . . . . . . . . . . . . . . . . 15

5.1 Possible binary rhythms for a textual line . . . . . . . . . . . . . . . 71

5.33 Initial tonality versus first relevant modulation . . . . . . . . . . . . 129

5.77 Vibrato Prevalence . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

6.1 Fado feeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

9.1 Alternative notation to better convey “fado feeling” . . . . . . . . . 219

12.1 Fernández and Vico classification system . . . . . . . . . . . . . . . 233

13.1 A fado model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250

17.1 Model Evaluation with human subjects . . . . . . . . . . . . . . . . 319

3.1 Classification with 10 taxonomies . . . . . . . . . . . . . . . . . . . 41

5.1 Distribution of Sections . . . . . . . . . . . . . . . . . . . . . . . . . 83

15.1 Most common forms in the corpus . . . . . . . . . . . . . . . . . . . 266

On the musics and sounds of fado

Fado is considered Intangible Cultural Heritage by UNESCO, since 2011, so we

We have complemented the analysis with the use of computational musicology

In this part we will show how fado is a multidimensional concept, represent-

might be followed by a certain oral or social chain of transmission, tangible fixa-

3.1 Ethnomusicology and Historiography

objective worldview. Furthermore, he claims that it is impossible to fully under-

“The capacity, variable among peoples as it is among individuals,

Many ethnomusicologists adopted this hermeneutic: thus, an experiential and

At present, this method is widely recognized in academia, and it is fundamental

According to Nettl, music scholars believe that

In any given culture it is virtually impossible to understand exactly how new

On a related discussion, Negus presents the idea that

“there is no simple correspondence between an ethnic group and